The Great UBEC Resource Thread - RC Groups
Thread Tools
Dec 15, 2009, 11:17 AM
TMO
TMO
Registered User
TMO's Avatar
FAQ

The Great UBEC Resource Thread


Dear Fellow RCGroupers,
Through many recent posts, it seems as if there are several well informed, yet conflicting reports about the installation and application of UBEC's in our power systems. It is clear to me that UBEC installations will become more and more popular in the coming years, especially in setups using 2.4ghz radio systems. I propose this thread as a resource so that when people are searching for info on these matters, they don't have to search through the many posts which have turned into debates, for which there are no clear-cut outcomes.

Please stick to information which you know authoritatively to be true, and you would submit as advice without any reservations whatsoever. Lets not debate arcane details or exotic applications.

The topics I'd like to address are:

1) Any UBEC's commercially available that deviate substantially from the installation norm- The Dimension Sport comes to mind for example. Please title your post with the product name.
2) Paralleling a UBEC with a BEC or using 2 BEC's: is it always ok? never? With a harness? on the receiver bus? Are there any specific esc's or receivers that wouldn't tolerate it?
3) Known RTF, BNF and PNP setups that definitely should have a UBEC installed, but dont.
4) Twin motor installations- tips, gotcha'a etc.
5) Switching versus Linear BEC's and their differences as they relate to the above, especially item 2
6) anything else I didn't think of.

-Tim
Last edited by TMO; Jan 11, 2010 at 04:23 PM.
Sign up now
to remove ads between posts
Dec 15, 2009, 11:44 AM
Check my blog-updated
Mayday!Mayday!'s Avatar
How to disable the bec onboard an esc-
https://www.rcgroups.com/forums/show...5&postcount=13

System wiring with UBEC-
http://scriptasylum.com/rc_speed/_wiring.html

Castle Creations UBEC-
http://www.castlecreations.com/produ...ec_wiring.html

A different way to connect your UBEC-
https://www.rcgroups.com/forums/show...29&postcount=4
Last edited by Mayday!Mayday!; Dec 15, 2009 at 08:20 PM. Reason: Playin nice
Dec 15, 2009, 12:36 PM
RIP Azarr - "Old age is not for sissies"
Azarr's Avatar
For some unknown reason, UBEC has become a generic term. Actually they're all just add on BECs and there is really only 1 UBEC, the trademarked Ultimate BEC from Kool Flight. Which, if I remember correctly was the first add on BEC on the market.

Azarr
Dec 15, 2009, 12:52 PM
F-OFF
imyohero's Avatar

dimension sport BEC


I run the dimension sport BEC, It's passthru feature means you dont have to remove the positive lead when connecting the esc to bec. Also if needed you can run two becs parallel. This bec has been rock solid for me on fm and 2.4 no interferance or glitching on either system.
Dec 15, 2009, 01:12 PM
TMO
TMO
Registered User
TMO's Avatar
Here's a link to a member's site which shows peak servo amp draws using an oscilloscope: http://homepages.paradise.net.nz/bhabbott/Servo.html
His results suggest peak draws occur around center.

It may assist those who are trying to calculate an "amp budget" on their setups. I would add 25% to each of the numbers for budgeting purposes, as a stalled servo or a binding linkage will draw even more. If there is more data around on this topic I think it would be useful. For those that may not know http://www.servodatabase.com/ is a useful resource in general, but lacks the sort of data as I linked above.

When your budget is complete, factoring in the aforementioned extra, I'd be sure your BEC or separate BEC (yes Azarr, i probably shouldn't use that UBEC term generically) has significant overhead still.

So for example, say I'm setting up a Taylorcraft 450. I like hs-55's on the wing, and hs81's for rudder and elevator. So my budget for the hs-55's is 500mA each (400 + my fudge factor of 25%) for 1A. Next the 81's need 875mA each (700+ 25%) for 1.75A. Total servo current budget is therefore 2.75A (unlikely to happen but.....). Given these assumptions, a 3A BEC on 2S voltage (which is what they are usually spec'd for) should be fine. On 3S, there's not enough overhead there for me to be real comfortable. This is an installation on 3S where I would be installing a 5A external BEC. This is just my way and I'm sure you'd be just fine on a typical 3A BEC, but if you want security-for sure and no foolin'- my setup gives it.
Last edited by TMO; Dec 15, 2009 at 08:53 PM.
Dec 15, 2009, 01:15 PM
TMO
TMO
Registered User
TMO's Avatar
...
Last edited by TMO; Dec 16, 2009 at 11:01 PM.
Dec 15, 2009, 01:21 PM
Warbird E-Pilot / Hack Builder
I am using Hobby City Hextronic brand 3A BEC

(http://www.hobbycity.com/hobbycity/s...idProduct=3735)

Are these worth using? Price is only 7 bucks.
Dec 15, 2009, 02:25 PM
Go Hawks!
cryhavoc38's Avatar
Quote:
Originally Posted by Larry076
I am using Hobby City Hextronic brand 3A BEC

(http://www.hobbycity.com/hobbycity/s...idProduct=3735)

Are these worth using? Price is only 7 bucks.
Its a good bec for the money.
you will see that very same bec sporting different names sold by different companies, for more money than what HC charges.

Same with this one and I prefer it over any other smaller bec on the market. They are bulletproof and is made by Hobbywing
http://www.hobbycity.com/hobbycity/s...oise_Reduction
Dec 15, 2009, 08:06 PM
TMO
TMO
Registered User
TMO's Avatar
Here's a tip on soldering a separate BEC together with the ESC leads onto a Deans. For anybody who has done it, you know that it takes a little dexterity. You can use soldering paste to hold the two leads together, but I never do. Here's my method: EDIT: See pictures on post 16.

1) usually when soldering a deans, I only expose about 1/4 inch of the wire lead, and this is usually the length that is exposed when they come out of the package for most ESC's. When doing a separate BEC, I expose enough wire to cover the whole spade- Approximately 3/8ths.

2) Next, I slide the (always too short imho) shrink that comes in the package over the two, and shrink just above the joint, so that the the bonded pair has about 1/4 inch showing. This holds them together.

3) Now slide a larger diameter shrink, like dubro white (1/4") or preferably 5mm over the joined lead and slide it up the joined pair and out of the way. Repeat for the other lead.

4) Now, solder to the Deans Ultra in the usual way. And shrink the connections as normal with the larger diameter shrink.

I always give the whole thing a few wraps with electrical tape right near the connector so the tape wrap ends up flush with the deans t block. I do this because: over long periods, I dont trust shrink, especially when you solder to the outside of the Deans pos (+) spade. Second, I have fat fingers and deans can be a pain, so the elec. tape gives you something to grab onto. Finally the tape wrap reinforces the area a little, so wire contortions and twists etc. dont exert their whole force on the solder joint.

I know people who use that rubberizing stuff, or Hot Glue to accomplish the same, but Elec. tape is fine with me.

PS. Even though I'm still a Deans guy, do it one or twice with an EC3 or EC5 and your faith will be tested. It's really easy, and if the leads are pretinned, you really dont even have to tin your iron.

PPS: If you missed it: look up at Mayday Mayday's post, the second one at the top. The last link shows a pretty neat way to add a separate bec without going to any trouble.
Last edited by TMO; Dec 16, 2009 at 10:32 AM.
Dec 15, 2009, 09:15 PM
Screw the wind, let's fly!
Naval Warfare's Avatar
Quote:
Originally Posted by cryhavoc38;
Same with this one and I prefer it over any other smaller bec on the market. They are bulletproof and is made by Hobbywing
[url
http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=4319&Product_Name=TURNIG Y_3A_UBEC_w/_Noise_Reduction[/url]
With regard to that one.I know the jumper changes it from 3a to 5a, but
I dont know if it (in it stock form is set to 3 or 5.) Anyone know how to set that to 5a? From what I've just read, my running 5 HXT900's, on 3a is not a good thing. Good thread btw.
Thanks.

Nav
Dec 15, 2009, 10:58 PM
TMO
TMO
Registered User
TMO's Avatar
Another thing I wanted to address on this thread: the use of capacitors on the RX bus. I've installed the spektrum branded one for pan car racers. For those who don't know, pan car/ carpet racers are limited to 4.8V packs and specific motor winds in competition. They don't typically use a separate RX pack. A capacitor is often installed directly on an open receiver channel which discharges at ~4V- thus allowing the rx to hold rf signal when line voltage temporarily spikes (troughs?) due to flooring the throttle. Here is a linky: http://www.horizonhobby.com/Products...ProdID=SPM1600

OK, so I like the concept. It won't power your receiver for any period of time per ce. If your BEC on your ESC goes into thermal shutdown- you're still toast. But, if you get a temporary dip in bec voltage, this **may** provide enough to the bus to keep your rx from rebinding or going to failsafe mode. It seems to work, but then again, I only have it one plane.

Is there enough evidence from the people who really, really understand this stuff to support the idea that this could be a cheap insurance policy on a marginally powered BEC system like 4 hs-55's w/ a 2A BEC? Could this be the ticket for the BNF and PNP owners who are reporting what seem to be brownouts on stock systems?
Last edited by TMO; Dec 16, 2009 at 08:21 AM.
Dec 16, 2009, 08:55 AM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by Naval Warfare
With regard to that one.I know the jumper changes it from 3a to 5a
Not true. The jumper is for 5v vs 6v. I found replacing the jumper with a pot converts it into a very high performace DC supply adjustable from 1v to 30-40v. Good for charging batteries, testing motors, circuit development, etc..

3a and 5a are the continuous and burst current ratings respectively. By opening the shrink I found they run 5-6a continuous no problem.
Dec 16, 2009, 10:03 AM
Screw the wind, let's fly!
Naval Warfare's Avatar
Quote:
Originally Posted by rich smith
Not true. The jumper is for 5v vs 6v. I found replacing the jumper with a pot converts it into a very high performace DC supply adjustable from 1v to 30-40v. Good for charging batteries, testing motors, circuit development, etc..

3a and 5a are the continuous and burst current ratings respectively. By opening the shrink I found they run 5-6a continuous no problem.
My bad. I meant to say the voltage 4.8-6. So, I just take the blue jumper off totally to increase it?
Dec 16, 2009, 10:03 AM
Go Hawks!
cryhavoc38's Avatar
Quote:
Originally Posted by Naval Warfare
With regard to that one.I know the jumper changes it from 3a to 5a, but
I dont know if it (in it stock form is set to 3 or 5.) Anyone know how to set that to 5a? From what I've just read, my running 5 HXT900's, on 3a is not a good thing. Good thread btw.
Thanks.

Nav
Rich has it correct. The jumper changes the voltage from 5v to 6v.
I like running my electronics at 6V for the increased speed and torque in the servo's.

If the jumper is ever removed accidentally in flight, the system defaults to 6v.

The bec is rated at 3A continuous, 5A max according to the specs.

Rich, that is a very interesting find. Adding a trim pot to adjust the actual output voltage. Thats a pretty cool find.
Dec 16, 2009, 10:22 AM
Screw the wind, let's fly!
Naval Warfare's Avatar
Quote:
Originally Posted by cryhavoc38
Rich has it correct. The jumper changes the voltage from 5v to 6v.
I like running my electronics at 6V for the increased speed and torque in the servo's.

If the jumper is ever removed accidentally in flight, the system defaults to 6v.

The bec is rated at 3A continuous, 5A max according to the specs.

Rich, that is a very interesting find. Adding a trim pot to adjust the actual output voltage. Thats a pretty cool find.
so I just remove the blue pin to up it to 6v?
Dec 16, 2009, 11:01 AM
Warbird E-Pilot / Hack Builder
I have a Turnigy Plush 25A ESC in my GWS Zero with 3 HTX900 servos on a 3 cell Lipo. The on board BEC is supposed to be 2 amps. Over 100 flights with no brown out issues. I am building a GWS Me-109 with the same power set up, and don't want to add the seperate UBEC if I don't really need it. Have I just been lucky with the Zero, or is it safe to run the 3 servos on the Turnigy?
Dec 16, 2009, 02:28 PM
TMO
TMO
Registered User
TMO's Avatar
Probably, maybe, etc. This thread is really about making sure, though. The peak draws on an hxt900 are around 750ma. 99%of the time, you're not going to peak all three at the same time.

But..........it could happen. And if it did for any length of time, the BEC would be stressed.

You may be a candidate for the capacitor setup as I described above in post #11

-Tim

Quote:
Originally Posted by Larry076
I have a Turnigy Plush 25A ESC in my GWS Zero with 3 HTX900 servos on a 3 cell Lipo. The on board BEC is supposed to be 2 amps. Over 100 flights with no brown out issues. I am building a GWS Me-109 with the same power set up, and don't want to add the seperate UBEC if I don't really need it. Have I just been lucky with the Zero, or is it safe to run the 3 servos on the Turnigy?
Jan 09, 2010, 11:34 AM
Registered User
Quote:
Originally Posted by rich smith
Not true. The jumper is for 5v vs 6v. I found replacing the jumper with a pot converts it into a very high performace DC supply adjustable from 1v to 30-40v. Good for charging batteries, testing motors, circuit development, etc..

3a and 5a are the continuous and burst current ratings respectively. By opening the shrink I found they run 5-6a continuous no problem.

Hi,
Do you have details of the connections to attach a potentiometer in place of the link?
Do you need to connect all three pins to the pot (centre pin to pot common)?
What range of pot?

Many Thanks.

Kev.
Jan 09, 2010, 12:42 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by Kev.E
Hi,
Do you have details of the connections to attach a potentiometer in place of the link?
Do you need to connect all three pins to the pot (centre pin to pot common)?
What range of pot?

Many Thanks.

Kev.
Center to pin 1 (FB) of thr 1510, ends to gnd and Vout. Here's a pic for the shielded one:
Jan 09, 2010, 10:18 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by joseph 1
Hello, I need some information, I noticed your posts on the turnigy UBEC and you say it can run 5-6 Amp continuous. Its marked what I thought 3A continuous and burst 5A. Did you do a test on the output ? You said you pulled back the shrink I assume you ran a test. Also what is the weight on it in grams. Any help appreciated thank you.

Yes I tested the output overnight with no moving air and although coil ran hot nothing failed. Exposed to airflow it should run cooler. Weight of the one in the picture was 9.5g with shield and toroid removed and other leads added.
Jan 10, 2010, 03:36 AM
Registered User
Aaah! I get it now.
It's time to go and tinker with the resistance box.

Thanks very much.

Kev.
Jan 10, 2010, 10:03 AM
Flapping Rudder
danamas's Avatar

Why use a UBEC?


TMO and others:

I am not well versed on this topic and hope not to clutter up your thread on this topic. I am just curious, what is the purpose of the UBEC? Is it to provide a backup for the ESC in case of ESC failure/overheating? Is it to provide a cooler ESC operation and better voltage supply to the servos? Would not a larger than needed ESC provide the same protection? If the ESC fails/overheats, does the ubec(additional bec) only provide battery power to the servos or does it continue to provide power to the motor?

I appreciate your response to a noob as I am considering using a ubec(extra bec) in my planes to supplement the esc(w/bec) if it is really necessary. Thanks!
Last edited by danamas; Jan 10, 2010 at 10:14 AM. Reason: Pose a question
Jan 10, 2010, 11:37 AM
TMO
TMO
Registered User
TMO's Avatar
Although there are some that work on the backup principle, the majority are an entirely separate power supply circuit for providing power to the receiver and servos. They step down and regulate the voltage from the main flight pack. BEC's on ESC's work the same way. The linear BEC's on typical ESC's generate heat, especially when stepping down higher voltages to 5V. As a result, when using 3S lipo's or higher, we often go to using a separate BEC. More expensive ESC's often have a switching BEC, which can usually tolerate higher voltage and produce higher current. In some high servo count applications, a separate BEC may be the only way to provide enough current for the servos. Separate BEC's usually take the place of the ESC's BEC entirely, and we disconnect the red power wire on the ESC lead when we install them. On a separate BEC system, typically, an ESC failure will not lead to loss of control, but it is usually the case for an ESC based BEC.

Quote:
Originally Posted by danamas
TMO and others:

I am not well versed on this topic and hope not to clutter up your thread on this topic. I am just curious, what is the purpose of the UBEC? Is it to provide a backup for the ESC in case of ESC failure/overheating? Is it to provide a cooler ESC operation and better voltage supply to the servos? Would not a larger than needed ESC provide the same protection? If the ESC fails/overheats, does the ubec(additional bec) only provide battery power to the servos or does it continue to provide power to the motor?

I appreciate your response to a noob as I am considering using a ubec(extra bec) in my planes to supplement the esc(w/bec) if it is really necessary. Thanks!
Jan 11, 2010, 04:09 PM
Registered User
Well, I stripped off the shrink wrap and de-soldered the metal shield to find that the converter/IC had had it's identification number ground off when it was manufactured (very fine, buffed, figure of eight marks all over the surface of the chip, totally obliterating the etched identifications).
There is no way to tell if it is a 1510 step down converter that is fitted on this board.
The board markings identify it as a Hobbywing 212v2.0
3A, max 6A, 2-6s li-poly.
I may still have a bash at soldering some wandering leads to the IC terminals to see if the output on this one can be adjusted, if I can keep my hand steady! It's really delicate.

If anyone has advice feel free to jump in now!
Jan 11, 2010, 04:19 PM
TMO
TMO
Registered User
TMO's Avatar
What's the prevailing wisdom on running parallel BEC's? Personally I don't see why it can't/shouldn't be done, but I'm no expert. As long as both are putting out the same voltage, why not? Maybe a diode if you want to be a boy scout about it. Why do the manufacturers of UBEC's (almost always) advise disabling the ESC's BEC? Aren't the majority of BEC's and UBEC's using the same chip anyway?
Jan 11, 2010, 04:30 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by Kev.E
Well, I stripped off the shrink wrap and de-soldered the metal shield to find that the converter/IC had had it's identification number ground off when it was manufactured (very fine, buffed, figure of eight marks all over the surface of the chip, totally obliterating the etched identifications).
There is no way to tell if it is a 1510 step down converter that is fitted on this board.
If it resembles my photo it is definitely a 1510. The circuit is identical to the app note which also tells you where to put the pot (Ra Rb). Apparently they don't bother to erase the chip in the cheap version because it's not cost effective. These can be had for $3 and change elsewhere.
Jan 11, 2010, 04:35 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by TMO
What's the prevailing wisdom on running parallel BEC's? Personally I don't see why it can't/shouldn't be done, but I'm no expert. As long as both are putting out the same voltage, why not? Maybe a diode if you want to be a boy scout about it. Why do the manufacturers of UBEC's (almost always) advise disabling the ESC's BEC? Aren't the majority of BEC's and UBEC's using the same chip anyway?
Actually UBECs (switchers) are completely different than BECs (linear). Diodes don't work for this. Also note that "prevailing wisdom" is often an oxymoron around here. Kinda like "military intelligence".

Other than that you answered your own question.
Jan 11, 2010, 05:58 PM
My hangar is,... "almost" full
No Step's Avatar
It appears Kool Systems UBECs are not available anymore, at least I can't find any online. I guess I'll have to try CC's BEC. ...... but they look so tiny
Jan 11, 2010, 09:09 PM
TMO
TMO
Registered User
TMO's Avatar
Could I parallel up a switching UBEC with a switching ESC BEC?

Quote:
Originally Posted by rich smith
Actually UBECs (switchers) are completely different than BECs (linear). Diodes don't work for this. Also note that "prevailing wisdom" is often an oxymoron around here. Kinda like "military intelligence".

Other than that you answered your own question.
Jan 12, 2010, 07:14 AM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by TMO
Could I parallel up a switching UBEC with a switching ESC BEC?
Yes.

Putting regulators in parallel is not only safe but allows for more current than either of them alone.
Jan 13, 2010, 05:38 PM
Registered User
Wrecked the regulator!
From various checks and tests I figured out that the IC/chip wasn't a 1510 on this one but I did work out the pin arrangement.
I had it working fine, feeding a receiver and two servos at 5.4V until I decided to load it up and connected it to the bench power supply at 11.1V then proceeded to load it at 3A via the charger/discharger connected to the same power supply - BANG!
Split the chip! Big crack on one side of the regulator chip.
Should have run the UBEC from a separate source.
I think I caused a short by connecting it all to the same source.
Anyway, It's a learning curve - and it's steep!
Gonna get another one or two and play again.

Tee Hee Hee.
More Scotch required!
Jan 13, 2010, 06:23 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by Kev.E
Wrecked the regulator!
From various checks and tests I figured out that the IC/chip wasn't a 1510 on this one but I did work out the pin arrangement.
I had it working fine, feeding a receiver and two servos at 5.4V until I decided to load it up and connected it to the bench power supply at 11.1V then proceeded to load it at 3A via the charger/discharger connected to the same power supply - BANG!
Split the chip! Big crack on one side of the regulator chip.
Should have run the UBEC from a separate source.
I think I caused a short by connecting it all to the same source.
Anyway, It's a learning curve - and it's steep!
Gonna get another one or two and play again.

Tee Hee Hee.
More Scotch required!
That'll do it. Next time use a bulb. I use #1157 brake lights set up with jumpers for 1a, 2a, 4a, and 6a. You should show us a photo.
Jan 14, 2010, 06:51 AM
Cajun-American
Boogie_'s Avatar
Great thread!

I'm relatively new to electrics & EFDs and run BEC in all my jets. But I was curious about something. I know to get high watts output in jets (or anything else) you can increase the cell count (voltage) and keep your amps lower.
Wouldn't switching the BEC to 6v effectively lower the amp draw the same way??

Julian
Jan 14, 2010, 07:13 AM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by Boogie_
Great thread!

I'm relatively new to electrics & EFDs and run BEC in all my jets. But I was curious about something. I know to get high watts output in jets (or anything else) you can increase the cell count (voltage) and keep your amps lower.
Wouldn't switching the BEC to 6v effectively lower the amp draw the same way??

Julian
A BEC (linear) will generate less heat knocking down to 6v instead of 5v but total watts is the same. 5v and 6v both run cool on a UBEC (switching) and total watts consumed is significantly less.
Jan 15, 2010, 08:10 AM
Kamikaze Ace
Glacier Girl's Avatar
Something else that I haven't seen mentioned, and should be for newbies not in the know.

UBEC's or stand alone BEC's output is related to input voltage. The higher your input voltage the lower the amp rating. And not knowing this can cause some serious problems if you think you have one rating when you actually have a much lower one.

Direct from Castle's site on their Castle 10 amp BEC.
Below 11 volt input you get 10 amp peak
12 volts, it drops to 7 amps
24 volts, it drops to 5 amps
And all of these ratings are with a 5 mph airflow across the BEC.

In my bigger birds I've gone away from hooking the UBEC into my flight packs.
I'm either running a receiver pack, or hooking the UBEC to a separate lipo pack on the ones running 9S power, as the CC unit's max input is 6S.

Also from Castle's site. The standard servo connector is rated for a maximum of a 5 amp load, so if you plan on running more then 5 amps to your rx, you will need multiple servo leads from the UBEC to do it.

Which is kinda funny as the CastleBEC rated for 10 amps, comes with a single lead, and the new Pro version at 20 amps, only comes with two.

If following Castle's info, you will need to double the amount of leads if using them at max ratings.
Jan 15, 2010, 10:27 AM
Registered User
I read the same thing on Castle's web site and I think there may have been similar info on Dimensionengineering's site.
I notice Turnigy has a 5A/7.5A max UBEC and it may be better suited for continuous 3A output when running on 11.1V input, it will certainly have an easier life and run cooler.
Jan 15, 2010, 06:09 PM
Martin - AKA mr.sneezy
PLMS's Avatar

Bench test on Turnigy 5A UBEC


Guy's,
I posted this mini-review on the HobbyCity forum.
https://www.rcgroups.com/forums/show....php?t=1175777

Far from being a definitive review though, somebody needs to verify my results before I'd say too much. My unit may be below par (read 'faulty').

Also interested in all your thoughts on a unified test procedure for all UBEC. Must use minimal test gear (CRO, adjustable load, multimeter, capacitors maybe).
Suggestions welcome.
Martin
Jan 16, 2010, 04:26 PM
We want... Information!
Bruce Abbott's Avatar
Quote:
Originally Posted by PLMS
My unit may be below par (read 'faulty').
Not 'faulty' as such, but definitely over-rated. The IC used in that unit is only rated for 3A continuous, and typically current limits at 4.8A (minimum 3.6A). Not surprising that you couldn't get the claimed 5A out of it!
Jan 16, 2010, 07:46 PM
Martin - AKA mr.sneezy
PLMS's Avatar
Quote:
Originally Posted by Bruce Abbott
Not 'faulty' as such, but definitely over-rated. The IC used in that unit is only rated for 3A continuous, and typically current limits at 4.8A (minimum 3.6A). Not surprising that you couldn't get the claimed 5A out of it!
Hi Bruce, good to hear from you. Nice detective work, thanks for the extra details. So, given it's been over-rated and very noisy it gets a thumbs down from me thus far.
Regards,
Martin
PS. It might get used as a power supply for some onboard LED lighting instead now.
Jan 17, 2010, 07:57 AM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by PLMS
Hi Bruce, good to hear from you. Nice detective work, thanks for the extra details. So, given it's been over-rated and very noisy it gets a thumbs down from me thus far.
Regards,
Martin
PS. It might get used as a power supply for some onboard LED lighting instead now.
I regularly get 6a long enough for readings on a test stand. The IC is temperature not current controlled in this situation so airflow is quite relevant. I get about half that with the shrink on.

I also get less current out of the shielded one which has a much smaller coil that gets hot very quick. I suspect it is gauge, core saturation, and heating of copper rather than the IC which is the real limit.
Jan 21, 2010, 07:02 AM
Registered User
Medvidek's Avatar
Will this cuase any problems?
Jan 21, 2010, 12:34 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by Medvidek
Will this cuase any problems?
No sense running micro-power RX off strong 6v and amp hog servos off weak 5v. Unless those are REALLY big servos UBEC is probably not needed at all. If you must put a UBEC in there at all cost then best bet is connect UBEC to RX and plug servos into RX.

Servos are a little faster off 6v. May be the only reason to actually pull the ESC red wire because voltages are different.
Last edited by rich smith; Jan 21, 2010 at 12:40 PM.
Jan 21, 2010, 03:28 PM
B Mac
mach7rc's Avatar
Quote:
Originally Posted by imyohero
I run the dimension sport BEC, It's passthru feature means you dont have to remove the positive lead when connecting the esc to bec. Also if needed you can run two becs parallel. This bec has been rock solid for me on fm and 2.4 no interferance or glitching on either system.
Hello imyohero. I have the Su-34 EDF with twin 45amp esc's. I ordered a sport bec. I was told that if I install the sport bec it would automatically turn off the onboard bec's in the esc's because they are in parallel and one plug to the receiver. I am pushing 10 servos, would it be a good idea to purchase another sport bec and install it in parallel that way I would have two sport's in place. If so would the second one install the same way but leaving the lead to the receiver uninstalled. Any words of wisdom would be appreciated. Thx.
Jan 21, 2010, 04:18 PM
Flying motor mount master
fly_boy99's Avatar
I would say a lot of folks just buy something not knowing what their load is (servo, lighting, etc).

You can start from either end but I would advocate starting from the known instead of the unknown. Meaning it would help greatly if you measured your current draw before buying that BEC.

Eagletree, Medusa (out of biz), BNB all provide products that will help in the quest for knowledge!

If you don't like "military intelligence" you might like "Genuine imitation"!

B
Jan 21, 2010, 05:08 PM
Martin - AKA mr.sneezy
PLMS's Avatar
Quote:
Originally Posted by rich smith
I regularly get 6a long enough for readings on a test stand. The IC is temperature not current controlled in this situation so airflow is quite relevant. I get about half that with the shrink on.

I also get less current out of the shielded one which has a much smaller coil that gets hot very quick. I suspect it is gauge, core saturation, and heating of copper rather than the IC which is the real limit.
Hi, hey are you talking about the old Hextronic 5A UBEC or the new Turnigy/Otter 5A UBEC ? The reason I ask is that I don't think there is a shielded version of the new one I reviewed ?

If you did test the older one, do you have other specs recorded like ripple and amps before regulation roll-off etc. Thanks for your input.
Martin
Jan 21, 2010, 06:35 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by PLMS
Hi, hey are you talking about the old Hextronic 5A UBEC or the new Turnigy/Otter 5A UBEC ? The reason I ask is that I don't think there is a shielded version of the new one I reviewed ?

If you did test the older one, do you have other specs recorded like ripple and amps before regulation roll-off etc. Thanks for your input.
Martin
As far as current tests I checked the original Hextronic (no LED), the new Hextronic (LED), and the shielded one in the picture. Testing involved hooking up a load and seeing how long it goes before shutting down. They were either on or off with no warning or sign of droop. I have the Turnigy 5a and 12a and $4 cheapies but no testing yet. Ripple is something I check on AC supplies but not switchers so can't help you there.
Jan 21, 2010, 06:54 PM
Martin - AKA mr.sneezy
PLMS's Avatar
Quote:
Originally Posted by rich smith
As far as current tests I checked the original Hextronic (no LED), the new Hextronic (LED), and the shielded one in the picture. Testing involved hooking up a load and seeing how long it goes before shutting down. They were either on or off with no warning or sign of droop. I have the Turnigy 5a and 12a and $4 cheapies but no testing yet. Ripple is something I check on AC supplies but not switchers so can't help you there.
I a bit confused now (it's Friday I guess). I don't see a picture of a shielded one in your post (or mine). Thanks.
Jan 21, 2010, 07:02 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by PLMS
I a bit confused now (it's Friday I guess). I don't see a picture of a shielded one in your post (or mine). Thanks.
I took the shield off. Maybe you can see the 4 holes where it was soldered in.
Jan 21, 2010, 07:18 PM
Martin - AKA mr.sneezy
PLMS's Avatar
Quote:
Originally Posted by rich smith
I took the shield off. Maybe you can see the 4 holes where it was soldered in.
Ah, now I'm with you. Yup, that's not the one I tested. That one is probably much better...
Jan 26, 2010, 03:22 PM
TMO
TMO
Registered User
TMO's Avatar
One thing I'd like to get cleared up:

does the capacitor- (which we all agree was originally designed for 4 cell carpet racers)- do anything to mitigate the brownout problem when installed on the bus of a 2.4gHz RX? We all know that it, nor any other capacitor that we could fit in a plane, cannot "power" the rx for anything but an instant. Yet despite urging from some people that its effect is placebo at best, I can say that in the single case in which I installed it, it really seemed to fix the problem. And at least in theory, if not this one, there should be some capacitor of some specification that could function in this way- i.e.: as an insurance policy against a momentary drop in RX voltage below the 3.8V critical threshold.

The specs on the Spektrum one (p/n spm1600) are listed as: 4700uf and 10V

I know I don't know enough to answer the question, but I'd like to know.
Jan 26, 2010, 10:15 PM
GTH NFL
Park_Flyer's Avatar

CC BEC Output Voltage


Can anyone provide a recommendation on the output voltage setting for a CC BEC? I am using a CC BEC for the first time on a relatively simple 4s model set up - two servos, LS SPJ pusher config. The servos are rated at 4.8v and I'm using a Spektrum 6100 RX, but could swap that out if recommended.

I know that the default voltage on the BEC is 5.1v. Is there any reason to change that, i.e., should it match the ratings of the servos? Would it be better to run it a bit higher in the event of voltage drop outs that might affect the Spectrum rx?

Any recommendations, thoughts?

Thanks,

Ben
Jan 26, 2010, 10:54 PM
TMO
TMO
Registered User
TMO's Avatar
Nope, leave it alone. 5.1v is just fine and shouldn't tweak the servos or anything. Most of them would be ok all the way up to 6v.
The CCBEC is a switching BEC capable of 7+ amps on 4s so your setup won't faze it one bit. You aren't going to see any voltage sag in that system at the RX. You're good. Go fly. Enjoy.

-Tim

Quote:
Originally Posted by Park_Flyer
Can anyone provide a recommendation on the output voltage setting for a CC BEC? I am using a CC BEC for the first time on a relatively simple 4s model set up - two servos, LS SPJ pusher config. The servos are rated at 4.8v and I'm using a Spektrum 6100 RX, but could swap that out if recommended.

I know that the default voltage on the BEC is 5.1v. Is there any reason to change that, i.e., should it match the ratings of the servos? Would it be better to run it a bit higher in the event of voltage drop outs that might affect the Spectrum rx?

Any recommendations, thoughts?

Thanks,

Ben
Jan 27, 2010, 09:21 AM
GTH NFL
Park_Flyer's Avatar
Quote:
Originally Posted by TMO
Nope, leave it alone. 5.1v is just fine and shouldn't tweak the servos or anything. Most of them would be ok all the way up to 6v.
The CCBEC is a switching BEC capable of 7+ amps on 4s so your setup won't faze it one bit. You aren't going to see any voltage sag in that system at the RX. You're good. Go fly. Enjoy.
Thanks Tim. This is a very helpful thread. Why not request to make it a sticky?

My next BEC project will be a twin-motor set up with retracts (P-38) and will likely have 7 - 8 servos. Based on what I've read here, I'm rethinking the configuration and will likely have a battery for each motor (installed in the nacelles) and a separate battery for the BEC installed in the fuse.

Thanks again.
Jan 27, 2010, 10:55 AM
TMO
TMO
Registered User
TMO's Avatar
Hey,
I had an idea for a new product, or mod project. What about making a tiny UPS for our RX's? This would make the capacitor debate moot. It's no better or cheaper than just going out and buying a UBEC, but I'm just going for concept here.

Here's the idea, but at this point I'm lacking any "science." We use one of those single cell lipo's that we are using in the Micro Helicopters- typically about 3g, 120-150 mAh in capacity, 3.7v Nominal. Fully charged it's 4.15-4.2V We find some sort of micro relay that opens the circuit at say 4.0V. I was thinking that one of those voltage alarms (like this one for example only: http://www.bphobbies.com/view.asp?id...1&pid=C1436801) or similar device could be modified to open its relay at the necessary voltage.

Then we would have a basic uninterruptible power supply to the RX bus. We wouldn't have the whole battery's capacity, since once it discharged a bit, the voltage would drop down to where we would brown out anyway, but we'd get a few seconds at least.

Would this work? Where's Rich? If I need a smackdown I want entertainment value at least.

-Tim
Jan 27, 2010, 03:36 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by TMO
One thing I'd like to get cleared up:

does the capacitor- (which we all agree was originally designed for 4 cell carpet racers)- do anything to mitigate the brownout problem when installed on the bus of a 2.4gHz RX? We all know that it, nor any other capacitor that we could fit in a plane, cannot "power" the rx for anything but an instant. Yet despite urging from some people that its effect is placebo at best, I can say that in the single case in which I installed it, it really seemed to fix the problem. And at least in theory, if not this one, there should be some capacitor of some specification that could function in this way- i.e.: as an insurance policy against a momentary drop in RX voltage below the 3.8V critical threshold.

The specs on the Spektrum one (p/n spm1600) are listed as: 4700uf and 10V

I know I don't know enough to answer the question, but I'd like to know.

The real drain that prevents a big cap from solving brownout is servos. If you use a diode with the cap to isolate from the servos then brownout is virtually eliminated. Even w/o the cap it helps big time.
Jan 27, 2010, 04:51 PM
TMO
TMO
Registered User
TMO's Avatar
Could you describe how to do this? Is it something we could all do on the cheap? You mean a diode from the BEC output, right?

Quote:
Originally Posted by rich smith
The real drain that prevents a big cap from solving brownout is servos. If you use a diode with the cap to isolate from the servos then brownout is virtually eliminated. Even w/o the cap it helps big time.
Jan 27, 2010, 06:02 PM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by TMO
Could you describe how to do this? Is it something we could all do on the cheap? You mean a diode from the BEC output, right?
Regardless of whether powering from ESC, UBEC, or battery pack connect servos as normal to the 5v but put a low power diode (1n914) between RX power pin and the 5v.

The RX which draws almost no power compared to servos will run for seconds or even minutes after battery is removed if big cap is added. 6v is even better than 5v because servos are snappier and hold time is longer.
Jan 27, 2010, 08:43 PM
TMO
TMO
Registered User
TMO's Avatar
If I understand you, the diode should go between the positive on whatever the power source may be, and the pin where you connect it on the RX's power bus. I presume an extra servo connector could be modified for the task. By extension, this diode would keep current from leaking back to the power source when and if the power source's voltage dropped beyond the critical level. Am i getting it?

Quote:
Originally Posted by rich smith
Regardless of whether powering from ESC, UBEC, or battery pack connect servos as normal to the 5v but put a low power diode (1n914) between RX power pin and the 5v.

The RX which draws almost no power compared to servos will run for seconds or even minutes after battery is removed if big cap is added. 6v is even better than 5v because servos are snappier and hold time is longer.
Jan 28, 2010, 02:44 AM
Dance the skies...
Tom Frank's Avatar

Servo Loads on BECS


Today I made two flights with the Dimension Engineering "ServoSense" amp draw logging device to measure the load on my ESC BEC. This device logs the Max. instantaneous amp draw (in .1A increments), the Average amp draw for the entire time it's powered (in .05A increments), and the Max. 15 sec. Average amp draw during the flight (in .05A incrments). The readout is done after a flight by counting flashes of two different colored LEDs which cycle through these three recordings until the device is unplugged. One flash equals one increment, so for example, 8 flashes for the Max. instananeous amp LED would equal 8 X .1A = .8 A. The ServoSense is connected in series between the ESC BEC RX plug and the RX itself. So it records amp draw for everything being powered by the BEC... in my case just the servos and RX (very small load and is being ignored).

My goal was to take these measurements on a plane with all HXT900 servos since they have been labelled as having one of the highest amp draws in a test using oscilloscope traces to measure peak amps. Unfortunately my planes that have just these servos were not easily accessed to connect the ServoSense, so I had to make do with the planes I had.

I made two measurement flights, one with a small 11 oz. HL YAK 3D plane, and one with a 19 oz. E-Flite Tribute FX 3D plane. Both planes have 45 deg. throws on all control surface surfaces, and the flights were similar, including some gentle pattern flying as well as some very agressive 3D maneuvers and continuous snap rolls (for 5 secs.) with all servos at full travel. MY goal was to give measurements reflecting how I normally fly, not necessarily as a torture test. Here's the tabular data for those flights:

.......................HL YAK....................E-Flite Tribute

Servos:.........(2) HXT900 (R/A)....(4) Futaba 3114 (R/E/2A)
...................(1) Futaba (E)

Max. Amps......... 1.2A .................... 1.6A
Instantaneous

Max. Amps........ .15A .................... .20A
15 sec.

Ave. Amps........ .10A .................... .10A
While powered

I was most interested in the Max. amp data. The Futaba and HXT900 servos have virtually identical torque and speed ratings, but the scope tests I mentioned above did not test the Futaba servos. Being of presumed higher quality, one might assume they have lower power consumption than the HXT servos. So waht does all this tell me?

For the Tribute flight, each Futaba servo drew an average of:

- 400 mA Max. Instantaneous
- 50 mA MAx. 15 sec. Average

For the YAK flight, each servo drew an average of:

- 400 mA Max. Instantaneous
- 50 mA MAx. 15 sec. Average

No difference just looking at averages!

The E-Flite plane is close to twice as heavy as the YAK with much bigger control surfaces, so one should expect the in-flight loads to be considerably larger. My conclusion is the Futaba servos do draw less power than the HXTs by virtue of the average amp draw for each servo on the larger plane being the same as the smaller plane, but how much lower is conjecture. Since one of the servos on the YAK was a Futaba, it's likely it was drawing less than the average and the HXTs were drawing more, but how much is conjecture.

The lab test with oscilloscope I mentioned above recorded peak instantaneous current for an HXT900 at 700 mA, with NO LOAD on the servo! I won't argue with an oscilloscope, so the significant difference in data must boil down to the accuracy and sampling method of the ServoSense device. Perhaps it's recording more of an "RMS" peak than a true instantaneous pulse maximum peak?

My main concern was for the loads being long enough and high enough to cause BEC overheating and shutdown. I have not experienced that (yet?), even with 4 servos on a 2A rated BEC on 3S voltage. Based on this, my conclusion is the ServoSense measurements are giving a more realistic representation of that concern than an oscilloscope instantaneous pulse trace.

I'm open to other interpretations!
Jan 28, 2010, 07:56 AM
7000mw of raw power!
rich smith's Avatar
Quote:
Originally Posted by TMO
If I understand you, the diode should go between the positive on whatever the power source may be, and the pin where you connect it on the RX's power bus. I presume an extra servo connector could be modified for the task. By extension, this diode would keep current from leaking back to the power source when and if the power source's voltage dropped beyond the critical level. Am i getting it?
Exactly. And also keeps the servos from dragging RX voltage down.

Cathode (black band) to RX, anode to 5v or 6v source (the ESC in my test). Remember servo red wires bypass the RX and go directly to the 5v source.
Jan 28, 2010, 08:54 AM
TMO
TMO
Registered User
TMO's Avatar
Dear Tom,
On behalf of everyone, thanks for sharing this data. I appreciate the systematic approach and consistency too. Even Bruce (the 'scope tester) in threads here on RCgroups felt that the scope readings should not be extrapolated too far.

Also, the highest peak traces result from that first turn of the servo motor ( Bruce's words here: "When a servo makes small movements it draws a higher peak current than when moving a larger distance. Why? When the servo motor is stationary, only its (low) armature resistance limits current flow. Once the motor starts turning it generates a voltage which opposes the power supply, so the effective voltage is less and current drops. ")

But it raised awareness, and that was good. Another thing about his data is that I wonder whether he was operating these servos on 6V. There's a great opportunity here: If someone with a 'scope was willing, I for for one would be willing to contribute servos to create a database using controlled conditions and have it be something that could be an ongoing project, like Dr. Kiwi does with motors.

It is also worth pointing out that you are a very experienced pilot/builder and I bet your linkages and setup is more or less a "best practices" datapoint. For the typical newbie arf or chinese factory RTF, I'd add at least 25% to those numbers that you got from servo sense even on the same equipment. I'm worried about people searching this thread and saying something like " well Tom got 400ma and therefore I can safely run 5 on my 2A BEC.

The difference between the instaneous and the 15sec amp draws is an eyeopener! It reinforces some of my thinking on this thread that what may be missing is not BEC's with bench power supply ratings, but rather the types of appraches like the one Rich and I have been working on- means of spike power draw attenuation.

Anyway, thanks again.

Tim


Quote:
Originally Posted by Tom Frank
Based on this, my conclusion is the ServoSense measurements are giving a more realistic representation of that concern than an oscilloscope instantaneous pulse trace.

I'm open to other interpretations!
Last edited by TMO; Jan 28, 2010 at 10:38 AM.
Jan 28, 2010, 03:22 PM
Dance the skies...
Tom Frank's Avatar
Quote:
Originally Posted by TMO
Dear Tom,
On behalf of everyone, thanks for sharing this data. I appreciate the systematic approach and consistency too. Even Bruce (the 'scope tester) in threads here on RCgroups felt that the scope readings should not be extrapolated too far.

Also, the highest peak traces result from that first turn of the servo motor ( Bruce's words here: "When a servo makes small movements it draws a higher peak current than when moving a larger distance. Why? When the servo motor is stationary, only its (low) armature resistance limits current flow. Once the motor starts turning it generates a voltage which opposes the power supply, so the effective voltage is less and current drops. ")

But it raised awareness, and that was good. Another thing about his data is that I wonder whether he was operating these servos on 6V. There's a great opportunity here: If someone with a 'scope was willing, I for for one would be willing to contribute servos to create a database using controlled conditions and have it be something that could be an ongoing project, like Dr. Kiwi does with motors.

It is also worth pointing out that you are a very experienced pilot/builder and I bet your linkages and setup is more or less a "best practices" datapoint. For the typical newbie arf or chinese factory RTF, I'd add at least 25% to those numbers that you got from servo sense even on the same equipment. I'm worried about people searching this thread and saying something like " well Tom got 400ma and therefore I can safely run 5 on my 2A BEC.

The difference between the instaneous and the 15sec amp draws is an eyeopener! It reinforces some of my thinking on this thread that what may be missing is not BEC's with bench power supply ratings, but rather the types of appraches like the one Rich and I have been working on- means of spike power draw attenuation.

Anyway, thanks again.

Tim
Hi Tim... thanks for the comments.

Yes, I can fully understand the spike when starting the servo motor, as well as the huge increase as a motor current draw when heavily load, or worse, stalled. I do get my linkages and hinging to be as low resistance as possible... I used tape "Z" hinging on the two test planes which allows the surface to rotate 180 deg. in both directions with virtually no resistance. Also, the two planes used had very short, straight, stiff external pushrods with no rubbing, so you make a good point about the servo static load being a starting concern in addition to flight loads. Maybe I should add a "disclaimer" warning to my posting about this?

I also am thinking of just doing some more bench testing with all HXT900 servos loaded with known torque demands (a simple weight on a string attached to the servo arm rotating in the vertical plane), not only to compare the ServoSense readings, but also to see if they can deliver the rated torque without stalling.

I didn't measure my BEC voltage, but both planes used Castle Creations ESCs, so I think it will be close to 5V with no load, not 6.

I had a question on the scope traces maybe you or Bruce could explain. The various servos not only had different voltage peaks, but also very different pulse widths. Am I correct in assuming the horizontal time scale is the same for all the photos? If that's the case, I noted the HXT900 had one of the longer pulse widths, about 13 times longer than the similar size (but slower and lower torque) HS-55. Then I see the similar sized GWS Naro has both a short and (4 times!) longer pulse width in the same trace photo. The servo speed and torque are obviously gearing dependent as well as motor dependent, but there's not as much difference between specs of the similar sized servos that might explain the wide variance in the pulse widths. Can you explain what is going on with this? Is it just due to differences in how far or how fast the servos were being stroked when the trace photo was captured? The pulse width (duty cycle) is a factor in the BEC heating issue in addition to the peak currents, so I'm just trying to better understand what the scope traces are showing.

FWIW, the Max. 15 sec. average current was something Dimension Engineering added at my suggestion when I was one of their beta testers for this device. I was interested in the BEC heating aspect then as well, not just the current peak value. Maybe the time period is a little long which brings the average down, but at the time I felt this was a reasonable time interval for an aggressive series of aerobatic maneuvers that might not give the BEC much time to cool between maneuvers. Kudos to Dimension Engineering for adding it... not sure why this neat little device is no longer being sold, must not have had much demand.

A final comment... I ALWAYS make sure I have airflow across my ESC, using internal baffling if felt necessary. Essential for minimizing BEC heating.
Jan 28, 2010, 08:24 PM
TMO
TMO
Registered User
TMO's Avatar
Dear Tom,
Regarding your question about the distinctive waveforms of the different servos: I believe that is an artifact of the motors themselves- each one with distinct character as a result of its interaction with the control circuit and potentiometer. He said he had a servo tester set on slow rotation and since analog servos are "power on" only once per pulse, it appears to me that the horizontal scale must be about .4ms. I arrive at this purely by assumption that a servo's neutral pulse width is 1.5ms, He said that he observed the peaks around center, and the divisions for most servos in the graphs seem pretty tightly grouped at about 3.5-4ms. Anyway, I suspect the reason why the traces are not a uniform length or form has to do with the power circuit-potentiometer interaction. I'm going to PM Bruce to see if he like to add some color here. I think if we were looking at a readout of a complete cycle for the servos is would be lots and lots of pictures (!) so I suspect what we are seeing here is a sort of "greatest hits" for each servo. I wouldn't assume these waveforms looked the same for the servos throughout their output range.

When seeing the waveform of digital servo its easy to see why they are so power hungry- they operate more or less continuously! I didn't really understand this before seeing the 'scope graphs. As your real life data has shown, maybe we can all just relax a bit in the analog world, but a full house digital setup can really tax a system.

Regarding Dimension, I suspect the eagletree stuff and telemetry based systems backed them into a corner a bit, and they said to themselves: "just how bad do want to be a player in this space at this pricepoint"? Just my thought and I don't know anybody over there....

Best,
Tim
Jan 29, 2010, 10:30 PM
We want... Information!
Bruce Abbott's Avatar
Quote:
Originally Posted by Tom Frank
I also am thinking of just doing some more bench testing with all HXT900 servos loaded with known torque demands (a simple weight on a string attached to the servo arm rotating in the vertical plane).
Good idea. Comparing your results to Servormances (a French servo test database) should be interesting.

Quote:
I had a question on the scope traces maybe you or Bruce could explain.... Am I correct in assuming the horizontal time scale is the same for all the photos?
Yes. The horizontal timebase was set to 5mS/division, which is 50mS total (a bit more than 2 frames).

Analog servos send one pulse to the motor per frame. The pulses get shorter as the servo output arm gets closer to the target position, thus slowing the motor and (hopefully) preventing overshoot. For the same purpose a Deadband region is also created, which removes power from the motor just before the arm reaches its target position. Mechanical inertia of the motor and drive chain causes it to keep going a short distance after power is removed.

Quote:
I noted the HXT900 had one of the longer pulse widths, about 13 times longer than the similar size (but slower and lower torque) HS-55.
The HS-55 has a coreless motor with a very lightweight armature, therefore it accelerates rapidly and only requires a short pulse to get up to speed.

The HXT900 motor's armature is of the heavier iron cored variety (do not believe any advert that says it is coreless!) which has high inertia, therefore requiring a longer pulse to get going or change direction.

Quote:
I see the similar sized GWS Naro has both a short and (4 times!) longer pulse width in the same trace photo.
The reason for this is that the scope is only showing a couple of frames which are asynchronous to the servo's positional error, so individual pulses will vary in length. I took several snapshots for each servo, and tried to select one that was representative. If the servo mechanism has high inertia then you would expect to see a greater proportion of long pulses.

You can also see what appear to be commutator switching spikes in the pulse, which get closer together as the motor picks up speed (both the GWS Naro and HXT900 show this quite clearly). Large spikes are an indication of high inductance (typical of iron-cored motors) perhaps combined with an undersized power supply smoothing capacitor.

Quote:
Originally Posted by TMO
Another thing about his data is that I wonder whether he was operating these servos on 6V.
All the servos were tested on a variable power supply set to 5V.

Quote:
He said he had a servo tester set on slow rotation
I should clarify this. My servo tester was set for a small deflection around center, continously reversing about 5 times per second. The idea was to capture the peak current caused by the motor being forced to change direction. This is the kind of stress that a servo often encounters when hovering a 3D plane or helicopter, when controlled through a gyro, or when flying with a standard (non-DSP) PPM receiver in weak signal conditions.
Jan 29, 2010, 11:00 PM
Dance the skies...
Tom Frank's Avatar
Bruce,

Thanks for the clear and concise explanations... it all makes sense now!
Jan 29, 2010, 11:20 PM
TMO
TMO
Registered User
TMO's Avatar
Yes, thanks Bruce. I just learned more about servos than in the last 3 years!

Bruce, would you be willing to go so far as to give some recommendations of some of your favorite analog servos in the basic 6g, 9g, and 15g ranges? I understand that your choices would generally be application specific, but for the purposes of this thread, could you make recommendations for good all-around performers with moderate amp draws.

Also, where do you stand as far as peak draw attenuation at the RX power bus? Should we just be installing UBEC's and forgetting about the issue, or do you see any merit in pursuing a capacitor/diode arrangement in a marginally powered linear BEC setup?

Thanks in advance,
Tim
Mar 22, 2013, 11:43 PM
RC Aircraft Enthusiast
Marvie100's Avatar
https://store.rc4wd.com/mobile/product.asp?itemid=1479

Anyone seen these for EC5 connectors?
Mar 25, 2013, 06:30 PM
Registered User

ubec and esc bec


my esc has a bec, do I just disconnect it and hook up the ubec
(this is what I normally do when I only have a receiver on board) but now I have added a guardian flight controller.
The guardian normally gets its power for operation of itself and the servos connected to it from a three wire plug on the aileron channel. All of the other signals are made with only a signal wire connected to the receiver.

My thought was to use the esc's bec to power the receiver and the ubec to power the guardian and the servos. I would pull the black and red wires from the guardians aileron connecter and connect them to the ubec.

My question is that since the signal from the receiver is originating from a different power source than what is being applied to the guardian could this be a problem?

I had always been taught (Marklin Manual)you should never parallel power sources, but that comes from model trains where there would be a danger of electrocution if one of the transformers was left on and back fed the other resulting in 120v coming out of the plug that could zap you should you come into contact with it.

Am I making a big to do out of nothing and should I just disconnect the esc's bec.
Mar 28, 2013, 02:51 PM
TMO
TMO
Registered User
TMO's Avatar
Yes, just remove the red pin from the esc's throttle connector plug, or if you want a more permanent fix, just cut the red wire. Signal and Ground are still attached, and this is plugged into the throttle channel. Then the UBEC is jumped off the battery, and the connector for it is placed into a different channel on the RX (usually the "batt/bind").

So I don't think you are paralleling your power source. My guess is that your guardian is drawing power, not providing it. It may have some Capacitors that provide drop out protection; but unless it has a true switching regulator onboard, that won't affect the BEC. Let's figure this out first.

So I think it will be just fine drawing power as you indicated before- while plugged into a servo channel on the RX. The only source of power into your system should be the UBEC. All the red and black wire/pins on an RX are already paralleled- they're set up in a straight parallel bus. Plug power into ANY channel and you power the whole bus. This is also true of negative (-) which is why you don't end up with a floating ground on the throttle (it's already common since the throttle and the ubec are both plugged into the RX.

I'm thinking you can stick with just powering it off of an unused channel. Make sure it can't feed switched power back in to the RX- Capacitors are OK, but a true BEC switching regulator should never be paralleled with another- even of the same type.


Hope it helps





[QUOTE=davovinch;24531480]my esc has a bec, do I just disconnect it and hook up the ubec
(this is what I normally do when I only have a receiver on board) but now I have added a guardian flight controller.
The guardian normally gets its power for operation of itself and the servos connected to it from a three wire plug on the aileron channel. All of the other signals are made with only a signal wire connected to the receiver.

My thought was to use the esc's bec to power the receiver and the ubec to power the guardian and the servos. I would pull the black and red wires from the guardians aileron connecter and connect them to the ubec.

My question is that since the signal from the receiver is originating from a different power source than what is being applied to the guardian could this be a problem?
/QUOTE]
Last edited by TMO; Mar 28, 2013 at 02:58 PM.
Mar 29, 2013, 11:15 AM
Registered User
E-Challenged's Avatar

Using the Hitec Optima Receivers SPC Port


Hitec thinks that powering your Optima receiver directly from the motor battery and seperately powering your servos from an add-on BEC, or a flight battery pack, is safer/more reliable, in terms of supply voltages than powering both the Optima receiver and servos from an add-on BEC. That may be correct, but if there is a break in power to the receiver or servos, you will still lose control regardless.

Note: Hitec Optima receivers are designed to operate on up to 35 volts so connecting the pigtail to the SPC port soldered to the ESC power leads along with the leads to the add-on BEC will powr but not hurt the receiver. Be sure to plug the servo-type connector into the SPC port with proper polarity. (Don't ask!!)

The attached diagram/drawing shows a special inter-connection setup using an old Astro Flight 25 geared brushed motor, old AF 211 speed control( no BEC or LVC circuit) , Kool Flight UBEC, and SPC port to seperately power my Optima 6 receiver from the SPC port and servos from the UBEC. Not shown is the Li-Saver, add-on low voltage cutoff (LVC) device, not included in the old AF 211 speed control. I am using this setup in an old Sig 1/6 scale Cub build project.

I have been using Dimension Engineering's Park BEC's, rated at 1.5 amps, with small to medium sized models using old brownout prone, voltage sensitive, 2.4GHz Spectrum AR6000 receivers, 3S lipos and four HS-55 servos for some 6 years, with no more unexplained loss of control crashes.

I have read that some more current receivers are more tolerant of low voltage dips and reset much quicker than older receivers which may explain why some claim to be using speed controls with linear BEC's, 3S lipos and four small servos with no brownout/loss of control problems and no need for a "steenking" add on switch mode BEC. Many still think that using a high amp-rated speed control guarantees that the BEC circuit will somehow power their receiver and X number of servos with 3S or 4S lipos. I keep advising that a tiny almost weightless $20 add-on switch mode BEC is cheap insurance.

I have been advising people to read up on add-on BEC's on Dimension's website. Many ESC sellers are not yet specifying what type of BEC is included in their speed controls and what the capacity is in terms of lipo cell count vs number of servos.
Last edited by E-Challenged; Mar 29, 2013 at 11:57 AM. Reason: clarification
Apr 08, 2013, 10:58 PM
RC Aircraft Enthusiast
Marvie100's Avatar
Quote:
Originally Posted by Marvie100
https://store.rc4wd.com/mobile/product.asp?itemid=1479

Anyone seen these for EC5 connectors?
Never mind, I just made my own


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Mini-HowTo E-Sky Lama V4 Resource Thread - Parts, Mods., Tuning, Upgrades, Etc. blkdog7 Coaxial Helicopters 12392 Feb 25, 2017 07:01 AM
Discussion thread emphasis via image/emoticons for communication enhancement resource thread!!!! Accu157 Life, The Universe, and Politics 4 Aug 01, 2006 03:16 AM
resource thread for Ausies woodsy Foamies (Kits) 0 Jul 27, 2004 12:37 AM
Great resource on airfoils used by full scale aircraft. Thomas B Scale Kit/Scratch Built 0 Mar 03, 2003 07:06 PM
A great resource for the foam flying wing addict. Thomas B Foamies (Kits) 0 Nov 17, 2002 03:36 PM