MicroScream: $5 Audible Lipo Alarm

[jimsky]

I have a few Plantraco DSP4-SC combination receiver/ESC in my planes. I really like these aside from one drawback, no low voltage lockout feature. So far no problems, but I'd like something to prevent an overdischarge lipo situation.

I didn't like the LED style lipo alarms, so I designed what I really wanted: a cheap, lightweight, very loud audible alarm that will tell me without a doubt it's time to land the plane and swap battery packs. Less than 5 bucks of parts and about 20 minutes with a soldering iron you can have one too. I've call this alarm the MicroScream.

I know many say, you will know when" time's up" due to preformance roll off if you don't have an ESC with a low voltage feature. But with this little alarm no need to even think about it. When you hear the alarm, land the plane.

I selected a small 85 decibel piezo alarm, it's pretty darn loud. Just a few simple parts are needed...it's all outlined on the attached schematic. The last one I built weighed only 2.9 grams and I wasn't even trying to do a lightweight build. See the attached photo of it on my scale.

I've tabulated the schematic for a 2 cell and 3 cell version, but there's no reason why it can be modified for additional cell usage.

I've learned a lot here and figured I'd like to give something back. So here you go: print the schematic, order you parts and have fun!

Jimsky

[Andy W]

Very cool..
I'll move this with redirect over to DIY .. I think a reference to this thread in the LiPo sticky thread would be worthwhile, however..
..a

[Larry3215]

Very cool! I would like to use this with hi voltage setups as well. What would the values be for 4S, 5S, 6S, 9S, 10S etc. Is there a formula to use to calc the values?

Thanks!

Larry

[jimsky]

There are two things that need to be adjusted for additional cells. The value of R1 to set the alarm trip level, and R4 to drop voltage so X1 doesn't see more than 7 volts maximum.

The voltage developed by the R1/R2 divider that connects to the "Refrence" pin of the TL431 wants to be 2.5 volts when the battery is it's warning point (that would be 3 volts per cell used). Select R1 to drop the appropiate voltage using Ohm's law.

The piezo I picked has a maximum voltage of 7 volts, so a dropping resistor is needed as the battery voltage goes above 2 cells. I used the specs on the piezo as a starting point, but then trimmed the resistor via testing.

I'll work up a more encompassing tabulation sheet and add it to this thread.

Jimsky

[Marty H.]

Very nice, I was looking for another little project, after building my lipo charger. This will work well with my 5A gws esc and experimental foam planes. Thanks jimsky.

[jimsky]

Regarding information for additional cells, I'd like to review this a bit more. There's an issue regarding the maximum voltage the gate of the 2N7000 can tolerate. With high cell count battery packs this would need to be addressed.

Presently I only own brushed motor 2 cell configuration planes. I know technically you can stack up as many cells as you like, but what it the practical maximum number of cells that are typically used (barring science project builds)?

Jimsky

[blueprints]

Would it be possible to hv adjustable voltage trigger e alarm like variable resistors or sort? for e.g setting it to trigger @6.2V for 2 cell configuration. Thanx.

Jem

[jimsky]

Sure you could add a pot. But that's weight, cost, size and is it really set right? When I was designing this alarm light weight and small size was my goal, so I went with fixed resistors. Personally I think 3 volts per cell is a good alarm setting, here's my logic.

Testing my SuperFly-E with a GWS 350 motor, 8060 prop and Tanic 2 cell 2200 battery pack I found the following. With a fully charged pack the battery voltage dipped 0.75 volts going from "motor off" to "motor full on". This can be thought about as the "load regulation" of the battery pack, the voltage change from "no load" to "full load" due to it providing current to the motor.

This needs to be taken into account when selecting the alarm trip point, otherwise you'll be restricting your flight time. You can always land earlier before the alarm sounds, and my guess is most people will, due to degraded preformance. The alarm is there to tell you must to land, but it still has some saftey margin to boot.

My thinking is as follows:
You've been flying quite a while (motor on - loading the battery) and the alarm sounds in the air indicating the battery voltage has hit 6.0 volts. You back off the throttle, the battery voltage increases (less loading) and the alarm will actually go off. Land the plane, disconnect battery. If you were now to measure the actual battery voltage you will find it's higher than 6.0 volts due to the load regulation of the battery. Happy, well exercised but not over depleted lipos.

As long as you don't fly around till the cows come home with the alarm sounding (pulling the battery well below 6.0 volts) you should be fine. When the plane starts to get sluggish, keep it close by and do frequent fly bys to listen for the alarm. When you hear the alarm, land the plane and you should be fine.

Jimsky

[blueprints]

Jimsky,

I muz say u've really gave me quite a run down. Right i guess ure right. This was designed with weight n size to keep in mind. Can't wait to build my own. Juz like to know wad does Yageo mean. 3.57K by parallel connection u mean? Also would it be absolutely critical for e resistor values to be exactly precise for the alarm to work. Thanx for sharing ure great idea. Appreciate it.

Jem

[blueprints]

Jim,

Kinda difficult to get fixed value resistor such as 3.57K here in singapore. hv to put them in parallel or something.... gonna be messy n bulky... where'd u got ure resistors?

[jimsky]

Yageo is the manufacturer of the 1% resistors that Digikey sells.
R1 & R2 form the voltage divider that sets the alarm trip point, that's why I used 1% resistors. You could cobble up values using series/parellel combinations, but it will increase size and weight. The 1% resistor values I specified are all industry standard values. Radio Shack doesn't carry 1% resistors, but any real electronics distributor should. As indicated in the notes section of the schematic, you can buy all the parts at:

www.digikey.com

Jimsky

[robert harik]

This would be great for hand launch gliders with li-polys, where you dont have any noticable perfomace drop(no motor, just servos).
Thanks, Rob

[HELModels]

That is neat!

I do have a question about controlling the cathode current on the TL431 with R3. How is the value of R3 selected?

It looks like 2N7000 Gate must be considered somehow? I realize Vout= (1 + R1/R2) * Vref and that is how R1 and R2 are picked. I'm just curious and also realize there is a minimum of 1mA cathode current, which R3 allows.

[jimsky]

Cathode current for the TL431 is specified as a minimum of 1mA and a maximum of 100mA, so pick R3 accordingly.

I did mention in a previous post that the maximum voltage the gate of Q1 sees must be considered for higher cell configurations.

Jimsky

[wildtech]

Hi
Is there some alternate part numbers that we can use for U1 and Q1?

I am having problems finding these devices in Australia.

regards

[clipclop]

Try http://www.farnellinone.com/ for parts in Australia
Stewart

[jimsky]

U1, the TL431 is the heart of the device. I don't think you can easily replace it and use the existing circuit. I did make an identical device using a separate zener reference and comparator. It used more parts, was larger, heavier and consumed much more current.

Q1 could be replaced with a BJT instead of a logic level FET, but you would have to address the high OFF voltage of U1 (assuming it was still used). This would require modifications to the circuit.

So yes other parts can substituted, but changes will most likely be required to address the differences.

Jimsky

[gedaso]

Futurelec (http://www.futurelec.com) stock the TL431CLP but not Q1.

To be honest, I'm not sure where these guys are, but they have an Australian office and offer domestic shipping prices. I've ordered components from them before without drama.

As clipclop mentioned, Farnell is another option but the postage tends to be a bit more expensive (quick though).

[imeins]

Ignore this - my brain was in neutral !!! Just saw all the other Farnell posts.

----------------------------------------

Try the highly overpriced but "have everything" people at Farnell here in Oz. www.farnell.com etc.

Ingmar

[pardus]

Hi
Brilliant Idea, don't like the guessing game much
I have a 6S2P LiPo pack and I'd like to modify this circuit for 25.2V (18v alarm), Unfortuantely my knowledge of electronics is a little limited.
What would I need to change?
Thanks in advance
Graham

[jimsky]

The 2 cell and 3 cell version of this alarm was my goal, and that's what I designed. The reference to higher cell counts was "just thinking out loud", I have not reviewed this, nor have I done any calculations, prototyping or testing. In fact, all I presently own are 2 cell battery packs.

The more I've thought about higher cell count versions of this alarm, the more issues of concern I have found with this simple circuit. I'll give this some though but I'm afraid more than just resistor values might have to change.

Circuits distilled down to utter simplicity are economically elegant, but these application focused creations tend so loss their ability to be flexible. In essence they were designed to do what they do...with limitations.

Things that were not issues before, become problems needing solutions. The problem is, you don't know what all the problems are until you make one to test it and see.

I'll try to find some time to give this some thought, and perhaps prototype a circuit.

Jimsky

[pardus]

Thanks Jimsky, I appreciate it. I'm about to head off to my garage to start construction of the 3S one. I was hoping that a few resistor values might be all that needed changing for 6S but I can see that at such high voltages more things might have to change. Why the big manufactures like FMA haven't done something like this, I don't know - it seems like the way to go

My ESC's highest cutoff is 12v so my LiPo's will be damaged by the time it gets anywhere near that but of course the motors performance will be severely degraded by the time the batteries reach 18v so I should notice but I would, (like many) like to know exactly when it gets there.
thanks again
Graham

[FlyByMike]

Will this do?

http://www.battsignal.com/

and only $24.95

-Mike

[wildtech]

Hi

I have tried building this circuit but the buzzer is always on.

How critical are the Voltage divider values?
What voltage does the Reference pin require to turn on/off?
Currently I have 2.53 for a 2 cell battery.

Any help appreciated.

[jimsky]

If you use the 1% resistor values shown, taking the reference tolerance and divider tolerance into account the trip point accuracy should be less than or equal to +/- 2.5% of 6.0 volts (2 cell) and 9.0 volts (3 cell). I think having accurate trip points are important, that's why 1% resistors were specified.

The internal reference is 2.495 volts, nonimal.

One issue might be is you didn't follow appropiate static procedures and blew the 2N7000. If the transistor is shorted, the piezo would always be on. It's important to follow static procedures when working with semiconductor components. Grounded wrist strap, soldering iron tip grounded, ...all that good stuff. Yes, it is important.

If that's not it, it's possible it might be another situation.

The TL431 does have a rather high OFF voltage. If you get a 2N7000 with a low gate threshold voltage (it's possible if you check the minimum specs, could be as low as 0.8 volts) that's not "typical" value of 2.1 volts...then you might have a problem. Basically the piezo would always be ON.

I've built a bunch of MicroScreams and yet to run into this scenario, but technically it could happen. I checked all the ones I've built, and had plenty of margin to prevent this issue.

My experience is most parts come in as "typical", but cheap component manufactures give themselves an out specifying wide tolerances. If the 2N7000 is not shorted, try replacing it in hopes of getting one with a higher gate threshold voltage.

Jimsky

[wildtech]

HI

Unsoldered the gate and manually shorted out Q1 and it works as expected.
So it is not the 2n7000 causing the problem

As I am unfamiliar with the TL431 at what volt should it short or go open to make the circuit work?
I take it it is biased on via Divider R1 and R2 and once the Ref exceeds a low voltage (2.495) it is triggered off and R3 pulls high the Gate.

If this is correct then it may be the TL431 that just won't pull R3 voltage down to ground. I'll try replace on Monday when I can get more.

regards

[jimsky]

Do a web search on the TL431 and download the datasheet, it will help explain things.

I had stated before, the TL431 has a high OFF voltage. When it's OFF it does not go to "zero volts" or ground, not even close. This voltage may turn a 2N7000 "ON" if it has a low gate threshold voltage. This means it's possible the 2N7000 would always be ON.

If a BJT (even lower ON voltage) was used for Q1, something (more parts) would have to be done to get rid of the high OFF voltage of the TL431. The logic level FET addresses this without the need for extra components. But like I said it's possible the tolerances can stack up against you.

Jimsky

[jimsky]

The MicroScream alarm posted in this thread works quite well. I have one in all my planes, it provides real peace of mind when flying. All my personally built units in my planes are ultra lightweight "dead bug" style builds, using magnet wire to connect to the battery. A couple of installations have the piezo buzzer remote from the electronics, optimizing the location of both the electronics and the piezo buzzer.

I posted the schematic so people handy with a soldering iron could build it for themselves for the cost of the parts. For those who would prefer to buy a complete and tested MicroScream low voltage lipo battery alarm; they can be purchased at:

www.like90.com

These units are fabricated using a more formal printed circuit board build technique. There are great for people who what the peace of mind having a battery alarm, but don't want to build circuits. Presently there are 2 cell and 3 cell versions being offered. Attached are a couple of photos of the finished alarm.

Jimsky

[**neons**]

Hi Jimsky,
Just purchased a unit. Can you tell me the best way to connect to the battery or whatever way that you suggest. I am using Sermos -Anderson connectors if that means anything. I plan to attach it to different planes as I fly them so it is portable. I use the Jeti ESC's that has a red cover and is not setable voltage for 3s lipos. Nice idea. I just happened to find this tonight.
Thanks **Neons**

[jimsky]

Well the MicroScream monitors the battery voltage, so it needs to be connected to the right to the battery wiring. Red wire to battery positive, Black wire to battery negative. That doesn't leave a lot of options on how to connect it. It has to be on one side the battery buss or the other...battery connector side, or ESC side.

The only way I see to make it "portable" would be to add wires & JST connector to one of these locations, and put the mating JST connector on the MicroScream. And then do this to your other planes. This way you can simply move the MicroScream from plane to plane using the JST connector to "tap" into the battery line.

Jimsky

[Friar_Tuck]

great looking idea! Thank you jimsky!
You did try to throw us though by including an R4 on your pictured two cell unit.

[jimsky]

Nope, not throwing nobody. :D

Beleive it or not, but they manufacture and sell zero ohm resistors...that's what it is. It's basically a jumper.

Jimsky

[Friar_Tuck]

What a hoot, a zero ohm resistor! Why didn't I question the band:{ I went to order up some components last night, and digikey is out of the alarms.
I think this is a list of Mouser's part numbers.

http://www.mouser.com/

595-TL431CLP
Texas Instruments Regulators
Adj Shunt


781-2N7000
Vishay/Siliconix N-Channel MOSFETs
TO-92 60V 0.2A 0.4W


660-CF1/4L151J
KOA Speer 1/4Watt Axial Leaded Carbon Film Resistors
150 OHM 5% BULK


271-3.57K
Xicon 1/4W 1% Metal Film Resistors
OR,GN,PR,BR,BR 50PPM


271-6.65K
Xicon 1/4W 1% Metal Film Resistors
BU,BU,GN,BR,BR 50PPM


271-2.49K
Xicon 1/4W 1% Metal Film Resistors
RD,YW,WH,BR,BR 50PPM


291-1.5K
Xicon 1/4W 5% Carbon Film Resistors
1/4W BR,GN,RD,GD


665-AI-155
Projects Unlimited Alerts
EM 85dB 2300Hz 5VDC

It was late when I was compiling these guys, so doublecheck!

[**neons**]

Thanks Friar Tuck,
You saved me a lot of time. I got an order in for the parts.
**Neons**

[utahrc]

What is the weight of the manufactured version? Any chance of going to SMT? I'd love an SMT kit myself, but I don't suppose there are many folks who are comfortable soldering SMT by hand.

[jimsky]

The guys at www.like90.com have both the 2 cell and 3 cell versions of the MicroScream in stock.

Since this was a "do it yourself" project, I think a surface mount version would be out of the skill realm of many individuals. But that doesn't mean you can't cross the parts and do it yourself.

Using the through hole components I've build this alarm "dead bug" style that weighed less than 3 grams, and I wasn't even trying to do a true light weight build.

Jimsky

[lovefool]

I guess correct address of futurlec is http://www.futurlec.com/. They have quite good price and cheap shipping cost. They charge my credit card with AU$, and shipping from USA. ( Address says Bankok sometime. ) It takes more than one week for delivery by normal shipping, so if you want to stock up some parts, try this.

[Friar_Tuck]

Your welcome Neons, but Jimsky gets all the credit, he came up with this little gem. All I did was dig up some numbers.

[**neons**]

Hi Jimsky,
Got my unit in today. I flew today and just missed the mailman. I will get to try it soon though. What a nice small unit from www.like90.com well packaged and shrinkwrapped neatly. You did a great job engineering this. I bench tested it and it is slick. I hope to build a few on my own in a week for my Jeti units that have no voltage adjustments. I was almost ready to put them on E-bay. Thank you & Good Luck'
**Neons**

[jimsky]

Thanks for the good words. I hope you will be happy with your purchase. It was fun working with Chandler and Colin from like90. It was a collaborative effort, I need to share the credit with them. Glad to hear this little device "has breathed life back into your Jeti ESC's."

The more I fly with this alarm, the more I like it. Just a nice audible indication that your time is up. No hard shutdown across the field, no modifying the power system...pulsing or burping of the motor. You hear the alarm (it's just about impossible to miss) and you know it's time to land.

I converted my Slow Stick to lipos and took it for a checkout flight last weekend. My wife stopped by the field and tried flying. After a while the SS got sluggish, I told my wife the battery was probably low. She said "What, no alarm in this plane?" I guess she's been paying attention to what I've been doing in my basement workshop.

When I got home, the battery was low...too low for my Apache to charge! I nursed the lipo back to health with a Lambda benchtop regulated power supply, and finished the charge on the Apache. I then installed a MicroScream on my Slow Stick. What was I thinking, flying without an alarm? :D

Jimsky

[Frans Bal]

All that stuff about Lipo-battery's and conventional controllers and motor shut-off made me think. A buzzer will do the job extremly well. But is it possible to switch off the motor when the lipo battery has dropped to a pre-set voltage. It is well kown that loss of signal to a controller will shut down the motor. i used this to make an automatic low-voltage switch-off. My English is't that good but i will try to explane. by means of an OpAmp the battery voltage is compared to a fixed voltage (any voltage can be made, 2S to 6S. When the voltage of the battery sinks below the fixed voltage the Opamp switches on. this Opamp signal drives a opto-coupler and an LED. The opto-coupler output signal is put into a CMOS AND port together with the impuls out of the reciever. this signal is then fed into the controller. And presto the motor will switch of when low battery is detected. At this moment if have a first test version working. But if anyone is interested in a diagram (or is it circuitdrawing) let me know.

[RookieOne]

Nice idea Jimsky.

I'll build mine with jumper setting to chose between 2S and 3S. "One size fits all":D:D

Thanks for sharing it with us.

[Friar_Tuck]

Hey Jimsky,
As I am far from an electrical engineer, I would like to ask you a question. I really like this setup, but would like to know what resistor would a guy use on R1, to get it to go off at 4.9V? I am thinking of putting one together for the RX pack too!
Thanks,
Jim

[jimsky]

That's easy. :D
Basically at 5.0 volts you want to create a 2.5 volt signal. Two resistors of equal value create a voltage divider of one half in the input voltage. So make R1 the same value of R2, 2.49K.

The piezo will be not as loud since it's supply voltage will be 5.0 volts, instead of the 6.0 volts as in the regular MicroScream. It should still be plenty loud enough to hear.

Jimsky

[Friar_Tuck]

Thanks again!
Now I will have one on every plane I own. This place is great, advice that works, what a concept!
Thanks,
Jim

[aeajr]

Interesting! Do you ever run this with NIMH?

[jimsky]

I've never tried but I don't see any reason why it wouldn't work. The issue with lipos is if you discharge them below 3.0 volts per cell you can damage the battery. That's the pressing reason for using a low voltage battery alarm when fly with lipos if you don't have a lipo friendly ESC.

I don't beleive NiMH or NiCAD's are adversely effected draining them empty. So with a plane using these types of battereis you could technically just fly until performance dropped, assuming your receiver and/or servos don't get weird due to the low voltage situation.

I assume the trip point of the MicroScream would want to be adjusted for the particular cell count of the NiMH or NiCAD pack in question.

Jimsky

[dparson55]

Thank you for such a simple alarm design. I'm just your average electric RC'er but may attempt to build this thing for my next project.

I wondered if it was possible to build it for a 10V cutoff? I'm hoping for this higher cutoff to allow a little more juice for landing and taxiing back a seaplane. The seaplane will be using an older Castle Creations Pegasus 35 and second generation Thunder Power 2100 3S1P Lipoly pack.

Maybe the default 9V for 3S packs is fine, or maybe the 10V figure I suggested is too low...I'm just not sure. Any advice would be appreciated, so thanks a million!

[jimsky]

Well if you use 3.7 volts as a nomimal lipo cell voltage, a "full" three cell pack would be 11.1 volts. To be setting the cutoff to 10 volts would be really limiting your flying time. You'd only going to be flying 1.1 volts worth of your battery pack, seems restrictive to me.

Personally I've gotten pretty good at predicting just when the MicroScream is going to go off on my planes. Power does start to roll off. I bring several battery packs to the field. Frequently I'll swap a battery BEFORE the MicroScream goes off. Why fly a doggy plane if you're not having fun, and you've got a full battery on hand? My SuperFly-E likes full batteries, my T-IFO is still very "flyable" when the alarm goes off.

Now with the MicroScream available I'd never just use this to protect my pricy lipo packs. I like the fail-safe MicroScream yelling at me, confirming it's time to land.

My guess would be the 9 volts cutoff will be fine. Like anything, it takes a little bit of time to get yourself accustom to something new...be it a new girlfriend, a new car, or a new plane setup. ;)

Jimsky

[dparson55]

Thanks for the information, hopefully the 9V (or 9.2V?) indication will be fine for a quick landing and taxi back.

I went ahead and ordered the 3S model from Like90, though their website says the Microscream is backordered until June 8. Hopefully it's in stock and they just haven't updated the site.

Sometimes I manage the throttle well and sometimes I just fly fast, so a cutoff (or indicator in this case) is absolutely needed for my variable flying style. This alarm will be great, though I'd pay twice the amount for a variable voltage one.

I wondered though: does the buzzer usually stop when you cut the throttle & the battery recovers a little voltage, or do you just have to unplug the battery. If it's super loud up close, I may have to invest is some earplugs. Just wondering...

[jimsky]

If you reduce the motor loading the alarm will stop sounding...until the voltage drops to the trigger point again at the new reduced throttle setting.

Initially dependent upon the particular motor, battery and loading, the alarm might start to "chatter", partially sounding at the threshold point, until the battery voltage is firmly below this threshold point. Keep you ears sharp, listen for the chatter or squawking...lower the throttle and fly home. That will provide you with the most "battery margin".

Only my T-IFO the alarm can chatter for quite some time, perhaps a minute or two, getting louder and more constant as the battery voltage slowly decreases. On larger motor planes which drain the battery at a faster rate, the battery voltage transitions through the threshold point quicker.

I could have added more circuitry to make a sharp alarm transition, but decided against it. I wanted to keep the alarm as light and simple as possible. And the squaking actually provides an advanced "heads up" you time is up.

The alarm is 85 decibels, it's not going to make you go deaf. And remember once you lower the throttle, the loading reduces and the alarm stops. You shouldn't have a plane on the ground with the alarm sounding...if you do, you flew WAY too long after the alarm initially went off.

Jimsky

[Shadowman95]

Hi,
I was looking to substitute a 9.1v zener doide for u1, avoiding the r1/r2 combination all together. OI stillm use the 2n7000 R3, and R4.

But mine always stays on, Have I missed something here?

Can you give a brief explanation of how the low votage alarm works.

I assume at > 9v (3 cell) the is a -v developed across R3 but when the battery V gets to 9V this voltage reaches 0v, and the fet turns on.

Cahn R1/R2 U1 be replaced with a 9V zener? or does U1 do something different. After reading the data sheet it looks like a temp stabilised zener , and R1/R2 set its breakdown to 9V?

:confused:

[jimsky]

Please review some of my prior posts, I have explained the circuits operation before. I don't see how what you have propose will work. You need a reference voltage, a battery voltage signal and something that's able to compare the two....how do you do that with just a zener?

And a traditional discrete zener (like a 1N751A, i.e. @ 5.1 volts) requires considerable more current. Then you need still need a comparator, so I don't get it? More parts, more current, more weight....I don't see the benefit of what you're proposing.

The cool thing about electronics is there are 1000's of ways to solve the same problem. If you want to play, go for it; you're bound to learn something in the process.

Jimsky

[Tentacle1]

Well here's another victim of a faulty device.. i've made my own PCB design, so that it could be very small, but it keeps buzzing all the way from 5V to 12 :(

This is the pcb i designed:
http://www.xs4all.nl/~ehorde/ff/PCBdesign.gif
Legenda:
TO92's: Source Drain Gate and Cathode Anode REF
Resistors (green): 2 = 2.49K 1%, 1 = 1.5K 5% and 6 = 6.65K 1%
Piezo buzzer: p- = negative and p+ = positive
Black line's are leads on the bottom and gray are on top (bridges)
the big - and + are the battery connection points ;) .

http://www.xs4all.nl/~ehorde/ff/top.jpg http://www.xs4all.nl/~ehorde/ff/side.jpg
http://www.xs4all.nl/~ehorde/ff/perspective.jpg http://www.xs4all.nl/~ehorde/ff/complete.jpg

Darn it for not working :mad: but is looks nice, dont it :p

[Zlatko]

I came across this "warnled" which is suposed to turn on at 2.4V and a standby curent of 5µA. I cant read German and havent been able to find it anywhere else.
One could build a simple voltage indicator by adding a zener diode and a resistor.
www.conrade.de and search for 185787
Datasheet http://www.produktinfo.conrad.com/datenblaetter/175000-199999/185787-da-01-de-warn-led_2_4v_5mm.pdf

Cheers

[**neons**]

Hi Jimsky,
I now have 5 of these Microscreams to play with. I used one on my Himax motor and I did have a problem hearing it on a flyby. Brushless motors tend to have a whine to them so it blends. I need further testing say with a glide by and I may hear it. While I was flying there was some guys doing some level landscsping with a sound transit that uses a laser beam and a loud screamer built in. It was very noticable because it had a sharp warbling tone to it.
Do you think down the road maybe not to far away it would be possible to apply a warbler to a Microscream in an updated Microscream II with the help of yourself or some one in the threads to help with a design? Maybe the whistle they use has the warbler it uses built in. I think the distinction of sound would make it stand out very dramatically like our Police, Fire and EMT depts. You still have been a credit to the E-flight hobby non the less and I thank you for your contribution.

[jimsky]

I've said it before, the cool thing about electronics is there are 100's of ways to solve the same problem. Each take will have it's own strenghts and weaknesses. My goals were: basic functionality, simplicity, low cost, and light weight.

But to me lightweight is paramount, this is aviation electronics. Every extra gram on a plane eats into performance and flight time. What can be simply done for "ground base electronics" just wouldn't fly (pun intended) for flight electronics.

The self-resonating piezo alarm I selected is pretty loud, fairly light weight, inexpensive and simple to drive. Audible alarms with varying output frequencies require more complicated oscillator circuits. Additional size, cost, complexity, and weight. Can it be done...sure it can.

What I've shared with the RC'ers here works well for my requirements, and I've got a ton of other projects keeping me busy. If I really do get that Corsair for my Bday at the end of the summer and set it up with the brushless motor like I've been dreaming about...maybe your problem will become my problem driving me to address it.

Want a real simple way to make the MicroScream more noticable with virtually no extra design work? Use 2 or even 3 piezo alarms. Try it, you might be suprised.

Jimsky

[**neons**]

Thanks Jimsky,
I think I will look for another larger Piezo whistle or give a group of some of these smaller whistles a try. I will expirement with it. As you said, I have the circuit to work with here. I have some reserve power with my brushless motor and it will handle it easily. I don't know why I thought it may be the whistle that makes a warble sound but also needs a driver. I'm not by no means very adapt at building and designing circuits. I can do the lower level stuff though. Your design is a godsend for us.

[RB]

jimsky, you tha man !!! My first electronics project !!! :) It even works !! I checked it with a nicad inline with my whattmeter. Now I need to work on component placement to look as good as yours. I mostly want the buzzer when I am using my lipos with a nicad esc but with the batt connectors I can put it in different planes.

Ordered the bits from Digikey on Friday afternoon and got the stuff today via first class mail!

Thanks
RB

[jimsky]

Glad to hear you are enjoying your home built electronics project. At only 3 grams each and a few bucks worth of parts, I just build and install one on each plane I have (is that 7 or 8?). Your connectors to allow portability is a cute twist.

Jimsky

[DPN]

Okay, I have all my parts and I'm about to start...

Dumb Question #1. I only intend to use this with 2 cell packs, so when you say R4 is not required, replace with short, this means I connect the Piezo + directly to BATT + and R3 and R1, right?

Dumb Question #2. Resisitors don't have a polarity, do they? I mean, there's no right or wrong end...

Thanks,

DPN

[RB]

Q1. Yes

Q2. Yes. Resistors have no polarity. One of the blue res. has yello and white stripes and the other has a green. Good way to tell them apart when all are sitting on the bench.

I made a ground strap out of wire that I placed under my wrist watch :)

I found it easiest to do the buzzer neg joint first since it only had the one connection. Your plan may vary :D

I finally got to try it on two different planes last night. Works great. Sounds cool doing vertical rolls. :eek:

RB

[DPN]

Wow!

Fired up the soldering iron, and made one that works first time! I like this much better than the LED low voltage device I was using.

Thanks Jimsky for this great little device!

DPN

[Fit]

Need some help.. the buzzer is on all the time.
I have:
R1 @ 6.5K for 3S lipos
Is this too far off 6.65K for the circuit to work properly?
How far off can i be from 6.65K for R1?

R2 @ 2.493k
R3 @ 1.492K
R4 @ 150.4 ohms

R2 onwards can be considered spot on. I also have tried two pairs of 2N7000
and TL431s'.
Any ideas on what went wrong?

Thanks :)

[jimsky]

Someone had a similar problem as listed in post #24, I a reply in post #25.

Jimsky

[Fit]

Hi Jim,

What voltage does the Reference pin require to turn on/off?
Currently I have 2.53 for a 2 cell battery.

Where do I place the probbes of the Voltmeter to get the voltage for the reference pin? If I'm building for a 9V Cutoff I should be getting 9V +/- 0.225V right?
Been troubleshooting the circuit on a breadboard for the past 2 hrs still can't get it to work..

I've got R1 to be 6.65.. but still no go..

[jimsky]

Typically voltage measurements are made referenced to the battery (-), otherwise know as the ground of the circuit. The meter minus should be connected to battery minus, the meter plus is used to make the desired voltage measurment.

I have described the theory of operation before, see posts number 4, 25, and 27. Additionally the TL431 datasheet has good application information.

Jimsky

[ozoneone]

Hey Jimsky, thanks! I made a MicroScream last night and tested it 3 different ways. It always seems to work. I really like the results so I plan on making one for my new heli. I was wondering if you have the art for the pcb? If you don't that's OK I'll make my own.

BTW- the only test it failed was on a Model Railroad Transformer. BUT it should fail as the transformer puts out a pulsed variable DC. The MicroScream reacted with a mid-level hum, but on other DC tests (transformer and battery) it worked perfectly! SCREEEEEEEM!

[jimsky]

All my own personal builds I've had great success doing the "dead bug" style. Lightweight is key for me. Typically mine come in at about 2.5 to 2.6 grams complete with about 12" of magnet wire. If you are afraid of components shorting/touching each other, a couple of very small dabs of hot glue does the trick.

I don't think a printed circuit board is warrented; it only adds cost, complexity and weight for the small scale hobby user. But if you WANT to make a pcb, by all means do it. I always end up learning something from just about every project I do.

Jimsky

[ozoneone]

My first was "dead bug" but as I now intend to make several for myself I think it would actually be easier on a board. Making the board is no problem after I have the art. It's the hole locations that are a pain, but once one is made the others are easy. I don't think it will weigh much more in my case, less wire leads and less solder. For a park flyer I would leave it off the board. THANKS AGAIN!!!!!

[Friar_Tuck]

Hello there,
I have had some made at PCB express, nice little two sided PCBs, if you want the file for having them made, just let me know, and I will send it to you.

[ozoneone]

Wow, I just looked at www.pcbexpress.com and boards 0-9 sq in costs $80 for two? What kind of file do you have? If that's the correct pricing then I'll go back to Radio Shack etchant, a drill press and a Sharpie and I'm good to go.

Thanks, Friar Tuck, and please let me know if I have the pricing correct and what type of file you have. Thanks again!!!

[Friar_Tuck]

I have the file for PCbexpress, two boards, which will give you 36 items, is $59.00 delivered.

[ozoneone]

Sure then, thanks Friar Tuck! My e-mail address is ozoneone@yahoo.com

Thank you, I'll check it out again. But I'll need to figure out what I'll do with 36 boards!

[jerahteo]

Hi jimsky very impressed with your project. I printed out your diagram and when to my local electronic store in Singapore to buy the parts I can get most of the components except R1 which you said need to be 1% 3.57K but I only managed to find 3.6K 1% will that work ?

Appreciate your value advise.

Many thanks :D

[jimsky]

jerahteo,
That is a very small value change, it should work just fine.

Jimsky

p.s. Turn on you RCGroups e-mail feature so people can send you private messages.

[jerahteo]

Thank you I will go and grab it today I have my own screammmmmmmmmmmmmm

[jerahteo]

But I notice that the buzzer which I see from my local store is all black and don;t have any indication of say 80db how to know if that buzzer is correct

[Mick Molloy]

Any body got a PCB drawing for this little beauty yet??

[Friar_Tuck]

Ozoneone, if you are wondering what to do with 36, just send me 10 for $20!
:)

[RB]

I finally got around to building my second microscream. It is much more compact than the first with just the transistors vertical. Getting the transistors horizontal will be the next project. I am kinda surprised that this one actually worked as I really abused the 431 during the several attempts :p to get it into position. I may add a little epoxy as a potting agent around some of the components. This one will be hardwired into one of my shockys.

RB

[Friar_Tuck]

Hello Jimsky,
Having a ball with these!
use them everywhere!
Thanks again, here are mine. They come out at .2 ounces with wire.

[RB]

Screamer layout #3. This is about as small as I think you can get them. The appriopriate legs of the 431 and the n7000 are soldered together and the resistors and mag wire added in. You have to think about the order of assembly. Soldering the N7000 to the negative side of the buzzer first off allows you to clamp the buzzer in a vise and have a stable work platform.

RB

[ozoneone]

I made two MicroScreams so far and both work perfectly, however when in use in my heli they are not loud enough to hear above the heli's normal sound at only five feet away. The only way to hear it is to turn the throttle off.

I was wondering if a different buzzer could be substituted to make it louder, and/or an LED (flashing LED or strobe maybe?) added into the circuit???

Thank you!

[RB]

It looks like you can add an LED. I have some flashing 3 to 15v LED's that NYblimp sells and I just hooked up a spare screamer to a 4.8v RX batt and then touched the ends of the LED wires to the appropriate terminals and the LED did flash. Jimsky would need to comment on any possible excess current drain that might affect the components. If it works out then I am going to add the LED to all my future builds. I have encapsulated all mine in hot glue. Why didn't you think of this sooner ??

RB

[htuong95]

...

[htuong95]

...

[jimsky]

The 2N7000 should not have any problems dealing with the nonimal current of an LED. I've made MicroScream versions with two peizo buzzers (LOUD!) and flashing LEDs. Just be careful where you hook these devices up. If you are making the 3 cell version that requires the dropping resistor for the piezo, pay attention.

If you connect the LED from Battery + and the drain of the 2N7000 you should be fine. This assumes the LED can deal with the voltage applied to it when the alarm trips (9 volts for 3 cell, or 6 volts for 2 cell). There are LEDs that have internal limiting resistors, others do not.

If the LED you have selected doesn't have an internal limiting resistor, you will need to add one. This would allow control of the current through the LED.

You might be able to get slick and just parallel the LED across the piezo. The current through the piezo buzzer dropping resistor will increase, causing a larger voltage drop. You'll want to readjust the value of the dropping resistor so both the piezo and LED are happy.

As long as you address the issues above you should have no problem modifying the MicroScream circuit.

Jimsky

[ozoneone]

Thanks Jimsky! I'll give it a try.
You'll want to readjust the value of the dropping resistor so both the piezo and LED are happy.
Any suggestions?

[jimsky]

Errr, I'm wrong on that......

Actually an LED by itself will clamp at about 2 volts, so you don't want to put an LED (without internal limiting resistor) right across the piezo. The LED will clamp at 2 volts and you will never hear the piezo. Sorry I haven't had my morning coffee yet.

An "naked" LED will require it's own limiting resistor. This will allow you to select the set the on current of the LED, while not influencing the piezo circuit.

An LED with an internal limiting resistor could be placed right across the piezo, but like I said the piezo dropping resistor value might need adjsutment.

Jimsky

[RB]

The 3 to 15v flashing LEDs that nyblimp sells definitely work on the 6v screamer. Just solder the LED to the piezo + & - and you are in business. I went back and added the LEDs to all 3 of the screamers I have built so far. Hot glue stinks when you touch the iron to it.

Also finally installed the Jimsky screamer on my Stringfly mini shock flyer.

Isn't this site great :p

[**neons**]

Hello Jimsky,
I have been using the MicroScreamer on a few of my planes outdoors now for a few flights. I am having some positive results with my planes if I cut the thottle on the flybys to listen for it. My hearing is not up to par as years ago. Yes it is a nice piece of equipment you have given to the E-world.
My question is this. I would like to add another Piezo whistle to all of my MicroScreamers to bring the Db's up some. I am using 3 lipos 11.1volts. I will also add a Led to them later after the 2 whistles work.. The main thing is the value of the R-4 150-Ohm might change as you suggest. How do I adjust the R-4 resistor for the additional Piezo whistle? Is it done by Db sound or a VOM meter reading? Do you have the value change for the R-4 resistor already from yours? This should be louder for outdoors. The system now will be plenty loud indoors. What do you estimate the 2 whistles put out in the newer Db sound level as a guesstimate?
Thank You,
**Neons**

[jimsky]

neons,
The dropping resistor value is selected with the maximum voltage of the piezo buzzer in mind (7 volts maximum). The 3 cell version fires the alarm at 9 volts, so connecting the piezo right to the battery voltage would overvoltage the buzzer....so a dropping resistor is used.

The resistor value is selected so that about 2.5 volts are dropped, putting about 6.5 volts across the piezo buzzer. Two piezo buzzers in parallel will draw twice the current, causing about twice the voltage drop across the dropping resistor. Reduce the value of the dropping resistor to get the voltage applied to the two piezos back to about 6.5 volts. You want the alarm LOUD right? :)

I've played around on the bench with two piezos and they're pretty loud. The internal oscillator frequency of the piezos are not held to a tight tolerance. Play around selecting them, I've gotten two piezos that were slightly off frequency from each other....and the noise the two of them made were just amazing. Kind of like a low and high note automobile horn, but not quite as dramatic.

By the way, if you lower your throttle during a flyby you reduce the battery loading, causing the voltage to rise. This could cause the alarm to stop sounding. So you might be fooling yourself thinking your batteries are OK when they are not. Just something to keep in mind.

I'm suprised people are having a hard time hearing that 85dB piezo alarm. My hearing has suffered from too much time playing in rock & roll bands years ago, and I can still hear the piezo quite a distance away on all my planes. In fact I usually hear it when it starting "ticking" before it goes into a full blown alarm. You guys must be flying some loud planes. :D

Jimsky

[Magnum9]

Just put together my first screamer. I had to use a 2.4k resistor as 2.5k was unavailable. Above roughly 7.5 volts the alarm is audible, but very quiet. As voltage drops slightly it turns off and then activates at full volume at 6.0 volts. Any suggestions?

[ozoneone]

You guys must be flying some loud planes.

Some Micro Helis can get pretty loud, & ESC cutoff gets ugly on helis! I will try the 3-15V LED across the buzzer. I etched my own board & it came out great. Very small, with 18 holes (16+2 for leads) but after I test the LED I will change the art to accomodate the LED. Thanks again.

[jimsky]

Magnum9,
This is a simple, lightweight, minimalist design. I tried to use the least number of components possible to keep the size and weight down, while providing basic functionality. There are some drawback going this route....

If you crunch the limits of all the component specifications, the possibility of this problem happening does exist. I haven't seen it myself, but I did post about the possibility of it happening. See posts number 25 and 27 in this thread for an explaination.

Jimsky

[RB]

I flew one of my screamers with the led last night and I don't know if the led is worth the effort for a plane. The piezo goes off at app 5.9v but the led doesn't really start to do its thing until about 5.7 or 5.8. The led might save you if several planes are flying and you can't hear the piezo but otherwise the scream will make you want to land before the led goes off.

For a heli the led will probably do great. I wonder if you could just leave the piezo off and only have the led or maybe two leds ????

RB

[gPearce]

Jimsky--Can you or someone tell me how to "turn on my e-mail feature". And what does that do.please?
G Pearce

[gPearce]

Jimsky--Although I am still sort of puzzeled about how this circuit can work I think if the buzzer is always on how about a signal diode betweenthe432 and gate to increase the firing voltage of the 2N7000?
G Pearce

[gPearce]

Jimsky---The reason I am puzzeled is after looking at both data sheets I see that V-out on TL431 is about 2V when Vin is>Vref. Isn't this the turn on voltyage of the N7000?? So why dosen't the buzzer go on allthe time since the resistor junction voltage is >2.5(Vref) when the batt V is above the design low point??