HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Nov 28, 2012, 03:50 PM
Micro Jet Pilot
Chris T's Avatar
Joined Jul 2005
2,091 Posts
Dan, if you make the red and black wire shorter and the rx power lines longer, I bet it's gonna get even better. Joe, can you tell if there is a difference in your wiring? I am thinking I should send Dan a different version of the blc, the one I send to Dave plus the big cap. I thought I made the code more robust in the end, but I might have made it worse. I have made the experience that a transition to the sync mode at lower rpm was better in the squall, but on the test motor it was opposite. I might have been misleaded. I have an extra blc lying here with the old code.


... I am a bit stupid. I was wondering why my servo wouldn't work and get hot, now I figured I loaded the blc code into it

Chris
Chris T is offline Find More Posts by Chris T
Reply With Quote
Sign up now
to remove ads between posts
Old Nov 28, 2012, 04:37 PM
Registered User
Carbondale PA
Joined Aug 2005
548 Posts
Chris,

My wiring is the same as Dan's. It is very strange that it works on my Cricket receiver and not his. I did not change the length of my wiring for my motor tests.

My motors seem to spool up slow to top speed. It sounds really cool! However, is this normal?

Joe.
Joe Malinchak is offline Find More Posts by Joe Malinchak
RCG Plus Member
Latest blog entry: Micro Laser 3D
Reply With Quote
Old Nov 30, 2012, 12:43 AM
Registered User
Oak Ridge, Tennessee
Joined Dec 2007
374 Posts
Chris-

A few more tests below- all with the cricket rx and hip hop tx combo.

They were all about the same, rpm-wise. I did not see any difference with any of the caps connected or with it wired direct to a lipo. (a larger 240mah cell) The motor always cut out- locked at the same throttle point, around 8500 rpm or so. (9200 at best) And again removing the prop didn't change things, or swapping the coil leads.

I wound a quick 9.5 ohm coil (#36 wire) on a bigger 4x5 mm magnet, and got the same results again. It stopped at 6700 rpms, and the same throttle setting point on the tx.

And I noticed on the tx, that with all the tests the motor does not start running until I reach slightly over the 1/4 throttle point. (stick-travel wise) Starts running at 3000 or so... Speed increases normally as throttle stick is raised- then the motor stops at the 1/2 throttle point. So there is very little stick travel involved. Each click changes the speed by 1000 or so revs. I remember the two other rx'es were similar in this respect.

I'm still wanting to hook up a ar6400 rx for a test... and I'll try to hook it up to my scope for a reading too. I will try to do this tomorrow.

Joe- I would like to use the cricket rx in a test plane, if I can get it working as well as you did. And I don't notice the speed spooling up as you mentioned, yet.
shagrug is online now Find More Posts by shagrug
Reply With Quote
Old Nov 30, 2012, 01:53 AM
Team30 Micro EDF
NitroCharged's Avatar
Adelaide, Australia
Joined Apr 2004
11,199 Posts
Dan, I dialled a lot of expo into my throttle to really help soft start the motor. Not sure if that's any help to you?
NitroCharged is offline Find More Posts by NitroCharged
Reply With Quote
Old Nov 30, 2012, 03:47 AM
Micro Jet Pilot
Chris T's Avatar
Joined Jul 2005
2,091 Posts
Hi,

The slow spooling up is a feature, Throttle changes are adapted slowly by code. Expo won't help, cause the internal spooling filter is not affected. Starting at about 1/4 throttle is a bit strange, I would expect it to start earlier. But this might be due to the trim. With trim in the center, the start point might be earlier. Could it be that your throttle is reversed? Not very likely, but what happens if you power up at max throttle and trim to the max? After all, I will send you a different verion and I expect it to run. The reason is that that code version will run in sync mode where you experience cut off so the situation will be much different. It will be very educative to me to see if the code makes the difference. As I said I had different behaviors in different systems too and I might have made it worse for your setups. Thanks for all the testing so far. Kicking in at 3000 rpm is probably what should be expected, I raised the start rpm to improve the startup and avoid the locking in the dead spot. The blc shouldn't lock once running, so something is strange with that particular controller. If the new controller is running for you, maybe you can send the old one back so I can examine it.

Chris
Chris T is offline Find More Posts by Chris T
Reply With Quote
Old Dec 01, 2012, 03:33 AM
Micro Jet Pilot
Chris T's Avatar
Joined Jul 2005
2,091 Posts
New controller for Dan and extra cap for Dave are going to the post office now

Chris
Chris T is offline Find More Posts by Chris T
Reply With Quote
Old Dec 01, 2012, 03:57 AM
Flying Dutchman!
robertboxhoorn's Avatar
Nederland, FL, Emmeloord
Joined Sep 2011
368 Posts
Any Progress with the squall himself yet?
robertboxhoorn is offline Find More Posts by robertboxhoorn
Reply With Quote
Old Dec 01, 2012, 04:24 AM
Micro Jet Pilot
Chris T's Avatar
Joined Jul 2005
2,091 Posts
Will sit down right now and start on the new blc/rx module. It will be about the size of the c3blc1. Later today, I will give the servo another try with the correct code. There's not much left to do. I expect that power will be at the low end for flying, but it was the same with the 6'' squall in the beginning. I probably need a hotter motor and maybe a better battery. I read an article, which promises that there will be new battery technologies exceeding the performance of lipos by far. I should get it to fly before those are out

Chris
Chris T is offline Find More Posts by Chris T
Reply With Quote
Old Dec 01, 2012, 06:24 AM
Micro Jet Pilot
Chris T's Avatar
Joined Jul 2005
2,091 Posts
Hi,

if you thought the attiny 3 x 3 mm chip is small, have a look at the 4wlcsp ldo (.76 x .76 mm) underneath (bottom left). The right chip is the new SI4355, the successor of the famous IA4320/Si4320. It's also 3 x 3 mm. The silver thing at the bottom right is a 30MHz crystal (1.6 x 1.2 mm). I think this is as small as it gets. The test board is huge, the final board will be tiny and packed. Now comes the hard part. Getting the rx configured so I have a radio link. This is a total blind flight, since there is no way of debugging. Either it works or it does not.

This is challenging

Chris

P.S.: BTW, the SI4355 is soldered in wrong orientation. I had to rotate by 90 after I made the photos...
Chris T is offline Find More Posts by Chris T
Reply With Quote
Old Dec 01, 2012, 04:38 PM
Team30 Micro EDF
NitroCharged's Avatar
Adelaide, Australia
Joined Apr 2004
11,199 Posts
Thanks Chris!
NitroCharged is offline Find More Posts by NitroCharged
Reply With Quote
Old Dec 01, 2012, 10:03 PM
Registered User
Oak Ridge, Tennessee
Joined Dec 2007
374 Posts
Chris-

My thanks too, I'll look forward to testing it soon.

This one I'm testing now I'll send back soon, after I try a couple more things. I have pulled a ar6400 from a worn out airframe to test next, probably later tonight. I don't have any other brushless signal rx'es to try other than a plantraco servo version.

I've got some screenshots posted below of the motor (the coil pulses) running on my usb scope. I was using the bigger coil/magnet for all of the test runs. When the motor locks I think it flatlines at 3.9v or so. (I forgot to get a shot of this- but will) I have no idea why there are gaps or irregular peaks, slowing down the frame rate shows them better. I don't hear or feel the motor skipping or pulsing any while running, it seems normally smooth. The 160mah lipo used was putting out 4.02 v or more (4.04) to the board while running.

I noticed a bit of the solder that looked to be bridging between traces on the board (at the fet), but re-soldering and fixing didn't help it any. I don't see any other problems with the board. If you can suggest anything else to try just let me know. It's all very interesing to me.
shagrug is online now Find More Posts by shagrug
Last edited by shagrug; Dec 01, 2012 at 10:12 PM.
Reply With Quote
Old Dec 04, 2012, 03:59 PM
Registered User
Oak Ridge, Tennessee
Joined Dec 2007
374 Posts
Below are some more shots, this time testing with the ar6400 rx.

I tried it with both dx5e and dx7 tx'es and still have the same throttle limitation. It starts running at a lower point (about three clicks up) but still locks when nearing the half-stick throttle point (again about 9500 rpm). It does self start a lot better with this rx for some reason, better than all the others I've tried...

The scope signal is a bit different too, sort of "cleaner" at the half and full speed running. Slowing the frames looks better too, with more evenness.

This has me wondering if there is a difference in the quality of signal on brushless outputs, sort of like what I found on testing brushed outputs on various rx'es. Some worked much better than others with the a1442.

I put the controller in a envelope and will mail it back tomorrow. (with the lipos)
shagrug is online now Find More Posts by shagrug
Reply With Quote
Old Dec 05, 2012, 01:47 AM
Micro Jet Pilot
Chris T's Avatar
Joined Jul 2005
2,091 Posts
Dan, no rush in sending it back. It might make sense to keep it until you have tested the new one.

Chris
Chris T is offline Find More Posts by Chris T
Reply With Quote
Old Dec 05, 2012, 06:57 AM
Micro Jet Pilot
Chris T's Avatar
Joined Jul 2005
2,091 Posts
Update

I have been working on configuring the SI4355. The first step was to get a save communication between the atmel running on 4.2V (since it is doing the motor driving as well (h-bridge driving and emf reading)) and the SI4355 running on 3.3V LDO voltage (since it is not qualified for voltages larger than 3.6V). I had some SI4311 burning down on 4.2V preveously,so I knew this is a requirement. Critical is that the SPI lines driven from the atmel side must not exceed 3.9V (maximum digital input voltage on the SI4355). I wanted to achieve this without external components and I did. Therefore, I am driving the pins in open drain style, switching between active 0 and tristate with pull up activated in the atmel. This way I am driving the SI4355 pins high with the atmel internal pullup resistance (20 - 50kOhm connected to 4.2V). The esd diodes in the SI4355 prevent the voltage from rising more than .6V above SI4355 supply voltage. The levels don't exceed 3.9V so this is working fine. Timing needs to be slowed down, so the signals can settle correctly. Next thing was to get an idea how the config pattern has to look like. The AN691 document describing the configuration is revision 0.2 and what they call easy config is pretty much unclear up to date. Anyway, they provide a GUI to generate about 202 byte of cryptic code and I can load that code into the atmel eeprom and send it to the SI4355 from there. This is supposed to do the complete configuration according to the radio settings chosen in the GUI. In the end, some GPIOs need to be set so I can receive the radio signal in direct mode and that should be it. I can chose from predefined configs that all run FSK on 30kHz deviation, trying to generate a new setting with 90kHz deviation which I have set my tx to seems to generate false channel bandwidth settings. There seems to be a bug in the GUI. All new stuff. So I will use one of the tested presets and adapt my tx to 30kHz (Carrier frequency is of course 915MHz like plantraco, Nick's and my stuff and others). They promise ranges over 1km with these settings and printed antennas. So for our needs this should be rock solid. Next week I will complete this step and see if I can establish the radio link. If so, this will be the smallest radio for a long time. I think it's comparable with Koichi Tanakas IR device which was the way to go for living room flyers years ago. This time we are talking about long range narrow band ism The rx chip itself is low power, consuming 10mA by default. Switching the Si4355 into standby between the ppm pulses (or maybe using an alternative protocol) should allow to get this below 1mA.

On the other side, I found new hall sensors for the servos which only consume 3mA and also have a shutdown pin. So one of the major drawbacks in scaling down this techology will not apply anymore. With the current designs the current consumption of the devices themselves would load a 10mAh battery already considerably, without driving any motor or servo coil. Using these new chips, I can see a pitts flying in my house already

Chris
Chris T is offline Find More Posts by Chris T
Last edited by Chris T; Dec 05, 2012 at 12:21 PM.
Reply With Quote
Old Dec 05, 2012, 02:08 PM
TheyreComingToTakeMeAway!
derk's Avatar
USA, ID, Coeur D'Alene
Joined Dec 2003
5,458 Posts
Sounds awesome Chris! What do you estimate the receivers footprint would be?
derk is offline Find More Posts by derk
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion 6'' Micro Squall (20mm EDF, Single Phase Sensorless Motor) Chris T Micro EDF Jets and Power Systems 263 May 08, 2013 10:31 PM
Sold New Phase 3 Squall! 40A 11.1v ESC + used 3s ducted fan setup jgalexander Aircraft - Electric - Power Systems (FS/W) 0 Mar 03, 2012 01:45 AM
For Sale Phase 3 Squall kit plus motor/fan combo SmoothCruizer2 Aircraft - Electric - Jets (FS/W) 5 Oct 05, 2011 11:50 PM
Sold NIB Phase 3 Squall EDF 4-cell Hi-Perf ARF w/Fan/Motor/ESC $85 each Shipped Insured jmrust Aircraft - Electric - Jets (FS/W) 6 Nov 18, 2010 09:41 PM
For Sale 63mm EDF combo 3S (fan,motor,esc) squall phase 3 crd Aircraft - Electric - Power Systems (FS/W) 2 May 08, 2010 08:19 PM