SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio
Reply
Thread Tools
Old Nov 12, 2011, 11:00 AM
zis
Registered User
Joined Mar 2009
11 Posts
Flash tgy-2011-10-12.hex, tgy-2011-10-12-ultrapwm.hex, tgy.hex not working on TowerPro 25A Rev. H
Can you finish them?
zis is offline Find More Posts by zis
Reply With Quote
Sign up now
to remove ads between posts
Old Nov 12, 2011, 12:50 PM
Jakub J's Avatar
Krakow/POLAND
Joined Dec 2009
1,743 Posts
Today fun flying

TowerPro 12A ESCs with alternative firmware.

Completly different experience
ZERO oscillation even during fast descend or fast climbing.
P(id) up to 12 without any problems (in this video P=6).

small quad fun flying (3 min 12 sec)


Jakub
Jakub J is offline Find More Posts by Jakub J
Site Sponsor
Reply With Quote
Old Nov 12, 2011, 07:39 PM
Suspended Account
Joined Dec 2010
11,764 Posts
Quote:
Originally Posted by zis View Post
Flash tgy-2011-10-12.hex, tgy-2011-10-12-ultrapwm.hex, tgy.hex not working on TowerPro 25A Rev. H
Can you finish them?
Sorry this is not one-click plug and play solution.
Measure the pin out, adjust the header file to match, then submit a patch.
Nobody will finish your work for you.
timecop is offline Find More Posts by timecop
Reply With Quote  (Disabled)
Old Nov 12, 2011, 08:05 PM
hacker
Canada, BC, North Vancouver
Joined Dec 2010
936 Posts
Quote:
Originally Posted by nikivan View Post
I now have a secong ESC (HK 18A) failed after flasing with tgy.hex It happened when trying to lift off, the motor screamed and the quad flipped. After that there was no response from the ESC. Simon, if you are interested I could pass you the units for inspection, just send me a PM.
Hello! D'oh! Sure, or we could also meet since you're so close, rather than shipping things. Maybe we can revive them. I've been sick with various things for a couple of weeks, but I'm better again now...

Which HK 18A board were these? Was this with tgy.hex or a newer build? Was the screaming sound medium or higher pitched whine or was it more of a juddering sound? Did you take off quicky from stopped, or were the motors already spinning (and sounding normal) for at least a second before you got off of the ground?

tgy.hex does not ramp start-up power, and this can cause false zero-crossings on _startup_ with the late demagnetization on some motors at high duty cycles. This is what caused the juddering on my tricopter motors when I had no radio plugged in to the KK board, as I mentioned previously.

The latest tree has ramped start-up power to try to avoid this problem, and also faster throttle response than even before (all loops now also check for RC pulse update and process, rather than once per commutation step). I also raised MIN_DUTY to hopefully make Nek happy with the "reduced power" since the clock rate increased, but I hadn't mentioned this yet because I keep meaning to find a way to compare it with stock Plush firmware, and my servo tester is so granular (4us steps) that it's difficult to compare. In other words, I think it's time to make a new build and make Nek happy with it so he stops recommending tgy.hex.
simonk is offline Find More Posts by simonk
Reply With Quote
Old Nov 12, 2011, 08:09 PM
hacker
Canada, BC, North Vancouver
Joined Dec 2010
936 Posts
Quote:
Originally Posted by zis View Post
Flash tgy-2011-10-12.hex, tgy-2011-10-12-ultrapwm.hex, tgy.hex not working on TowerPro 25A Rev. H
Can you finish them?
What happens now? Does it just keep resetting on startup before the beeps? You may hear a repeating faint "click" from the motor instead of the beeps, if this is happening. You probably have the 2009+ TowerPro versions which have inverted nFETs (low side), and need the "tp_nfet" config. I haven't uploaded a .hex for this build because there are so many variants, and because it hasn't been tested yet (I'm not sure if the pinout is modified or not for i2c). Actually, I'll probably upload a .zip with all builds, or make a web page with selectable options that just builds the correct version on demand. Since you even tried the UltraPWM version, you are obviously just trying things randomly, which is generally not the best way to get it working.

Which makes me really want to make an auto-configuration build... Hmm. We can detect inverted high side FETs because the sense lines will be high before initializing the ports, but I'm not sure about inverted low side FETs, because the outputs will be low if inverted or not. And we can't flip a single high or a single low FET without causing a short if either the high or the low side is inverted. None of this is something we want to be doing if the motor is already spinning (eg: brown-out reset), either. Hmm...
simonk is offline Find More Posts by simonk
Reply With Quote
Old Nov 13, 2011, 03:49 AM
If it flies, I will crash it
mxspode's Avatar
USA, CO, Westminster
Joined Jan 2008
3,040 Posts
Quote:
Originally Posted by bas10an View Post
Thanks Jim,

I realize that I read this like a 100 times already, somehow it just slips right through

I always run long ESC wires and short motor wires, I like to keep the ESC as close to the motor as possible for cooling, it'll extend both flight time and ESC life.
Speaking of which, new kids on the block for micro multis, comes with Atmel chip and yes, it has programming pads

Sebastian
Quote:
Originally Posted by bas10an View Post
I wish someone would produce something like this in the 50-60gr. class.

Check out the ESC, still Atmel and programming pads.
It'll fit like a glove under a good old Blue wonder.

Pancake motors I like the idea of. The ESC mounted directly under the motor like that, I don't know. Call me a pessimist but all I can see is the downside. Not enough upside to justify for me. I have scene way to many motors with bent mounts and such at work and it seems that they would be a bit to vulnerable.

It is an interesting notion and I am sure they will probably sell plenty though. People love gimmicks and flash.

Rob...
mxspode is offline Find More Posts by mxspode
Reply With Quote
Old Nov 13, 2011, 04:01 AM
Registered User
tigar's Avatar
Athens, Greece
Joined Sep 2001
1,167 Posts
Hi,
i have flashed with the turnigh firmware these esc. It works but having a difficult start-up. After starting the motor runs very smooth. I have tried different motor types, all with the same behavior.
Is there any other build i should try to make things better?
The elements on the back side are named: LR7843 708P and 50P03 8H01
if it means anything to you...

Thanks
tigar is offline Find More Posts by tigar
Reply With Quote
Old Nov 13, 2011, 04:43 AM
hacker
Canada, BC, North Vancouver
Joined Dec 2010
936 Posts
Quote:
Originally Posted by tigar View Post
Hi,
i have flashed with the turnigh firmware these esc. It works but having a difficult start-up. After starting the motor runs very smooth. I have tried different motor types, all with the same behavior.
Is there any other build i should try to make things better?
The elements on the back side are named: LR7843 708P and 50P03 8H01
if it means anything to you...
Thanks
Hello! Which ESC is that? What input voltage and which motors (kv at least)? Which build/tree version? Start-up has not really changed for a while, but I have a bunch of un-pushed stuff that does change it, and currently I actually just merged startup and running modes (uncommitted). However, with the current tree, you may notice a problem at higher input voltages or at too-low duty cycles, depending on fet timing and the motor. That looks to be a normal "type 2" pinout, unless I'm missing something, and the sense resistors seem to be the usual 33k (and 10k to gnd?), so I don't have any other ideas at the moment.
simonk is offline Find More Posts by simonk
Reply With Quote
Old Nov 13, 2011, 04:54 AM
Registered User
tigar's Avatar
Athens, Greece
Joined Sep 2001
1,167 Posts
Quote:
Originally Posted by simonk View Post
Hello! Which ESC is that? What input voltage and which motors (kv at least)? Which build/tree version? Start-up has not really changed for a while, but I have a bunch of un-pushed stuff that does change it, and currently I actually just merged startup and running modes (uncommitted). However, with the current tree, you may notice a problem at higher input voltages or at too-low duty cycles, depending on fet timing and the motor. That looks to be a normal "type 2" pinout, unless I'm missing something, and the sense resistors seem to be the usual 33k (and 10k to gnd?), so I don't have any other ideas at the moment.
Hi,
unknown esc, 3s battery and 750 to 1100 kv motors, i dont know about the version but i downloaded the latest file i could find.
You are talking about builds, is there any way i can change or build the hex file?
tigar is offline Find More Posts by tigar
Reply With Quote
Old Nov 13, 2011, 04:57 AM
hacker
Canada, BC, North Vancouver
Joined Dec 2010
936 Posts
Quote:
Originally Posted by tigar View Post
You are talking about builds, is there any way i can change or build the hex file?
Grab the code from git at git://github.com/sim-/tgy.git or from https://github.com/sim-/tgy/downloads eg https://github.com/sim-/tgy/zipball/master
simonk is offline Find More Posts by simonk
Reply With Quote
Old Nov 13, 2011, 05:30 AM
manuLRK
Belgium
Joined Sep 2004
1,533 Posts
Can this be confirmed?

Thanks,
manu

Quote:
Originally Posted by zis View Post
The file read tgy.asm:
;-- Fuses -----------------------------------------------------------------
;
; Old fuses for internal RC oscillator at 8 MHz were lfuse=0xa4 hfuse=0xdf,
; but since we now set OSCCAL to actually run at about 16 MHz, we'd better
; set brown-out detection to 4.0V. This code should work without changes on
; boards with external 16MHz crystals / resonators; just set lfuse=0x3f.
;
; Suggested fuses with 4.0V brown-out voltage:
; Without external crystal: avrdude -U lfuse:w:0x24:m -U hfuse:w:0xd7:m
; With external crystal: avrdude -U lfuse:w:0x3f:m -U hfuse:w:0xd7:m
;
; Testing fuses with 2.7V brown-out voltage (unsafe at 16MHz):
; Without external crystal: avrdude -U lfuse:w:0xa4:m -U hfuse:w:0xd7:m
; With external crystal: avrdude -U lfuse:w:0xbf:m -U hfuse:w:0xd7:m
alll is offline Find More Posts by alll
Reply With Quote
Old Nov 13, 2011, 06:58 AM
hacker
Canada, BC, North Vancouver
Joined Dec 2010
936 Posts
Quote:
Originally Posted by alll View Post
Can this be confirmed?
It's a correct quote, if that's what you mean? ( https://github.com/sim-/tgy/blob/master/tgy.asm )

The fuses normally don't need to change if you are re-flashing an existing ESC, with the exception that now because we run OSCCAL at 0xff, we probably should set the brown-out detection voltage to 4.0V. I haven't been able to reproduce any problems with it left at 2.7V, but I believe this is out of the specification for the Atmega8A or 8L running at 16MHz, and so a brown-out reset is preferred before lockup or other fault. This would only happen if, for some reason, VCC drops lower than the 5V it should be at anyway, so it's usually not a problem. Again, I've specifically tried to reproduce problems with the "testing" fuses (BODEN set with BODLEVEL cleared), and haven't been able to, but this doesn't mean it can't happen.

Bascially, any fuses that allow the oscillator to work should work, so if you don't know what you're doing, don't mess with them. Boards with oscillators should work even if you select intrc. However, boards without oscillators will break if you select external oscillator, and boards with oscillators will break if you select a low speed oscillator, at which point you have to drive the clock manually with an external source. I did this recently when I typoed a fuse setting and so I flashed another ESC to emit 1MHz from the SCK pin, then jumpered it to XTAL2 while flashing the fuses back. All of the start-up delay stuff is unimportant, and CKOPT should probably be (already) set for boards with an external 16MHz oscillator. Everything else doesn't matter.
simonk is offline Find More Posts by simonk
Reply With Quote
Old Nov 13, 2011, 07:07 AM
Registered User
tigar's Avatar
Athens, Greece
Joined Sep 2001
1,167 Posts
Quote:
Originally Posted by simonk View Post
Grab the code from git at git://github.com/sim-/tgy.git or from https://github.com/sim-/tgy/downloads eg https://github.com/sim-/tgy/zipball/master
What and how can i do with this?
I used the .hex file from the download area..

Edit:

If someone could post the hex that is made from this files?
tigar is offline Find More Posts by tigar
Last edited by tigar; Nov 13, 2011 at 07:46 AM.
Reply With Quote
Old Nov 13, 2011, 12:31 PM
manuLRK
Belgium
Joined Sep 2004
1,533 Posts
program plug

Hi,

I had some (+10) of these esc laying around without ISP pads, so i made a programming plug. Tested and i am able to read the atmega!

Used a silicon past. But this is still surgical work.
I have other SS-HK, with the smaller atmega8, will try that too, but i don't think i will make it, it is so small...

Still have to try the firmware.
manu
alll is offline Find More Posts by alll
Reply With Quote
Old Nov 13, 2011, 02:33 PM
Registered User
Joined Dec 2005
681 Posts
Quote:
Originally Posted by jesolins View Post
Hi Nick,
Also Ohm test the motor windings to make sure they all show about 30 Ohms in any combination. Less than that on any winding set or different values could indicate a motor shorting issue that is taking out the ESCs. Do you run your motor wires or ESC power wires over 12 in/30cm?
Cheers,
Jim
Quadrocopter and Tricopter Mega Link Index
Thanks, Jim. It happened with two different motors, but I will check them anyway. The lines are less than 30cm.

Quote:
Originally Posted by simonk View Post
Hello! D'oh! Sure, or we could also meet since you're so close, rather than shipping things. Maybe we can revive them. I've been sick with various things for a couple of weeks, but I'm better again now...

Which HK 18A board were these? Was this with tgy.hex or a newer build? Was the screaming sound medium or higher pitched whine or was it more of a juddering sound? Did you take off quicky from stopped, or were the motors already spinning (and sounding normal) for at least a second before you got off of the ground?

tgy.hex does not ramp start-up power, and this can cause false zero-crossings on _startup_ with the late demagnetization on some motors at high duty cycles. This is what caused the juddering on my tricopter motors when I had no radio plugged in to the KK board, as I mentioned previously.

The latest tree has ramped start-up power to try to avoid this problem, and also faster throttle response than even before (all loops now also check for RC pulse update and process, rather than once per commutation step). I also raised MIN_DUTY to hopefully make Nek happy with the "reduced power" since the clock rate increased, but I hadn't mentioned this yet because I keep meaning to find a way to compare it with stock Plush firmware, and my servo tester is so granular (4us steps) that it's difficult to compare. In other words, I think it's time to make a new build and make Nek happy with it so he stops recommending tgy.hex.
Thanks, Simon. The ESCs are this type http://www.hobbyking.com/hobbyking/s...idProduct=6548
running tgy.hex

The motors initially spinned OK, then during the takeoff motor #4 just quit. I actually have onboard video to show, if you want. It could be that I was running 4S battery. Previously on 3S it was flying well. I haven't opened the ESC yet. Prevois one had visual damage to one row of fets and also the transistor driving them. I will PM you my phone #.

Cheers,
Nick
nikivan is offline Find More Posts by nikivan
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Rctimer 30A esc cutting out with KK board Efliernz Multirotor Talk 2 May 26, 2011 03:20 PM
Sold FS: Tricopter Parts- HK401b, Pentium 30A ESC, rctimer 750kv motors pwu_1 Aircraft - General - Miscellaneous (FS/W) 2 Jan 18, 2011 10:58 PM
Sold RCTimer 2826-6 + ESC 30A New in packaging will trade as well Pics Added! planebldr Aircraft - Electric - Power Systems (FS/W) 4 May 25, 2010 10:54 PM
Discussion RCTimer 30A Programmable ESC hunter7193 Power Systems 8 Jan 02, 2010 03:27 PM
For Sale Several New ESC's - 10a, 20a, 30a - Reduced Callen9345 Aircraft - Electric - Power Systems (FS/W) 0 Apr 20, 2009 03:31 PM