Voltage drop in 9x

Rad · Post by **Rad** » Sat Jul 18, 2020 4:29 pm

Hi guys, strange issue here:

2 Radios, one flysky and one turnigy 9x, but with erskyTxR223A3 flashed, 2 three position switches added, hobbyking backlight, the resistor for module protection, and the mosfet for battery reverse protection. The turnigy has also a smartieparts programmer.

Both radios work good, but I think I might have an issue with the turnigy.
With a full battery, lets say the hobbyking lifepo, it says it has 10.4v when fully loaded, but if I enter in a menu, or create a mix or something, when I go back to the main screen the voltage drops a lot, until 5v or so, but in two seconds goes back to normal. It does not happen always, but relativelly often.

I wonder if that is normal or if it means any kind of issue. I did all the mods in both radios and they look exactly the same to me.

Any info would be appreciated! Hope it is fine.

And yeah, I tried the same battery in both radios. And also with another 18650 battery I did myself, and I have the same issue.

Post by **MikeB** » Sat Jul 18, 2020 7:42 pm

Please confirm these a 9X radios without any upgrade boards and you have er9x on them (not erskyTx), or otherwise, what upgrade boards are you using to be able to run erskyTx?
If er9x, what setting do you have, in the hardware menu, for "Bandgap", on or off?

Mike

Rad · Post by **Rad** » Sat Jul 18, 2020 8:46 pm

MikeB wrote: ↑Sat Jul 18, 2020 7:42 pm Please confirm these a 9X radios without any upgrade boards and you have er9x on them (not erskyTx), or otherwise, what upgrade boards are you using to be able to run erskyTx?
If er9x, what setting do you have, in the hardware menu, for "Bandgap", on or off?

Mike

Hi Mike, it is er9x, the test version you uploaded yesterday.
Using a smartieparts v1.1 programmer
Bandgap is on.

Both radios are the same, except one is flysky AFHDS and the one with this little issue is turnigy, AFHDS 2A, but I fitted a multiprotocol module. Both with the 128 processor

Rad · Post by **Rad** » Sat Jul 18, 2020 8:52 pm

The voltage drop happens like once of every 10-15 times or so I go to the mixer menu or radio/model setup menu. Sometimes I even see the splash screen for a fraction of second.

Post by **MikeB** » Sat Jul 18, 2020 9:21 pm

Sounds like you may be getting a watchdog reboot. On a normal start, the battery voltage is read, then new values displayed are the average of the current value and the new value. On a watchdog reboot, the voltage is likely left at 0, so the first average is half the battery voltage (5.2?), and will rise over time to the correct value.

Have you tried with the modules swapped over?

Mike

Rad · Post by **Rad** » Sat Jul 18, 2020 9:27 pm

MikeB wrote: ↑Sat Jul 18, 2020 9:21 pm Sounds like you may be getting a watchdog reboot. On a normal start, the battery voltage is read, then new values displayed are the average of the current value and the new value. On a watchdog reboot, the voltage is likely left at 0, so the first average is half the battery voltage (5.2?), and will rise over time to the correct value.

Have you tried with the modules swapped over?

Mike

I just noticed it is exactly half battery!
I fitted the flysky afhds module in the turnigy radio and it does the same, so I guess we can discard the module. What can be the problem and why does this happen? And how do I correct it?

Rad · Post by **Rad** » Sat Jul 18, 2020 9:43 pm

Ok, this might be a bit more serious than I thought. Now the radio just did not switch on. I had the backlight on, but nothing else, and a couple of days ago two models dissapeared (model 4 and 5)

Now I managed to recover it flashing the software again with eepe and coping the models to the radio and so far it works again.

Not sure what is going wrong... I can only think about the wires to the microprocessor. Id did it very carefully and seems fine to me but do you know if one of those two pins shorted can cause this issue?

Post by **MikeB** » Sat Jul 18, 2020 9:53 pm

If it only happens around 1 in 10 times, it may be difficult to locate.
Initial suggestions:
1. Backup the EEPROM from both radios, then put the EEPROM from the working radio on the Turnigy one and see if the problem is solved.
2. Try with no module and create a new model and see if the problem is solved.

Which two pins?
Have you done the telemetry mod, or just added the three positoon switches.

Mike

Rad · Post by **Rad** » Sat Jul 18, 2020 9:59 pm

Ok, I will try swaping the EEPROMs tomorrow morning.
I tried with no module and new model, and problem persisted.

The pins: The ones you told me here, here are my own pics:
viewtopic.php?f=26&t=12742&p=150437&hil ... in#p150437

No telemetry mods.

Here is the EEPROM of the faulty radio.

Model files radio..eepe: (7.39 KiB) Downloaded 229 times

Hope we can solve it, this radio is for a friend and that is a compromise...

Rad · Post by **Rad** » Sun Jul 19, 2020 7:25 am

Hi, tried swaping EEPROMS and the problem persist

Post by **MikeB** » Sun Jul 19, 2020 9:19 am

I'll give this some thought today, and possibly create a special test version of the firmware to try to find the problem.

Mike

Rad · Post by **Rad** » Sun Jul 19, 2020 9:39 am

MikeB wrote: ↑Sun Jul 19, 2020 9:19 am I'll give this some thought today, and possibly create a special test version of the firmware to try to find the problem.

Mike

Thanks! Very appreciated!

Post by **MikeB** » Sun Jul 19, 2020 7:29 pm

Are you flashing the "standard" er9x.hex (so I know which version to add debug into)?
I don't think a short on the added wiring would cause the problem you are seeing.
As a test, could you flash the release 821 (from er9x.com) and see if that has the same problem (the added 3-pos switches won't work).

Mike

Rad · Post by **Rad** » Sun Jul 19, 2020 9:21 pm

Hi, I was flashing V1, er9x psr822 Mod FRSKY 128

I just flashed the standard er9x 821 and after a short while I experienced that issue. I switched off the radio, then ON, and now the radio froze in the switch alert. After one reset, all the eprom were resetted
After a closing inspection I am pretty sure those pins are not shorted.

Considering one radio works and the other does not, I was pretty sure it had to be a mistake from my side. Discarding the pins to the processor, it had to be the mosfet. So I opened both radios and swapped the PCB with the switch, and after testing it for a couple of minutes, it seems to work good.

I attached a picture of the PCB, not the best pic and not the cleanest job but I guess I created an issue there. Or maybe is a faulty mosfet.

Do you see something? I guess I will just solder the diode again.

Rad · Post by **Rad** » Sun Jul 19, 2020 9:35 pm

And this is the PCB that works fine.

Post by **jhsa** » Sun Jul 19, 2020 10:17 pm

Hmm, can't see the connections and the cuts very well. Also what Mosfet arte you using?
On the board that doesn't work, make sure the Mosfet tab is not touching the ground plane just under it.

João

Rad · Post by **Rad** » Mon Jul 20, 2020 5:57 am

jhsa wrote: ↑Sun Jul 19, 2020 10:17 pm Hmm, can't see the connections and the cuts very well. Also what Mosfet arte you using?
On the board that doesn't work, make sure the Mosfet tab is not touching the ground plane just under it.

João

The mosfet is: https://www.aliexpress.com/item/3288405 ... hweb201603

I have some news, I started testing the faulty radio with the switch from the working radio, and I experienced the same issue. Then I tested with the other half of the working radio, and same issue. I guess it narrows the search to the main PCB.

If it helps, the microprocessor says:
"GREEN ED040502-H16D 1105,

No idea if that is good or not, but all the other radios I had were ATMEGA 128 or 64.

Rad · Post by **Rad** » Mon Jul 20, 2020 5:59 am

Uh oh, I have seen some issues with that "green" thing

https://openrcforums.com/forum/viewtopi ... 126&t=9387

Rad · Post by **Rad** » Mon Jul 20, 2020 6:57 am

Tried the switch from the defective radio in the working radio, and the working radio worked perfectly.

I bet the issue might come from this green processor then

Rad · Post by **Rad** » Mon Jul 20, 2020 7:03 am

jhsa wrote: ↑Sun Jul 19, 2020 10:17 pm Hmm, can't see the connections and the cuts very well. Also what Mosfet arte you using?
On the board that doesn't work, make sure the Mosfet tab is not touching the ground plane just under it.

João

I think the tab is indeed touching that (not 100% sure, because I loctited it). but I thought that tab was just for cooling heat, isnt it? so there is no way to make shorts.

Post by **Kilrah** » Mon Jul 20, 2020 7:21 am

It's for cooling but it IS connected to one of the 3 terminals and will definitely make shorts.

Rad · Post by **Rad** » Mon Jul 20, 2020 7:42 am

Kilrah wrote: ↑Mon Jul 20, 2020 7:21 am It's for cooling but it IS connected to one of the 3 terminals and will definitely make shorts.

Good to know, thanks. I removed it but still the problem persists...

Post by **MikeB** » Mon Jul 20, 2020 10:26 am

I've just posted a special test version, for the '128, with the watchdog set to 2 seconds instead of the normal 0.5 seconds. I've done this in case the "Green" processor watchdog is running fast, causing a watchdog reboot when it shouldn't happen.

Mike

Rad · Post by **Rad** » Mon Jul 20, 2020 11:03 am

MikeB wrote: ↑Mon Jul 20, 2020 10:26 am I've just posted a special test version, for the '128, with the watchdog set to 2 seconds instead of the normal 0.5 seconds. I've done this in case the "Green" processor watchdog is running fast, causing a watchdog reboot when it shouldn't happen.

Mike

Flashed! I think we might be on the right way... I think the voltage drop issue (and sometimes reboot) happens less often now with this firmware.. Maybe rising it to 4 seconds is even better?

I am going to flash my eeprom and I will update if something goes different.

EDIT: Definitelly this goes much better now... Maybe 4 seconds is too much and 3 does the job, but you are the expert here.

bob195558 · Post by **bob195558** » Mon Jul 20, 2020 11:26 am

Replace 5V regulator would may be a help.
https://openrcforums.com/forum/viewtopi ... =95&t=3819

Post by **jhsa** » Mon Jul 20, 2020 12:47 pm

I think this GREEN chip thingy might not be an 100% clone of the m128, and that might be the problem.

João

Rad · Post by **Rad** » Mon Jul 20, 2020 12:52 pm

jhsa wrote: ↑Mon Jul 20, 2020 12:47 pm I think this GREEN chip thingy might not be an 100% clone of the m128, and that might be the problem.

João

That is what I think... I have seen a bunch of poor souls with the green chip who are not able to flash Open TX... Hope it can be fixed somehow with the software...

One thing is clear for me, I wont be getting more turnigy radios.

bob195558 · Post by **bob195558** » Mon Jul 20, 2020 12:54 pm

Maybe then upgrade to the ATMEGA2561-16AU CPU Chip
https://openrcforums.com/forum/viewtopi ... 120#p74105

Rad · Post by **Rad** » Mon Jul 20, 2020 1:41 pm

bob195558 wrote: ↑Mon Jul 20, 2020 12:54 pm Maybe then upgrade to the ATMEGA2561-16AU CPU Chip
https://openrcforums.com/forum/viewtopi ... 120#p74105

Doing that I am afraid is way out of my skills, but thanks for the tip!

Post by **MikeB** » Mon Jul 20, 2020 2:05 pm

I've got a disassembly of the original firmware for the Green chip and it doesn't appear to set the watchdog at all.
For the Atmel 2 Seconds is the maximum for the watchdog timeout setting.
I'm wondering if the Green chip is clocking the watchdog faster, and using some unused bits in the configuration register to set up the longer timeouts. Either that, or the watchdog is just not working properly.

I've just posted a new test version that includes a 300mS software driven watchdog, while leaving the hardware watchdog set to 0.5 seconds (nominal).
If that works, I'll do another test version with something in that tests the hardware watchdog timeout to see what it actually is.

Mike

openrcforums

Voltage drop in 9x

Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x

Re: Voltage drop in 9x