Discussion of Mark's firmware

Forum for announcements and discussion of beta firmware.
Paul.Allen
Site Admin
Posts: 98
Joined: Tue Aug 06, 2013 5:33 am
Location: Utah, United States

Re: Discussion of Mark's firmware

Post by Paul.Allen » Mon Nov 25, 2013 5:02 pm

You know, there is a very small possibility that it is the wrong sensor. I used a different one in the beginning that started at 400mv and had a 15mV per deg C (from memory could be wrong). Can you look carefully with a magnifier and a flashlight to see if the text on both temperature sensors match exactly?

Tcepsa
Posts: 46
Joined: Tue Aug 06, 2013 5:26 pm

Re: Discussion of Mark's firmware

Post by Tcepsa » Mon Nov 25, 2013 9:36 pm

Paul.Allen wrote:You know, there is a very small possibility that it is the wrong sensor. I used a different one in the beginning that started at 400mv and had a 15mV per deg C (from memory could be wrong). Can you look carefully with a magnifier and a flashlight to see if the text on both temperature sensors match exactly?
Will do!

Mark
Site Admin
Posts: 415
Joined: Sun Jul 28, 2013 6:47 am
Location: Brisbane, Australia

Re: Discussion of Mark's firmware

Post by Mark » Tue Nov 26, 2013 12:19 am

Paul.Allen wrote:You know, there is a very small possibility that it is the wrong sensor. I used a different one in the beginning that started at 400mv and had a 15mV per deg C (from memory could be wrong). Can you look carefully with a magnifier and a flashlight to see if the text on both temperature sensors match exactly?
It appears that they're giving similar readings, so if they're wrong, they're probably still the same as each other.

Would be helpful to know what the text actually is maybe though?

Paul.Allen
Site Admin
Posts: 98
Joined: Tue Aug 06, 2013 5:33 am
Location: Utah, United States

Re: Discussion of Mark's firmware

Post by Paul.Allen » Sat Nov 30, 2013 11:10 pm

Hey Mark, Is there a way you could do some hysteresis on the temperature sensors. Like for example, maybe it has to get a reading up or down from the current reading two times in a row before it changes and then one it changes, samething, it has to get a differnt temp two times in a row before it changes again. I only suggest this to get rid of the occasional flicker up and down when a reading is right on the cusp of two different temperatures.

Also, Tcepsa, so what is the text on your temp sensors?... They are the 3 pin little black things right above the negitive battery terminal.

Mark
Site Admin
Posts: 415
Joined: Sun Jul 28, 2013 6:47 am
Location: Brisbane, Australia

Re: Discussion of Mark's firmware

Post by Mark » Sun Dec 01, 2013 1:01 pm

Paul.Allen wrote:Hey Mark, Is there a way you could do some hysteresis on the temperature sensors. Like for example, maybe it has to get a reading up or down from the current reading two times in a row before it changes and then one it changes, samething, it has to get a differnt temp two times in a row before it changes again. I only suggest this to get rid of the occasional flicker up and down when a reading is right on the cusp of two different temperatures.
Yes, I could do something along those lines. The simplest thing would be just to over sample the temperature and ignore the highest and lowest readings.

I'm not sure if there is really that much point though? I think it's useful to be able to see when the temperature reading is borderline between 2 values...

Tcepsa
Posts: 46
Joined: Tue Aug 06, 2013 5:26 pm

Re: Discussion of Mark's firmware

Post by Tcepsa » Wed Dec 04, 2013 4:56 am

Hello, sorry to be so late with this; Thanksgiving and family visiting and all.
Paul.Allen wrote:Also, Tcepsa, so what is the text on your temp sensors?... They are the 3 pin little black things right above the negitive battery terminal.
1) Looks like the text on the temperature sensors is "AL8J" on both versions of the board that I have. Does that match, or am I looking at the wrong part?

2) I've gone back and checked against version 0.512 and I am still getting weird readings. I'm going to go back through the earlier posts in this thread and try to find the earliest version of the firmware that supports this revision of the board to make sure I check as far back as possible.

Mark
Site Admin
Posts: 415
Joined: Sun Jul 28, 2013 6:47 am
Location: Brisbane, Australia

Re: Discussion of Mark's firmware

Post by Mark » Thu Dec 05, 2013 5:26 am

Tcepsa wrote:1) Looks like the text on the temperature sensors is "AL8J" on both versions of the board that I have. Does that match, or am I looking at the wrong part?
Looking at mine, it looks more like ALBJ although there is at least one where it looks more like an 8 than a B.
2) I've gone back and checked against version 0.512 and I am still getting weird readings. I'm going to go back through the earlier posts in this thread and try to find the earliest version of the firmware that supports this revision of the board to make sure I check as far back as possible.
So you're getting strange readings on both of your chargers?

Tcepsa
Posts: 46
Joined: Tue Aug 06, 2013 5:26 pm

Re: Discussion of Mark's firmware

Post by Tcepsa » Thu Dec 05, 2013 6:01 am

Mark wrote:
Tcepsa wrote:1) Looks like the text on the temperature sensors is "AL8J" on both versions of the board that I have. Does that match, or am I looking at the wrong part?
Looking at mine, it looks more like ALBJ although there is at least one where it looks more like an 8 than a B.
You're right, it's ALBJ on all four of mine.
Mark wrote:
Tcepsa wrote:2) I've gone back and checked against version 0.512 and I am still getting weird readings. I'm going to go back through the earlier posts in this thread and try to find the earliest version of the firmware that supports this revision of the board to make sure I check as far back as possible.
So you're getting strange readings on both of your chargers?
I'm getting strange readings on the newer one (Rev. 7). The older (purple) charger is coming up with values that make sense, though one seems to be reading a degree (C) or so cooler than the other.

Mark
Site Admin
Posts: 415
Joined: Sun Jul 28, 2013 6:47 am
Location: Brisbane, Australia

Re: Discussion of Mark's firmware

Post by Mark » Sun Dec 08, 2013 3:10 am

Tcepsa wrote:
Mark wrote:So you're getting strange readings on both of your chargers?
I'm getting strange readings on the newer one (Rev. 7). The older (purple) charger is coming up with values that make sense, though one seems to be reading a degree (C) or so cooler than the other.
I've had a look back through your previous posts and found the one where you posted your first results with your revision 7 charger, and the readings then were what would be expected:
viewtopic.php?f=3&t=37&p=222#p222

Doing a few calculations, the values that you're getting now are consistent with the analog readings being taken with a 5V reference instead of a 3.3V reference. I suspect that the 3.3V regulator has failed such that it's passing 5V straight through.

This shouldn't be a problem for the 328P MCU since they're rated to run at 5V, but it could well destroy any SD card that you plug into the charger since they're supposed to work at 3.3V.

You could confirm my diagnosis by measuring the voltage between pins 1 and 3 on the FTDI connector - counting from the right. It should measure 3.3V, but if I'm right with the above, you'll see 5V there. If you try this be very careful not to short any of the pins on that connector as there is the potential to fry the MCU if you do.

You might also notice that the LEDs are slightly brighter now on your Rev 7 board compared to the older one.

If you were happy to not use the SD slot, I could give you a modification to the source code so that you would get temperature readings that make sense again, however you would have a resolution of about 0.5 degrees instead of 0.3 degrees C.

It's probably best to post your board back to Paul so that he can confirm the above and replace the 3.3V regulator for you.

Tcepsa
Posts: 46
Joined: Tue Aug 06, 2013 5:26 pm

Re: Discussion of Mark's firmware

Post by Tcepsa » Sat Dec 14, 2013 5:18 pm

I think I might have figured it out. I took the reading off of pins 1 and 3 on the FTDI connector as you suggested and it was the expected 3.3 V, so my 3.3 V supply regulator is still good. However, in order to get the temperature (and other value) readings, I have been leaving my FTDI cable connected to that header so that I can get the serial values in the terminal. Then I remembered that some FTDI cables are 5V rather than 3.3V, so I also measured pins 1 and 3 of my FTDI cable and sure enough it's a 5V cable. So whenever I have the FTDI cable plugged in, it is getting 5V. I'm counting myself lucky that I have not burned out any of the components!

Should the board maybe be redesigned so that the power pins from the FTDI cable are also on the 5V side of the 3.3V regulator, just in case?

If it's just a matter of getting incorrect temperature readings if you have a 5V cable connected, then it's maybe not worth it. But if there's other damage that I (and other unsuspecting users) might be doing, then it's probably worth at least considering, or putting a warning or something (and if there was an "only use 3.3V FTDI cables" warning somewhere and I just missed it then that's on me--ideally such warnings wouldn't be necessary, but it sounds like it is late enough in the design process, at least for the 2-cell model, that making that change would probably be overly onerous).

In the meantime, I'll try adding two 1N2004 diodes between pin 3 of my FTDI cable and pin 3 of the FTDI connector on the board and seeing whether my temperature readings get a little more realistic.

Post Reply