finding short on motherboards with a shorty (with display)

[kripton2035]

@jdobry: please make a new thread with your device details... this thread is for the shorty-with-display. thanks.

[Miti]

Hi jdobry,

Our troubleshooters liked your project so we created a 10 up panel and assembled 10 boards on the SMT line. I programmed them with v1.02.02 and everything looks good but I have an issue with the I2C communication, I think. The OLED goes crazy after some time, between few seconds and about a minute. The display and sound freeze with a broken display or not. The SW is still running because I can turn it off with a long press and it respond with a long beep. My guess is that the I2C is too fast for some OLEDs. Is it possible to slow it down a bit?

Thanks ,
Miti

[Kean]

Hi jdobry,

Our troubleshooters liked your project so we created a 10 up panel and assembled 10 boards on the SMT line. I programmed them with v1.02.02 and everything looks good but I have an issue with the I2C communication, I think. The OLED goes crazy after some time, between few seconds and about a minute. The display and sound freeze with a broken display or not. The SW is still running because I can turn it off with a long press and it respond with a long beep. My guess is that the I2C is too fast for some OLEDs. Is it possible to slow it down a bit?

Thanks ,
Miti

Without seeing them, I doubt the I2C bus is "too fast" or the ones Jiri made would not have worked.  I guess the OLEDs could be marginal.
As a troubleshooting step, I would look at the I2C signals on a scope.  That would confirm the speed and shape of the I2C signals.
But even before that, I see on the schematic that the I2C pullups are 10k which is a bit high for 3.3V.  Maybe try decreasing R10 and R11 to something in the range of 3k3 to 4k7.

I haven't made any of these yet, but I am about to order some of the PCBs and parts.

[Miti]

I doubt the I2C bus is "too fast" or the ones Jiri made would not have worked.  I guess the OLEDs could be marginal.

Yeap, another way of saying the same thing. Too fast for the OLED. He may have been lucky with very fast OLEDs, mine may be a bit slower.
I did play a bit with the pull ups, I soldered 4.7K in parallel with the 10K. First I soldered one on the SCL and the display didn't come on at all. Then I soldered on both SCL and SDA and it came back on but crashing again. I will post some scope screen shots.

[Kean]

I assume you must have used SSD1306 based OLEDs to be compatible.  The SSD1306 is specified to support up to 400kHz I2C.

Do you have a link to the OLEDs you bought?
I'm not a fan of TZT stores on AliExpress, but was planning to buy the suggested ones at https://www.aliexpress.com/item/32927682460.html

[Miti]

I assume you must have used SSD1306 based OLEDs to be compatible.  The SSD1306 is specified to support up to 400kHz I2C.

Do you have a link to the OLEDs you bought?
I'm not a fan of TZT stores on AliExpress, but was planning to buy the suggested ones at https://www.aliexpress.com/item/32927682460.html

That’s where I bought from. I’m pretty sure that the SCK is around 800kHz so I may be right about “too fast”.

Edit: Added SCK scope shot.
Edit1: The start condition Hold Time is too short as well, minimum is 600ns.

[Kean]

@Miti OK, it looks like you might be right then and thanks for checking the signals.
@jdobry will need to look into this.  His firmware must be enabling I2C Fast-mode Plus (~1MHz).
I didn't dig too deep into the code, but I saw it being enabled in the HAL defs file.

[Miti]

I don’t think Jiri is still maintaining this project.
I installed STM32 Cube and tried to change the I2C mode from FastModePlus to FastMode but I can’t figure out how to do that. 
Moreover, the project doesn’t seem to be the latest one. I guess I’m 

[Miti]

Ok, I’ve modified the I2C timing variables and the clock is now 400kHz but the display is still breaking. My displays may be junk or there’s something else wrong with the software.

[Kean]

Ok, I’ve modified the I2C timing variables and the clock is now 400kHz but the display is still breaking. My displays may be junk or there’s something else wrong with the software.

I will be assembling mine in the next week, so I will report back when I've done that.
Maybe you can share your modified firmware with 400kHz I2C?

[Miti]

Here it is.

[Lagros]

Try unsoldering R10 and R11 and turn it on without these resistors.

[Miti]

Try unsoldering R10 and R11 and turn it on without these resistors.

How would it work without pull-up resistors? What do you suspect?

[Lagros]

Pull-up resistors are present in OLED.
Try it!

[Ocela]

I don’t think Jiri is still maintaining this project.
I installed STM32 Cube and tried to change the I2C mode from FastModePlus to FastMode but I can’t figure out how to do that. 
Moreover, the project doesn’t seem to be the latest one. I guess I’m 

Miti, this is not enough. It is also necessary to change the frequencies inside the microcontroller buses
The pull-up resistors for the I2C bus and 3.3V voltage are not selected correctly. They should be equal to 1k. You can calculate this using an online calculator.

[wacd]

I built this Shorty on breadboard it works relatively good but there is one thing that i don't understand.

When i measure various shunt resistors it shows me strange results

0.0200R - 0.0189
0.0250R - 0.0245
0.0330R - 0.0333
0.0400R - 0.0430
0.1000R - 0.1100

Something wrong but i don't know what.

[kripton2035]

did you build MY shorty-with-display with arduino nano, or the "clone" with ST controller ?
the shorty-with-display is not a precision instrument, so the error is possible. it's only to measure relative to another point
also I suspect the device to change the value when it becomes hotter with time. it could change values

[Antonio90]

Hello, all results are within 5% tolerance, except the last one, which is 10%.
What is the tolerance of the shunts? 5-10%? Even if the resistors are 1% the results are well within expected values for a continuity tester IMHO.

[wacd]

I removed these shunts from laptop motherboards

There is something about the results , its like a pattern for example

-10
-5
-2
-0
+2
+5
+10

Sorry for not being able to explain myself i'm just a begginer at this kind of work and a complete stranger to the software side of the project.

[wacd]

Yes its your version Kripton. It's a clone nano. MCU has Atmel writing on it. 328P.

@Antonio90

All of them are %1


If i play around with these configs , i can get it to display the exact value of the resistor that i'm measuring. But when i change the resistor with a smaller or with a bigger value , it just does not measure it correctly. How can this be improved any suggestions?

[wacd]

Closest one is 0.025 , it measured 0.022

Others are way off and it gets worse with bigger value resistors.

[kripton2035]

so the inverse calc of the resistor is not as linear as the resistor value ? it can be...

[wacd]

I really have no idea about the programming side but the problem is, it displays lower value than it actually is under certain value resistor , and displays higher value than it actually is above certain resistor value. i hope i was able to explain myself with my broken English 

[kripton2035]

this line could be modified to adjust to the measured values

[wacd]

Yea i played with those values for 2 days and got no real results , i guess i'll have to make do with what i have now.