i2c bus micro temp sensor problem

[mcgyvr]

see attached schematic (only a portion to protect the design)

Just started production runs on a microprocessor unit that goes into one of our products.
We got 5 boards done on a quick turn for orders we have in house already.
We fired them up to test them and found that after about 2 minutes the U7 chip (TC74 temp sensor "onboard") seems to stop functioning. (It displays the temperature on the LCD screen for about 2 minutes then the display turns to "--" (which indicates an error or no input from the temp sensor).

We think we have narrowed it down to a hardware issue as the prototype boards (same pcb layout/components) when programmed with the same new firmware do not have this problem. eeprom/temp chip and 2 external temp sensors (same chip) are all on the same I2C bus. The 2 external temp sensors do not work at all. All other analog inputs work fine. SCL1 and SCL2 on the schematic go to 2 other I/O pins on the micro. If you power down the unit and wait for about a minute or so and fire it back up the temp is displaying correctly again until the 2 minutes or so is up and it cuts out again. If you turn it off and right back on again after failure the issue is still there.

I was just wondering if you could think of any issue that might cause this. It just seems funny that it works fine for about 2 minutes then just stops.
Also the 2 external temp sensors do not display a reading at all anytime even when the onboard temp is working.

I don't have access to a scope/LA right now but should have one in a few days.

The company that designed this for us has gone belly up and is of no help.

 

[MacGyverS2000]

So many things this could be, without a scope/meter it's difficult to narrow down.

Is sclk still getting a signal after it fails? Is the SDA line being dragged high/low permanently by another chip? Can you separate those other chips from the control lines for testing? Is the voltage stable? Is the I2C controller hardware-based or software, and if software-based, are you sure of the design? Do you see current draw increase after it fails? Is your power supply strong enough?

Dan - Owner

http://www.Hi-TecDesigns.com

 

[mcgyvr]

I'll answer what I can.
1-voltage seems stable 3.298v right at startup 3.278v around the time the device stops.
2-IIC controller is hardware based built into the micro I believe (micro is Freescale MC9S12NE64VTU)
3-Do not see current draw after failure.
4-Power supply is strong enough.

Like I said all the hardware side (power supply,etc..) was working just fine in proto's. No power issues that I know of.

I will try and see if I can remove all but the TC74 and eeprom from the bus for testing.

 

[MacGyverS2000]

Could the EEPROM be hanging up any of the bus lines? Is it continually accessed throughout operation or intermittently? Is there a pattern between EEPROM access and failure of the temp units?

Dan - Owner

http://www.Hi-TecDesigns.com

 

[mcgyvr]

I found the problem (with my very limited electronic background)
R22 was a 100 ohm resistor and not a 1K as it should have been.
I am very proud of myself..

I feel bad for the assembly house who now has to go back and replace 3500 SMT resistors.. Seems their purchasing person missed a 0 when purchasing the parts.

Thanks for your help

 

[MacGyverS2000]

Well, that certainly qualifies as an answer to my second question... ;-)

Dan - Owner

http://www.Hi-TecDesigns.com

 

[mcgyvr]

Yes thats how I figured out the problem.
I was trying to answer/understand your other questions and BAM there it was... wrong pull up resistor

 

[itsmoked]

Don't feel sorry for them.. Put one check in the 'these slobs messed up' column.

Keith Cress
kcress -

http://www.flaminsystems.com