LM86: Sometimes get temperature 127 degree C from RTHB

Jim Tsan

Prodigy 115 points

Part Number: LM86

Dear Sir,

We polling the LM86 to get the remote temperature one times per second.

But sometimes it returns an error temperature with 127 degree C.

Are there any setting, limitation or errata we should take care of it?

Thanks.

over 5 years ago

0 David Vaseliou over 5 years ago

TI__Expert 6820 points

Hi Jim,

Polling one time per second should not be an issue. Getting a reading of 127 °C could indicate an open or short on the D+ pin. Can you please check the Status register and any alert pins to see if this is the case?

Thanks,

David

0 Blake Kuo over 5 years ago in reply to David Vaseliou

Intellectual 300 points

Hi David,

Did you mean that might be short or open issue if no alert signal is triggered ?

Thanks.

Blake

0 Jim Tsan over 5 years ago in reply to David Vaseliou

Prodigy 115 points

Hi David,

Due to our threshold of Remote temperature is 100.C, alert pins should be triggered if 127 is a true temperature. How can we judge if LM86 is normal by alert pins?

If the D+ pin is short, what can we detected on the STATUS_REGISTER (03h)?

Thanks,

Jim

0 David Vaseliou over 5 years ago in reply to Jim Tsan

TI__Expert 6820 points

Hi Jim,

No, the status register will not indicate a short condition. Can you please send an oscilloscope waveform of the D+ and D- pins for the conversion that results in a 127 °C reading? This should be the conversion right before the Alert pin is asserted and can be used as your trigger for the oscilloscope.

The D+ pin should look similar to the figure below.

Thanks,

David

0 Blake Kuo over 5 years ago in reply to David Vaseliou

Intellectual 300 points

Hi David,

We have checked the D+/ D- pin and no open or short issue found.

Do you have other suggestions or comments for calrifying why we get 127C from remote sensor ?

I think that this is not a real temperature, this is a impossible temperature if we just turn-on our system for few seconds.

Thanks

Blake

0 David Vaseliou over 5 years ago in reply to Blake Kuo

TI__Expert 6820 points

Blake,

Can you please read the status register and send me the result when you see 127C reading. Also can you send the oscilloscope waveform of the D+/- pin right before the Alert pin is triggered.

Thanks,

David

0 David Vaseliou over 5 years ago in reply to David Vaseliou

TI__Expert 6820 points

Blake,

Can you also send me a pdf of the schematic?

Thanks,

David

0 Blake Kuo over 5 years ago in reply to David Vaseliou

Intellectual 300 points

Hi David,

We found that migh be D+ open issue on thermal diode side.

But I still have a question about abnormal waveform from D+ pin. How to interpret the abnormal or abnormal waveform ?

Thanks.

BRs.

Blake

0 David Vaseliou over 5 years ago in reply to Blake Kuo

TI__Expert 6820 points

Hi Blake,

From the waveform it looks to be open as well. These remote pins inject different currents and read back a voltage at the different current levels. The change in voltage with respects to the current injection is dependent on the temperature of the thermal diode.

Since the thermal diode is open you can see that the D+ is railing at 3.3 V because the device cannot inject current into an open circuit. If you would like to learn more about how remote temperature sensors work you can read or watch some of the material I have posted below.

Application Note: Optimizing Remote Temperature Sensor Design

Video on die temperature monitoring: TI Precision Labs - Temperature Sensors: How to Monitor Die Temperature

Thanks,

David

Sensors

Sensors forum

LM86: Sometimes get temperature 127 degree C from RTHB