I am using 4 TMP05 sensors with ADT7470 in daisy chain mode. We are currently testing it on our proto boards, where we found that in two boards, they are giving erraneous readings. In one board the values show large fluctuations in the reading and at times going to negative values. On debugging, we found that the first and last IC of the chain had gone bad. Because the O/P pin is giving a strobe signal without any output pulse. But even after replacing we found that the ICs are going bad repeatedly.
Can you please explain on what could be the issue?
On the second board, it is showing negative readings continuously. Here too, we found that the ICs have gone bad. These are newly mounted ICs.
I have attached the waveform of the O/P pin of the first device in the chain of both the cases. And the schematics too.
May I know at what environment you are testing - do you have a reference temperature? You mention have erroneous data, how often do they appear?
The IC with negative readings screenshot, which TMP05 in the chain is it? If you have 4 TMP05s and probing the last in the chain, there should also be 4 temp measurements as seen on Figure 27 of the TMP05. Do you have scope shots of when the daisy chain is operating properly? The Strobe signal without any reading in the OP, at what pin is this as well?
By the "IC have gone bad", do you test them individually and find that they don't have correct output?
Hi Karen, (KarenNE)
We have followed 4 TMP05s at 4 different locations of the board. The ambient is around 25 degrees. But the 4 TMP05s are showing negative readings. At times it goes to +13 degrees or so. It changes drastically and randomly. But never shows the expected reading. It is always erraneous readings.
I am attaching the scope shot of the I/P and O/P pin of the FIRST TMP05 in the chain, where it is not working.
I do not have scope shots of when the daisy chain is working properly. But, in the non working system, it is observed that the pulse at the O/P pin has a fixed Th and Tl irrespective of the START pulse. This is not expected as per the info given in the datasheet. And also there is no STOP pulse after the O/P pulse. But, the START pulse can be captured as expected at the Input pin of the TMP05 which is issued by the ADT7470.
And this is how I debug if any of the TMP05 of the 4 has gone bad, individually: I unmount the first one and short the I/P and O/P pin and see if the remaining 3 shows correct readings. And continue with the same procedure. Because unless the first one gives a START pulse for the second one, it can't work properly right?
Also, can you please tell us about the soldering temperature to be used and ESD sensitivity of the device?
This is the only issue holding up the product release now.
The signal in green in the second scopeshot shows the START pulse at the I/P pin.
There is information on the absolute maximum ratings section of the datasheet about peak temperatures for reflow soldering.
On the process of unmounting the first one, that would break the chain unless I misunderstand the process. And which TMP05 input is shorted to which output?
It would be helpful to know if each TMP05 is working fine when not in the chain, then we will be able to say that they're not working - try each in continuously converting mode and check what the output is. If they are working fine, connect them in daisy chain but use a pulse/pattern generator instead of the ADT7470 for the start pulse.
Also, what supply are you using?
I'm checking the ESD classificatioon for this as well and will get back to you as soon as I get this information.
Thanks for the help. We found that it was a soldering related issue. The ICs are working fine when we reduced the soldering temperature.
Sorry for the trouble.
Glad to hear this is sorted out Rasif.