I ordered 3 ea TMP36 (equivilent to LM35) from Spark Fun, and they were all crap. Two were DOA and the other worked for awhile before crapping out. Might i expect better reliability from the LM35?

 

TMP36 does not take 35 volts like LM35 does.

Also. Reading the comments... The datasheet shows bottom view, so maybe you connected the chip incorrectly.

 

TMP36 does not take 35 volts like LM35 does.

Also. Reading the comments... The datasheet shows bottom view, so maybe you connected the chip incorrectly.

Pinouts are the same on TMP36 and LM35, but absolute max. supply voltage on the TMP36 is only 7V. Not good!

 

Pinouts are the same on TMP36 and LM35, but absolute max. supply voltage on the TMP36 is only 7V. Not good!

Well. We don't know what Brownout used to power it. All we know is that they have a problem with it.

Another thing. Reading more comments... Another person said that they got TM335 instead of TMP36. So maybe it is not what Brownout did, but what Spark Fun sent out.


I have worked with TMP36. I ordered mine from Adafruit and I followed Adafruit tutorial about it, so I never had a problem with it.

 

I followed the datasheet.

 

Might i expect better reliability from the LM35?

In direct answer to your question, I have used LM35s occasionally and never had a problem, but I have used the LM34 (F version) extensively, and never had a problem with it. (Although I did get some counterfeit LM34s once from Jameco.)

 

Maybe that's what I'm dealing with.

 

LM35 can have issues with oscillation/latchup etc.

The LM35 datasheet mentions a fix, by adding a cap and a resistor if I remember right.

If your TMP36 is a close copy of the LM35 it might have that same problem, and depending on your circuit layout/wiring etc it would explain why the new devices failed to work.

 

Thanks. Could be latchup, the output is near zero.

 

Brownout,

Have you considered the Dallas One Wire devices? I am a big fan.

 

I just ordered a batch of LM35's. I think they should be more robust than the TMP36's. For other devices, eg. the One Wire, I plan to use as many standard devices and test them with the PIC as I can. I only just got the serial port working.

 

I'm NOT a big fan of dallas 1-wire temp sensors.

They are overly expensive, slow and clunky to get a temperature reading from, and I don't like the idea of placing a tiny micro inside a temperature sensor that will get cooked...

The ONLY good point I can see is that you can string a lot of them on one signal wire. Everything else is bad.

 

I'm NOT a big fan of dallas 1-wire temp sensors.

They are overly expensive, slow and clunky to get a temperature reading from, and I don't like the idea of placing a tiny micro inside a temperature sensor that will get cooked...

The ONLY good point I can see is that you can string a lot of them on one signal wire. Everything else is bad.

Good points but many irrelevant from a hobbyist view. You can get them pretty cheap from China, and most hobbyists aren't buying bunches of them anyway.

I will admit they can be a real challenge to interface to, but then again that is half he fun.

Ironically I am having just those issues right now. I have some code that works but I realized the clock setting that I used for my delays was wrong, but somehow the code works. Changing to the right setting causes the code to fail! I am sure it is all delay related, it is trying to figure out just where is what the issue is.

 

The LM 35's arrived. I plugged on in, and the project spraing back to life. Don't know what the deal was with the TMP36's, but no longer care.