MAX31855 temperature error (reads 25% low)

[David_AVD]

Sample from Maxim?

That would take weeks.  I'll get a breakout board from Wiltronics in the mean time as I need to get this prototype running ASAP.

[Tinkerer]

It really does look like the chip is the wrong curve.

I bought the chips from Digikey; 2 pieces originally then another 2 a month later.  The two prototypes have one from each purchase, but they are probably the same batch.

Unfortunately neither RS or Farnell have stock, so can't easily get another chip to see if that's the issue.   

I'll see if I can get one somewhere else and go from there.

I have heard that Digikey will substitute 'equivalent' parts when they can if what they sold isnt actually in stock(and without telling you; happened to my company a few times). Are you sure the chip you ordered and the chip you got are the same?(obvious question, I know)

[free_electron]

boiling water ? water is conductive.... make sure you have a thermcouple that is fully electrically isolated ...

[CrabxCore]

Sample from Maxim?

That would take weeks.  I'll get a breakout board from Wiltronics in the mean time as I need to get this prototype running ASAP.

This is how I got my 2 chips for my project. They shipped from the philippines to the USA in about a week. Much faster than what I was expecting!

[mg326]

Hi David_AVD,

I have been using the MAX31855 in a project and have produced a number of prototype boards using it and have had no problems up until this point. I have just however produced a new batch of these boards and am seeing the exact same problem you have described on 75% of them (i.e. reading ~78°C at 100°C input from an Omega thermocouple calibration unit). The new board has some very minor alterations from previous versions, none of which should be relevant to this behaviour from the MAX31855.

If you make any progress with finding out whats going on, please post it up as it would be hugely helpful to me too, but from my perspective it definitely looks like a problem with the component itself.

[amyk]

We now have 2 reports of MAX31855s behaving wrongly in the same way, so a mismarked batch seems likely.

[David_AVD]

boiling water ? water is conductive.... make sure you have a thermcouple that is fully electrically isolated ...

Yes, the thermocouple is fully isolated and suitable for liquid immersion.   

[David_AVD]

I have been using the MAX31855 in a project and have produced a number of prototype boards using it and have had no problems up until this point. I have just however produced a new batch of these boards and am seeing the exact same problem you have described on 75% of them (i.e. reading ~78°C at 100°C input from an Omega thermocouple calibration unit). The new board has some very minor alterations from previous versions, none of which should be relevant to this behaviour from the MAX31855.

Interesting.  I've wasted way too much time on what should have been such a simple circuit.  I'll update here once I get another chip.

EDIT: I've emailed Maxim support to see what they say.

[mg326]

Have just had confirmation from MAXIM that it was a faulty batch, the affect chips are:

Date Code: 1352A2
Lot Code: 253AL

[David_AVD]

That batch is what two of the chips I have are.  I *think* the other two were different, but will double check when I get into work.

[David_AVD]

Just checked the other prototype and it has batch:
Date code: 1435A2
Lot code:  253AK

Both batches of chips are giving me the same result, so maybe there is more than one bad batch ?

Maxim did reply to my enquiry, but only to say they are looking into it and will reply again ASAP.

[David_AVD]

Just got word from Maxim:

Replacement parts should not be from lots ending in either "253" or "107"

[David_AVD]

Well, Digikey sent me 2 replacement chips as soon as I verified when I purchased them.  They arrived today and they are batch code 107AE.

I wasn't confident that they would be any better than the 253xx batch, but just replaced the chip in both prototypes and the temperature is spot on.  Whoo!

So, not entirely sure what to make of Maxim's email saying the 107xx batch are also bad.  The 253xx batch do seem to have been proven bad of course.

[dannyf]

not entirely sure what to make of Maxim's email saying the 107xx batch are also bad.

Maxim has poor documentation of its production process? That's scary.

[amyk]

They probably mean that some of the 107xx batch could be bad, and they're not sure exactly which ones, so to be safe they're saying to avoid the whole batch.

[David_AVD]

Still somewhat annoyed at the (unchargeable) wasted hours on this project due to something out of my control and should never happened.   

[dannyf]

they're not sure exactly which ones,

That is not tolerable for production purposes.

[48X24X48X]

Hi guys,

I experienced the same problem of the same batch of the MAX31855KASA+ bought from Mouser.
It's painful because I already shipped a handful of them to customer. 

In my case, the differences of reading started to show more when the temperature goes up higher.
I got a toasted PCB as a result. Now, I'm wondering whether Maxim did a random test on each batch.

[48X24X48X]

So it looks like the trim issue is only on MAX31855KASA+ units with lot codes 253AL and 253AK with date code 1352A2. I also notice that data code 1345 is also with the same issue. other data code should be ok.

That is the reply from Maxim.
No apologies given by them on top of that.