TEMPDIODE reading low

[Altera_Forum]

I'm reading the temp diode on a stratix IV fpga, and it reading about 50C below what it should be. 

 It is changing in regards to thermal heat and cooling, but it always seems about 50C too low.   Thoughts?

[Altera_Forum]

What reading are you getting? What temperature are you expecting? 

 Jake

[Altera_Forum]

When I power up in the morning I get somewhere between -15 and -10C....seems a little cold for room temperature!

[Altera_Forum]

What value are you reading from the sensor (Digital value)? Are you using the internal ADC or an external ADC. If external, which sensor are you using? 

 Jake

[Altera_Forum]

I'm using the internal ADC. I think I'm getting the value 0x414....but the issue might be in my conversion.

[Altera_Forum]

I'm confused. I thought the internal diode only provided 8 bits of resolution as indicated by the alttemp_sense megafunction user guide: 

http://www.altera.com/literature/ug/ug_alttemp_sense.pdf  Jake

[Altera_Forum]

Sorry, I was confused too. We are using an external ADC with a 12-bit resolution. That is why we are getting 0x414, and read -12C. 

 I'm still not sure why we are gettin 0x414 though. In the megafunction manual, table 4 shows a sign bit. Could this be screwing us up in the 12-bit read?

[Altera_Forum]

Which external ADC are you using? 

 Jake

[Altera_Forum]

An Actel Fusion AFS600...we're using the AT temp monitor pins.

[Altera_Forum]

Can you confirm that you are connected as shown in figure 2-77 on page 2-96 of the AFS600 datasheet? 

 http://www.actel.com/documents/fusion_ds.pdf  Jake

[Altera_Forum]

Truly I don't know how to use the ADC on an Actel Fusion AFS600 to monitor the temperature on a Stratix IV. You really shouldn't be looking at Altera's megafunction user's guide to solve this because you are not using the megafunction. I think you need to be talking to the Actel FAE to solve your problem. 

 Jake

[Altera_Forum]

I've got TEMPDIODEP connected to AT and TEMPDIODEN connected to ATRTN.

[Altera_Forum]

Just an update on this in case anyone else has the same problem... we had a 2200pF cap placed between the terminals that didn't have enough time to charge up during the reading. We probably could've went with a smaller cap, but unpopulating it seemed to work very nicely.