TF536 Rev2 Gets Hot

[PaulJ_2.0]

I expected 7 to be the best but the phase shouldnt cause a temp difference.. that *IS* weird.

Ok on reflection it might not have got too hot and crashed after 15 mins, I just assumed that was the case. It was probably cooler than the original firmware.

7 is certainly fine here it's still going strong inside an enclosed 500 with no cooling/heatsinks.

[terriblefire]

7 is certainly fine here it's still going strong inside an enclosed 500 with no cooling/heatsinks.

Ok thats the firmware that crashed for AlenPPC. If you have setup that reproduces heat easily i can send you a series of firmwares to narrow down the pins causing the heat.

[PaulJ_2.0]

I can send you a series of firmwares to narrow down the pins causing the heat.

Sure thing, I have plenty of time this evening to go through them.

[go0se]

I haven't noticed any heat issues with the 536Rev2 at all.

Here's a quick test with the 536R2 running the previous/standard tf536r2_main_top_test.jed firmware against a TF330 running the non-NVRAM-fix firmware.

Both CPLDs have a thermocouple placed in the centre and were allowed to idle on workbench for 10mins before running "Hard Drivin" demo loop from WHDLoad for a further half hour.

The ambient temperature in the room is between 20 and 25 degrees. Both cards show very similar results ~40C, allowing for variations in my cheap thermocouples.

I appreciate I'm comparing apples and oranges here as one CPLD is a 144 and the other is a 288 but both seem nominal.

I haven't noticed any CPLDs on any of the 536 cards I've built getting overly warm to the touch. Other than that you would know that they were on.

Now an hour later the 330 is consistently a couple of degrees hotter than the 536, but both are running fine.

Cards.

DSC_0115.resized.JPG

Ambient

DSC_0134.resized.JPG

CD32

DSC_0140.resized.JPG

A500

DSC_0143.resized.JPG

[PaulJ_2.0]

Those temps seem about the same as I am getting in the open air but in an enclosed case, they get quite a bit warmer. The Rev1 board is barely warming over ambient for me.

[go0se]

OK, I'm not seeing any significant difference between the R1 (tf536r1_main_top_FULL_PHASE_7.jed) and R2 (tf536r2_main_top_test.jed) back to back. Both are within a couple of degrees on the Hard Drivin demo and on a fast ram test using AmigaTestKit the R1 is consistently a few degrees hotter. I've swapped the thermocouples round and repeated the tests to ensure that the readings were fair.

I appreciate you said that the tests were performed in the case but putting these motherboards in cases would be a pain for me just at the moment. I can't see how being in the case would cause the R1 to run just above ambient but the R2 to run significantly hotter, The A500s revolutionary plastic fin vent technology wasn't that efficient :)

What software are you using to stress the cards?

EDIT: I will retest in cases tomorrow if I get the chance.

R1

DSC_0119.resized.JPG

R2

DSC_0115.resized.JPG

[PaulJ_2.0]

Ok well this is strange then, as I was one of the ones that were struggling with the Rev1 and had to supercool the CPLD with freeze spray to make it run stable.
After that was fixed the CPLD now runs at around 30 at its hottest.

The rev2 will run 35+ when idle and upto 70 when pushed.
I have gotten around this by using some 5V low profile heatsinks and fans, but I am more curious as to why they are running hotter, it's odd that your Rev1 is that hot, maybe I just have an odd case in my my rev1 running so cool.

I am using frontier on loop to stress.

[alenppc]

I haven’t noticed any heat issues on my builds of the rev2 536 at all...

[PaulJ_2.0]

Maybe I have a batch of CPLD that run hot. I will order some more and see how they go.
I should also point out that if my Rev1 wasn't running so cool, I probably wouldn't have thought much about it.

[dalek]

My rev 1 runs hot too - adding the pullups didn't seem to make a difference. I just put a heatsink on it as there weren't that many people complaining about heat, but it seems like a bit of a band-aid. Unfortunately I have nothing other than my finger to measure heat with. The CPLD was the bigger variety from digikey so no chance of fake.