red-ray wrote on 2020-06-14, 12:22:

Thank you, it transpires that the Analog ADM9240 is not enabled by the BIOS so needs to be enabled by the monitoring software wh […]
Show full quote

AlephZaku wrote on 2020-06-13, 22:52:

Here you go!

Thank you, it transpires that the Analog ADM9240 is not enabled by the BIOS so needs to be enabled by the monitoring software which is why such as Everest did not initially display any readings.

I suspect I have adjusted SIV to enable then report the ADM9240 so please try the attached SIV32X V5.49 Test-09 test SIV and post new save files.

Looking at [SMB Bus] no ADM1021 get's reported so either the MP440BX does not have one or it's broken. I will wait and see what happens on other MP440BX systems.

So it looks like the ADM9240 was giving me a reading of 29C when I took that. That also corresponds to what SpeedFan was picking up. This is the temperature that the CPU is running at then?

AlephZaku wrote on 2020-06-14, 14:00:

red-ray wrote on 2020-06-14, 12:22:

Looking at [SMB Bus] no ADM1021 get's reported so either the MP440BX does not have one or it's broken. I will wait and see what happens on other MP440BX systems.

So it looks like the ADM9240 was giving me a reading of 29C when I took that. That also corresponds to what SpeedFan was picking up. This is the temperature that the CPU is running at then?

Thank you for running the test and posting the results. Almost all the code worked as I intended, but the temperature was not on the initial screen to the right of Sensor Analog ADM9240 when it should be. This should be fixed in the attached SIV32X V5.49 Temp-09 test SIV, is it?

Given there is no ADM1021 maybe the ADM9240 reports the CPU temperature, but 29°C seems rather low to my thinking. To check I would 100% load the CPU and see if it increases. One way to 100% load the CPU is to use Menu->Machine->CPU Stress, as the panel hides the main window press [Main] to restore it.

red-ray wrote on 2020-06-14, 15:51:

AlephZaku wrote on 2020-06-14, 14:00:

red-ray wrote on 2020-06-14, 12:22:

Looking at [SMB Bus] no ADM1021 get's reported so either the MP440BX does not have one or it's broken. I will wait and see what happens on other MP440BX systems.

So it looks like the ADM9240 was giving me a reading of 29C when I took that. That also corresponds to what SpeedFan was picking up. This is the temperature that the CPU is running at then?

Thank you for running the test and posting the results. Almost all the code worked as I intended, but the temperature was not on the initial screen to the right of Sensor Analog ADM9240 when it should be. This should be fixed in the attached SIV32X V5.49 Temp-09 test SIV, is it?

Given there is no ADM1021 maybe the ADM9240 reports the CPU temperature, but 29°C seems rather low to my thinking. To check I would 100% load the CPU and see if it increases. One way to 100% load the CPU is to use Menu->Machine->CPU Stress, as the panel hides the main window press [Main] to restore it.

I want to get a better PSU in there before I run the stress test, so I might have to get back to you on that (which... might take awhile, unless I settle on getting something used from eBay. New PSUs don't seem to be in stock anywhere right now!)

AlephZaku wrote on 2020-06-14, 16:33:

I want to get a better PSU in there before I run the stress test, so I might have to get back to you on that (which... might take awhile, unless I settle on getting something used from eBay. New PSUs don't seem to be in stock anywhere right now!)

OK, but please can you check that the temperature is now on the initial screen. No fan speeds were reported, do you have any fans connected?

There are PSUs available here, but the range is rather limited. What size do you plan to get? My Dual P-III system power usage is as below.

Buy one of those point n shoot infra red thermometers to get an idea of temp

red-ray wrote on 2020-06-14, 21:07:

OK, but please can you check that the temperature is now on the initial screen. No fan speeds were reported, do you have any fan […]
Show full quote

AlephZaku wrote on 2020-06-14, 16:33:

I want to get a better PSU in there before I run the stress test, so I might have to get back to you on that (which... might take awhile, unless I settle on getting something used from eBay. New PSUs don't seem to be in stock anywhere right now!)

OK, but please can you check that the temperature is now on the initial screen. No fan speeds were reported, do you have any fans connected?

There are PSUs available here, but the range is rather limited. What size do you plan to get? My Dual P-III system power usage is as below.

- It's showing the new temperature on the main screen now.
- I did briefly run the stress test, but it didn't seem to increase the temperature any. So I'm assuming that's not the CPU temperature, but how long do I need to let that run before I'd notice? (I did open the case and poke the heatsink afterwards, and it felt cool. So either heat's not transferring to it or I didn't leave the thing on long enough).
- There's a case fan but I think it might be directly hooked up to the PSU and not a fan header.
- I went with a 500W PSU that has 20A on the 5V and, as far as I can tell, is not group regulated. I tried running the math and I think what I'm running now should cap out around 16A.

AlephZaku wrote on 2020-06-15, 00:24:

- There's a case fan but I think it might be directly hooked up to the PSU and not a fan header.
- I did briefly run the stress test, but it didn't seem to increase the temperature any. So I'm assuming that's not the CPU temperature, but how long do I need to let that run before I'd notice? (I did open the case and poke the heatsink afterwards, and it felt cool. So either heat's not transferring to it or I didn't leave the thing on long enough).

Thank you for checking and OK, I would have also expected a CPU fan.

Below you should see that after 10 minutes my CPUs were at least 14°C warmer and once the load stopped they cooled down, but it took quite a while and with hindsignt I should have set a 600 second Cool Time. With modern CPUs the temperature changes happen faster.

red-ray wrote on 2020-06-15, 05:57:

Thank you for checking and OK, I would have also expected a CPU fan. […]
Show full quote

AlephZaku wrote on 2020-06-15, 00:24:

- There's a case fan but I think it might be directly hooked up to the PSU and not a fan header.
- I did briefly run the stress test, but it didn't seem to increase the temperature any. So I'm assuming that's not the CPU temperature, but how long do I need to let that run before I'd notice? (I did open the case and poke the heatsink afterwards, and it felt cool. So either heat's not transferring to it or I didn't leave the thing on long enough).

Thank you for checking and OK, I would have also expected a CPU fan.

Below you should see that after 10 minutes my CPUs were at least 14°C warmer and once the load stopped they cooled down, but it took quite a while and with hindsignt I should have set a 600 second Cool Time. With modern CPUs the temperature changes happen faster.

So I ran it for around 8 minutes and the monitored temperature went from ~30C to ~42C. Touching the CPU cartridge and heatsink, the back of the cart was definitely warm and the heatsink was fairly hot (but not to the point where I couldn't touch it. I have no idea how warm the heat sink should be in relation to the CPU to which it's attached though).

I recently got a 700Mhz coppermine Pentium 3 in a slot 1 cartridge, with one of those heatsinks that's held in place with the little plastic pins that are supposed to be really hard to remove without breaking.

Eh, depends. For example, I pulled plastic pins from IBM PIII with Foxconn heatspreader without any problems.

AlephZaku wrote on 2020-06-16, 22:32:

So I ran it for around 8 minutes and the monitored temperature went from ~30°C to ~42°C.

Given it went up by 12°C I expect the reported temperature must be the CPU temperature so I adjusted the attached SIV32X V5.50 Temp-00 test SIV to report it as the CPU temperature.

This also means it should now get reported and plotted on the [CPU Burnin] plots, does it? Has the new PSU arrived?

red-ray wrote on 2020-06-16, 23:15:

AlephZaku wrote on 2020-06-16, 22:32:

So I ran it for around 8 minutes and the monitored temperature went from ~30°C to ~42°C.

Given it went up by 12°C I expect the reported temperature must be the CPU temperature so I adjusted the attached SIV32X V5.50 Temp-00 test SIV to report it as the CPU temperature.

This also means it should now get reported and plotted on the [CPU Burnin] plots, does it? Has the new PSU arrived?

Think it's showing up correctly on the CPU Burnin screen now. Thanks!

New PSU came, and I've got a case fan coming tomorrow so I'll check temps again once that's installed.

AlephZaku wrote on 2020-06-17, 02:02:

Think it's showing up correctly on the CPU Burnin screen now.

Thank you for the confirmation, I have now released SIV V5.50 Beta-00 so please use that.

SIV should report the case fan speed if you connect it to the Fan 1 (Chassis Fan) Header (J8L1) motherboard fan header with Fan 2 (Chassis Fan) Header (J4M1) being reported as the CPU fan, Note: both can supply a maximum of 300mA. I expect the speed of the Fan 3 (Active Heatsink Fan) Header (J4F1) can't be reported as pin-3 is connected to ground 🙁.

red-ray wrote on 2020-06-17, 18:45:

AlephZaku wrote on 2020-06-17, 02:02:

Think it's showing up correctly on the CPU Burnin screen now.

Thank you for the confirmation, I have now released SIV V5.50 Beta-00 so please use that.

SIV should report the case fan speed if you connect it to the Fan 1 (Chassis Fan) Header (J8L1) motherboard fan header with Fan 2 (Chassis Fan) Header (J4M1) being reported as the CPU fan, Note: both can supply a maximum of 300mA. I expect the speed of the Fan 3 (Active Heatsink Fan) Header (J4F1) can't be reported as pin-3 is connected to ground 🙁.

Fan's hooked up to header 3, unfortunately. Header 2 isn't soldered on but I could see if I could hook it up to header 1 -- it's mostly a question of whether the cord'll reach.

One thing I did want to point out is that, at startup the temperature was reading 26C. ~10 minutes of stress testing it got it up to 38C, and eventually it hit 40C after stopping the stress test, but after that it's just been hovering around between 39C and 42C, averaging about 40.5C. When I was running it *without* a fan for about 40 minutes of some fairly modest DOS gaming it was hitting 53C based on SpeedFan (which I believe is giving me the same value SIV is) so there's that. Not entirely sure what to make of these readings.

When I had my dual p3 500 system together I used speedfan to read the temps and fan speeds. It worked without issue.