Would the TPM header on the MSI A520M-A PRO be compatible with this adapter?

Duffman wrote on 2023-09-08, 13:14:

Would the TPM header on the MSI A520M-A PRO be compatible with this adapter?

I'm afraid most 500 series motherboards have switched to use SPI TPMs so the LPC bus is no longer exposed anymore.

To find out whether the board's TPM slot is SPI or LPC, check the motherboard manual and look for TPM-related sections. If it's SPI TPM, the pinout should mention some SPI pins instead of LPC ones.

400 series chipsets appear to be the last to still use LPC TPMs, though they may be of different form factors depending on the vendor.

Ah I see, OK.

I see the MSI B450M-A PRO MAX uses LPC for TPM.
But from rasteri's testing AMD boards might not work? is that right?

Duffman wrote on 2023-09-08, 15:45:

But from rasteri's testing AMD boards might not work? is that right?

I'm sure there must be a way but we haven't figured it out yet.

If someone will build own version and its willing to build one more for money let me know.. Im not soldering type.. I still thing that it would be nice to add it to https://www.serdashop.com/ as besides DreamBlaster etc..

ruthan wrote on 2023-09-21, 00:30:

If someone will build own version and its willing to build one more for money let me know.. Im not soldering type.. I still thing that it would be nice to add it to https://www.serdashop.com/ as besides DreamBlaster etc..

I'm in the same boat.

Regarding the presence of DMA on the LPC bus. There are STILL motherboards shipping with LPT headers (Intel H610 based!) that appear to support ECP mode, which requires the use of a DMA channel.

I finished my design of LPC2ISA started years ago. I added switched inverter 5V -> -12V and LDO for -5V so it could be powered by a single Molex (3V3 is taken from LPC header). So I'll prepare a batch with some other PCBs for JLC...

Built a v0.2 version of the board and unfortunately it's still not working on my test boards. I'm observing the same behavior as the v0.1 board I built previously.

- The AMD test board is not detecting the bridge.
- Intel test boards refuse to boot unless I disconnect the TPM cable (whether dISAppointment is powered or not does not matter). I'm not testing the LDRQ there.

Also, in my case, -12V and -5V LEDs are not lit. I did make an inspection of the board after soldering and could not find any apparent issue with the related parts.
I put a debug card on the ISA slot. While not getting any POST code, the -12V LED on the card was lit, however.

I'm suspecting the 12V-ATX adapter I'm using to power the dISAppointment may not be working as intended. I'll be getting an ATX 24-pin splitter (so it gets power directly from my PSU) and test it again when I have time...

By the way, I wonder if there were ever any desktop motherboard model confirmed (at least per boardview) to have actually exposed the LDRQ pin on the TPM header.

NJRoadfan wrote on 2023-09-24, 03:54:

Regarding the presence of DMA on the LPC bus. There are STILL motherboards shipping with LPT headers (Intel H610 based!) that appear to support ECP mode, which requires the use of a DMA channel.

Interesting... but does ECP really work there? I wonder if there is still any actively used appliance that requires ECP.

LSS10999 wrote on 2023-10-21, 04:43:

I'm suspecting the 12V-ATX adapter I'm using to power the dISAppointment may not be working as intended. I'll be getting an ATX 24-pin splitter (so it gets power directly from my PSU) and test it again when I have time...

That wouldn't cause the system to not boot, though....

By the way, I wonder if there were ever any desktop motherboard model confirmed (at least per boardview) to have actually exposed the LDRQ pin on the TPM header.

Yeah I've found some server boards (eg supermicro) that have LDRQ. Also I have a PC Engines ALIX that does.

By the way I also had a disappointment that was failing to boot on an intel motherboard, I have no idea why but I built a second one and that worked perfectly. Possibly the ICs from aliexpress have a high failure rate or something.

rasteri wrote on 2023-10-21, 11:06:

That wouldn't cause the system to not boot, though....

This non-boot issue is only specific to Intel test boards and I'm not sure about the real cause. My AMD test board boots fine, just it won't detect the bridge no matter what.

Both v0.1 and v0.2 I built exhibited identical behaviors so I doubt it's merely soldering issue. Will be testing using an ATX splitter as that's the only thing I could suspect right now.

rasteri wrote on 2023-10-21, 11:06:

Yeah I've found some server boards (eg supermicro) that have LDRQ. Also I have a PC Engines ALIX that does.

Being mentioned in the manual alone won't be enough... If the boardview suggests the LDRQ is indeed wired to the TPM header then there's a good chance it'll work out-of-box.

rasteri wrote on 2023-10-21, 11:47:

By the way I also had a disappointment that was failing to boot on an intel motherboard, I have no idea why but I built a second one and that worked perfectly. Possibly the ICs from aliexpress have a high failure rate or something.

For this particular chip... no idea, but I hardly ever encounter faulty ICs, though it's true that some ICs can be more delicate than usual ones and require careful handling.

There is also W83626 which is pin-compatible with this F85226 bridge, but its documentation regarding configuration registers is rather vague (like CRxx instead of actual address) and I don't think it's register-compatible. Also, W83626 operates at 5V (with some pins requiring 3.3V) though I'm not sure if it can be driven using a single 3.3V source...

Yeah I've found some server boards (eg supermicro) that have LDRQ. Also I have a PC Engines ALIX that does.

I have the Supermicro X10SAE ready and waiting 😂😎

LSS10999 wrote on 2023-10-21, 12:20:

This non-boot issue is only specific to Intel test boards and I'm not sure about the real cause. My AMD test board boots fine, just it won't detect the bridge no matter what.

OK this is weird. I just tested my disappointment board and it refused to boot, just like yours. Then while I was fiddling with the board, it magically started booting.

I tried your LPCEXP program but it hung while trying to detect the ISA bridge. Then I touched the disappointment board again, on some of the ISA pins underneath the ISA slot - and it detected it!

So I tried to reboot the board, and it again failed. Then I rebooted while touching some of the ISA pins, and it booted!

I know this sounds crazy - but does touching the ISA pins while booting cause it to work? The ones on the AT-specific part (with the 16-bit lines). If it works on two boards then I must be missing a pullup resistor or something...

edit : specifically it seems to boot if I'm touching the MASTER# pin...

edit 2 : oh ****... master is supposed to have a pullup resistor on it and I forgot that. My guess is that's why it doesn't work....

OK, installing an 8.2K resistor between MASTER# and 5V makes it boot perfectly every time again.

I have no idea why it ever worked, but at least the fix is easy :

RayeR wrote on 2023-09-27, 22:48:

I finished my design of LPC2ISA started years ago. I added switched inverter 5V -> -12V and LDO for -5V so it could be powered by a single Molex (3V3 is taken from LPC header). So I'll prepare a batch with some other PCBs for JLC...

Does your design use the same bridge chip?

rasteri wrote on 2023-10-23, 20:54:

OK, installing an 8.2K resistor between MASTER# and 5V makes it boot perfectly every time again.

I have no idea why it ever worked, but at least the fix is easy :

Excellent. Glad to know that the cause of the issue has been found. 😀

Will be testing the fix when I have time.

Yes, I already have bought Fintek F85226FG from Aliexpress years ago, waiting to be soldered.
I checked my schematics and I have a pull-up 8k2 on master signal. I drawn it by reference sch. from the DS.

RayeR wrote on 2023-10-24, 02:23:

I checked my schematics and I have a pull-up 8k2 on master signal. I drawn it by reference sch. from the DS.

🤣, ok rub it in 😜

Unfortunately this doesn't appear to make it work any better on my AMD board.

rasteri wrote on 2023-10-24, 09:33:

RayeR wrote on 2023-10-24, 02:23:

I checked my schematics and I have a pull-up 8k2 on master signal. I drawn it by reference sch. from the DS.

🤣, ok rub it in 😜

Unfortunately this doesn't appear to make it work any better on my AMD board.

Did you test some other I/O ports back then? What I/O ports the LPC controller does not abort (like 2E/2F) may give out a clue about the chipset's behavior... whether the functionality of the LPC controller is still intact or have been crippled but not reflected in the documentation. On the other hand, I suspect this "abort" behavior might be the reason why I was able to boot even without the resistor on master signal.

The values I put in LPCEXP was mainly for AWE32/64. Maybe consider changing the values to somewhere else... or try an address covered by the ones I set by default (like 620H). It still remains a question to me why AM4 motherboards would default to a single wide I/O range starting from 230h when the register 44H already provided switches for almost every known ISA appliance within that range.

I'll set up an AM3+ testbed to do a comparison when I have the chance...

LSS10999 wrote on 2023-10-24, 12:23:

Did you test some other I/O ports back then? What I/O ports the LPC controller does not abort (like 2E/2F) may give out a clue about the chipset's behavior... whether the functionality of the LPC controller is still intact or have been crippled but not reflected in the documentation. On the other hand, I suspect this "abort" behavior might be the reason why I was able to boot even without the resistor on master signal.

I wrote code to just spam reads to every port below 400, after running your program.

I don't know what flavour of LPC sigrok's decoder expects but it sure ain't the LPC that any of my motherboards speak. So these are transcribed by hand, and there may be mistakes.

If the port doesn't appear in this list then it doesnt get forwarded to LPC at all.

12E-2F - not aborted -- SUPER IO
24E-4F - aborted -- ISA BRIDGE
3
460 - not aborted -- KEYBOARD
562 - aborted
664 - not aborted -- KEYBOARD
766 - aborted
87E - aborted
97F - aborted
1080 - aborted
11EE - aborted
12EF - aborted
13
14200-20F - aborted
15
16220-28F - aborted
17290-297 - not aborted -- PART OF AudioPortEnable3?
18298--2ff - aborted
19300-308 - not aborted -- MIDIPortEnable0 PLUS A FEW MORE PORTS?
20309-3AF- - aborted
21
223BC-3BF - aborted
23
243E0-3F7 - aborted
253f8-3ff - not aborted -- SerialPortEnable0

I might try changing the address of the serial port in the BIOS to see if that causes a different range of ports to not be aborted.