Don´t worry about the capacitors at the moment.
Do these settings match your BIOS chip?
If there´s a sticker on the BIOS chip you must remove it to find the typenumber / manuf.
Yes they do.
I assume you have the VRM module installed and set to 3.3V - or is it onboard @ version 1.1?
Have you tried with all cache and TAG-chips removed?
A picture would still be welcome / useful!
The VRM is onboard.
Somewhere I've read, that this revision only allows 5.0v or 3.45v.
At least it says so on the silkscreen. I have set JP31 to pin 3-4.
On stason.org, it says pin 1-2 of jp31 should set the voltage to 3.3v, but on the silkscreen it says it's 5.0v.
I'm confused.
Tomorrow, I'm going to take some high-res pictures of this board.
I have the very same board, but rev 2.1 and BIOS on that thing also gave me issues.
Is your CPU detected correctly? Does the machine boot past POST?
AMD 486dx4-100NV8T is a 3V processor, not sure if board gives it right voltage without VR.
Manual does not say if it supports AMD's 486DX4.
What I would do at this point is to try another 486 in it, since I experienced same problem on the other board when I tried unsupported CPU - halts on POST.
Edit: is it that board: https://theretroweb.com/motherboards/s/ecs-um … 10p-aio-rev-1-x
Edit2: what it comes to motherboards manuals, I'd give more trust to theretroweb.org than stason.org
Agreed! Maybe try some good old 5V 486DX / DX2 first.
The onboard VRM of Rev. 1.1 only does 5V or 3,45V while the VRM add on card for Rev. 2.x (VR-102) can provide 3.3, 3.45, 3.6 and 4V.
The difference 3.45 / 3.46 is irrelevant and 3.3V CPU *should* work with 3.45V Vcore.
(There are some mainboards that monitor the core voltage very precisely and won´t POST when they consider the difference too big.)
smoke86 wrote on 2022-09-15, 18:57:I have the very same board, but rev 2.1 and BIOS on that thing also gave me issues. Is your CPU detected correctly? Does the mac […]
I have the very same board, but rev 2.1 and BIOS on that thing also gave me issues.
Is your CPU detected correctly? Does the machine boot past POST?
AMD 486dx4-100NV8T is a 3V processor, not sure if board gives it right voltage without VR.
Manual does not say if it supports AMD's 486DX4.What I would do at this point is to try another 486 in it, since I experienced same problem on the other board when I tried unsupported CPU - halts on POST.
Edit: is it that board: https://theretroweb.com/motherboards/s/ecs-um … 10p-aio-rev-1-x
Edit2: what it comes to motherboards manuals, I'd give more trust to theretroweb.org than stason.org
It does not boot past POST. It prompts me to run the SETUP menu, which I can't do for obvious reasons.
The CPU is reported as an 486DX4, running at 100 mhz.
Are you sure this is a 3V processor? The datasheet claims it has a V core of 3.3V.
I have an A80486DX2-66 in my stash. This will hopefully do the trick.
Well, yes, you are correct. I generally refer to 486s roughly as 3 or 5 volts just to make a distinction.
Check on this other Am486 and let us know.
From what you say it occurs to me that keyboard obviously gets power and initalizes, but it fails to input any data.
It may be a broken /cut data trace on a motherboard or malfunctioning keyboard connector (try doing some measurments with multimeter to check for contiunuity, both connector and traces).
What keyboard are you using?
In the past I ran into problems with some keyboards (especially more moder ones, plugged through adapter) that the onboard controller didn't like.
I guess you did check the keyboard in another machine, but did you check another keyboard with that ECS board?
smoke86 wrote on 2022-09-15, 20:18:Well, yes, you are correct. I generally refer to 486s roughly as 3 or 5 volts just to make a distinction. Check on this other Am […]
Well, yes, you are correct. I generally refer to 486s roughly as 3 or 5 volts just to make a distinction.
Check on this other Am486 and let us know.
From what you say it occurs to me that keyboard obviously gets power and initalizes, but it fails to input any data.
It may be a broken /cut data trace on a motherboard or malfunctioning keyboard connector (try doing some measurments with multimeter to check for contiunuity, both connector and traces).
What keyboard are you using?
In the past I ran into problems with some keyboards (especially more moder ones, plugged through adapter) that the onboard controller didn't like.
I guess you did check the keyboard in another machine, but did you check another keyboard with that ECS board?
I tested 2 other keyboards with this board. Neither of them worked.
The traces look fine.
Perhaps you're right and the keyboard connector itself is broken. Fortunately, replacing it shouldn't be that big of a deal.
I'll let you know once I've tested the other CPU (it's an Intel btw).
Do you have a POST code card? It would be interesting to see where it is hanging.
I also ran a late AMD 486 CPU detection issue on an ECS PhoenixBIOS board, but it was the UMC 498 VLB AIO board, not this PCI one. I was using an Am5x86 and got different results depending on whether I ran it at 3x or 4x. From what I remember the BIOS stashes information about the type of CPU in some of the high CMOS registers, and they fail to get updated if it was a 5x86 at 4x. So you may end up with some very hard to debug situations where the board suddenly works if you put in a 486DX2-66 and then switch to a 5x86-133, but would stop working if the CMOS battery is left disconnected overnight. It would be very hard for someone to figure that out without looking at the BIOS image itself. Resetting CMOS defaults through setup (if you were able to get to it) wasn't enough to reproduce it either; you have to punch in code into DEBUG to set every register to zero or leave the battery disconnected.
I've got some fantastic news.
The DX2 fixed my keyboard problem.
Good to know that the problem was only caused by the wrong jumper configurations and not an actual defect.
Sorry, if I wasted your time.
I feel so stupid for not figuring it out on my own.
Still, this begs some more questions.
What are the correct jumper configurations for the DX4?
Should I update the BIOS?
Many thanks to everyone who has helped me. I really appreciate it.
You're the best.
I don't think your problem was caused by wrong jumper setting.
This motherboard does not support DX4s made by AMD, which means BIOS does not recognize its microcodes properly, causing system to halt.
As it turned out that your system is working just fine, I'd give it a chance and update bios, although - just to be safe - I recommend sourcing another compatible EEPROM (not EPROM!) chip and flash it using 'hot swap' method (as it is already socketed).
Edit: Intel, AMD and Cyrix DX4s are not 100% compatible with each other, they run at different settings. So there is not such thing as general 'jumper setup for DX4".
Great idea. I'll give it a go on another day.
For now, I'm going to stick with my DX2.
Or look for Intel's DX4, as it's officially supported.
Although I said that I'd stick with the DX2, I still find it funny that the board was bundled with the DX4 when I purchased it.
Also, when I got this board, the jumpers were set differently.
Sadly, I didn't memorize the "old" jumper settings.
On the bottom side, there's even a sticker with "Trend 486DX4-100" printed on it.
This is very strange.
You're right.
I should probably get an Intel 486DX4.
However, for now I put this on my To-Do list.
Earlier Cyrix and AMD 486 CPUs unfortunately can have different pinouts than "standard" (in quotes since they couldn't necessarily read Intel's mind when they were designed). That is why 486 motherboards have all those jumpers. Your NV8T CPU is an "original" Am486DX4 in contrast to an "enhanced" one, and lacks half the internal cache and write-back internal cache support (and SMM support) as compared to the enhanced one. The final 486 CPU pinout is called P24D and was introduced by the i486DX2-WB and that is what all the final-generation processors use (the only difference being the CLKMUL of 2x/4x or 3x).
If you really want to you could figure out the jumper settings with a multimeter, or by carefully studying the jumper chart and someone with the background of all the different pinouts can eventually reason what they are. Or, if you get an Intel DX4-WB, or an "enhanced" AMD one, you can use the "P24D" pinout.
Did you see this? Have to click through cert warning.
https://www.dosreloaded.de/BIOS/Mainboards/FI … /UM8810PAIO.pdf
jakethompson1 wrote on 2022-09-17, 04:14:Did you see this? Have to click through cert warning.
https://www.dosreloaded.de/BIOS/Mainboards/FI … /UM8810PAIO.pdf
This is manual for rev 3.0, while topic is about rev 1.1, they are a bit different.
Also the problem is not about jumper settings, but about lack of BIOS support for that CPU.
Here´s one jumpered for Intel DX4 (detailed pictures):
https://www.ebay.ca/itm/185550652355?
Supposing this setup was working you might try it.
