That board is unable to run 3.3/3.45/3.6V ("low-voltage") 486 processors. That SiS 461 chipset also doesn't support write-back L1 cache, as far as I understand. Furthermore, that board doesn't support power saving techniques. This means all the features that require complicated CPU configuration are unsupported anyway. Every 486DX and 5V 486DX2 processor should work fine at the "486DX" jumprt setting.

There are adapter sockets with voltage regulators and the multiplier jumper. Such adapters make 3.45V processors usable in 5V-only-boards. With an adapter like that, your processor, the DX4 processors and the 5x86 processors (both AMDs clock-quadrupled 486-class processor as well as Cyrix's technically more advanced, but not necessarily faster processor) should also be usable, at the well-known DX/DX2 jumper settings. So you need either an adapter (that seem to be quite expensive nowadays if available at all), or you need to get a 5V CPU.

On most boards, the VL jumpers serve no actual function. They provide an informational signal to the VL slots to tell cards whether they should request an extra wait state on write and whether they get an extra clock for responding to bus cycles. Most cards ignore these jumpers anyway. "Proper" way of setting the VL jumpers is JP16,JP17 = both 1-2 for 20, 25 and 33 MHz bus clock, and both 2-3 for 40 and 50 MHz bus clock. OTOH, at 50MHz bus clock, you are not supposed to add plug-in cards to the VL anyway (just on-board local-bus components are in spec).

I have no idea what the "486 new version" settings is supposed to mean.

AT boards like the one you have don't support a power push button. The board doesn't control the power supply, but the board just always operates when the supply is turned on. AT power supplies are switched on/off by interrupting mains power. Just try the LED orientation. If you get it wrong, the LED will just not work - as long as you connect an LED to pins supposed to be connected to an LED, you can't fry it.

So what you're implying is that it's much easier and cheaper to just get an Intel 5V processor and that going with a lower voltage third party one (that can be more powerful, in line with a first-gen pentium) requires an adapter with an included VRM that can be more difficult to find and significantly more expensive?

I'm using an SFX PSU with an AT to ATX 20-pin adapter with a 20 to 24-pin bridge. There are two cables on one end of the AT adapter that are for wiring to a push switch. My baby AT case has one of the old switches that you push to turn on and the button remains depressed, and then you push again to release and power off.
In this image you can see the cables on the left and the switch on the right.

lavadrop wrote on 2022-03-01, 22:31:

So what you're implying is that it's much easier and cheaper to just get an Intel 5V processor and that going with a lower voltage third party one (that can be more powerful, in line with a first-gen pentium) requires an adapter with an included VRM that can be more difficult to find and significantly more expensive?

Exactly. Checking on a well-known auction site, I get a multitude of offers for 486DX2/66 processors below 30€ delivered to Germany (might be different for your location), but I can't find a single 486 VRM adapter.

lavadrop wrote on 2022-03-01, 22:31:

I'm using an SFX PSU with an AT to ATX 20-pin adapter with a 20 to 24-pin bridge. There are two cables on one end of the AT adapter that are for wiring to a push switch.

OK, I get it. Switch the switch into the on position. Put one wire of your AT-to-ATX adapter to any tab (doesn't matter which one) of the switch. Connecting the second wire to one of the three other spade tabs will make the supply turn on (and that's the right one to use), connecting the second wire to one of the other two remaining tabs will have no effect. Do this while a board and a hard drive is connected, so the supply will have sufficient load. It's highly unlikely that an ATX supply breaks due to missing load, but it might refuse to turn on.

Thank you for clearing that.
Do you happen to know if the Am80486 Rev. E can handle 5v? According to amd's archive.org page:

Product Description
The Enhanced Am486 microprocessor family of products integrates high-performance design techniques with improved cache structure and high on-chip integration. It is binary compatible with all Am486DX2 and Am486DX4 microprocessors. This 3-V core (with 5-V tolerant I/O), Enhanced SMM, and clock control provide lower power consumption for energy efficient "green" desktop designs and longer battery life for portable PCs. The Enhanced Am486DX4-75 and Enhanced Am486DX4-100 CPUs are available in a SQFP package. The Enhanced Am486DX4-100 and the Enhanced Am486DX4-120 CPUs are available in a PGA package.

lavadrop wrote on 2022-03-02, 01:51:

Do you happen to know if the Am80486 Rev. E can handle 5v? According to amd's archive.org page:

5V tolerant I/O only tells you that the rest of the board can talk to the CPU with 5v signals, while the CPU itself uses 3V.
It will work with 5V - but way outside its specs, much hotter and most likely die earlier.

Thank you FeFlash.

TI dx2-66, I assume?
I think those were just relabled Cyrix chips.

Didn't even know there were 3.x Volts ones in that speed until I recently discovered I own an Intel -66 that actually is rated at 3.3V...
It's right there in the fine print on the front, 3rd line: &E3V, where most read &E5V or &EW5V for posers.

The general rule was, anything in history up to and including the AMD or Cyrix DX2-80 is 5 V, kids don't use swear words and Ukraine is part of the Soviet Union.

From there onwards, the world went to shit.

So, below 100 MHz, 3.x V CPUs are rare and unusual.

Reasons to _not_ use a 3 V DX2-66:
a) possibly too collectable to be used (if in better condition than mine...)
b) too few boards around that have 3 V support and those are sought after to use with 100 MHz+ CPU, so basically a waste of a more capable board.

Have a sharp look at this board:
http://www.amoretro.de/wp-content/uploads/201 … motherboard.jpg
- at the bottom left and upper right corner of the socket, there are two little black boxes standing upright on the pcb.
Those are voltage regulators. Usually they're less phallic and flat on the board, easier to spot.
That there is one of the earlier boards that had them - non-ZIF socket, 30 PIN simms, VLB slots.
You'll find that board or others of that time under the same designation, but lacking the regulators. Whether for marketing reasons or to save 50 cents of semiconductors, I don't know.
Most boards for 3 V are later PCI ones and all of them do have 3 V, afaik.

My recommendation for your board:
Go outside to the garden, lift a stone or two, you're bound to find a DX2-66 with the usual 5V

And, rather don't bother with a DX2-80 since the 40 MHz system clock makes it much harder to get it to run stable.
It's not a bad board, I assume, but there was a SIS 471 chipset that must have brought some improvement.

Those frontside connectors, that manual you linked to is all you need. Click on "Jumpers", that includes the pinout of the respective headers. CN4 onwards. It's highschool grade physics stuff, you'll figure it out.

Have fun!

Hey Cuttoon! Sorry for the late reply, I think the email alerts are not turned on.

Yes, that is correct, TI 486, it's actually in almost pristine condition. I'm attaching a picture.

My motherboard does not have VRMs around the socket so I'm assuming there's no support for lower voltages.
I'm still puzzled by the setting:
JP10 : CPU Version Selector
1-2 : Normal, 2-3 : for 80486 New Version

I got an Intel one, DX2/33 like the first one. I think I figured out all the jumpers and I'll try tomorrow powering it up. I still have to figure out how to install/run BIOS overlay software to handle my CF card.

lavadrop wrote on 2022-03-12, 08:00:

I'm still puzzled by the setting:
JP10 : CPU Version Selector
1-2 : Normal, 2-3 : for 80486 New Version

I did not quite dive into the subject, but there is a later version of the intel DX2-66 with "write back" cache. It's rather rare and can be identified by the "&EW5V" marking in the third line. Look it up here, plenty of info bound to be found.
A usual DX2-66 should behave exactly like a DX-33, to the board. Intel and AMD DX2-66 were really common. Some actions don't make it past ten Euro.

That TI chip is quite the show piece!

A bit OT but I have a AMD Am486DE2-66V8TGC (DX2-66) that is 3.3v and does not require a heatsink. Runs very cool.

Yes, that’s because of the lower voltage. The original Intel 486 processor are 5v and generate more heat.

lavadrop wrote on 2022-03-19, 21:30:

Yes, that’s because of the lower voltage. The original Intel 486 processor are 5v and generate more heat.

True ! wow have you been hiding (member since 2016 and 19 posts) 🤣

Horun wrote on 2022-03-20, 00:12:

lavadrop wrote on 2022-03-19, 21:30:

Yes, that’s because of the lower voltage. The original Intel 486 processor are 5v and generate more heat.

True ! wow have you been hiding (member since 2016 and 19 posts) 🤣

Yes, I’ve been a lurker of this website for a long time, I remember back then it was more of a board about running LucasArts and Sierra compilations on Windows Vista.

lavadrop wrote on 2022-03-20, 21:54:

I remember back then it was more of a board about running LucasArts and Sierra compilations on Windows Vista.

Yeah, that was a friendly takeover/silent coup. It used to be all about running very old games on newer systems and "VOGOVOS" didn't quite have the same ring to it, so...

Oh, freddled gruntbuggly!

Cuttoon wrote on 2022-03-20, 22:04:

lavadrop wrote on 2022-03-20, 21:54:

I remember back then it was more of a board about running LucasArts and Sierra compilations on Windows Vista.

Yeah, that was a friendly takeover/silent coup. It used to be all about running very old games on newer systems and "VOGOVOS" didn't quite have the same ring to it, so...

Oh, freddled gruntbuggly!

Ah the days when you could buy an SC-55 and an MT-32 for 100 USD BOTH.