Out of curiosity... How did K6-2 350 run on this board? Which voltage was it running on?
Just another beat to the dead horse... Did you try reseating the BIOS chip?

As it's been sitting around for a very long time I would try reaseating the cpu and ram about 10 times just to shake off any oxidisation
in the sockets themselves.
The fact that You get beeps at all is a good sign of life.

lazibayer wrote:

Out of curiosity... How did K6-2 350 run on this board? Which voltage was it running on?
Just another beat to the dead horse... Did you try reseating the BIOS chip?

Well, according to the user guide I found for a similar model, the manual CPU selection dip switches were set for "AM K6-233/266" which is supposedly dual voltage selection. In another manual I found for a smaller version of the board says it is I/O volatge 3.3 V and core voltage 2.2V so I assume this is a Super Socket 7 mainboard and It was set for 2.2vols. The voltage can be selected in BIOS too according to that same manual.

I will try to do that with the BIOS chip.

Imperious wrote:

As it's been sitting around for a very long time I would try reaseating the cpu and ram about 10 times just to shake off any oxidisation
in the sockets themselves.

Yeah, I'm thinking on washing it, but I dont know if the SDRAM slots can be dirty or not making good contact with the memory modules, maybe the pins are bended inwards.

Have any of you ever seen this problem?

Welcome to Vogons! (or at least to active use hehe 😁)

Yours looks like a lovely board!
In computing, 4 seconds is something of a magic number (holding down the power button for 4 seconds will usually result in a hard shutdown with ATX boards, I like this bit of knowledge as one of many tools to troubleshoot problems).

My suggestions would be to get the board out of any case and put it on a work bench (motherboard box (no ESD bags near it!), PSU and PCI graphics card, speaker hooked up and laying besides the motherboard box, monitor attached, single DIMM or double SIMM in your case, no HDD/ODD/FDD).
Doublecheck the jumpers, test using a Pentium or Pentium MMX andset the jumpers for a Pentium or Pentium MMX, heatsink the CPU, then test.

Elanus wrote:

Well, according to the user guide I found for a similar model, the manual CPU selection dip switches were set for "AM K6-233/266 […]
Show full quote

lazibayer wrote:

Out of curiosity... How did K6-2 350 run on this board? Which voltage was it running on?
Just another beat to the dead horse... Did you try reseating the BIOS chip?

Well, according to the user guide I found for a similar model, the manual CPU selection dip switches were set for "AM K6-233/266" which is supposedly dual voltage selection. In another manual I found for a smaller version of the board says it is I/O volatge 3.3 V and core voltage 2.2V so I assume this is a Super Socket 7 mainboard and It was set for 2.2vols. The voltage can be selected in BIOS too according to that same manual.

I will try to do that with the BIOS chip.

Imperious wrote:

As it's been sitting around for a very long time I would try reaseating the cpu and ram about 10 times just to shake off any oxidisation
in the sockets themselves.

Yeah, I'm thinking on washing it, but I dont know if the SDRAM slots can be dirty or not making good contact with the memory modules, maybe the pins are bended inwards.

Have any of you ever seen this problem?

It is NOT a Super 7 board, no support for 100MHz FSB.

SiS5598 is an alright chipset. Or at least I have a weakspot for it (do not underestimate the SiS 5598! 😜). It has some interesting features that Intel TX is missing and to top it all off, your board is a s7-compatible board in ATX. I'd not be careless with that board, even if you think it's junk.
Like someone already mentioned, try reseating everything, doublecheck every jumper settings, clean the contacts of your memory modules as dirty contacts (ime) often seem like they are clean, but after cleaning you wouldn't believe the amount of dirt remaining on my paper towels after cleaning the contacts (like literally totally black stains on the paper towels, even though it seemed perfectly clean before cleaning the memory modules).

Tetrium wrote:

Welcome to Vogons! (or at least to active use hehe 😁)

Yours looks like a lovely board!
In computing, 4 seconds is something of a magic number (holding down the power button for 4 seconds will usually result in a hard shutdown with ATX boards, I like this bit of knowledge as one of many tools to troubleshoot problems).

Yes, It seems that there was something shorting the motherboard, as soon as I got it out of the box it stopped shutting itself off.

Tetrium wrote:

My suggestions would be to get the board out of any case and put it on a work bench (motherboard box (no ESD bags near it!), PSU and PCI graphics card, speaker hooked up and laying besides the motherboard box, monitor attached, single DIMM or double SIMM in your case, no HDD/ODD/FDD).
Doublecheck the jumpers, test using a Pentium or Pentium MMX andset the jumpers for a Pentium or Pentium MMX, heatsink the CPU, then test.

Well I have been doing some tests,

I dont have a PCI graphics card, so I'm stuck with onboard video. I see something interesting. There is a setting to disable onboard video, when it is set to "Enable" I get no signal on the monitor, but if I set to "Disable" the monitor gets a signal that activates it but shows only a black image. Something else, I connected the integrated soundcard output to a pair of speakers and they produce the same beeps, they are internally connected to the PC Speaker connector.

Removed BIOS Chip, checked contacts, cleaned them, and reseated.

1. Put the board on a wooden base connecting only PSU, 1X64Mb DIMM, processor (AMD K6-2 350, selected for "AM K6-233", FSB 75MHz, and the 1.5/3.5 multiplier settings). I get a single continuous beep. Change the memory for EDO RAM on BANK 0, same result.

2. Tested a Pentium 120, selecting "P54C", FSB to 60MHz and the multiplier to 2.0X and then I got long beeps (about 2 seconds duration each), not a single continuous one. Put EDO RAM, same result.

3. Tested AMD K6 233, same settings, again single continuous beep.

The only thing I can think of is that this mainboard is not detecting any RAM I insert on it. 😢

Elanus wrote:

Yes, It seems that there was something shorting the motherboard, as soon as I got it out of the box it stopped shutting itself off.

I've made it a habit to (whenever testing stuff that has problems or is untested) run everything on a workbench. It prevents stuff like this from wasting time and possibly even working parts 😜

Elanus wrote:

Well I have been doing some tests, […]
Show full quote

Well I have been doing some tests,

I dont have a PCI graphics card, so I'm stuck with onboard video. I see something interesting. There is a setting to disable onboard video, when it is set to "Enable" I get no signal on the monitor, but if I set to "Disable" the monitor gets a signal that activates it but shows only a black image. Something else, I connected the integrated soundcard output to a pair of speakers and they produce the same beeps, they are internally connected to the PC Speaker connector.

Removed BIOS Chip, checked contacts, cleaned them, and reseated.

1. Put the board on a wooden base connecting only PSU, 1X64Mb DIMM, processor (AMD K6-2 350, selected for "AM K6-233", FSB 75MHz, and the 1.5/3.5 multiplier settings). I get a single continuous beep. Change the memory for EDO RAM on BANK 0, same result.

2. Tested a Pentium 120, selecting "P54C", FSB to 60MHz and the multiplier to 2.0X and then I got long beeps (about 2 seconds duration each), not a single continuous one. Put EDO RAM, same result.

3. Tested AMD K6 233, same settings, again single continuous beep.

The only thing I can think of is that this mainboard is not detecting any RAM I insert on it. 😢

I can really advice you to get at least a couple PCI cards. I ran a 5598 board at one time and decided to use a dedicated PCI card, if only because that way I have more control over the way it's put together (and makes troubleshooting easier).

Which CPU voltages are you setting? Setting the K6-2/350 at the same voltage as a K6/233 is supposed to run at, will probably kill the K6-2/350 (k6/233 ran at 3.2v or something, from top of my head. K6-2/350 ran at 2.2v or so, much lower as the K6-2 (and K6/266 and up) had a die shrink).

Of the CPUs are all in working condition, swapping them out may not help you a lot (but might aid in accidentally making a mess when switching CPU + jumpers around). I prefer to stick to some fodder CPU I can spare, but know is in working order, and is pretty much universal (like a Pentium MMX 166 or a Pentium classic 133 (but yours is good too, but the 133 should be much more common)).

It might be the RAM, sometimes some RAM sticks will not work with a particular board. Do you have any single sided 8-chips 64MB SDRAM or 128MB 16-ships SDRAM laying around? Those are from about the same time period and I think such ones should work (especially if these are PC-100).

And you mentioned you already removed the BIOS chip, but I was gonna mention trying a BIOS reset 😜

Tetrium wrote:

I can really advice you to get at least a couple PCI cards. I ran a 5598 board at one time and decided to use a dedicated PCI card, if only because that way I have more control over the way it's put together (and makes troubleshooting easier).

Yes, I have been searching for one, however, in my country these are almost impossible to find. I'm looking on ebay to see which one can I import from USA.

Tetrium wrote:

Which CPU voltages are you setting? Setting the K6-2/350 at the same voltage as a K6/233 is supposed to run at, will probably kill the K6-2/350 (k6/233 ran at 3.2v or something, from top of my head. K6-2/350 ran at 2.2v or so, much lower as the K6-2 (and K6/266 and up) had a die shrink).

Well, I'm not 100% sure, the settings I use are described on this manual: http://stason.org/TULARC/pc/motherboards/D/DT … S-VER-1-10.html
That seems to be extracted from the manual, which I'm afraid, is not online anymore.

Digging in forums I found other names for, similar or the same mainboard from other companies (Gemlight and Vextrex) that apparently were in the same conglomerate as DTK. Under one of those names I found this manual http://stason.org/TULARC/pc/motherboards/V/VE … X-VER-1-20.html

The closest complete manual I could find is for the smaller version on the board and I got it from The Internet Archive. The manual is here https://drive.google.com/file/d/0B9czMn73hJgo … iew?usp=sharing On that manual the voltages set for "AMD K6 233/266" is 2.2V.

Tetrium wrote:

Of the CPUs are all in working condition, swapping them out may not help you a lot (but might aid in accidentally making a mess when switching CPU + jumpers around). I prefer to stick to some fodder CPU I can spare, but know is in working order, and is pretty much universal (like a Pentium MMX 166 or a Pentium classic 133 (but yours is good too, but the 133 should be much more common)).

I don't have access to other Socket 7 mainboard to test them, at least non functional since I have a completely dead M571 V3.2 with one bulging cap, It was my main system with the K6-233 eons ago. All three CPU's were working the last time they saw life.

Tetrium wrote:

It might be the RAM, sometimes some RAM sticks will not work with a particular board. Do you have any single sided 8-chips 64MB SDRAM or 128MB 16-ships SDRAM laying around? Those are from about the same time period and I think such ones should work (especially if these are PC-100).

The memory I used in my tests is a single sided 64Mb module with 8 chips it is labelled PC-100, and it was tested successfully on a socket370 system I have, CPU-Z reported PC-100 and PC-66 compatibility on that module. I also tested other single sided 64Mb module and a confirmed working 128 MB double sided PC-100 module.

Something else I'm thinking is on hotswapping the BIOS chip and flash the latest BIOS I could find using the Socket370 system, maybe It got corrupted. However I want to get a PCI, graphics card because If I ever get it it to work, I'm not sure If I want to use the 5598 integrated graphics, I remember it was very limited. Maybe a Matrox Millennium, they are cheap on Ebay

Have you done a CMOS reset? Removing the Flash ROM doesn't reset anything, the battery backed NVRAM isn't stored there.

I wasn't aware the SiS5598 supported SDRAM, but I do see it mentioned in the link you posted for a similar board.
Even so, this chipset is early for SDRAM so I wonder about how picky it might be. The wording in that link does imply that the density of your modules is probably acceptable (it allows max of 128MB per module, which implies a 16-chip 128MB and 8-chip 64MB would probably be in valid configurations). But the behavior you're getting seems like a possible RAM issue.
If you can find any, I'd try EDO memory on this board. It should have good compatibility with EDO modules.

There is probably a 3.3v/5v jumper for the RAM. Make sure it's set correctly. Your PC100 SDRAM should be 3.3v, but most EDO is 5v, so it must have a provision for selecting between those. If it's been set to 5v, then unfortunately the health of your SDRAM modules will be questionable.

I'd stick with the Pentium 120 for troubleshooting. It's the easiest to get a system running with IMO. The configuration is potentially a little simpler and it's not as demanding on the board.
If you hear beeps, that at least means the CPU is doing something. The BIOS chip is just a Flash ROM with code that is executed by the CPU.

If you have a multimeter, you could check some voltages and see if anything is out of whack. That includes the CMOS battery, actually, which you could try replacing if there's any doubt about it.

shamino wrote:

Have you done a CMOS reset? Removing the Flash ROM doesn't reset anything, the battery backed NVRAM isn't stored there. […]
Show full quote

Have you done a CMOS reset? Removing the Flash ROM doesn't reset anything, the battery backed NVRAM isn't stored there.

I wasn't aware the SiS5598 supported SDRAM, but I do see it mentioned in the link you posted for a similar board.
Even so, this chipset is early for SDRAM so I wonder about how picky it might be. The wording in that link does imply that the density of your modules is probably acceptable (it allows max of 128MB per module, which implies a 16-chip 128MB and 8-chip 64MB would probably be in valid configurations). But the behavior you're getting seems like a possible RAM issue.
If you can find any, I'd try EDO memory on this board. It should have good compatibility with EDO modules.

Yes I did reset the BIOS using the jumper and removing the battery, it was my first step. Two of those modules (64Mb and 128MB PC-100) came with the board, and reportedly were working on the system years ago. Those were confirmed working on my Socket 370 system. I used 2 EDO RAM modules, I tested them months ago in a M717 Slot 1 mainboard (wich started dying recently). Here are some pics of them. When I used them in Bank0 the result was the same.

I think they came with the P120 from a Packard Bell Synera socket 7 system. They have 8 chips on each side.

shamino wrote:

There is probably a 3.3v/5v jumper for the RAM. Make sure it's set correctly. Your PC100 SDRAM should be 3.3v, but most EDO is 5v, so it must have a provision for selecting between those. If it's been set to 5v, then unfortunately the health of your SDRAM modules will be questionable.

The board does not have this jumper, the memory slots have "3.3V" engraved on them, I guess it only supports 3.3V modules. I don´t know the voltage used by the EDO sockets and modules I have.

shamino wrote:

I'd stick with the Pentium 120 for troubleshooting. It's the easiest to get a system running with IMO. The configuration is potentially a little simpler and it's not as demanding on the board.
If you hear beeps, that at least means the CPU is doing something. The BIOS chip is just a Flash ROM with code that is executed by the CPU.

Yeas that's why I thing there is something wrong with the ability to detect RAM, maybe due to some damaged component.

shamino wrote:

If you have a multimeter, you could check some voltages and see if anything is out of whack. That includes the CMOS battery, actually, which you could try replacing if there's any doubt about it.

Yes the battery is new, the multimeter reads 3.0V on it, the old one was dead. I tested the voltages produced on the fan connectors. The readings were 12V in J17 and 5V on JP11, I don't know were else can I measure voltages, the ones I got seem normal to me.

You should be able to get that board right but please do take care of it as these boards are only getting more expensive and sure it is not SS7 but still very nice to have. ATX S7/SS7 do save a good amount of money by not being AT where it would have needed extra adapters, headers, and extra work drilling cases or finding increasingly hard to find AT cases. On board DOS compatible audio is always a huge bonus for people on a smaller budget not needing to cough up for a ISA sound card right away.

nforce4max wrote:

You should be able to get that board right but please do take care of it as these boards are only getting more expensive and sure it is not SS7 but still very nice to have. ATX S7/SS7 do save a good amount of money by not being AT where it would have needed extra adapters, headers, and extra work drilling cases or finding increasingly hard to find AT cases. On board DOS compatible audio is always a huge bonus for people on a smaller budget not needing to cough up for a ISA sound card right away.

Yes!, I fell in love with the system as soon as I opened the case. It came with a nice case too, a 4Gb Samsung hard disk loaded with Windows NT, and Office 2000, a PCI USB card and an old Analog Devices ISA modem sound card combo. I really wanted to Use my AWE 64 on this but as soon as I saw the ESS sound chip I became obsessed with getting it to work. I'm certainly taking all precautions, I want it to work an last.

I looked up the chips on the EDO SIMM that you showed. According to the datasheet they run on 5V.
http://www.ic72.com/pdf_file/h/11446.pdf

Perhaps your motherboard always sends 5v to the SIMM sockets and always sends 3.3v to the DIMMs.

When you look at the SDRAM DIMM sockets, they have a key that is offset far to one end of the socket.
If you count starting from that end, the 6th pin inward will have the SDRAM supply voltage, which should be 3.3v. It's the same on both sides (at the keyed end).
Without even needing to power the board, you may find with a resistance test that there is continuity from that pin to the 3.3v supply at the ATX connector. If so, then you know it's hardwired to 3.3v.
If necessary, and if you're careful, it's possible to stick the positive multimeter probe against that pin and measure the DIMM voltage while it's running. If you do that then I recommend wrapping the probe with tape so that only the tip is exposed.

On the EDO SIMMs, the 5v is at one of the corner pins of each chip. I've attached an edited version of your picture showing which pins have +5v on them. I'm not sure what pin in the socket this corresponds to. Without powering the board, but with a SIMM installed, you might find continuity from the +5v pin to the +5v supply pins at the ATX connector.

You can measure the power supply's 5v and 12v rails easily using the molex drive connectors. You mentioned already measuring those rails using some points on the board. The tolerance is supposed to be +/- 5%, but in practice you may have problems if any rail is in the lower end of the allowed range.

You can measure the PSU's 3.3v rail by backprobing the ATX connector. The orange wire(s) are 3.3v. You can stick the positive probe in the back of the connector and make contact with it to get a reading.
The same can be done with +5Vsb (purple I believe).

Very useful information, I'm going to check this out as soon as I have some spare time. Thanks.