I had one of these boards that did something very similar. I swapped around multiple different IDE cables, drives etc with different configurations.
Eventually I landed on the conclusion that the primary IDE channel would ONLY support one drive at a time.
It didn't seem to matter if it were master or slave, only one would work at a time. I didn't have the same problem with the secondary IDE channel.

Might not be the fix for your machine but it was for mine.
I will also say that my board supposedly supported 83mhz FSB, but it didn't work properly and would corrupt the hard drive if ran at that speed for too long.
SiS made a good chipset but PC Chips is PC Chips.

Do you have memtest86+ floppy? Try booting the PC with it. If it boots, wait for the memtest to complete. Let's see if that works.

Hey people thanks for replying! I was still trying to solve the problem, but the problem still persists. I tried to change the external clock to 83MHz, and it really got scarry with a temporally black screen, for a moment the system didn't want to boot anyway. I thought I had corrupted the HD or damaged the CPU 🤣, but it came back to work again...In real, I had messed with the jumpers. I take almost all jumpers from this board to put in another boards, motherboards, sound card, etc... I don't remember how all the jumpers was set. But I perceived that slot IDE 2, seems to be more responsible, in IDE 1 the any ISA cards are not recognized, while in IDE 2 so. In respect to CPU jump settings, I don't if it could be the issue, I was trying many combinations of external clock and internal clock, until nothing. One thing I can't remember well is the CPU core voltage specific correct for my CPU. My CPU for this board is an intel Pentium MMX 233, so I don't even remember what was the internal clock and external clock it was when working, in fact, I don't know if this is relevant, beside the fact the 83MHZ, as Jasin pointed, can damage something...About the memtest86+ it's a nice idea, I'm gonna try it and see if it will work.

Unfraternally, I could not create a floppy memtest86+, the reason? I use WINNIMAGE to create my floppies, but I always had trouble writing images to floppy. I bought 2 new boxes of floppies, and I can say I couldn't write a single disk. It always gives that error of head track, this is really annoying, because those floppies are new. I remember that to write the 4 disks for MS-DOS (Besides the supply disk), was much sacrifice, I expend many hours by exhaustively only forcing the error in the track, until the process is done. The funniest part is that I can copy and paste files to the floppy normally without any error, but just copying the files to the floppy won't make simply the memotest86+ bootable, ok? There are times it can inexplicably write nice without any trouble. These are reasons why I abhor floppies, they really drive me crazy. I actually think that my entire set of retro computers needs to be replaced because they are already starting to go down the drain🤣.

Just to see if it boots at all, create a FreeDos boot image and try that.

I had the exact same issue, the motherboard would sit at Starting MS-DOS... but it loaded FreeDos perfectly fine.

JeffeSilva wrote on 2023-12-18, 23:16:

Hey people thanks for replying! I was still trying to solve the problem, but the problem still persists. I tried to change the external clock to 83MHz, and it really got scarry with a temporally black screen, for a moment the system didn't want to boot anyway. I thought I had corrupted the HD or damaged the CPU 🤣, but it came back to work again...In real, I had messed with the jumpers. I take almost all jumpers from this board to put in another boards, motherboards, sound card, etc... I don't remember how all the jumpers was set. But I perceived that slot IDE 2, seems to be more responsible, in IDE 1 the any ISA cards are not recognized, while in IDE 2 so. In respect to CPU jump settings, I don't if it could be the issue, I was trying many combinations of external clock and internal clock, until nothing. One thing I can't remember well is the CPU core voltage specific correct for my CPU. My CPU for this board is an intel Pentium MMX 233, so I don't even remember what was the internal clock and external clock it was when working, in fact, I don't know if this is relevant, beside the fact the 83MHZ, as Jasin pointed, can damage something...About the memtest86+ it's a nice idea, I'm gonna try it and see if it will work.

Ok, so if I am understanding you correctly then you are currently running a Pentium MMX at a FSB speed of 83mhz? And you aren't sure at what core voltage you have it set?
Firstly, a Pentium MMX (other than Tillamook) should be ran with a stock voltage of 2.8v. Unless you are overclocking the CPU to a higher clock speed rating, you should leave it at 2.8v.
Secondly, Pentium MMX only officially supports FSB speeds of up to 66mhz. They can take higher FSB speeds with a certain degree of success but I wouldn't personally run one at anything past 75mhz.
Thirdly, while 83mhz FSB speeds are SOMEWHAT supported by the SiS 5598 chipset, and by extension certain revisions of the PC Chips M571.....it is spotty at best.
Technically speaking it's overclocking the chipset and is more trouble than it's worth.

All of that is to say that in order to troubleshoot this adequately, I would recommend setting all clocks and voltages to their default values.
Get the machine stable and then you can work on adjusting one setting at a time until you find the ragged edge of stability, then you take it back one step and you have found your issue.
My board was buggy even at 75mhz FSB until I found the right ram stick, with the proper timings, cache settings, GPU etc.
If you need more performance than a Pentium MMX at 200-233 you should look into a K-6, K6-2, or K6-III (classic, non plus). They both support up to 100mhz FSB and can clock upwards of 450mhz.

You COULD use a patched BIOS and for the K6-III you would need to.
http://www.steunebrink.info/k6plus.htm
I did with my board and it worked decently, but I ended up having to over volt my chip a little more than I liked.
In the grand scheme if you need more grunt, a Pentium III or Athlon platform is probably the better choice.

Hi! It was quite long since last time in this post.

To avoid make unnecessary replies from my part, I decided to expand lot of time trying to fix this problem and unfortunately the problem is still occurring. I tried everything I could do. I tried many types of boots, from floppy to CD-ROM, still same result...Regards Jasin and Zuldan, thanks a lot for your help guys, I tried the recommended and even with a FreeDos it got stuck too. I tried also even with different memory modules, like DIM, SIM, many HDs, different cables, floppies, CD-ROM. Also, tried different options in the BIOS, defaults, etc...nothing worked.

Now, I had searching in the internet for problem like this, and seems like it's related to problem in a "CHIP" interface, that's right a hanging during the boot time. The motherboard apparently is working nice but It really sounds a hardware problem. It just became like this inexplicably. Now I will let this board alone for a while, but before, there's something that I'd try for one more time: To re-flash the bios again. The problem is that I never flashed this kind EEPROM before, but I know they can be written with ultraviolet light and I don't have this kind of device. There's a jumper option 12v to select when you will flash it. I can't imagine how could I flash it, as the motherboard don't boot from any resource I select.

And about the AMD K-6..Wow, I have about 3 of these guys and I didn't even know they could reach this frequency...The sad part of this is that I have a single cooler, and it's a simple one, I don't know if it could deal with such high frequency, but It sounds really nice and interesting. I remember also that I used these AMDs for few times. I mostly use a P54C or P55C, in this case the P55C is in the M571, while the P54C in a MLR598(sometimes I change to the P55C).

I hardly ever use this board, I decided to because I wanted to try to recovery my SB16 with board by doing some tests. This board always shows right off the bat the ISA information when a card is plugged, while the MLR598 doesn't show it, only when it's inside windows....Ok, that's all, and thanks again.

Currently I'm running K6-2+ 500 at 6 * 83 MHz (500 MHz) on Elpina BA E-VO 9825 which is rebranded PCChips M571 version 3.2A, so the problem is not in "overclocked" SIS chipset. The only problems with this mobo are RAM-picky AMI BIOS and dual drive primary ATA channel.

JeffeSilva wrote on 2023-12-23, 22:36:

Hi! It was quite long since last time in this post. […]
Show full quote

Hi! It was quite long since last time in this post.

To avoid make unnecessary replies from my part, I decided to expand lot of time trying to fix this problem and unfortunately the problem is still occurring. I tried everything I could do. I tried many types of boots, from floppy to CD-ROM, still same result...Regards Jasin and Zuldan, thanks a lot for your help guys, I tried the recommended and even with a FreeDos it got stuck too. I tried also even with different memory modules, like DIM, SIM, many HDs, different cables, floppies, CD-ROM. Also, tried different options in the BIOS, defaults, etc...nothing worked.

Now, I had searching in the internet for problem like this, and seems like it's related to problem in a "CHIP" interface, that's right a hanging during the boot time. The motherboard apparently is working nice but It really sounds a hardware problem. It just became like this inexplicably. Now I will let this board alone for a while, but before, there's something that I'd try for one more time: To re-flash the bios again. The problem is that I never flashed this kind EEPROM before, but I know they can be written with ultraviolet light and I don't have this kind of device. There's a jumper option 12v to select when you will flash it. I can't imagine how could I flash it, as the motherboard don't boot from any resource I select.

And about the AMD K-6..Wow, I have about 3 of these guys and I didn't even know they could reach this frequency...The sad part of this is that I have a single cooler, and it's a simple one, I don't know if it could deal with such high frequency, but It sounds really nice and interesting. I remember also that I used these AMDs for few times. I mostly use a P54C or P55C, in this case the P55C is in the M571, while the P54C in a MLR598(sometimes I change to the P55C).

I hardly ever use this board, I decided to because I wanted to try to recovery my SB16 with board by doing some tests. This board always shows right off the bat the ISA information when a card is plugged, while the MLR598 doesn't show it, only when it's inside windows....Ok, that's all, and thanks again.

You could try boot with a EPROM attached to a network card. It bypasses the BIOS boot sequence.
https://www.youtube.com/watch?v=ofZLDAOmZYU

You can also use a dedicated XTIDE card instead of an EPROM on a network card https://youtu.be/5NRHMTGXd94?si=GBCq2u_1kc6cWXzQ

analog_programmer wrote on 2023-12-25, 08:29:

Currently I'm running K6-2+ 500 at 6 * 83 MHz (500 MHz) on Elpina BA E-VO 9825 which is rebranded PCChips M571 version 3.2A, so the problem is not in "overclocked" SIS chipset. The only problems with this mobo are RAM-picky AMI BIOS and dual drive primary ATA channel.

Not necessarily, simply because one board is stable at that frequency it doesn't mean that all will be.
Also, you say that yours is the version 3.2A, that is the first version that "officially" supported the 83mhz FSB to begin with. I put official in quotes, because even though PC Chips/Elpina/Amptron called it official support for their re-branded TXPro chipset, the chipset was actually an SiS 5598 and never supports 83mhz in any revision, officially or otherwise.
So yes, by definition it is overclocking the chipset, whether PC Chips claimed otherwise or not.
My board would do it to some degree, I ran have a whole thread dedicated to the journey to getting it stable.
I never fully did, even though I was able to run all the benchmarks and even game on the PC for sometimes hours at a time.
It still wasn't 100% stable, with the k6-3+, the Cyrix MII, the K6-2 or the Pentium MMX chip that I used in it.

But you are correct about the ram sticks, it's super picky with pretty much any hardware.
Once you find the right combo it can be a very stable board, but it's no in the league of a good MVP3 or Ali board.

Just for the record: I'm not saying anything, I wrote what's the situation with this SIS 5598 chipset based motherboard.

Stability of the system also depends on power supply, which also depends on the (old) electrolytic capacitors (both in the PSU and on the MB).

There are no different revisions of this chipset to (not) support this chipset's "overcklocking" frequencies (75 and 83 MHz), the chipset supports them out of the factory. I think, the early board revisions do not have suitable PLL's schematics to produce them.

These 75 and 83 MHz FSB frequencies are not unofficial for (super) socket 7 mobos, but neglected in manuals. These "odd" FSB frequencies are used for some models Cyrix MII CPUs and for example i430TX chipset also supports them perfectly fine (I have one very underrated socket 7 Zida Tomato 5STX mobo with i430TX and this very same K6-2+ 500 also runs rock stable at 6 * 83 MHz on it).

analog_programmer wrote on 2024-01-02, 19:23:

Just for the record: I'm not saying anything, I wrote what's the situation with this SIS 5598 chipset based motherboard. […]
Show full quote

Just for the record: I'm not saying anything, I wrote what's the situation with this SIS 5598 chipset based motherboard.

Stability of the system also depends on power supply, which also depends on the (old) electrolytic capacitors (both in the PSU and on the MB).

There are no different revisions of this chipset to (not) support this chipset's "overcklocking" frequencies (75 and 83 MHz), the chipset supports them out of the factory. I think, the early board revisions do not have suitable PLL's schematics to produce them.

These 75 and 83 MHz FSB frequencies are not unofficial for (super) socket 7 mobos, but neglected in manuals. These "odd" FSB frequencies are used for some models Cyrix MII CPUs and for example i430TX chipset also supports them perfectly fine (I have one very underrated socket 7 Zida Tomato 5STX mobo with i430TX and this very same K6-2+ 500 also runs rock stable at 6 * 83 MHz on it).

Yes and no, and I don't mean to offend I just want to make sure that any information that is perpetuated is correct.
The 5598 chipset only officially supported bus speeds of 55/60/66/75. Any faster speeds were a spec included by the motherboard manufacturer, but it was by very definition overclocking.
The first SiS chipset to officially support 83/90/95/100 speeds was the 530.
Even that is questionable if you happened to have a board like a PC Chips M598LMR, ask me how I know.

It wasn't just PC Chips or SiS that was doing this, plenty of Slot 1 boards did this as well, the i440BX is a rather famous example. Many if not all 440BX chipsets were rock stable at 133MHZ and the various board partners built in support for FSB speeds of up to 150mhz.
But they were technically also overclocking the chipset as Intel never officially supported anything faster than 100mhz FSB. Just like with the 5598 there were no dividers either so almost all of the time one system bus or another was running out of spec. This is why these boards can be so hard to troubleshoot as stability and performance is literally a sum of all the parts.

FWIW I agree with you completely on the capacitor issues.
I actually think that is why my crappy board didn't work worth a damn when I first built the PC, but over time it became more and more stable as the caps were allowed to cycle.
I bet a recap would have done the thing wonders, but I gave it to a friend who is happily using it at 66mhz operation.

Jasin Natael, I agree with everything, except the term "officially supported". What does this mean? Is something supported by the chipset, but omitted from most of the users manuals - yes! So it is supported. And I gave you an example where to dig for these rarely used "not official" 75 and 83 MHz (S)S7 FSB frequiencies source - Cyrix MII CPUs. A separate question is whether a particular motherboard as design, build and quality covers the chipset parameters for stable operation. In this case it does even if it is so hated PCChips brand (or some of its clones).

Most of these (S)S7 chipsets also work with 55 or even 50 MHz FSB, but again this was mostly made for specific early models Pentiums (and some other Pentium-class) CPUs and this was also omitted from the newest revisions of SS7 board's manuals as P54C was already considered obsolete at the end of '90s. Does this mean that 50 and 55 MHz FSB are "not supported"or "not official" at EOL of (S)S7 platform? 😀

(Super) socket 7 was unique platform in therms of compatibility and still my personal favorite.

analog_programmer wrote on 2024-01-02, 21:57:

Jasin Natael, I agree with everything, except the term "officially supported". What does this mean? Is something supported by the chipset, but omitted from most of the users manuals - yes! So it is supported. And I gave you an example where to dig for these rarely used "not official" 75 and 83 MHz (S)S7 FSB frequiencies source - Cyrix MII CPUs. A separate question is whether a particular motherboard as design, build and quality covers the chipset parameters for stable operation. In this case it does even if it is so hated PCChips brand (or some of its clones).

Most of these (S)S7 chipsets also work with 55 or even 50 MHz FSB, but again this was mostly made for specific early models Pentiums (and some other Pentium-class) CPUs and this was also omitted from the newest revisions of SS7 board's manuals as P54C was already considered obsolete at the end of '90s. Does this mean that 50 and 55 MHz FSB are "not supported"or "not official" at EOL of (S)S7 platform? 😀

(Super) socket 7 was unique platform in therms of compatibility and still my personal favorite.

By official support I am referring to the technical data sheet of the chipset that was provided by the manufacturer.
Each piece of hardware is tested and validated to certain specifications. What these are and why they are supported or not is up to the manufacturer themselves. Usually guided by a certain average of what each piece silicon is capable of. In other words if 100% of tested samples are capable of say 500mhz but only 92% of the samples will do 550mhz, then the official max speed is published as 500mhz. Now of course it is more complex than this, binning is a thing and chipsets aren't a single piece of silicon with a single die, so it's not apples to apples. But that is the general idea.

To your second point, Cyrix supporting a different FSB speed isn't a problem at all but simply because your CPU is capable and supports a certain bus speed it doesn't mean that your motherboard automatically supports that speed as well. One has zero to do with the other. The original Socket 7 spec dictated a bus speed of UP TO 66mhz. This was largely guided by the Intel standard, of course other manufacturers came along and released both CPUs and chipsets capable of running at higher speeds. Cyrix is a prime example. SiS is another as their 5598 was designed to operate specifically for the Cyrix (and others, IDT, Rise etc) at 75mhz. But it unequivocally absolutely wasn't designed to operate at 83mhz. 70-80, maybe even 90% of them would do so, but it wasn't designed for it. It wasn't published by the engineers of the chipset and SiS never endorsed doing so. The 90s was a bit of a wild west when it came to this sort of thing. The tech was burgeoning and changing rapidly. PC Chips slapped a different name on the chipsets and claimed they would do things they weren't ever meant to do. Who was going to stop them?

Back to the datasheets; you can easily Google them up if you would like. But I will link to a few if you would care to argue the point further.

The supported bus speeds are listed on page 2 of this PDF.
https://www.datasheetarchive.com/datasheet?id … =M&term=sis5598

Here is an old chipset guide from Anandtech that lists the official and "unofficial" FSB speeds of the 5598 as well.
https://www.anandtech.com/show/72/14

There are dozens more if you feel the need to look them up.

I know this AT's article and it's good for a quick references of socket 7 chipsets.

As an engineer I can easily get your strict point of view, but why you're insisting so much on what's written in some official datasheets? There's a tons of technical documentation with not officially documented (or just barely mentioned) options hidden as "for future usage" or "for test purposes" or... whatever. And the only thing that determines whether this "for future use" will become an officially usable and documented function soon after the release of the product to the market in most cases turns out to be this very same market. I'm leaving it up to you to draw on your own correlation between the "unofficial" (I'll call them "unpopular" or intentionally undocumented) FSB frequencies for these (S)S7 chipsets and the market share of processor models that used them when they were current products.

analog_programmer wrote on 2024-01-02, 23:25:

I know this AT's article and it's good for a quick references of socket 7 chipsets.

As an engineer I can easily get your strict point of view, but why you're insisting so much on what's written in some official datasheets? There's a tons of technical documentation with not officially documented (or just barely mentioned) options hidden as "for future usage" or "for test purposes" or... whatever. And the only thing that determines whether this "for future use" will become an officially usable and documented function soon after the release of the product to the market in most cases turns out to be this very same market. I'm leaving it up to you to draw on your own correlation between the "unofficial" (I'll call them "unpopular" or intentionally undocumented) FSB frequencies for these (S)S7 chipsets and the market share of processor models that used them when they were current products.

Fair enough, I don't wish to beat a dead horse. I think we essentially agree on all of the points that truly matter.

Ok, this is what I call overclock and it's out of specs:

rmay635703 wrote on 2020-12-07, 00:08:

I ran m569s/m571s for many years usually at 500mhz with a 2.2volt k6-2 […]
Show full quote

I ran m569s/m571s for many years usually at 500mhz with a 2.2volt k6-2

If you “run” a k6-2 make sure you only use the pci/2 setting as async is unusable

There is a k6+ modded bios and a modded bios for 128gb support

The “video” is by its lonesome on a bracket and the audio on another, both are used on many pcchips all in one systems and yeah pcchips/amptron/elite group specific

Personally on AT motherboards I STRONGLY recommend the m569 bios over the m571 bios because you can use 256mb SDRAM
The m571 bios is extremely unstable with most sdram modules and can’t really use most modules larger than 64mb unless you use edo
The m571 system bios is also even more picky about the module used when you run the onboard video, what was strange is that the ram speed and ram quality/bus speed seemed to have less to do with it than chip count and the modules size. AKA A stable module would always work with any bus speed/memory timings an unstable one was always unstable no matter what bus or timing settings I used.

I also found that the m571 runs 83mhz rock stable especially if you use the slow onboard video with a fast compatible memory module, I’ve had mine overclocked as high as 105mhz by modding the clock crystal.

Good Luck

Unfortunately this fellow forum user turned out to be very non-communicative and non-cooperative, but I'll try this mod on my own, when I found suitable crystal oscillator.

analog_programmer wrote on 2024-01-03, 16:24:

Ok, this is what I call overclock and it's out of specs:

rmay635703 wrote on 2020-12-07, 00:08:

I ran m569s/m571s for many years usually at 500mhz with a 2.2volt k6-2 […]
Show full quote

I ran m569s/m571s for many years usually at 500mhz with a 2.2volt k6-2

If you “run” a k6-2 make sure you only use the pci/2 setting as async is unusable

There is a k6+ modded bios and a modded bios for 128gb support

The “video” is by its lonesome on a bracket and the audio on another, both are used on many pcchips all in one systems and yeah pcchips/amptron/elite group specific

Personally on AT motherboards I STRONGLY recommend the m569 bios over the m571 bios because you can use 256mb SDRAM
The m571 bios is extremely unstable with most sdram modules and can’t really use most modules larger than 64mb unless you use edo
The m571 system bios is also even more picky about the module used when you run the onboard video, what was strange is that the ram speed and ram quality/bus speed seemed to have less to do with it than chip count and the modules size. AKA A stable module would always work with any bus speed/memory timings an unstable one was always unstable no matter what bus or timing settings I used.

I also found that the m571 runs 83mhz rock stable especially if you use the slow onboard video with a fast compatible memory module, I’ve had mine overclocked as high as 105mhz by modding the clock crystal.

Good Luck

Unfortunately this fellow forum user turned out to be very non-communicative and non-cooperative, but I'll try this mod on my own, when I found suitable crystal oscillator.

I think I remember reading that posting. It would be interesting to play around with crystals just to see what it's capable of. SiS chipsets are pretty interesting, they were really pushing the envelope of integrating anything they could. Many of these boards needing nothing but CPU/ram and drives to be ready to go.