VOGONS


Reply 40 of 60, by retardware

User metadata
Rank Oldbie
Rank
Oldbie

I can second what @RaiderOfLostVoodoo said regarding this forum.
Btw, @Stiletto did a great collection, found this very useful collection page when looking for slow down utilities.

RaiderOfLostVoodoo wrote on 2021-11-04, 17:59:

But I want it as a DIP switch on the front. And only for the turbo settings.
I did wonder if there is a plug and play solution. [...] So it's DIY time.

Apropos DIY...
This is a 5.25" turbo wheel with nn.n MHz display which plugs into the crystal oscillator socket.
Range is 2-33.3 MHz.
Only works with AMD processor because Intel CPUs hang when clocked below ~16MHz.

Reply 41 of 60, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

My ga-5aa has a turbo led and switch. When I push it, the light turns on, then off when button released. I guess turbo buttons had a locking feature?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 42 of 60, by Stiletto

User metadata
Rank l33t++
Rank
l33t++
retardware wrote on 2021-11-05, 02:43:

Btw, @Stiletto did a great collection, found this very useful collection page when looking for slow down utilities.

Thanks! It's pretty old (I originally posted it in 2002, updated the links in 2012!) but still perfectly serviceable! Might be missing some newer utility, IDK.
And as the sierrahelp.com page states, it's still on VOGONS!

List of Slowdown Utilities / List of Slowdown Utilities (Beta)

Some of the links may have broken again, I'll try to find the time to check it out!

"I see a little silhouette-o of a man, Scaramouche, Scaramouche, will you
do the Fandango!" - Queen

Stiletto

Reply 43 of 60, by Disruptor

User metadata
Rank Oldbie
Rank
Oldbie

The clock generator on my ASUS PVI-486SP3 changes FSB frequency that slow so that I can abuse it to switch between 40 and 25 MHz just with a 3 pin turbo switch.
However, it has a regular turbo switch connector too.

Reply 44 of 60, by shamino

User metadata
Rank l33t
Rank
l33t
retardware wrote on 2021-11-05, 02:43:
Apropos DIY... This is a 5.25" turbo wheel with nn.n MHz display which plugs into the crystal oscillator socket. Range is 2-33.3 […]
Show full quote

Apropos DIY...
This is a 5.25" turbo wheel with nn.n MHz display which plugs into the crystal oscillator socket.
Range is 2-33.3 MHz.
Only works with AMD processor because Intel CPUs hang when clocked below ~16MHz.

A long time ago I got caught up in the idea of somehow designing a variable deturbo that could be soldered into a motherboard. I bought a few Cypress oscillator chips that were programmable, and even found the obscure programmer on eBay at one point and bought that too. Then I started fiddling with the software, realized I didn't know what I was doing and lost interest, never touching it again.

This picture is what I thought I was trying to achieve back then, complete with the rotary dial. Looks awesome.

Reply 45 of 60, by j^aws

User metadata
Rank Oldbie
Rank
Oldbie
RaiderOfLostVoodoo wrote on 2021-11-04, 17:59:
I had exactly the same idea. But I want it as a DIP switch on the front. And only for the turbo settings. I did wonder if there […]
Show full quote
j^aws wrote on 2015-02-15, 19:05:

Yes, definitely cumbersome with changing jumpers. However, I have plans to build an external jumper switch-box for all controllable jumpers (audio/ video/ motherboard) when everything is finalised, hopefully. Otherwise, I personally wouldn't bother to open the case to change jumpers.

I had exactly the same idea. But I want it as a DIP switch on the front. And only for the turbo settings.
I did wonder if there is a plug and play solution. But guess there isn't. So it's DIY time.
Did you finish it? I'm curious and would like to see the result.

Hmm, I remember this post. This idea was for my AT beige cases, and having an external switch box that I could move around between them, and not having to modify the cases. The plan was to standardise internal wiring in the cases, then using something like a DB25 connector, hook up the external switch box.

The DIP switch idea is also great and did cross my mind as well. It's discrete, too.

My original plan took a back burner as I converted a modern ATX case to AT format instead. This case also has an external flap that hides four 5 1/4 drive bays, and I'm prepared to modify the case, and use the drive bay fascia to house some funky toggle switches. They'd be discrete behind the case flap. I have all the parts - just need to finish the project.

PS: I'm amazed a six year old post reply still managed to send me an email notification!

Reply 46 of 60, by asdf53

User metadata
Rank Member
Rank
Member

On my 486 board (FIC 486-VIP-IO) the turbo function does nothing. The switch makes the turbo LED go on and off, but that's it. There's an option "Boot up system speed high/low" in the BIOS, it also makes the turbo LED go on/off, but does nothing either. No difference in any of the CPU/memory benchmarks. I reset the bus speed to the default 33.3 MHz in case it only works with that, but no. I'm using a Cyrix CPU - could it be that it's trying to manipulate CPU registers to enable or disable the cache and this only works on Intel CPUs? This is the only CPU I have at the moment, so I can't test it.

Reply 47 of 60, by Mr.Hunt

User metadata
Rank Newbie
Rank
Newbie
j^aws wrote on 2015-04-10, 08:54:
I don't think I've come across a software tool that does this, but maybe one exists and is lost in time...? […]
Show full quote
philscomputerlab wrote:

Could you please share how you do this, meaning what tool, what commands?

This could be of interest for "slow down" projects and helpful in getting old games to work a bit better 😀

I don't think I've come across a software tool that does this, but maybe one exists and is lost in time...?

Basically it's done in hardware, more specifically, there is a jumper on my Socket 7 board that switches between L1 cache as Write-back or Write-thru. And the K6-III+ is installed via an adapter, so I'm not using a Super Socket 7 board. I'm running the board with L3 cache (external) disabled, so that I can use 6x multi without issues. There aren't any BIOS options for Write-back or Write-thru modes - only the jumper on the board. Ironically, the board also has a Turbo Switch header that does nothing.

The K6-III+ can run between 50MHz to 480MHz using 2x25 to 6x80, and with both L1/L2 enabled using Write-thru, I'm getting between 23-544 scores in Speedsys (544 in WB). So that's around as low as a 486DX-33, and will go even slower by disabling L1/ L2 caches too.

I'm not sure how this is exactly working - maybe there are some undocumented registers that could be enabled to do this in software?

Hello! I understand that I am writing quite late, but maybe you still remember. What is your motherboard for this K6-iii+? And maybe you remember what clocker was on it? (chip)

I make the assembly on the GA-5AA and, unfortunately, the minimum is that I can do 60MHz, and the maximum that I can connect to the turbo button is 95Mhz. Otherwise, I can’t get everything to it, there are only 4 contact groups, so it turns out (independent ) 4 contacts, which are not enough for a large scatter, even 100 bus, not a guards already about 110, etc.

Reply 48 of 60, by j^aws

User metadata
Rank Oldbie
Rank
Oldbie
Mr.Hunt wrote on 2024-10-23, 19:18:
j^aws wrote on 2015-04-10, 08:54:
I don't think I've come across a software tool that does this, but maybe one exists and is lost in time...? […]
Show full quote
philscomputerlab wrote:

Could you please share how you do this, meaning what tool, what commands?

This could be of interest for "slow down" projects and helpful in getting old games to work a bit better 😀

I don't think I've come across a software tool that does this, but maybe one exists and is lost in time...?

Basically it's done in hardware, more specifically, there is a jumper on my Socket 7 board that switches between L1 cache as Write-back or Write-thru. And the K6-III+ is installed via an adapter, so I'm not using a Super Socket 7 board. I'm running the board with L3 cache (external) disabled, so that I can use 6x multi without issues. There aren't any BIOS options for Write-back or Write-thru modes - only the jumper on the board. Ironically, the board also has a Turbo Switch header that does nothing.

The K6-III+ can run between 50MHz to 480MHz using 2x25 to 6x80, and with both L1/L2 enabled using Write-thru, I'm getting between 23-544 scores in Speedsys (544 in WB). So that's around as low as a 486DX-33, and will go even slower by disabling L1/ L2 caches too.

I'm not sure how this is exactly working - maybe there are some undocumented registers that could be enabled to do this in software?

Hello! I understand that I am writing quite late, but maybe you still remember. What is your motherboard for this K6-iii+? And maybe you remember what clocker was on it? (chip)

I make the assembly on the GA-5AA and, unfortunately, the minimum is that I can do 60MHz, and the maximum that I can connect to the turbo button is 95Mhz. Otherwise, I can’t get everything to it, there are only 4 contact groups, so it turns out (independent ) 4 contacts, which are not enough for a large scatter, even 100 bus, not a guards already about 110, etc.

Hi! It's okay if the reply is late as I still get notifications.

I still have the motherboard somewhere in storage, but don't recall the name offhand as it's been a while. I'm pretty sure this one was a 430FX chipset socket 7 board from what I can recall. The exact model and manufacturer is something I'd need to pull the board out to get.

If you say you have 4 pins to control the clock speed for your board, that should give you plenty of flexibility if they are all active. May need to dig up the clock specs to get an idea as there maybe undocumented speeds from pin combinations not advertised in motherboard manuals.

I tend to use socket 7 adapters with onboard VRMs and bypass motherboard VRMs to get better clock stability for unusual clocks, so that might help as well.

Reply 49 of 60, by Mr.Hunt

User metadata
Rank Newbie
Rank
Newbie
j^aws wrote on 2024-10-25, 14:52:
Hi! It's okay if the reply is late as I still get notifications. […]
Show full quote
Mr.Hunt wrote on 2024-10-23, 19:18:
j^aws wrote on 2015-04-10, 08:54:
I don't think I've come across a software tool that does this, but maybe one exists and is lost in time...? […]
Show full quote

I don't think I've come across a software tool that does this, but maybe one exists and is lost in time...?

Basically it's done in hardware, more specifically, there is a jumper on my Socket 7 board that switches between L1 cache as Write-back or Write-thru. And the K6-III+ is installed via an adapter, so I'm not using a Super Socket 7 board. I'm running the board with L3 cache (external) disabled, so that I can use 6x multi without issues. There aren't any BIOS options for Write-back or Write-thru modes - only the jumper on the board. Ironically, the board also has a Turbo Switch header that does nothing.

The K6-III+ can run between 50MHz to 480MHz using 2x25 to 6x80, and with both L1/L2 enabled using Write-thru, I'm getting between 23-544 scores in Speedsys (544 in WB). So that's around as low as a 486DX-33, and will go even slower by disabling L1/ L2 caches too.

I'm not sure how this is exactly working - maybe there are some undocumented registers that could be enabled to do this in software?

Hello! I understand that I am writing quite late, but maybe you still remember. What is your motherboard for this K6-iii+? And maybe you remember what clocker was on it? (chip)

I make the assembly on the GA-5AA and, unfortunately, the minimum is that I can do 60MHz, and the maximum that I can connect to the turbo button is 95Mhz. Otherwise, I can’t get everything to it, there are only 4 contact groups, so it turns out (independent ) 4 contacts, which are not enough for a large scatter, even 100 bus, not a guards already about 110, etc.

Hi! It's okay if the reply is late as I still get notifications.

I still have the motherboard somewhere in storage, but don't recall the name offhand as it's been a while. I'm pretty sure this one was a 430FX chipset socket 7 board from what I can recall. The exact model and manufacturer is something I'd need to pull the board out to get.

If you say you have 4 pins to control the clock speed for your board, that should give you plenty of flexibility if they are all active. May need to dig up the clock specs to get an idea as there maybe undocumented speeds from pin combinations not advertised in motherboard manuals.

I tend to use socket 7 adapters with onboard VRMs and bypass motherboard VRMs to get better clock stability for unusual clocks, so that might help as well.

On my GA-5AA I have ICS9148AF-75, by datasheet i have from 60 till 140 Mhz. I want to replace for more lover clocker, but i did not find what clocker can I use for replace. Try to find datasheet for PLL52C66-05 for understand pinouts but have no luck. As i understand PLL52C66-05 must be from 33 till 100 Mhz But we didn't know pinouts.

Reply 50 of 60, by j^aws

User metadata
Rank Oldbie
Rank
Oldbie
Mr.Hunt wrote on 2024-11-05, 10:09:
j^aws wrote on 2024-10-25, 14:52:
Hi! It's okay if the reply is late as I still get notifications. […]
Show full quote
Mr.Hunt wrote on 2024-10-23, 19:18:

Hello! I understand that I am writing quite late, but maybe you still remember. What is your motherboard for this K6-iii+? And maybe you remember what clocker was on it? (chip)

I make the assembly on the GA-5AA and, unfortunately, the minimum is that I can do 60MHz, and the maximum that I can connect to the turbo button is 95Mhz. Otherwise, I can’t get everything to it, there are only 4 contact groups, so it turns out (independent ) 4 contacts, which are not enough for a large scatter, even 100 bus, not a guards already about 110, etc.

Hi! It's okay if the reply is late as I still get notifications.

I still have the motherboard somewhere in storage, but don't recall the name offhand as it's been a while. I'm pretty sure this one was a 430FX chipset socket 7 board from what I can recall. The exact model and manufacturer is something I'd need to pull the board out to get.

If you say you have 4 pins to control the clock speed for your board, that should give you plenty of flexibility if they are all active. May need to dig up the clock specs to get an idea as there maybe undocumented speeds from pin combinations not advertised in motherboard manuals.

I tend to use socket 7 adapters with onboard VRMs and bypass motherboard VRMs to get better clock stability for unusual clocks, so that might help as well.

On my GA-5AA I have ICS9148AF-75, by datasheet i have from 60 till 140 Mhz. I want to replace for more lover clocker, but i did not find what clocker can I use for replace. Try to find datasheet for PLL52C66-05 for understand pinouts but have no luck. As i understand PLL52C66-05 must be from 33 till 100 Mhz But we didn't know pinouts.

Your GA-5AA is a Super Socket 7 board, right? What CPU are you using?

Are you trying to get your CPU to clock half as slow by replacing the clock generator IC with slower clock options? I'm not aware of pin compatible ones.

There are speed gaps in SS7 boards for different CPUs. I use Speedsys to measure target application speeds - these targets can be reached by tweaking various hardware parameters. Do you have a particular target Speedsys range?

If you have a particular target Speedsys that you cannot get with your setup, there is likely an easier way to get to that rather than replacing the clock generator on your board. E.g. on a SS7 board, a P54c Pentium has extra parameters for slowdown that you can manipulate using Setmul compared to K6 CPUs on a SS7 board.

Reply 51 of 60, by Mr.Hunt

User metadata
Rank Newbie
Rank
Newbie

Yes, I use K6-3+ (or rather K6-2+ with a soldered resistor), I know that I can use Setmul but I want to use hardware button and I try to find clocker with identical pinouts. Right now i have two positions: 1. 60MHz x2, so I have 120MHz; 2.95MHz x6, so I have 570MHz. I can't use more high because my button have 4 independed connectors (I haven't seen more than 4), so one for multiplier (x2 or x6), second and third for frequency (60MHz or 95MHz), and the last for power indicator. Unfortunately I can’t use more than 95MHz because need another free one independent channel on the button for 100MHz or 110MHz.

Reply 52 of 60, by asdf53

User metadata
Rank Member
Rank
Member
Mr.Hunt wrote on 2024-11-05, 10:09:

On my GA-5AA I have ICS9148AF-75, by datasheet i have from 60 till 140 Mhz. I want to replace for more lover clocker, but i did not find what clocker can I use for replace. Try to find datasheet for PLL52C66-05 for understand pinouts but have no luck. As i understand PLL52C66-05 must be from 33 till 100 Mhz But we didn't know pinouts.

Where did you get the information that the PLL52C66-05 supports 33 MHz FSB? Are you sure that's not just the PCI clock? The only info I could find about it is that it was used on this single board computer with a Pentium MMX. If it was made for these kinds of computers, it seems unlikely that it has AGP clocks.

Another idea, use a crystal resonator with lower frequency. You could solder two or more crystals on a little PCB that can be switched, and solder that to the clock generator's input pin. I don't know if that would work though, I have never done this.

Reply 54 of 60, by asdf53

User metadata
Rank Member
Rank
Member
PARKE wrote on 2024-11-10, 15:24:

That Phase Link chip is mentioned in an old Vogons thread about clock chips: Re: Clock generators on Socket 7 and older motherboards

Thanks, I have no idea how I missed that. It seems though, according to the frequency table, that it does not have AGP clocks and an unusable PCI frequency for 33 MHz, so it would probably not work well for slowing down a Super Socket 7 board.

Another thing I've heard about is a binary counter / frequency divider. A chip that takes an input frequency and divides it by powers of two. For example: 100 / 2 = 50 MHz, 100 / 4 = 25 MHz, 100 / 8 = 12.5 MHz. Could it be possible to solder such a chip in series with the clock generator's CPU output?

I had thought about this concept a few years ago when I tried to match the speed of a 4.77 MHz CPU on a Socket 7 motherboard. A Pentium 75 with disabled cache and chipset throttling almost got there, the speed was 90% correct, but not quite, so additional fine tuning is necessary.

Reply 55 of 60, by PARKE

User metadata
Rank Oldbie
Rank
Oldbie
asdf53 wrote on 2024-11-10, 17:12:

Thanks, I have no idea how I missed that. It seems though, according to the frequency table, that it does not have AGP clocks and an unusable PCI frequency for 33 MHz, so it would probably not work well for slowing down a Super Socket 7 board.

Much of the details of this sort of enterprise is above my head but I think that his best chance is switching to an earlier version of the clock chip that sits currently on his board.
His GA-5AA board has an ICS9148AF-75 and he could make a step back to an earlier version with the same 48 pin layout and similar functionality. I remember a vogons member upgrading an ASUS board with chips from this family so it is probably/maybe a viable route to success.
The ICS9148-26 listed here:
https://www.alldatasheet.com/view.jsp?Searchword=ICS9148
covers a lower (but also smaller) range

The attachment ICS.jpg is no longer available

Reply 56 of 60, by Mr.Hunt

User metadata
Rank Newbie
Rank
Newbie
PARKE wrote on 2024-11-11, 11:14:
Much of the details of this sort of enterprise is above my head but I think that his best chance is switching to an earlier vers […]
Show full quote
asdf53 wrote on 2024-11-10, 17:12:

Thanks, I have no idea how I missed that. It seems though, according to the frequency table, that it does not have AGP clocks and an unusable PCI frequency for 33 MHz, so it would probably not work well for slowing down a Super Socket 7 board.

Much of the details of this sort of enterprise is above my head but I think that his best chance is switching to an earlier version of the clock chip that sits currently on his board.
His GA-5AA board has an ICS9148AF-75 and he could make a step back to an earlier version with the same 48 pin layout and similar functionality. I remember a vogons member upgrading an ASUS board with chips from this family so it is probably/maybe a viable route to success.
The ICS9148-26 listed here:
https://www.alldatasheet.com/view.jsp?Searchword=ICS9148
covers a lower (but also smaller) range

The attachment ICS.jpg is no longer available

You are right, this is exactly the chip I was thinking about.

Reply 57 of 60, by Mr.Hunt

User metadata
Rank Newbie
Rank
Newbie
asdf53 wrote on 2024-11-10, 17:12:

Another thing I've heard about is a binary counter / frequency divider. A chip that takes an input frequency and divides it by powers of two. For example: 100 / 2 = 50 MHz, 100 / 4 = 25 MHz, 100 / 8 = 12.5 MHz. Could it be possible to solder such a chip in series with the clock generator's CPU output?

Where could I read about this?

Reply 58 of 60, by PARKE

User metadata
Rank Oldbie
Rank
Oldbie
Mr.Hunt wrote on 2024-11-17, 11:48:

You are right, this is exactly the chip I was thinking about.

Not exactly the right one for lack of AGP function. Try to compare with a higher number than 27.