Reply 40 of 78, by Kahenraz
- Rank
- l33t
I've added some more detailed information, to help people avoid methods which did not work with this onboard video.
I've added some more detailed information, to help people avoid methods which did not work with this onboard video.
Sweet! So what’s next?
Talk to jan for k6-3+ support? :p
DIY Bios Modding guide Jan Steunebrink k6-2+/3+ 128gb
http://web.inter.nl.net/hcc/J.Steunebrink/k6plus.htm
With the 3+ and central tweaking unit in windows you can I believe clock it all the way down to 2x and disable all the caches from windows. Or raise it all the way to 6x with L1 L2 and L3 enabled
Edit: oh…. I only see one voltage fet
You have a single plane voltage board. Gotta get a interposer
Socket 5/7/SS7 (Voltage Interposer) Tweaker. (Released) Here is one if you can solder well.
I think I need to build that interposer. This board would be another good candidate.
How much memory can this board support?
With two cache tag chips 512mb
Without two cache tag chips 512mb but slow or 64mb but fast
But….
K6-3
K6-2+
K6-3+
And 512mb fast no matter the onboard cache. In fact you can disable onboard cache and it may get faster.
For Wing Commander, a 386 DX-25 looks like a good match. I read that Socket 5 CPUs will work in Socket 7. Will this CPU work in this board?
Kahenraz wrote on 2022-10-29, 04:33:For Wing Commander, a 386 DX-25 looks like a good match. I read that Socket 5 CPUs will work in Socket 7. Will this CPU work in this board?
Socket 5 CPUs will work in Socket 7 boards, no idea if this also applies to Super Socket 7 but I don't see why not if the board supports the right voltages. Socket 4(PGA,1st gen P5) and Socket 5(SPGA, 2nd gen P5) was also only used on early Pentium CPUs, 386/486 were Socket 2/3 which used PGA and are not pin compatible with the SPGA used in Socket 5/7. (BIOS might be the limiting factor for Super 7 boards running Socket 5 CPUs)
As for 386/486 CPUs working in Socket 4 boards . .thats a negative ghost rider, I havent been able to find any reports of Socket 4 boards being able to run 486 class CPUs.
So dont go trying to put that socket 3 CPU into this Socket 7 board it wont fit without some serious pin modding with a grinder.
I may end up owning at least one AT board then.
Socket 4 is socket 4 only.
It’s basically a 5v version of socket 5
Socket 5 cpus will work in s7 and ss7 if correct settings are applied.
Socket 7 and ss7 cpus usually work in one another sometimes a bios update is needed
Socket 7 and ss7 cpus will work in socket 5 with a interposer and bios update.
Note: some motherboards with sockets labeled socket 7 are really socket 5 mobos pretending to be socket 7. (Single plane voltage) OP’s board seems to be such a board. As I only see one mosfet.
wrote:OP’s board seems to be such a board. As I only see one mosfet.
Wrong.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
I can use Socket 7 CPUs in it, and it clearly states support for up to 233 on the board. It's definitely Socket 7.
The Serpent Rider wrote on 2022-10-29, 06:15:wrote:OP’s board seems to be such a board. As I only see one mosfet.
Wrong.
Ya know I normally don't have an issue with one word replies but I think here some further elaboration might help your reply actually add to the discussion .
As it is your reply is rather unhelpful and adds nothing to the discussion.
Where´s that 2nd voltage regulator supposed to be? On the backside of the heatsink?
Or does this board use the 3.3V from the ATX PSU for IO-voltage?
majestyk wrote on 2022-10-29, 06:51:Where´s that 2nd voltage regulator supposed to be? On the backside of the heatsink?
Or does this board use the 3.3V from the ATX PSU for IO-voltage?
Are there boards that do this ?
Doesn't sound terribly safe to be using a main supply voltage in that manner for CPU IO Voltage without some regulation and filtering/clamping.
wrote:Or does this board use the 3.3V from the ATX PSU for IO-voltage?
Well, obviously Intel would fully use the standard they've made. Unlike certain motherboard brands, which didn't bother to use +3.3v rail for anything on ATX boards, including PCI slots.
wrote:Doesn't sound terribly safe to be using a main supply voltage in that manner for CPU IO Voltage
Yeah, sure. Just few years prior they've powered the whole Socket 4 directly from PSU. Literally shocking!
I must be some kind of standard: the anonymous gangbanger of the 21st century.
This requires the PSU to fulfill these standards (and the ATX-standard) of course.
Still there must have been reasons why most mainboard manufacturers provided a second onboard vrm (despite the additional costs):
The following come to mind:
1. Being able to offer different I/O voltages and jumper accordingly
2. Mainboards with both ATX and AT power supply
3. Non-Intel CPUs like AMD, Cyrix, IDT... could have been less tolerant to transients and ripple, or have higher demands regarding to stability / tolerance and manufacturers wished to be versatile + on the safe side by providing a local VRM near the CPU socket.
The Serpent Rider wrote on 2022-10-29, 07:15:Well, obviously Intel would fully use the standard they've made. Unlike certain motherboard brands, which didn't bother to use + […]
wrote:Or does this board use the 3.3V from the ATX PSU for IO-voltage?
Well, obviously Intel would fully use the standard they've made. Unlike certain motherboard brands, which didn't bother to use +3.3v rail for anything on ATX boards, including PCI slots.
wrote:Doesn't sound terribly safe to be using a main supply voltage in that manner for CPU IO Voltage
Yeah, sure. Just few years prior they've powered the whole Socket 4 directly from PSU. Literally shocking!
There is always regulation and filtering components the power is routed through before it even gets near the CPU socket, the VRM on this board certainly doesn't have nearly enough components to be handling both 5v and 3.3v, Intel is hardly likely to just shunt the 3.3v straight from the ATX power socket either, that's just stupid.
wrote:This requires the PSU to fulfill these standards (and the ATX-standard) of course.
Obviously, these motherboard came from OEM PC brands - PSUs had to met requirements set by Intel.
Being able to offer different I/O voltages and jumper accordingly
Correct, but Intel had no concern for Cyrix, AMD and Winchip CPUs.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
One way to be sure :p
Find the votlage setting jumper and measure the middle corner pin on the socket.
Can’t say much more without the board in front of me other than most single fet boards are single plane. (Regardless of what socket says)
But using the fet for vcore and atx for 3.3 is possible.
One way to be sure
Intel ATX motherboards officially support Pentium MMX. End of the story.
I must be some kind of standard: the anonymous gangbanger of the 21st century.