Seems like those were just overdrive present and sx/dx

They should be set to match if it’s an overdrive chip vrs std non-overdrive but my machines never had them.

Generally overdrive CPUs have a hard coded multiplier and the sx disable pin which probably overrides the jumper anyway

For Intel overdrive CPU’s leave the Jumpers for a 33mhz CPU.
The Overdrive CPU will control the CPU speed to what ever the CPU made for.
If its a non-overdrive CPU then you need to set the jumpers on the motherboard to match the CPU speed.

Intel486dx33 wrote on 2025-02-11, 12:48:

For Intel overdrive CPU’s leave the Jumpers for a 33mhz CPU.
The Overdrive CPU will control the CPU speed to what ever the CPU made for.
If its a non-overdrive CPU then you need to set the jumpers on the motherboard to match the CPU speed.

I think that was the assumption, but as epictronics discovered, they had no impact on the FSB speed (25Mhz or 33Mhz), it was the resistors being present/not in J174/175/176 that dictates that. He was unable to use a non-ODP CPU without modifying the socket (due to the soldered on 486 SX25 and the 169 pin unit needed to disable it) so it appears there wasn't a provision for ODP vs non-ODP on these boards, the IBM manual seems to only state 487 Co-Pro or ODP as upgrade options.

My board (and his) were both 25Mhz FSB boards. We both did the resistor mod (J174/175/176) to change the board to 33Mhz.

It's unclear as to what these jumpers actually DO, at least on this configuration.

For the SX25 machines, it looks like they are all supposed to be 1-2, 1-2, 2-3 regardless of whether you have an ODP or not, because they assume 25Mhz FSB (no way to change it). For the 33Mhz FSB machines, the jumpers are the same, except apparently, for the DX33 machine, which requires you to move the J25 jumper to 1-2 if you fit an Overdrive DX66 to it. But the SX33, the DX66 Overdrive upgrade, you keep the jumper in 2-3. With the table again not showing non-ODP as an upgrade option. There are no paths that involve the FSB changing in any of the tables I've seen at this juncture either.

Hence my curiosity 😀

rmay635703 wrote on 2025-02-11, 04:51:

Seems like those were just overdrive present and sx/dx

They should be set to match if it’s an overdrive chip vrs std non-overdrive but my machines never had them.

Generally overdrive CPUs have a hard coded multiplier and the sx disable pin which probably overrides the jumper anyway

Yeah, I'm wondering if the ODP completely disables them (so they do nothing at all)? The only thing that causes me to take pause is the the DX33 system in the table in my previous post, which shows moving J25 to 1-2 with the ODP DX266, but that seems to only apply to that very specific system: 2133 28A.

As I said, I moved mine, and it didn't seem to do anything at all.

OVERK|LL wrote on 2025-02-11, 14:26:

rmay635703 wrote on 2025-02-11, 04:51:

Seems like those were just overdrive present and sx/dx

They should be set to match if it’s an overdrive chip vrs std non-overdrive but my machines never had them.

Generally overdrive CPUs have a hard coded multiplier and the sx disable pin which probably overrides the jumper anyway

Yeah, I'm wondering if the ODP completely disables them (so they do nothing at all)? The only thing that causes me to take pause is the the DX33 system in the table in my previous post, which shows moving J25 to 1-2 with the ODP DX266, but that seems to only apply to that very specific system: 2133 28A.

As I said, I moved mine, and it didn't seem to do anything at all.

Technically the 486 had a multiplier pin for 1x or 2x but in reality nothing used it until the late dx4/dx5 CPUs

Knowing IBM they probably had a dummy jumper for overdrive / standard cpu / disable SX and left it on all the boards but as a dummy on certain units.

You would need to scope out if the jumper actually attaches to something on both sides
Ground an a pin on the socket being likely suspects.

rmay635703 wrote on 2025-02-11, 21:59:

Technically the 486 had a multiplier pin for 1x or 2x but in reality nothing used it until the late dx4/dx5 CPUs […]
Show full quote

OVERK|LL wrote on 2025-02-11, 14:26:

rmay635703 wrote on 2025-02-11, 04:51:

Seems like those were just overdrive present and sx/dx

They should be set to match if it’s an overdrive chip vrs std non-overdrive but my machines never had them.

Generally overdrive CPUs have a hard coded multiplier and the sx disable pin which probably overrides the jumper anyway

Yeah, I'm wondering if the ODP completely disables them (so they do nothing at all)? The only thing that causes me to take pause is the the DX33 system in the table in my previous post, which shows moving J25 to 1-2 with the ODP DX266, but that seems to only apply to that very specific system: 2133 28A.

As I said, I moved mine, and it didn't seem to do anything at all.

Technically the 486 had a multiplier pin for 1x or 2x but in reality nothing used it until the late dx4/dx5 CPUs

Knowing IBM they probably had a dummy jumper for overdrive / standard cpu / disable SX and left it on all the boards but as a dummy on certain units.

You would need to scope out if the jumper actually attaches to something on both sides
Ground an a pin on the socket being likely suspects.

I may try to trace it if I get sufficiently motivated, but that may not be for a while, 🤣. I'm waiting on my EPROM for my 3C509 to switch over to XT-IDE from OnTrack.