Perhaps the board uses the same every jumper adds a value scheme as PC Chips M577 MVP3 board?

PC Chips M577 sets the voltage to 2.0V with no jumpers, the first jumper adds 0.1V, the second 0.2V, the third 0.4V and the forth 0.8V so the maximum voltage is 2.0V +0.1V +0.2V +0.4V +0.8V = 3.5V. The all jumpers off setting isn't documented but when comparing the different voltage settings it's easy to see how it works.

I have no idea if Chaintech uses a similar scheme but look if you can see a pattern in the jumper settings for the voltage. 2.2V is pretty much the maximum voltage the K6-3+ can handle long term and that is with a good cooler, 2.0V is fine with less than optimal cooling.

Thanks, skyscraper!

Looking at the jumper settings it looks indeed like 2.0V could be the "base voltage", and then you add 0.1, 0.2, 0.4, and 0.8 to that with various jumper combinations.

Is there any way to measure this voltage with a multimeter?

Your one stop K6-x+ BIOS stop:
http://web.inter.nl.net/hcc/J.Steunebrink/k6plus.htm

I'm pretty sure removing all jumpers will get you 2.0V as the jumper scheme seems to be exactly like the one used on PC Chips M577 board but there is probably a good place to measure with a DMM to make sure.

If you post a close up of the area around the socket I'm sure someone more qualified than me can help out, I mostly poke around until I find what I want.

I'm curious what that last jumper not used in any of the settings is for... Probably something totally unrelated. 😀

Skyscraper wrote:

I'm pretty sure removing all jumpers will get you 2.0V as the jumper scheme seems to be exactly like the one used on PC Chips M577 board but there is probably a good place to measure with a DMM to make sure.

If you post a close up of the area around the socket I'm sure someone more qualified than me can help out, I mostly poke around until I find what I want.

I'm curious what that last jumper not used in any of the settings is for... Probably something totally unrelated. 😀

Yeah, unfortunately the manual I've found for the 5AGM2 does not 100% correspond with my board. That jumper is called JP4 on my board, and the last pair of pins on it is not mentioned in the onboard descriptions.

Would there be any benefit of getting the 1.6V 400 MHz over the 2.0V 450 MHz?

gdjacobs wrote:

Your one stop K6-x+ BIOS stop:
http://web.inter.nl.net/hcc/J.Steunebrink/k6plus.htm

Thanks! Turns out the board was already updated with the latest bios.

What's the model number of the voltage regulator chip? I guess it's the analog devices chip near the CPU socket. The chip's datasheet might give us some clue about voltage settings.

These chips are just binned for lower voltage. So 2.2V will work fine, just makes the cooling is good, at 1.6V you get away with a small cooler.

c0keb0ttle wrote:

Thanks, skyscraper! […]
Show full quote

Thanks, skyscraper!

Looking at the jumper settings it looks indeed like 2.0V could be the "base voltage", and then you add 0.1, 0.2, 0.4, and 0.8 to that with various jumper combinations.

The attachment voltage.png is no longer available

Is there any way to measure this voltage with a multimeter?

Points of interest highlighted in red.
Vcore is probably going through that large inductor (in between the MOSFETs and the row of caps). If that's not easy to safely reach, then you can also try that pair of MOSFETs. Try the tabs first, they should be easy to reach and one of them probably has Vcore on it.
Try setting the jumpers to a known value and see if you can get Vcore at any of those points. If not, then (carefully) try the other 2 legs on each of those mosfets. Somewhere on those mosfets you should be able to find it. Just put the negative probe on an easy ground like the side of the PSU.

Once you've found the expected Vcore, then you can test undocumented settings.

c0keb0ttle wrote:

Skyscraper wrote:

I'm pretty sure removing all jumpers will get you 2.0V as the jumper scheme seems to be exactly like the one used on PC Chips M577 board but there is probably a good place to measure with a DMM to make sure.

If you post a close up of the area around the socket I'm sure someone more qualified than me can help out, I mostly poke around until I find what I want.

I'm curious what that last jumper not used in any of the settings is for... Probably something totally unrelated. 😀

Yeah, unfortunately the manual I've found for the 5AGM2 does not 100% correspond with my board. That jumper is called JP4 on my board, and the last pair of pins on it is not mentioned in the onboard descriptions.

Would there be any benefit of getting the 1.6V 400 MHz over the 2.0V 450 MHz?

It's kinda different for each generation of chip. Personally I tend to prefer the ones binned for lower voltage if the stock frequency is about the same.
Wasn't 2.0v already the highest voltage the +'s were binned for? Again, it may not matter at all, but a chip rated at its maximum voltage while these chips are binned for 550MHz at the same voltage may indicate that 2.0v 450MHz chip was binned lower due to it not running stable at 550MHz.

The +'s were released very late and manufacturing facilities may have been mature enough to run virtually all +'s at 550MHz at 2.0v. In the end, the best way is to try.

Just get both chips, I heard that German seller ran out of chips.

Finally got my K6-3+ / 400MHz 1.6V chip from Germany and could test this out.

If I remove all jumpers in the hope of making it 2.0V the computer won't POST at all.

The manual I have (which seems to be for a different revision) does not give any hint about what removing all jumpers will result in, nor does it mention what happens if you short pins 9-10 of JP4.

Fortunately, I can set the core voltage to what should be 2.1V by shorting only pins 1-2 of JP4 (also undocumented) which seems to work fine.

With 2.1V you should be able to reach 6x100 MHz. 2.1V is totally safe so the lack of 2.0V setting is no big deal but already at 2.2V these chips needs decent cooling. 😀

I know it's an old post, but just for reference ct-5agm2 uses ADP3152 voltage regulator, jp4 sets the 5 bits used by the VR to correctly output the desired voltage. ADP3152 can drive the 2 mosfets to output from 1.8V to 3.5V in 0.05 increments up to 2.1V and then 0.1V increments up to 3.5. Chaintech old website FAQ (which can be accessed trough WebArchive) about this motherboard incorrectly reports that 2.0V setting is achieved removing all 5 jumpers from jp4, which instead sets all 5 bits to 1 and forces the ADP3152 in shutdown state, so the mobo will not boot. The correct jumpers position for 2.0V is 1-2 open 3-4,5-6,7-8,9-10 closed.

NOTE: This is referred to 5agm2 rev. F150 please check your revision as stamped on the eprom chip and your VR chip markings before using the jumpers settings provided here. Other revisions might use different Voltage regulators. Revision D150 uses EL7571CM which from the datasheet could range from 1.3 to 3.5V if I'm not mistaken, 2.0V remains the same as ADP3152 (1-2 open 3-4,5-6,7-8,9-10 closed). Since they all seems to be Pentium2 compatible VRs the 2.0V setting should be always the same since on slot1 the 5 bits is supplied by the cpu and not jumpers selectable, but please check before applying.

Madowax wrote:

I know it's an old post, but just for reference ct-5agm2 uses adp3152 voltage regulator, jp4 sets the 5 bits used by the VR to correctly output the desired voltage. Adp3152 can drive the 2 mosfets to output from 1.8V to 3.5V in 0.05 increments up to 2.1V and then 0.1V increments up to 3.5. Chaintech old website FAQ (which can be accessed trough WebArchive) about this motherboard incorrectly reports that 2.0V setting is achieved removing all 5 jumpers from jp4, which instead sets all 5 bits to 1 and forces the adp3152 in shutdown state, so the mobo will not boot. The correct jumpers position for 2.0V is 1-2 open 3-4,5-6,7-8,9-10 closed.

Just for augmentation... here is the list of all voltage settings.