VOGONS


Reply 20 of 33, by criz_me

User metadata
Rank Newbie
Rank
Newbie

OK - so just to triple-check - which pins should be connected?

I will go ahead and find GND via the method suggested (minus from HDD molex pin).

Attachments

  • Screenshot.jpg
    Filename
    Screenshot.jpg
    File size
    321.37 KiB
    Views
    1059 views
    File license
    Fair use/fair dealing exception

Reply 21 of 33, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

You have the answer in front of you. Pull low with low value resistor at first for test from signal to ground. Like 10 ohms for example.

I had same situation with compaq proliant 1500 boards, I had to find correct pin to pull low to prevent fan error on startup.

Cheers,

Great Northern aka Canada.

Reply 22 of 33, by criz_me

User metadata
Rank Newbie
Rank
Newbie

Alright. As you may tell, I am not good at this, and I am trying to avoid frying my "new" CPU 😉

@snufkin You were right. I measured the pins wrong. In fact, the middle pin is connected to the case, so that is GND then. I switched GND and tach, and got a 84 point benchmark. Regarding resistor color coding, I used the multimeter to determine it has exactly 10k Ohm.

@pentiumspeed Following your advice, I added a 10 Ohm resistor (glad I ordered a whole box 😉 ) between the "real" GND now that I know it, and tach. I am running at full 134 points now! Excellent, thank you!

So, as a last step, you are suggesting to connect GND to tach, no resistor in the mix - right?

Reply 23 of 33, by maxtherabbit

User metadata
Rank l33t
Rank
l33t

So if you had the pins wrong, why not just try connecting tach from the fan directly to the tach pad without mucking around with resistors or forcing it low?

Reply 25 of 33, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

Also 3 pin fans in Dell is same way, ground the signal pin low. For this one I used loop of wire from ground pin to this.

Early OEM eg Proliant 1500, Deskpro XL, etc, computers does not use tach, and is always grounded low and if the fan fail, the signal is allowed to float high. Dell had not used tach on newer chassis (dark grey/odd purple). The ones in beige used tach. Then later on, when PWM fan technologies got introduced, Dell went to 5 pin using rearranged PWM pinout. I used homemade adapter to use standard PWM fan in Dell.

Cheers,

Great Northern aka Canada.

Reply 26 of 33, by criz_me

User metadata
Rank Newbie
Rank
Newbie

Success 😀

See attached, this is what I went with instead of graphite. I now have fan and CPU at 100% speed. Thank you all for your help! Also attached the correct pinout, in case others are looking for it.

Cheers
Christian

Attachments

  • P200MMX_PINOUT.jpg
    Filename
    P200MMX_PINOUT.jpg
    File size
    51.57 KiB
    Views
    956 views
    File license
    Fair use/fair dealing exception
  • IMG_6352.jpg
    Filename
    IMG_6352.jpg
    File size
    279.61 KiB
    Views
    996 views
    File license
    Fair use/fair dealing exception
Last edited by criz_me on 2021-04-30, 22:56. Edited 2 times in total.

Reply 27 of 33, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie

Nice, that should be ok. Only thing I'd worry about would be if the fan stalls you might not know and the CPU might overheat. Could be worth getting a small temp display and attaching the sensor to the heatsink. Something like this (I just did a quick search on ebay):

Thermometer.jpg
Filename
Thermometer.jpg
File size
33.27 KiB
Views
989 views
File license
Fair use/fair dealing exception

Reply 28 of 33, by pentiumspeed

User metadata
Rank l33t
Rank
l33t
criz_me wrote on 2021-04-30, 18:18:
Success :) […]
Show full quote

Success 😀

See attached, this is what I went with instead of graphite. I now have fan and CPU at 100% speed. Thank you all for your help! Also attached the correct pinout, in case others are looking for it.

Cheers
Christian

Please edit that graphic again to prevent misinformation. The overdrive does not use tach, which means to measure rpm (tachometer), correct term I suggest is fan failure signal, normally pulled low when working.

Cheers,

Great Northern aka Canada.

Reply 29 of 33, by Ydee

User metadata
Rank Oldbie
Rank
Oldbie

@criz_me: I'm sorry about the offtopic, but I have the same mouse as you - Primax and I can't run it in DOS 6.22. If you have it working under this OS, could you please write down what driver you're using? It would be very helpful, thank you.

Reply 32 of 33, by B24Fox

User metadata
Rank Member
Rank
Member

@criz_me, sorry to butt-in, but you ignored what could have been probably the best advice, given by maxtherabbit.. which could have had the potential of making everything actually behave like from the factory.

maxtherabbit wrote on 2021-04-30, 16:19:

So if you had the pins wrong, why not just try connecting tach from the fan directly to the tach pad without mucking around with resistors or forcing it low?

Not to mention that the CPU has different increments(steps) of fan-speed sensing and down-throttling, not just on&off (as specified in the intel POD documentation, and also proved by your testing).. and having that circuit in a 0(zero)Ohm short (between Tach and GND) , might not be a very good idea.
The fan might work perfectly fine (including Tach/Sense), if all 3 wires were just connected properly..!
That would have been the most interesting thing to benchmark! As it would have meant that basically any 3-wire 5V PWM Fan, could be a direct replacement for any OverDrive CPU Fan.

If you still have the time & energy, please make this right! I am genuinely interested (as i also have a fanless P24T P.O.D.), and i'm sure there might be others that are too 😃

Reply 33 of 33, by Caluser2000

User metadata
Rank l33t
Rank
l33t

You my want to turn the end of that lock wire up and towards the cpu so you don't a puncture a finger then bleed all over the mobo.

There's a glitch in the matrix.
A founding member of the 286 appreciation society.
Apparently 32-bit is dead and nobody likes P4s.
Of course, as always, I'm open to correction...😉