[jan-jaap]

Hi fellow Dutchie :-)

If you look in the PFC375 Datasheet, you'll see "Isolated V3 and V4 Can Be Used as Positive Or Negative Output". So yes, you can create a negative output by strapping the V3+ or V4+ to ground: https://www.artisantg.com/info/PowerOne_...asheet.pdf

[rutger]

That’s what I read but this is not for the faint of heart. Thank you for confirming, will post some nice pics when it is up and running again!

[rutger]

Thank you again jan-jaap for this great resource.

Had a few quick questions that I hope you remember after 6 years or so ;-)

- What did you use as connectors for the sense+/sense- and pwr-good terminals?
- The power switch connector has white (hot/230) and return (black). Did you connect the 230V L to white, and black to the L terminal on the new PSU?
- Do you need to load the new PSU in your experience to get accurate 5V/12V output?

thx

[rutger]

Think I got it... missed that in jan-jaap's post he mentioned molex for the sense+/sense- and power fail pins. Gutted a floppy drive power connector, voila. pwr-good I connected to J3-PWR GOOD.

Only caveat is that I'm not measuring 12V but 10.8V without load and it drops to 10V with a very small load (fan). Adjusting does not do a lot. Same for the -12V, it sits at -9.98V or so. 5V is at 5.2V.

I've got everything done, will sleep on it and see if I can put it in the machine tomorrow...

   

[robespierre]

Only caveat is that I'm not measuring 12V but 10.8V without load and it drops to 10V with a very small load (fan). Adjusting does not do a lot. Same for the -12V, it sits at -9.98V or so. 5V is at 5.2V.

That is consistent with a group regulated supply that derives its feedback from +5 V

[vishnu]

Not too interested in following the color coding convention on the wiring I see... 

[rutger]

Not too interested in following the color coding convention on the wiring I see... 

I really tried but failed. The crimp connectors only came in red, blue, yellow and black and weird sizes. Cables from the molex connector were yellow etc. and reverse with the red/black of sense+/sense-.

With the case back on later nobody will know   but I did check every power pin individually (and will do so again tomorrow).

---

Only caveat is that I'm not measuring 12V but 10.8V without load and it drops to 10V with a very small load (fan). Adjusting does not do a lot. Same for the -12V, it sits at -9.98V or so. 5V is at 5.2V.

That is consistent with a group regulated supply that derives its feedback from +5 V

Thx, that explains. I'll try to connect a load on 5V to see if that improves the readings!

[vishnu]

Thx, that explains. I'll try to connect a load on 5V to see if that improves the readings!

If you're going to load it, you might as well load it to its full rated output and let it sit there for a while to see what happens. Watch for voltage fluctuations, preferably with an oscilloscope, if you have one. Otherwise, a peak hold DVM.

[rutger]

She lives again!

Only ran for a few minutes as I need to put the case together again properly, don't want to fry any boards...

Also, since the last time this machine booted the drive has died, it does not manage to spin up anymore. I do have an image of it somewhere...

       

[jan-jaap]

You should ask member CB_HK about the 3D printed imitation he created. It basically mounts on the original drive brackets and slots into the space of the original disk, but it's an adapter for one of those SCSI-SD adapters. He even added a clicky device to imitate the sound of the original disk  

I printed one, but in Europe the SCSIknife (SCSI to SD as well) is much more affordable so I have one and it doesn't fit the mounting studs. Need to modify it. If you've got access to a 3D printer with the textured bed plate (I've got a BambuLab at work), the result is stunning