G4 Cube VRM

[RickM]

According to the specs the LTC7819 has a maximum output voltage of 0.8 Volt to 99% of the input voltage Vin (4.5 - 40 Volt). So I consider 12 Volt not " high voltage". = 12 / 0,99*40 = about 30% of max.

 

[jasperman]

Any suggestions for the best way to test an original VRM board? I have two replacements now for my “magic smoke” VRM I posted a bit ago. Should I look for shorts on the connector voltage pins in case there is a short on the motherboard? Or any way to test a board in isolation? I’d rather not plug it in and just hope for the best. Thanks everyone.

 

[JDW]

Unfortunately, I don't have a schematic of the VRM to know which pins are what. Knowing that is very important if one wishes to test the VRM outside the Cube. But you're right, it does make good sense to test it outside the Cube if you are able to find out how to properly identify the voltage and ground INPUTS and then identify the OUTPUTs.

 

[jasperman]

Thanks for the reply JDW. I believe evanboonie's start of a VRM replacement has the pinouts for the connector but need to reinstall Kicad to check.

 

[François]

View of the DC-DC board's connector with the plug holes facing towards you. A1 is in the top left corner of the drawing.

__A1_________connector___________ | 2 0 0 3 3 3 3 0 0 0 0 5 5 0 1 | | 2 2 0 3 3 3 0 0 0 0 5 5 5 0 e | ================PCB===========B15

Legend:
0: GND
1: 12V
2: 28V
3: 3.3V
5: 5V
e: enable

 

[jasperman]

Look for a replacement for my blown C3. It looks to be a 150 uF 6.3v capacitor? Does anyone have a digikey/mouser/newark link for the proper replacement?

Tested out my blown VRM and the 5v rail is shorted. Not sure which MOSFET is bad since they are in shorted with a DMM. Hopefully it didn't nuke my motherboard or some other component.

Thank you!

 

[jasperman]

someone on Reddit answered my question:

Panasonic TPB Series 150uF 6.3V Polymer Capacitor:

https://docs.rs-online.com/4155/0900766b814dbdb1.pdf

Kemet PS/K Series Neocapacitor, jE8 = 150uF 6.3V:

https://web.archive.org/web/2018012...uctCatalog/Attachments/728/neocapacitor_e.pdf

 

[indibil]

Hi everyone, I'm sure you're already familiar with this link, but it's a cheap and easy modification.

https://forums.macrumors.com/threads/open-source-power-mac-g4-cube-vrm.2405896/

I did it on a Cube and it works very well. It's powering a Sonnet Encore ST 800 and a Radeon 7500. For European users, you can find the parts at TME, and the heatsinks I used are these:

22x8x7mm With Pad
https://es.aliexpress.com/item/1005...st_main.5.21ef194dYANSFA&gatewayAdapt=glo2esp

It's worth the modification to prevent damage to the stock VRMs.

 

[jasperman]

Thanks indibit! Do you have a surface mount soldering set up with a hot plate? Or did you manage to make these modifications with just a regular soldering iron. Evan talks about how hard it is to get the board hot enough. I was going to try with just a soldering station since the SMD parts are so large comparatively. I did order all the parts from Digikey recently to try to fix my smoked one.

 

[indibil]

Thanks indibit! Do you have a surface mount soldering set up with a hot plate? Or did you manage to make these modifications with just a regular soldering iron. Evan talks about how hard it is to get the board hot enough. I was going to try with just a soldering station since the SMD parts are so large comparatively. I did order all the parts from Digikey recently to try to fix my smoked one.

Hi, I initially wanted to do it with a preheater and a hot air soldering iron, but I didn't want to overheat the components I'd be keeping, so in the end I used a technique that might not be as professional, but one I've used many times to remove components. I set the soldering iron to about 310-320°C. I applied flux to all the pins and added leaded solder until the four legs on each side were joined. Then I grabbed the MOSFET with tweezers and quickly switched the soldering iron from one side of the MOSFET to the other while gently pulling upward until it pops out. I cleaned the pads thoroughly and finally added the new MOSFETs, also using the soldering iron.

The photo shows a different heatsink, the one I bought first, but when I saw that a taller one would fit, I opted for the one I showed you in the previous answer, but I don't have a photo.


Just a detail, from what I've read, when a Cube VRM fails, it usually takes down the motherboard, and I don't know if the processor as well. I upgraded preemptively to avoid that. I'm mentioning this because you might recover the VRM but find the motherboard isn't working. I hope this isn't your case and it works perfectly.


.