Question I bought a broken r9 390x for cheap has the 6 pin connector grounded

Richez

Distinguished
Jul 1, 2015
139
3
18,715
I bought a broken asus strix r9 390 for 30 bucks on ebay. The card itself is in great physical shape but doesnt boot when in pc. When there is no pcie power plugged in it boots blind, when the 8 pin pcie connector is plugged in it does the same, But if i plug just the 6-pin pcie connector in the psu trips. The fans in the pc just turn on for about half a second. Im fairly good at soldering on pcb's and i have a rework station if it is something i can replace. I could use a hand identifying the issue with the card
 
When there is no pcie power plugged in it boots blind, when the 8 pin pcie connector is plugged in it does the same, But if i plug just the 6-pin pcie connector in the psu trips.
It might be proper behavior or a bad sign, it really depends to which off the 8 pins you plug your 6 pin power - The ground wires are L shaped on the 8 pin connector and only 3 pins are +12V.

Im fairly good at soldering on pcb's and i have a rework station if it is something i can replace. I could use a hand identifying the issue with the card
In that case, I can guide you through some diagnostics... No guarantee of course. Post some close pictures of the front and back of the card when it is fully naked and cleaned. Close up on memory, VRM, the Core so it is readable. It is pretty sensitive to static, try not to kill anything on the way :)
You will need:
  • A 50MHz scope with an x10/x100 divider
  • Well calibrated multimeter
  • Laboratory power supply (0~14V, 0~25A, precisely limitable by both)
  • Open bench bootable motherboard/CPU/PSU with onboard graphics and installed OS
  • PCIe riser card is strongly recommended but you may get by without it with some Kamasutra practice
  • SPI flasher (maybe)
  • Rework station (with preheater if you have one)
  • For memory or core, you will need an appropriate BGA matrix
  • Nice to have: Thermal imaging device (FLIR, will make some tests so much easier)
We will make some measurements but I need to see the board first and draw the key test points.

In most cases, it funnels down to dead impulse generators, burned Mosfet pairs, shorted or blown capacitors, shot logical elements, shot cores or memory modules, interlayer short-circuits, etc... But you may be lucky and it is a burned fuse or bad bios.

P.S this card 99% not worth the effort in terms of time/parts vs final value with 1% a bad fuse or a bad vbios
 
Last edited:

Richez

Distinguished
Jul 1, 2015
139
3
18,715
It might be proper behavior or a bad sign, it really depends to which off the 8 pins you plug your 6 pin power - The ground wires are L shaped on the 8 pin connector and only 3 pins are +12V.


In that case, I can guide you through some diagnostics... No guarantee of course. Post some close pictures of the front and back of the card when it is fully naked and cleaned. Close up on memory, VRM, the Core so it is readable. It is pretty sensitive to static, try not to kill anything on the way :)
You will need:
  • A 50MHz scope with an x10/x100 divider
  • Well calibrated multimeter
  • Laboratory power supply (0~14V, 0~25A, precisely limitable by both)
  • Open bench bootable motherboard/CPU/PSU with onboard graphics and installed OS
  • PCIe riser card is strongly recommended but you may get by without it with some Kamasutra practice
  • SPI flasher (maybe)
  • Rework station (with preheater if you have one)
  • For memory or core, you will need an appropriate BGA matrix
  • Nice to have: Thermal imaging device (FLIR, will make some tests so much easier)
We will make some measurements but I need to see the board first and draw the key test points.

In most cases, it funnels down to dead impulse generators, burned Mosfet pairs, shorted or blown capacitors, shot logical elements, shot cores or memory modules, interlayer short-circuits, etc... But you may be lucky and it is a burned fuse or bad bios.

P.S this card 99% not worth the effort in terms of time/parts vs final value with 1% a bad fuse or a bad vbios
Hey thanks for the reply i ended up checking the contacts on the 6 pin connector and the 8 pin the 8 pin checks out fine but the positive pins on the 6 pin connector are running to ground the only pin that is clear is the sense pin. I will be able to get you some pictures tomorrow i have tested the card with the cooler off and the die doesnt get warm to the touch but all the lights power up on the card. The fans on this asus strix 390x do not spin until it reaches a certain temperature as well. I used to have a very good fluke meter but it was stolen out of my work truck along with other tools when i was in the city so i am stuck without a good meter sadly and dont have a scope. I am planning on buying a new decent meter if there is any you would recommend i do have a multimeter but isnt all that great of a meter
 
Hey thanks for the reply i ended up checking the contacts on the 6 pin connector and the 8 pin the 8 pin checks out fine but the positive pins on the 6 pin connector are running to ground the only pin that is clear is the sense pin. I will be able to get you some pictures tomorrow i have tested the card with the cooler off and the die doesnt get warm to the touch but all the lights power up on the card. The fans on this asus strix 390x do not spin until it reaches a certain temperature as well. I used to have a very good fluke meter but it was stolen out of my work truck along with other tools when i was in the city so i am stuck without a good meter sadly and dont have a scope. I am planning on buying a new decent meter if there is any you would recommend i do have a multimeter but isnt all that great of a meter
That sounds like a shot mosfet.
The fact that 12v PCIe is shorted but the fuses are not blown is a good sign. Unless they did not put any fuses... lol.
You maybe have a good core. You will still have to use a scope to check PWMs once you take the mosfets off. You also need to check the capacitors behind those mosfets.
Trying to plug the card in without running it through a lab power supply may lead to a fireshow. Get a powered reiser, and a lab PSU so you can generate voltage levels and limit the current.
As for multimeter, I use Fluke 115 or cheaper 77 but anything calibrated will do the job.

Post some pictures so I can point you in the right direction.

Edit: just noticed that you posted that you do not have a scope.... with no scope, get ready to replace PWM ICs on the dead VRM phases as you can’t verify that they do emit proper impulses. There is a chance the quartz is bad too but if things are shortened that does not seem to be the issue. Just get a good resistance meter, reiser and lab psu. For reiser, pick one with all the PCIe voltages generated on the pcb...
Edit: Here is a good link to get a general idea of how VRM is arranged.
In your case, it looks like a shorted MOSFET link, shorted capacitor, or (worst case) shorted core.
 
Last edited:
Here is some pictures of the card also looks like the person who owned it was in here before i had already known it was taken apart but not a failed repair if you could even call it that looks horrible
Here is a picture of the failed repair
That ball of solder im assuming is where something is supposed to be and nothing is there
It may be just testing. Soldering skills are not at the highest level :)

Ok, so first off, let's get accurate resistance readings from all the chokes to ground. Card should be disconnected from any cables but your probes. Get a good grounding point and note your probe resistance. We need accurate readings on the Core as it has a low resistance. Every 0.1Ω matters here.


My guess about arrangement:

 
Last edited:

Richez

Distinguished
Jul 1, 2015
139
3
18,715
It may be just testing. Soldering skills are not at the highest level :)

Ok, so first off, let's get accurate resistance readings from all the chokes to ground. Card should be disconnected from any cables but your probes. Get a good grounding point and note your probe resistance. We need accurate readings on the Core as it has a low resistance. Every 0.1Ω matters here.


My guess about arrangement:

What would you like me to start testing first?
 
I decided to start checking resistances and here i just took some pictures. Wasnt able to get a good meter but this is all i have for the time being. Also im definitely not an expert when working with stuff this small and sensitive thanks for your help
R9 390 View: https://imgur.com/gallery/CGrfTGh
Well, it is something to start with...
What is the probe resistance? When you touch each other, what does it read? Also go through all of the coils, not just this row.
 
Measured the resistance in the meter and i would subtract 3 from that last digit
Pls be more specific, 3mΩ? 30mΩ? 300mΩ?

here are the resistance readings if need be i will hold off for a better meter
R9 390 testing View: https://imgur.com/gallery/3PqtuMN
I do not like the grounding point you picked on the bezel, it is coated and might give you erratic readings, try the PCIe 8 pin connector (on the pcb).
The one coil that you had a reading 0.0Ω to ground, should come off so you can read resistance to each point on it separately to identify where the short resistance comes from, the load or the power source. Be careful not to blow up those capacitors nearby.

P.S You skipped two phases on the top left in your pictures.
 
Last edited:
Pls be more specific, 3mΩ? 30mΩ? 300mΩ?
We need to know the probe resistance too... Pls specify the number and the measure

The one coil that you had a reading 0.0Ω to ground, should come off so you can read resistance to each point on it separately to identify where the short resistance comes from, the load or the power source. Be careful not to blow up those capacitors nearby.

Let's take this coil off the board and measure two of its' contact pads individually.
jasiOIS.jpg
 

Richez

Distinguished
Jul 1, 2015
139
3
18,715
I do not have a laboratory psu nor a flir thermal camera if my school was open i would have access to that equipment as well as a better meter lol also the internal resistance of the meter is .3 ohms sorry i thought i posted it. The part i removed showed .3 ohms which should be technically zero with this meter
 
Last edited:
The short is not there i had already checked that so that means the short is isolated from that circuit correct?
Now at this point, you should be supplying low voltage (~.8v) to that left pad with safe current (~1A) from your lab psu and checking what gets heated via your thermal imaging device.

From here you are blind.
I would check those capacitors, one might be shorted out.
Then (if not shorted) the ceramic capacitors (6) next to them off and check.
Then (if not shorted) the MOSFET (3) pairs further left.
Also, the 7805 (left side on the photo) gets shot a lot on those cards, check it for shorts with a diode ring.