Sapphire HD 7770 possible short

[manthos1969]

Hi
My 7770 oc edition wont let my pc to start when 6-pin power cable plugged on.
http://www.sapphiretech.com/productd...nload&lang=eng
The cpu and case fans spin only for 1/4 of a second and the system wont start,I'm using a 430w psu at this moment,I know there is a little bit of undervoltage but I don't push the gpu with extreme games e.t.c.
I tried my 7770 to another pc system (550w psu) and i have the same symptoms.
My old PCIe powered gpu works fine on the actual system.
I've been using the 7770 on the specific system about 6 months for light gaming without any problems before.I do not have visual markings, burning board or damaged capacitors or mosfet's and this was the point I started taking measuremets on capasitors, mosfets e.t.c
Coming up to a similar situation to mine on this forum was the reason of starting taking measurements on some components on my board.
I've been measuring with polymeter on 200 except of 3 times I set it on 20k.
Can someone help me find the bad mosfet(I think ) or any other problematic component on my gpu?
Sorry for my poor english. If you consider helping please use simpler English.
I appreciate taking your time to read this and I'm looking forward to your replies.

[manthos1969]

Re: Sapphire HD 7770 possible short

I forgot to upload my power connector measurements

[drag]

Re: Sapphire HD 7770 possible short

I have r9 270x that explode yesterday. It has similar problem that PC won't boot while graphics card is plugged in. I tried Seasonic 500w and nothing happens. Then i tried Nexus
1000w and 270x one little mosfet explode...

[manthos1969]

Re: Sapphire HD 7770 possible short

I have r9 270x that explode yesterday. It has similar problem that PC won't boot while graphics card is plugged in. I tried Seasonic 500w and nothing happens. Then i tried Nexus
1000w and 270x one little mosfet explode...

Hi,your gpu have a dangerous short if an mosfet exloded after the psu's power safe mode.
First take a look the Nexus 1000w psu health because an mosfet explosion happens on gpu and the psu's power safe mode won't work properly to stop the shorted circuit.
My opinion is to give it (gpu) to an technician because in your case the gpu is dangerous for your personaly health and for other pc components.
Sorry for my poor english.

[momaka]

Re: Sapphire HD 7770 possible short

Hi manthos1969,
welcome to the forum!

It's very nice that you posted the resistance readings on your pictures - thank you!
Basically, we will start by following the short-circuit resistance of 18.4 Ohms you got on the power connector.
In this picture...
https://www.badcaps.net/forum/attach...3&d=1461180102
... two "R25" inductors show 17.3 Ohms, and the capacitors close to them show 18.1 Ohms. That is very close to the 18.4 Ohms reading of the power connector. So I suspect the problem is with the MOSFETs around those "R25" inductors. That would most likely be Q623, Q624, Q613, Q614, and two more between those. Remove them from the board (one by one) and see if/when 18.4 Ohms short-circuit on power connector disappears. When it does, let us know. Then we can proceed and check the MOSFETs. Most likely, only one is bad.

I hope you are able to understand my post . If not, feel free to ask and I will try to explain in a different way. Your English is not bad at all - I was actually able to understand you 100% .

[manthos1969]

Re: Sapphire HD 7770 possible short

Hi thank you for replying to my problem.
I'm out of city for a few days and I can't do your suggestions right now.
Actually I have already started trying some fixes on my own (it was a wrong assumption of mine,nothing changed) and I removed the MOSFET Q703 (I damaged it )from the board because the source was shorted to the 12v rail and I was not getting any resistance between the gate and 12v rail when I charged the source,so I ordered a new MOSFET and it's on the way.
I'm looking forward your suggestions when I'm back home and I will post the results.
Please be patient because I'm just a cook,but I have good experience on iron soldering and hot air works.
I started the thread to get skilled on these types of workarounds mostly as a hobby and less for my GPU awakening,so, in the end, I could be able to find out how to search for damages. This is the most important thing to me.

[manthos1969]

Re: Sapphire HD 7770 possible short

Hi,I'm back again.
I started to remove the mosfet from Q621 because the resistance values(source & gate) were similar to the resistance from 12V rail.
After i got Q621 removed, the 12V rail to ground resistance changed to high values.
I gave the gpu a try powering it with an old pc,the pc starts now but enters power safe mode in 3 seconds,I think that is progress .
I have 2 suspicious mosfets (Q601,Q611)for the next step because the resistance values looked wrong to my old nooby mind.
Actually the mosfet (SO-8 FL type) removal was very painfull for my experience.

Q623 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q624 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q613 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q614 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q603 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q604 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q611 source to ground 23.4Ω ,gate to ground 23.4Ω,drain to ground high values
Q601 source to ground 23.4Ω ,gate to ground 23.4Ω,drain to ground 01.1Ω

Please advise if you see something more about this case for the next step.
I have changed the values with the new ones on the photo.

[momaka]

Re: Sapphire HD 7770 possible short

Sorry for the delay in response. I do this for a hobby as well, so I am not the fastest with responding either (so no problem if you don't reply right away - I never expect it anyways)

Actually I have already started trying some fixes on my own (it was a wrong assumption of mine,nothing changed) and I removed the MOSFET Q703 (I damaged it )from the board because the source was shorted to the 12v rail and I was not getting any resistance between the gate and 12v rail when I charged the source,so I ordered a new MOSFET and it's on the way.

You should *never* get low resistance between Gate and Source of a MOSFET. Same goes between Gate and Drain.

Now, you cannot charge the Gate in circuit to test the MOSFET. The best you can do is check resistance - G to S, G to D, and D to S (or S to D). Should not get a short-circuit on those. If you do, remove MOSFET and test out of circuit again.

As for the damaged MOSFET - do NOT run the video card without it.

Hi,I'm back again.
I started to remove the mosfet from Q621 because the resistance values(source & gate) were similar to the resistance from 12V rail.
After i got Q621 removed, the 12V rail to ground resistance changed to high values.

Good!

I gave the gpu a try powering it with an old pc,the pc starts now but enters power safe mode in 3 seconds,I think that is progress .

Do NOT run the video card with removed MOSFETs. You can kill the RAM and GPU core when you do that. And that is something we can't fix once done.

Q623 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q624 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q613 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q614 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q603 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω
Q604 source to ground 01.1Ω ,gate to ground 01.1Ω,drain to ground 23,4Ω

I think you measured Gate on the wrong pin. Should NOT be 1.1Ω to ground. But other than that, these MOSFET measurements appear correct. All of these MOSFETs above are "lower" type MOSFETs. That is, their Source is connected to ground. And their Drain is connected to GPU V_core supply as well as those "R25" inductors. GPU V_core supply is what provides power to the GPU chip, by the way. The 23,4Ω resistance you see is of the GPU chip itself.

Q611 source to ground 23.4Ω ,gate to ground 23.4Ω,drain to ground high values

That appears to be an "upper" MOSFET. Source is likely tied to GPU V_core (hence why you get 23,4Ω on it). And Drain should be tied to 12V rail (possibly on power connector), hence the high resistance measurement.

Q601 source to ground 23.4Ω ,gate to ground 23.4Ω,drain to ground 01.1Ω

Might be an "upper" MOSFET like Q611. But check Drain to ground measurement again. I don't think it should be 1.1Ω

Please advise if you see something more about this case for the next step.

Well, looks like you found one bad MOSFET (Q621). Replace it (and all other missing MOSFETs), then test the video card again. You can use MOSFETs in TO-252 case too, I think. Should be common to find on other dead motherboards and Xbox 360 consoles if you want "free" replacements .

Please be patient because I'm just a cook

Wow, nice!

[manthos1969]

Re: Sapphire HD 7770 possible short

Sorry for the delay in response.

Hey,dont be sorry,I'm just a noob with a nice puzzle of physics in front of me and I dont know yet how to recognize the puzzle's pieces,so that when asking for help,if you or some other guy wants to help it's welcome, when and how doesn't really matter,it's a hobby for me and I don't have time limitations,only the pleasure of learning.

Do NOT run the video card with removed MOSFETs. You can kill the RAM and GPU core when you do that. And that is something we can't fix once done.

I think you measured Gate on the wrong pin. Should NOT be 1.1Ω to ground.

You caught me ,an old man's eyes make mistakes.
Here are the right measures:
Q623 source to ground 01.1Ω ,gate to ground high values,drain to ground 23,4Ω
Q624 source to ground 01.1Ω ,gate to ground high values,drain to ground 23,4Ω
Q613 source to ground 01.1Ω ,gate to ground high values,drain to ground 23,4Ω
Q614 source to ground 01.1Ω ,gate to ground high values,drain to ground 23,4Ω
Q603 source to ground 01.1Ω ,gate to ground high values,drain to ground 23,4Ω
Q604 source to ground 01.1Ω ,gate to ground high values,drain to ground 23,4Ω
Q611 source to ground 23.4Ω ,gate to ground high values,drain to ground high values
Q601 source to ground 23.4Ω ,gate to ground high values,drain to ground 01.1Ω

Might be an "upper" MOSFET like Q611. But check Drain to ground measurement again. I don't think it should be 1.1Ω

Stays to 1.1Ω ,I spent an hour only for that but I dont know how to explain,the measure varies from high values to 1.1Ω like measuring a capacitor together with it.when I get 1.1Ω from drain to ground same second I get high values from drain to 12V rail.

You can use MOSFETs in TO-252 case too, I think. Should be common to find on other dead motherboards and Xbox 360 consoles if you want "free" replacements.

I have a bunch of motherboards left,but they shouldn't have exactly the same (type and number)MOSFET to run the gpu,so why all those different types of MOFSETs would in the first place?

I will post when I place the new MOSFETs and take new measures,thanks.

[momaka]

Re: Sapphire HD 7770 possible short

Q601 source to ground 23.4Ω ,gate to ground high values,drain to ground 01.1Ω

Stays to 1.1Ω ,I spent an hour only for that but I dont know how to explain,the measure varies from high values to 1.1Ω like measuring a capacitor together with it.when I get 1.1Ω from drain to ground same second I get high values from drain to 12V rail.

Hmm...
Okay, this may not be an upper MOSFET then. But let's leave it like that for now.

First, I would say check Q621, since you have removed it from the video card now. See if Q621 MOSFET shows low resistance between Sound and Drain. Also check for D-G and S-G resistance - when removed from the circuit, the resistance on the Gate should be infinite.

If bad, replace it with a new MOSFET, then test the video card.

I have a bunch of motherboards left,but they shouldn't have exactly the same (type and number)MOSFET to run the gpu,so why all those different types of MOFSETs would in the first place?

Yes, the MOSFETs on those may be quite different, especially if they are older motherboards (probably TO-263 or TO-252 - Google those to see what they look like). But often you can still re-use these MOSFETs as a replacement. All you really have to do is find the datasheet for the old and new MOSFET and compare the characteristics. Mainly, the Ampere rating of the replacement MOSFET should be the same or higher than the original. Same goes for voltage. And the Rise and Fall time should be somewhat close.

But otherwise, it is possible to do. The result may not look beautiful, but as long as it works, why not

I will post when I place the new MOSFETs and take new measures,thanks.

Sounds good.

[AMD_Ripper]

Re: Sapphire HD 7770 possible short

I'm curious if you fixed it ?

I have a shorted 7950 but can't post about it.

[manthos1969]

Re: Sapphire HD 7770 possible short

I'm curious if you fixed it ?

I have a shorted 7950 but can't post about it.

I'm still waiting the rite parts from China.

[momaka]

Re: Sapphire HD 7770 possible short

I'm still waiting the rite parts from China.

It's getting close to a month now, if not longer. I think you need to contact seller and let them know.

I have a shorted 7950 but can't post about it.

You can.
Then only time you cannot make a new thread is when you have 0 (zero) posts - i.e. when you just have registered. This is why new folks are usually asked to make a post here:
https://www.badcaps.net/forum/newrep...reply&p=668248

Keeps spam bots in check .