SOLVED: EK FC on HD4870 X2 no POST

[hellcamino]

I had this same issue with one of my 4870X2's using the Ek fc blocks, one card would boot and one wouldn't. After saying several bad words I started thinking about a safe way to test the card to cooler short. The best way I found was to use a multimeter and test for continuity between the block and power circuit to the card ( back solder points behind PCIe plugs), I initially tested mine and found about 2.2 ohms from block to ground and 5 ohms from power to block on one card and 13k ohms from power ckt to block on the one that worked.

I went one by one through the screws trying to find a single culprit for the short and while I would get fluctuations in resistance continuity was still present, I removed all the short screws and put fiber washers under them ( I didn't get washers with either of my kits). I finally got it solved by removing the scew that goes through the card to brace the I/O plate, once it was removed continuity went to 13k from power ckt to block and it boots.

[Dangals]

great to hear you got them up and running CyberDruid

Just out of curiosity was it the washers that was the issue or one of the screws being overtightened?

[CyberDruid]

Washers. I barely tightened the screw in the first place since if you do more than that the card starts to look like snakes on a plane. My usual process is to install all the screws then crisscross them until quite snug then to let off on the perimeter screws until the card is straight again. I usually have the ram/GPU area tight..but this time I loosened each screwand just barely let it tighten since the TIM is doing the job...

Getting great temps and everything is working as I expect it should.

Thanks again.

[Dangals]

Thanks - that looks awesome

I wish I had the funds to buy a second

[CyberDruid]

Slots are real close on the P5E3...going to have to get jiggy with some BP fittings.

[NaeKuh]

CD how about using an aquatuning or koolance bridge in between those cards?

Koolance has 1 space 2 space and 3 space hard mount fittings.

http://www.koolance.com/water-coolin...product_id=400





and yes you can mount instu, the fitting is rotary.

I use the larger ones: <here i hard piped 2 DDC-2>


Here i used them for my HD3870's:

[IanY]

Naekuh,

Do you know the ID of the passthrough? Looks kind of skinny to me.

[NaeKuh]

Naekuh,

Do you know the ID of the passthrough? Looks kind of skinny to me.

most likely 1/4 like the G1/4 thread.

Its a straight passthough iany.

[Giannis86]

just installed mine too using the washers from the package and everything works fine.
personally i didnt really have any problems with it
performs as great as it looks too

ps: putting a block on this card has increased my case temps a LOT (its a little messy in there maybe thats the reason but who cares? )

[RLM]

I'm trying to solve a problem with mine. I've installed the block three times now.

First off I did get the paper washers and also instructions I believe to not over tighten.

The score is that I'm getting lockups. On the second time of removing the block, I was intending to take the card back. I stuck the stock air cooling back on, and put the card back in the machine just to check some stuff and the lockups had gone.

There's so many parameters that I don't know if it's drivers or not. Anyway have stuck the block back on and I'm getting lockups again. Can't make it through 3dmark06 at 750/900. The key thing is that I am hearing that high pitch sound. More intense when 1.25v are going through the card.

I wonder if that's due to the screws being over tightened? I'm going to try and loosen them up a bit. Fun'n'games.

RLM

[CyberDruid]

I had this same issue with one of my 4870X2's using the Ek fc blocks, one card would boot and one wouldn't. After saying several bad words I started thinking about a safe way to test the card to cooler short. The best way I found was to use a multimeter and test for continuity between the block and power circuit to the card ( back solder points behind PCIe plugs), I initially tested mine and found about 2.2 ohms from block to ground and 5 ohms from power to block on one card and 13k ohms from power ckt to block on the one that worked.

I went one by one through the screws trying to find a single culprit for the short and while I would get fluctuations in resistance continuity was still present, I removed all the short screws and put fiber washers under them ( I didn't get washers with either of my kits). I finally got it solved by removing the scew that goes through the card to brace the I/O plate, once it was removed continuity went to 13k from power ckt to block and it boots.

I'm trying to solve a problem with mine. I've installed the block three times now.

First off I did get the paper washers and also instructions I believe to not over tighten.

The score is that I'm getting lockups. On the second time of removing the block, I was intending to take the card back. I stuck the stock air cooling back on, and put the card back in the machine just to check some stuff and the lockups had gone.

There's so many parameters that I don't know if it's drivers or not. Anyway have stuck the block back on and I'm getting lockups again. Can't make it through 3dmark06 at 750/900. The key thing is that I am hearing that high pitch sound. More intense when 1.25v are going through the card.

I wonder if that's due to the screws being over tightened? I'm going to try and loosen them up a bit. Fun'n'games.

RLM

It's not fun at all and shows that there is a design flaw in the block. Despite it's great looks and great cooling if it shorts out your $550 card you are going to be PISSED.

Try painting the screw heads and part of the thread with nailpolish a few coats to make a dielectric barrier. I think that if the edge of the screw touches the ring of solder on the hole it's bad.

And try taking out that screw highlighted in the quote above yours.

[CyberDruid]

CD how about using an aquatuning or koolance bridge in between those cards?

Koolance has 1 space 2 space and 3 space hard mount fittings.

and yes you can mount instu, the fitting is rotary.

I use the larger ones: <here i hard piped 2 DDC-2>


Here i used them for my HD3870's:

Hey Pardner I will look into that. It's clearly the cleanest solution. I am using 3/8ID-1/2OD Feser tubing in this build so once I get all the fittings I'll be set.

I was thinking of using Koolance for compression fittings since they have a larger barb and DD Rotaries since I don't need to advertise the MFR name on every single part I install...unless they pay me.

Hint hint.

[RLM]

Try painting the screw heads and part of the thread with nailpolish a few coats to make a dielectric barrier. I think that if the edge of the screw touches the ring of solder on the hole it's bad.

That's a bloody good idea.

I'd like to know, is anyone else hearing a high pitch sound coming from their card? I'm guessing you wouldn't notice with the stock fan on. I tend to get it during loading in between the 3dmark06 benches. I know from experience that sound is ominous. lol.

RLM

[CyberDruid]

A cap? There's like one capacitor

EDIT

Doh

There's a mess of caps...I just was not looking at the other side of my card lol.

I betcha that's the sound you are hearing. There are no coils that I can see...

[NaeKuh]

Hey Pardner I will look into that. It's clearly the cleanest solution. I am using 3/8ID-1/2OD Feser tubing in this build so once I get all the fittings I'll be set.

I was thinking of using Koolance for compression fittings since they have a larger barb and DD Rotaries since I don't need to advertise the MFR name on every single part I install...unless they pay me.

Hint hint.

ROFL..

sandpaper and scratch it off, or paint it.



The passthough is a 1 fit, meaning it fits directly on the G1/4.

And the cards rotate so installation isnt as bad since your dropping 2 cards in at the same time and there locked like that.

CD Also Quick Discconects on the GPU line. I dont think your client will care about the reduction flow in his gpu line, expecially when he wont see a noticable hit on temps, and will make YOUR life easier when you need to service it.

ps: putting a block on this card has increased my case temps a LOT (its a little messy in there maybe thats the reason but who cares? )

that makes no sense because the full card block should take brunt heat of the card and move it to your radiator. If your case temps went up, that is confusing. Where does your rad exhaust to?

[RLM]

Well you live and learn. The over tightening was the cause of my problems. Running through 3dmark just fine now. It does say on the instructions to use a small amount of force when tightening the screws. Doesn't take much to over do it.

Your

the card starts to look like snakes on a plane

certainly came to mind when I checked it.

Thanks.

RLM

[CyberDruid]

SWEET. Glad this issue is easily resolved.

Still sucks to have an issue at all.

[RickCain]

Those blocks look hella sexy!

Glad you got the problem fixed.

[CyberDruid]

On another Forum where I started a thread like this a guy said he used some electrical tape where he thought the card was () inappropriately the touching the block and tightened it down as much as he thought it needed.Worked for him. More than one way to fix this issue

[twwen2]

Hmm, i have the HD4870 EK full cover block, it didn't come with any washers... Let's hope it works when i turn the system on sometime next week...

[NaeKuh]

Hmm, i have the HD4870 EK full cover block, it didn't come with any washers... Let's hope it works when i turn the system on sometime next week...

STOP!

go out to the hardware store and get a pack of nylon washers.

Use Washers and then turn your system on live. Dont mickey mouse the end twwen.

[IanY]

They really should pay attention and make follow up amends. Otherwise, its not being very responsible.

Rushing to market isn't a sustainable strategy over the long run.

[Utnorris]

Naekuh,

Do you know the ID of the passthrough? Looks kind of skinny to me.

It's 3/8" if I remember correctly. I had someone else ask that before and since I had one laying around I measured it. I am 99% sure it is 3/8" ID.

Utnorris

[IanY]

It's 3/8" if I remember correctly. I had someone else ask that before and since I had one laying around I measured it. I am 99% sure it is 3/8" ID.

Utnorris

Thank you very much, UTNorris

[RickCain]

Rushing to market isn't a sustainable strategy over the long run.

+1

Way too much of this going on in the watercooling market lately.