Looking for SLI bridge signal speed, signal impedance, definition of the signals and pinout of the 26 pin connector information. Only. Any help would be appreciated! Best! jippa
Today, I did the testing with an "off-the-shelf" 90mms long SLI bridge. SLI enabled, no artifacts at all . Even more eager to find SLI bridge signal speed, signal impedance, definition of the signals and pinout of the 26 pin connector information. Best! jippa
Update! A finally got my SLI bridge ready. It is 400mms long. Testing is expected to happen the forthcoming weekend. Here is a picture of the bridge: Best! jippa
Wow! Hope this works, ofc you could put the cards back to back so the cards are closer together and the heatsinks are on the outside, you could probably reach with a standard bridge then, maybe 5" or so
Hi Jippa. Some time last year I looked into PCIE ribbons. I read that high powered GPUs such as the 570 and so on needed an extra 12v. If you look at this picture you will see what I mean. After looking at your log I don't see those 12v lines fitted to your ribbons so I could hazard a guess and say that it could be that causing the interference? I can't be sure obviously because I never did get around to using them but I did read that due to the constraints of ribbon cable there could be drops in PCIE bandwidth without the extra 12v line. Something to look into any way and good luck with the project.
Just finished testing. It is a no-no. Even worse than with the previous version. I am close to crying a bit. This is what I have done to her: AlienwareAndy: Thanks for the suggestion. I have seen this picture of a riser with a molex earlier on. What I can tell you is that I am running SLI flawless with my risers and a 90mms long factory bridge. Best for now! jippa
Hmm, is that 140 mm sli bridge working? Those wires are little bit too thick and have too big resistance so that coud make problems with signals. Best solution would be to make custom flat cable. Check this one: http://www.ebay.com/itm/FLAT-FLEX-CABLE-50-PIN-500mm-LONG-0-50-mm-PITCH-/281031529699?pt=LH_DefaultDomain_0&hash=item416ec8d4e3
This one was already soldered from a very special flat ribbon cable, and it did not work out for me. I am pretty much stuck with flat cables. However, I am expecting Mr. Postman to bring me something...big. I'll keep you, guys, informed.
Is there a special reason you are using SLI? You could sell those 570s and buy a single, more powerful GPU. That keeps you from your SLI connector dilemma... and not to forget, a singel GPU, e.g. a GTX670, is going to give off less heat than your two GTX570s, that might be worth considering in a passively cooled build!
Hmm... About the only thing I can think of is repositioning the left card so that It's pcb is lying parallel to the back of the case. Same vertical orientation though. This would position the two sli ports much, much closer than they are now. This should allow you to use the standard bridge. And looking at her state; Trust me when I say that it can get much, much worse when trying to fix something. God, the last time a connector had a loose wire my rig looked like a war zone!
Another thing to try - bumping up GPU voltage, maybe it will help by increasing SLI signal strength . ALS used that trick to get a stable system when he used a PCIe riser in his modding project.
elAwesome: The special reason for me using SLI is that I want to check if it's possible. siliconfanatic: When designing the setup, I even had dummy cardboard VGAs connected to each other with a flat piece of string to check possible positioning of the cards. Believe me, this present positioning of the cards is the only possible one in this particular PC case. These risers are 90mm in width, they are very hard to bend. As for the bridge, I have tried each and every possible setup with my 140mm bridge and there was no way I could connect to two cards with it. Neither side by side, nor twisted. You could do the side by side setup, however, with a 160mm long bridge. I have contacted NVIDIA, ASUS and WIESON Co. (and some further minor companies, too) for an extra long factory bridge, without any success. N!ck: I am up to something. Should this last effort of mine turn up to be a failure, I'll start to play with voltages. In one or two days, I hope to know the answer. Best! jippa