Reply 20 of 37, by hilram
Dothan Burger wrote on 2024-04-27, 16:46:I like it! How much did she cost?
US $44.64 + shipping another 10 dollars.
Dothan Burger wrote on 2024-04-27, 16:46:I like it! How much did she cost?
US $44.64 + shipping another 10 dollars.
Installation
Motherboard installed
PCI Riser
Fits snugly round the back, perfect!
PCI Graphics Card
OOPS! Looks like we hit a snag!
Anybody spot the problem?
It seems like the case orientation of the graphics card is "Upside down" if we were to route the extension cable the shortest, most logical route. As it stands, I can't secure the cards to the case.
This is one I did not see coming! Anybody out there who has faced (and overcome) the same?
Ya!! 🤪 Maybe you can find a longer extension? That sucks though...
That's what you get for trying to install those cards with the heatsinks pointing inwards.
Tevian wrote on 2024-04-28, 11:21:Ya!! 🤪 Maybe you can find a longer extension? That sucks though...
Exactly! To make matters worse, my extension cable cost almost as much as the case, and they don't accept returns!
Hey at least the cables aren't soldered, which would be annoying. They seem to be repurposed IDE cable connectors, which are readily available and you can easily DIY the longer cables - just crimp them on, e.g. in a vice.
elszgensa wrote on 2024-04-28, 11:33:Hey at least the cables aren't soldered, which would be annoying. They seem to be repurposed IDE cable connectors, which are readily available and you can easily DIY the longer cables - just crimp them on, e.g. in a vice.
I guess that is certainly a viable option. Only question is if the signal integrity will hold if I quadruple the length of the original cables.
I guess I have three options:
1) Try to replace cables with longer ones
2) Try to re-orient the PCI-card mounting bracket 180 degrees
3) Hack the case and the mounting bracket (pliers + drill)
I regard option 1 as the most complex, but in turn that would give the most satisfying result. Still there is a risk I will stop the PCI riser from working, either because the cables simply cannot be that long, or because of my own ineptitude.
Option 3 is the easiest fix, but will of course be ugly. And I will always know!
hilram wrote on 2024-04-28, 11:48:I guess that is certainly a viable option. Only question is if the signal integrity will hold if I quadruple the length of the o […]
elszgensa wrote on 2024-04-28, 11:33:Hey at least the cables aren't soldered, which would be annoying. They seem to be repurposed IDE cable connectors, which are readily available and you can easily DIY the longer cables - just crimp them on, e.g. in a vice.
I guess that is certainly a viable option. Only question is if the signal integrity will hold if I quadruple the length of the original cables.
I guess I have three options:
1) Try to replace cables with longer ones
2) Try to re-orient the PCI-card mounting bracket 180 degrees
3) Hack the case and the mounting bracket (pliers + drill)I regard option 1 as the most complex, but in turn that would give the most satisfying result. Still there is a risk I will stop the PCI riser from working, either because the cables simply cannot be that long, or because of my own ineptitude.
Option 3 is the easiest fix, but will of course be ugly. And I will always know!
I would first test the riser outside of the case and see how everything works. If signal integrity becomes an issue maybe you can attempt the shielding technique that ahmadexp used in the ITX Llama thread where he wrapped the riser cable in copper tape.
Yes, I suppose that is how I will go about it. But I have decided to order a new extension cable:
20 cm 32-bit PCI M to F connector
Ordered it from AliExpress, I hope it will work with daisy-chaining my one-to-two PCI extender. If not, I will have to use this one on its own, and sacrifice a second PCI-card.
That means giving up on EITHER
1) Positional audio, and instead rely on the Southbridge on-board sound instead OR
2) A Vooodoo 2-card for native Glide support.
Will also have to probably 3D-print a custom support for some kind of PCI-holder. My job has a couple of 3D-printers we can use for personal use if we ask nicely.
hilram wrote on 2024-04-28, 11:48:2) Try to re-orient the PCI-card mounting bracket 180 degrees
A 3D-printed bracket would probably be the most elegant solution for mounting the bracket upside-down
edited: Just saw you mentioned 3D printing in your most recent post. 😅
Of course, I could always 3D-print new brackets for the cards! Why didn't I think of that? Anyway, I have ordered that 20 cm extension cable, will try that first. Will keep you posted how things are going. PSU will arrive in a few days, will power on and connect it then.
hilram wrote on 2024-04-29, 11:31:Of course, I could always 3D-print new brackets for the cards! Why didn't I think of that? Anyway, I have ordered that 20 cm extension cable, will try that first. Will keep you posted how things are going. PSU will arrive in a few days, will power on and connect it then.
So how did this turn out? Can two devices share the same PCI port and not cause issues?
I bought that DELL P1917S and it turns out it's huge piece of crap. The screen doesn't look like anything like an IPS, a TN panel would look better in my opinion. Couldn't get any refresh other than 60hz to work. I took a chance and I wouldn't recommend it.
I'd be interested to see how well that riser works, I've got a similar system that I want to move into one of these sandwich cases so full height cards can be used with it.
Sorry for the abscence! Work was crazy up until the start of the summer holiday, but now I am back home, and free.
YES everything works! Even with the PCI Riser and the PCI Extension cable Daisy-chained! I am very happy with that! Tomorrow I will plough on with attempting to install Windows 98 on the machine. So stay tuned!
Dothan Burger wrote on 2024-06-03, 03:25:I bought that DELL P1917S and it turns out it's huge piece of crap. The screen doesn't look like anything like an IPS, a TN panel would look better in my opinion. Couldn't get any refresh other than 60hz to work. I took a chance and I wouldn't recommend it.
Oh My! Really? Thanks for the warning, I will reconcider then... I mean, even if you use the analog VGA input, still only 60 Hz?
hilram wrote on 2024-07-09, 15:57:Dothan Burger wrote on 2024-06-03, 03:25:I bought that DELL P1917S and it turns out it's huge piece of crap. The screen doesn't look like anything like an IPS, a TN panel would look better in my opinion. Couldn't get any refresh other than 60hz to work. I took a chance and I wouldn't recommend it.
Oh My! Really? Thanks for the warning, I will reconcider then... I mean, even if you use the analog VGA input, still only 60 Hz?
Yes even the VGA could only handle 60hz from my 7950gt in 98 and XP. I might have to bring it out again for a round 2 of testing as Dells website says I'm wrong. The big issue for me was that it looked so bad compared to my CRT.
It could be that it is not communicating its VESA modes properly to your graphics card. It might need actual monitor drivers for the right modes to be available. With "Generic PnP monitor" driver it might never give you anything other than 60 Hz. See if you can download an appropriate monitor driver from Dell, and then force the driver in Device Manager.
hilram wrote on 2024-07-10, 13:46:It could be that it is not communicating its VESA modes properly to your graphics card. It might need actual monitor drivers for the right modes to be available. With "Generic PnP monitor" driver it might never give you anything other than 60 Hz. See if you can download an appropriate monitor driver from Dell, and then force the driver in Device Manager.
75hz does work, what was confusing me is that the info section of the OSD says 1280x1024 60hz max but to the left it shows 75hz is working. Doesn't seem to make it feel much faster, just moving the window around feels terrible even with fast pixel response enabled. DOS gaming is working so I might just bring it to the next LAN party. I certainly won't be using it day to day.
Viewing angles seem poor compared to my other IPS panels.