I am going to guess that the host slot for the riser can't provide enough current to drive multiple cards, or that the riser is defective.

Thank you. Mmmh...currently trying another motherboard, and it seems like the same issues don't happen. Could different socket 3 motherboards send varying amount of current per slot? I was kinda expecting some kind of standard.
continuing my tests.

Care to share, some photos of the mainboard and riser you are having problems with ? Thank you.

Sure, motherboard is QDI V4P895P3 v4.0

ISA riser attached as well.

ISA riser card ends up in the top ISA/VLB slot.

I have since switched the previous 486-DX25 that was in there for a 486-DX33 (and changed the necessary jumpers).
I have less issues than before, which is good, but now the PicoGUS only gives me sound but music in Epic Pinball and Jazz Jackrabbit in GUS mode, sound and music both fine in War2 setup, no sound or music working in Duke3D setup... bleh.

FastTracker does not do a General Fault Protection anymore on startup, but doesn't play, probably because of the GUS problem above I suppose?

Still get the occasional B&W on reboot...

Your main issue is likely the riser itself. It has no ground connection to its screws etc. either, and solid ground connection is absolutely necessary for proper operation. I do see some capacitors which may have to be sized larger than they already are.

Other issue is that a board with all slots on it already isn't really meant to have another extension to already extended bus, signal integrity issues can happen but I think in this current case it is the riser itself to blame, and in particular its lacking ground connectivity.

I had the same problem with my AT clone. You have to solder electrolytic caps on the riser, 47uF 16V. You have to find the ground pin with a multimeter.

Thanks, a photo of the riser from the other side would be nice as well.
From this point of view it looks like it has some spots were capacitors can or are located on the other side.

As a test, I ask to stick the riser in the complete other end of the motherboard from where it is now, the bottom ISA slot, where the bios is, Yes, I know that it won't fit in the case in any way like this, that's why it's a test, and run your programs again for signs of instability & crashes.

Agree on ground and caps......I have a P.Bell ISA riser from a 486, it has 8 x 10uF/16v caps and looks similar to this (except mine labeled is Pk2109-01):
https://twitter.com/Foone/status/1369101935396089862/photo/1

Very interesting, now I didn't expect that! I don't see any capacitor anywhere on the riser...guess the bottom-left and bottom-right holes/pads look like there were some that were removed...?
Here are two new pictures of both sides, had to use the flash because of the low light, but hopefully it shows what's important well enough.

I unfortunately don't have the capacitors at hand (closest I have would be something like...10K uF/10V 😁) but happy to order some online to try and fix this.

Would someone be kind enough (with a drawing on my pictures or something maybe?) to confirm the capacitors should be placed, what kind of capacitors, and how proper grounding could be achieved? Of course I can measure/test anything necessary.

Thanks a lot!

weedeewee wrote on 2024-02-04, 16:51:

As a test, I ask to stick the riser in the complete other end of the motherboard from where it is now, the bottom ISA slot, where the bios is, Yes, I know that it won't fit in the case in any way like this, that's why it's a test, and run your programs again for signs of instability & crashes.

I did test other slots with no notable difference...

Found an eBay listing for a very similar riser in a somewhat big stock and a lot of sales, no capacitors or anything either..it's kinda surprising to me that they were/are popular if they were so fundamentally flawed, am I right to wonder?

Boohyaka wrote on 2024-02-04, 18:58:

Found an eBay listing for a very similar riser in a somewhat big stock and a lot of sales, no capacitors or anything either..it's kinda surprising to me that they were/are popular if they were so fundamentally flawed, am I right to wonder?

s-l1600.jpg
s-l1600-2.jpg

You're right, they didn't solder any caps on them to reduce the cost.

Here is my ISA riser, it has capacitors, they are two different types, although quite similar in appearance.
I hope it could be useful to you, if you have any doubts, just ask.

I will do a trace to check which V.lines are capped on mine when I get home (am sure +5v and +12v but could be others). Another thing I notice is lack of good ground plane compared to mine and PC@LIVE's. Both of ours (though different) and have huge ground plains on one or both sides where yours does not. That does help with interference....another factor in "poor" design along with the lack of cap's IMHO.

Ok total of 9 caps (missed one) 6 of the 10uF are on the +5v (three at bottom and three at top), one each 10uF on -5, +12 and -12.

6 isa slots + riser is a lot of reflections going around ISA bus. It might turn out to be impossible with this board in the end. Looking at the bus with a scope would tell more.

>it's kinda surprising to me that they were/are popular if they were so fundamentally flawed,

they work absolutely fine in dedicated motherboards (LPX?)

I have a pizza-box style case with a riser card that looks very similar to the one on this post, and it works without problems for me with regular 8-slot motherboards. And the case is definitely intended for Baby AT form factor boards, not LPX.

My riser card has 3 16-bit slots on one side and 2 8-bit slots on the other. I've stuffed it full of cards (VGA, Sound, MIDI, HDD & I/O controller) and it doesn't miss a beat, never had to add any caps or extra grounding... only used it with 386 motherboards though, not sure if there would be a difference with a 486.

Maybe the OP is running an overclocked ISA bus or something that in combination with the riser card can end up affecting signal integrity?

EDIT: just checked my riser card, it does have large ground planes on both sides around the slots, so maybe this makes a difference as others have mentioned. It is labeled "CTK-007".

Good points. Not sure if the riser in question would work ok even in a LPX like my P.Bell 486, with the lack of gnd shield and caps...
Here is my bad deduction on the board in question pic1, I see no simple way to add caps. And to compare pic 2 and 3 are of mine. Big diff between them....just an observation..

Interesting input, thank you all. My goal was in the end to build a 386 in this case, waiting on a new motherboard. I guess the option would be to find an alternative riser with a proper ground plane, if all else fails... it seems to be a pretty standardized format.

The case is absolutely beautiful, would be a pity to not be able to use it 🙁