• Hello, Guest! Welcome back, and be sure to check out this post for more info about the recent service interruption and migration.

x86 card revival thread

Nathan_A

Well-known member
Well, I stupidly missed out on an OrangePC 660 that I desperately wanted on eBay, but I did end up otherwise adding to my collection with an Apple 586 7" PCI card and an Apple Pentium (100 Mhz I think) 12" PCI card.

That brings me to:
  • Apple Houdini II 6100 PDS card (Intel 486 DX2 66)
  • Apple 586 PC Compatibility PCI card (Cyrix 5x86 100)
  • Apple Pentium PC Compatibility PCI card (Intel Pentium 100)
  • Reply DOS on Mac 7100 PDS card (Intel Pentium Overdrive 83)
  • Reply DOS on Mac 8100 PDS card (AMD Am5x86 133)
  • OrangePC 290 Nubus card (Intel 486 DX4 100)
  • OrangePC 620 PCI card (AMD K6-III 400)
I'd still love to find an OrangePC 340, 550, and 660 someday, but still not too shabby of a collection.

One thing I'd absolutely love to find is the Serial + Parallel port PCI companion card for the Apple PCI cards. For the same reason I want to find an OrangePC 340 or 550. I have some old DOS & Windows software that require HASP hardware keys for licensing.
 
Last edited:

Byrd

Well-known member
You’ve done well! One of those upgrades that didn’t register as cool about five years ago, now sought after.
 

Nathan_A

Well-known member
You’ve done well! One of those upgrades that didn’t register as cool about five years ago, now sought after.
Even when I was a kid I loved these things. I was a dyed in the wool Mac user, but I had so many friends that played PC games that either took ages to be ported or never made their way to the Mac back in the day, and so nearly every Mac I ever owned as a teenager and through my early 20's had some kind of PC Compatibility card in it.

It wasn't really until Bungie's Marathon and Myth series that I started to care less and less about blending my Mac and PC worlds together. I still think Myth TFL and II are two of the best and most innovative games every made, and I have dreamed of finding continuations of the genre "real-time tactics", but alas. By that time I was buying dedicated PC workstations for my CAD/CAM work.
 

NJRoadfan

Well-known member
I always had a thing for "PC on a card". My high school had a whole bunch of 7600s with OrangePCi cards (I think it was a 550) in them which I found to be pretty neat and expensive at the time.

Also, there was another Apple II card called the "Rana PC". Not too much out there for it outside of some grainy video stills of the card.
 

jessenator

Well-known member
I know jeremywork already posted these in the initial revival post, but thought I'd share the full-res versions. They are quite large, so I'll do my best to trick xenforo into not actually embedding them in the post.

This is the Apple 12" PCI Pentium 100 card, which I scanned, so the DoF might be wonky on the front scan:
Front of the card here.
Back of the card here.

And here's the back side of the Pentium 100 and the Pentium 166 cards compared:
Comparison image here.

I think I didn't want to scan the 166, because by that time I had already put the heatsinks on the Opti and LSI chips (because they need them IMO) and didn't want to get new adhesive pads for them.

I think somewhere we speculated that it had something to do with the LPX-40 PSU's beefier "auxiliary" voltage supply for it "killing" 7200+ PSUs getting overdrawn? Also, don't know if this is the correct place to post that, but I wonder what each rail's rating is on the two supplies. A vanilla LPX typically doesn't have these other two board power connectors, just a few 4-pin Molex. IBM's had a single aux power connector, but the Apple/Motorola had two.
This wasn't well-written 😖

But the theory I had started to formulate doesn't really hold much water at all... I speculated that since the LPX-derivative-PSU's output has more amperage on the 3.3V rails, using a card designed to theoretically take advantage of it it would cause issues with a 7x00 PCI PowerMac's PSU...

7200's PSU
3.3V = 10A max
5V = 20A max
Max draw combined = 25A

4400's PSU
3.3V = 20A
5V = 20A
Max draw combined = 100W (IDK why this one is displayed as Watts)

IDK if that differential is enough to cause problems, or why the card would need to draw that much...I mean it has 3.3V RAM, but RAM does not operate on the same level as a PC on a card... guess we won't know until someone finds the Cyrix card. If anyone does have one, shoot me a PM : P
 

Nathan_A

Well-known member
I've got a couple of the Apple PCI card cables on the way. Someone around here was making a breakout box PCB for the PDS variant cards, but I can't find the thread anymore despite my having posted to it.

Whoever that was, I've got some cards and cables I could lend to the effort of making PCI card versions of the box.
 

aapl2c

Member
I found this one, which is apparently for the 6100 varients:


I have a 12" PCI P166 card, but no cables so am looking into options, whether it's making a cable for the external connector (which is NOT the same cable as used on say the 6100 DOS card) or one that'd connect internally via a GIMO connector.
 

jessenator

Well-known member
making a cable for the external connector
You might've already seen this, but here's a DIY write up for both: https://68kmla.org/bb/index.php?thr...and-pci-590-0905-590-2104-and-590-4539.34150/

The nice thing is that the joystick port is separate on the PCI cards, so it's really just splitting video connections. I guess if you cared about a joystick, anyway :D

connect internally via a GIMO connector
This is a superb option if you've got the breakout board for it. It was much nicer than the bulky Y cable. Sadly, I could never get audio over the one I used to have. Still had to use the CD-ROM analog audio cables to pass through via the card. The alternate name Apple used, in the one or two locations it actually appears at all, made it sound like it connects both video and audio, but just doesn't :/ I wonder if that's simply because you can only monitor CD or the analog audio input in the Sound/Monitors & Sound control panel.

But hey, at least it's all internal wiring at that point.
 

aapl2c

Member
Thanks! I had seen that before and that's what I'm likely going to do, but I was wonder if anyone had ever made a PCB for the PCI board's cable like the one that was made for the PDS card, I imagine it'd be much easier to make and in theory less interference noise.

And yup, ideally I'd rather have one of the GIMO connector cables, as I even have a Power Macintosh 4400 this'd go into nicely 😀
 

jeremywork

Well-known member
Sadly, I could never get audio over the one I used to have. Still had to use the CD-ROM analog audio cables to pass through via the card. The alternate name Apple used, in the one or two locations it actually appears at all, made it sound like it connects both video and audio, but just doesn't :/ I wonder if that's simply because you can only monitor CD or the analog audio input in the Sound/Monitors & Sound control panel.
I haven't tried a 7200 or 4400, but on my 6500 I do get both audio and video via GIMO, which is a relief as there's no easy way of accessing the CD audio cables in those machines. It's possible they only implemented it on those systems for that reason, and on the machines with easy access to CD audio headers they skipped that functionality.

Even on the 6500, there's no discreet software sound source for GIMO port. IIRC it's CD-TV-Video which enables PC sound. PC sound works fine while listening to the radio on the mac side too.
 

Big Ben

Active member
I did put mu 12’ 100MHz PC compatibility card out from storage and did some testing with my 4400/200.
Despite being « not compatible » with this computer it worked fine.
Using a Compactflash as an hard disk thing is quite fast! At least for a 100MHz card.

I did order VRAM chips to upgrade the VRAM up to 2MB, and I’m truing to get a 64MB ram stick.

I’m wondering if it’s possible to upgrade the CPU by replacing the current one with a socket and if it could go up to 166MHz or more, but I don’t want to risk my card.
The 166 or the 200 overdrive draw much more power than the 100MHz.
Maybe if I find another board I will give it a try.

I got two other cards, one with my 6100 and a loose LC630/DOS motherboard, I hope I’ll have some time to mess with those soon.
 

Byrd

Well-known member
@Big Ben Ive been tempted to replace the CPU in my 100Mhz card as well but that’s a lot of pins, layers and traces to deal with. Perhaps a bus or multiplier overclock might be better - we should look into the resistor differences with the P166 cards. A boost to 133 would be decent.
 

Nathan_A

Well-known member
I am fairly close to completing my collection now. I acquired an OrangePC 660 a few weeks ago for only $100-more than I planned on paying for one a few months ago. $450 + shipping. I feel a little bit silly for buying it since I have two OrangePC 620's now, but I weighed the probability of ever seeing a functioning 660 again for sale at all against the loss of $100, and I think I'll be glad I snatched it up.

The last one on my list is an OrangePC 340. I've never ever seen one of those for sale, and I suspect they barely produced any of them since they would have come out at the worst possible expansion card manufacturer time, namely very, very late in the Nubus lifecycle when everyone was jumping to PCI very, very quickly.

I'll probably sell one of the OrangePC 620s, since I have no use for two of them.
 

Big Ben

Active member
Perhaps a bus or multiplier overclock might be better - we should look into the resistor differences with the P166 cards. A boost to 133 would be decent.
That’s… just brilliant ! For two reasons

1) P54C seems to be quite a beast when doing overclocking, I saw some reports of 100MHz P54C running up to 180MHz, early P5 seems to go up to 100MHz.

2) Even with an other faster CPU I totally forgot we had to deal with the clock multiplier !

Ramping up the system bus won’t be an option I guess, so I guess it’s not risky to try it. We just have to identify and understand the circuit responsible for this.
That shouldn’t be that hard I guess?
 

Byrd

Well-known member
That’s… just brilliant ! For two reasons

1) P54C seems to be quite a beast when doing overclocking, I saw some reports of 100MHz P54C running up to 180MHz, early P5 seems to go up to 100MHz.

2) Even with an other faster CPU I totally forgot we had to deal with the clock multiplier !

Ramping up the system bus won’t be an option I guess, so I guess it’s not risky to try it. We just have to identify and understand the circuit responsible for this.
That shouldn’t be that hard I guess?

Wow I didn't know that P100 revision clocked so well - normally I'd get 20 - 30% at best overlocking early Pentiums however maybe we can get this easy to 166Mhz like the other Apple card? With the 1MB VRAM upgrade it's then exactly the same card at the stock P166.

All we would need is someone to scan the front and back of their P166 card (@Nathan_A any chance?) and we can compare resistor settings and make an educated guess on multipliers - start with 166 and if not clock down to 2x multiplier for 133. The speed boost will be helpful as I often struggled to find a home for these cards when a fast G3, G4 would emulate much the same as the slow Pentium card.

JB
 

Big Ben

Active member
I suppose the bus to be 66MHz with 1.5x multiplier but it also could be 50MHz with 2x multiplier.

I can’t check it right now, my 4400 is taking a bath (literally) but I would check on mine asap.
 
Top