fjalvingh
Experienced Member
Found today on a Dutch second hands site: https://www.marktplaats.nl/v/comput...geheugen-kart-four-fase-system-kart-nr-1van-2
| VCF Latam | Apr 24 - 26, 2026, | Bahía Blanca, Argentina |
| VCF Pac. NW | May 02 - 03, 2026, | Tukwila, WA |
| VCF Southwest | May 29 - 31, 2026, | Westin Dallas Fort Worth Airport |
| VCF Southeast | Aug 01 - 02, 2026, | Atlanta, GA |
| VCF West | Aug 01 - 02, 2026, | Mountain View, CA |
| VCF Midwest | Sep 12 - 13, 2026, | Schaumburg Convention Center, IL |
| VCF Montreal V2.0 | Nov 07 - 08, 2026, | Saint-Lambert, Montreal, Canada |
| VCF SoCal | See you in 2027, | Southern CA |
| VCF East | Apr TBD, 2027, | InfoAge, Wall, NJ |
-
Please review our updated Terms and Rules here
Four-Phase Systems IV/90
- Thread starter NeXT
- Start date
NeXT
Veteran Member
Ooh! That would probably be helpful, however isn't marketplaats a domestic-only auction site?
fjalvingh
Experienced Member
It is, but I can do that as I’m Dutch. What would be a reasonable amount you think? Because the current price is already 120 euro..
MattisLind
Veteran Member
There are actually two separate auctions with one card each if I understand it correctly. One is currently at 130 and the other at 140. Is it just the gold that makes it so expensive?
fjalvingh
Experienced Member
One assumes it is either collectors or indeed gold collectors. Hate those.
MattisLind
Veteran Member
I was lucky outbidding a goldbug when bidding for 9 old HP boards last autumn. Even those guys has a limit when it isn’t worthwhile I guess. I had to pay 17 usd per board. Curious what the gold value was. It takes a bit to process them.
fjalvingh
Experienced Member
Well, I’m bidding the 130,- on one of the boards which represents about 1 gram of gold, apparently… Let’s hope we can save at least one.
NeXT
Veteran Member
I suspect there's a lot of pyramid scams at play. It costs nothing to point out the value of gold per troy ounce and the untold wealth found in extraction from E-waste. This has been a thing since what, the mid-2000's when the economy collapsed? Probably longer?
The idea is you overstate the gains from E-salvage both as a business and as a backyard economy, then you release "gold rich" material to the market for prospective buyers to purchase. You make a tidy sum and the buyer realizes that per ounce means nothing when the extraction process often far exceeds the value of the gold both in labor and equipment costs. The buyer wants out and sells it to the next person, either with a mark-up or no loss on their own investment and so on and so on until someone takes a financial hit. Ultimately you end up at the bottom where it enters the established high-volume gold reclamation market or neato EPA reports about properties you can probably consider superfunds from process contamination. I'm assuming ebay is once such outlet to this, as are the various scrapper facebook groups. Everyone's in it for the money, so when someone says they have a lot of high-yield scrap and they themselves are about as loaded down as Chicago Scrap Trucks, why are they trying to make a quick buck like this? You clearly have your connections. Why's this pile of sound cards with the edge connectors hacksawed off so special?
fjalvingh I've sent you a DM.
1ajs
Veteran Member
hopefully this pans out
NeXT
Veteran Member
Well in the meantime I've been slowly chipping (pun) away at trying to develop pinouts for the various IC's Four-Phase developed to try and tackle the total lack of datasheets that are/were available. The idea is that you look at the datasheets and compile a list of what each pin is designated as. Doesn't give you a descriptive map but it still gives you an idea.
It's tedious. Since nothing is arranged how you would expect it you have to verify multiple times. The other is that some chips have the same chip number but are in fact part of a family to which the pin labels change according to the schematic. For example ROM2-03 differs from ROM2-04 and ROM2-05.
I'd love to find a better way to do this as there's still a lot of chips left to do but if this is pointless I'd love to hear you chime in.
The other is I started mapping the pinout of every slot on the backplane. This equally isn't as simple as "it's a passive backplane, why bother?"
Because it's not. Look near the start of this thread and you will see it's a combination of a single backplane with a lot of parallel traces but those parallel traces are in fact broken into three major sections (I/O options, main processing and memory/video) and the entire video/memory section is carpeted with fine white wires that are further defining how memory is interacting with the character generators. You also get some other nasty suprises like the four character generator slots have an additional VEE -5v rail found nowhere else in the machine and some slots have pins flipped.
The map started off as a spreadsheet and is again being developed using pinout information from the various schematics. We are unfortunately missing the pinout for the parity board but so far the rest of the boards including their top edge connectors are available.
Edited: oh yeah, while working on the ram chip I found a fairly high resolution die photo for the RAM-9 hiding in Wikipedia's image cache. It looks like it was once part of a larger die photo catalog but at some point got de-referenced and is just sitting orphaned.
commons.wikimedia.org
I also determined that the die photo of the AL1 that Ken Shirriff has on his blog is the highest quality photo that exists. I think I mentioned that last year I inquired with the CHM about sourcing a higher quality image of the photo since he cites them as the source and they both sent me a lower quality image and told me not to share it. It's not as great as the RAM-9 but it's still over 1200 pixels across, so I sat down and verified the labels on the bond pads at least are correct even if the rest of the barely readable labels are apparently not.
It's tedious. Since nothing is arranged how you would expect it you have to verify multiple times. The other is that some chips have the same chip number but are in fact part of a family to which the pin labels change according to the schematic. For example ROM2-03 differs from ROM2-04 and ROM2-05.
I'd love to find a better way to do this as there's still a lot of chips left to do but if this is pointless I'd love to hear you chime in.
The other is I started mapping the pinout of every slot on the backplane. This equally isn't as simple as "it's a passive backplane, why bother?"
Because it's not. Look near the start of this thread and you will see it's a combination of a single backplane with a lot of parallel traces but those parallel traces are in fact broken into three major sections (I/O options, main processing and memory/video) and the entire video/memory section is carpeted with fine white wires that are further defining how memory is interacting with the character generators. You also get some other nasty suprises like the four character generator slots have an additional VEE -5v rail found nowhere else in the machine and some slots have pins flipped.
The map started off as a spreadsheet and is again being developed using pinout information from the various schematics. We are unfortunately missing the pinout for the parity board but so far the rest of the boards including their top edge connectors are available.
Edited: oh yeah, while working on the ram chip I found a fairly high resolution die photo for the RAM-9 hiding in Wikipedia's image cache. It looks like it was once part of a larger die photo catalog but at some point got de-referenced and is just sitting orphaned.
File:ALI RAM-9.jpg - Wikimedia Commons
I also determined that the die photo of the AL1 that Ken Shirriff has on his blog is the highest quality photo that exists. I think I mentioned that last year I inquired with the CHM about sourcing a higher quality image of the photo since he cites them as the source and they both sent me a lower quality image and told me not to share it. It's not as great as the RAM-9 but it's still over 1200 pixels across, so I sat down and verified the labels on the bond pads at least are correct even if the rest of the barely readable labels are apparently not.
Last edited:
fjalvingh
Experienced Member
We got the card, so at least that went well.
NeXT
Veteran Member
Went to drop the power supply in and hesitated because the two main fans are in the same compartment and if they go bad, you gotta pull the PSU out agian.
They are a pair of Nidec TA500 family AC fans that are rated for 130CFM and run on 120 volts. Interestingly the AC power harness for the fans already has a 2-pin molex plug for each fan, but the fans are the style with a pair of terminals and a rubber plug, instead of hardwired leads. I'm assuming at one point Four-Phase used fans with hardwired leads and they connected at the molex plug but later switched to these and had to make a 6" adapter cable for each fan.
Anyways, both fans when I got them sounded like junk. Many pages back I mentioned oiling them and this both sort of worked but also made a mess and slung oil everywhere. Both fans now worked better but one of them was rattling on the sleeve bearing, so I decided to replace both fans first. I selected a pair of Orion OA119AP-11-1TB fans as suitable replacements (127mm square, 38.5mm deep, 115v and 130CFM, with terminals) which came to about $96 on Digi-key, so I ended up spending another $5 on parts to rebuild my unrelated capacitance meter and get free shipping. I can't believe how much more expensive the 127mm fans were over the 120mm but whatever, I'll only ever need to replace them once.
Meanwhile I was looking at the terminal cable I made last year and I still don't like it because it's kinda hacky and only five feet long, which isn't enough for a machine that's supposed to live in a rack, even if I put the terminal on a desk next to it. While I had that on my mind I came across a pile of decommissioned plenum rated VGA cable for connecting classroom ceiling-mount projectors to a desk at the front of the room. The cables are shielded and very long but then it hit me that while the wiring diagram asks for 22awg cable, I think this should work perfectly as we will not be running 1000 feet from the computer as opposed to 12 or 50 feet. If anything we can also then put the sync and clock signals on their own coaxial lines for additional shielding. I'll have to bring a sample home and test.
They are a pair of Nidec TA500 family AC fans that are rated for 130CFM and run on 120 volts. Interestingly the AC power harness for the fans already has a 2-pin molex plug for each fan, but the fans are the style with a pair of terminals and a rubber plug, instead of hardwired leads. I'm assuming at one point Four-Phase used fans with hardwired leads and they connected at the molex plug but later switched to these and had to make a 6" adapter cable for each fan.
Anyways, both fans when I got them sounded like junk. Many pages back I mentioned oiling them and this both sort of worked but also made a mess and slung oil everywhere. Both fans now worked better but one of them was rattling on the sleeve bearing, so I decided to replace both fans first. I selected a pair of Orion OA119AP-11-1TB fans as suitable replacements (127mm square, 38.5mm deep, 115v and 130CFM, with terminals) which came to about $96 on Digi-key, so I ended up spending another $5 on parts to rebuild my unrelated capacitance meter and get free shipping. I can't believe how much more expensive the 127mm fans were over the 120mm but whatever, I'll only ever need to replace them once.
Meanwhile I was looking at the terminal cable I made last year and I still don't like it because it's kinda hacky and only five feet long, which isn't enough for a machine that's supposed to live in a rack, even if I put the terminal on a desk next to it. While I had that on my mind I came across a pile of decommissioned plenum rated VGA cable for connecting classroom ceiling-mount projectors to a desk at the front of the room. The cables are shielded and very long but then it hit me that while the wiring diagram asks for 22awg cable, I think this should work perfectly as we will not be running 1000 feet from the computer as opposed to 12 or 50 feet. If anything we can also then put the sync and clock signals on their own coaxial lines for additional shielding. I'll have to bring a sample home and test.
Last edited:
NeXT
Veteran Member
New display station cable is done. I'm much happier with this. I can probably make another 8 foot cable at some point but I've run out of AMP shells and need to order more.
Slob
Experienced Member
Gold recovery is a waste of time unless you have some unlimited free source of 60’s/early 70’s PCB’s. OTOH, electric company scrapped circuit breakers and some transformers contain a LOT of silver.
I knew a guy in Atlanta who was pulling a half ounce of gold per Spring weekend in the early 90’s from the Dahlonega mountains equipped with the largest permissible ameteur goldmining equipment. Occasionally, he’d pull out a nugget, which was worth a lot more than its weight in gold, used in artistic jewelry. Wonder what’s going on there now!
I knew a guy in Atlanta who was pulling a half ounce of gold per Spring weekend in the early 90’s from the Dahlonega mountains equipped with the largest permissible ameteur goldmining equipment. Occasionally, he’d pull out a nugget, which was worth a lot more than its weight in gold, used in artistic jewelry. Wonder what’s going on there now!