Gorgon

Part 4: System Testing

Part 3: Disk ][ Refurbishment

Part 2: Power Supply Repair and Testing Disclaimer: If you don't know what you are doing DO NOT open and attempt to fix or modify a Power Supply. Even disconnected from wall power they can contain hazardous voltages. new Replacement Power Supplies for Apple 2s can be found even to this day. The Power Supply in the Apple 2 was revolutionary as the "Woz" and Jobs decided to go with a much smaller Switched Mode Power Supply rather than a Linear Mode models commonly found in other 8-bit systems of the day such as from Commodore and Tandy Corp. In this case it is an Astek Model AA 11040B; Apple P/N 605-5703: First we need to get into it. This is accomplished by drilling out the rivets at the center of the two side panels using a number drill slightly larger than the diameter of the hole and working up a few sizes until the head separates from the body and spins, usually getting wrapped around the drill bit. Once opened and the strain relief bushing for the DC Leads removed (squeeze with a pair of needle-nose pliers if you don't have a releaser tool) remove the six machine screws and washers holding the circuit board to the top of the case: and expose the board: The board was examined looking for components with any signs of overheating and especially looking at the Electrolytic Capacitors for any bulging or signs of leakage. The bottom of the board was also inspected to check for any bad solder joints. Of especial interest is the "X" Capacitor, a 0.1uF, 250V Metalized Paper "RIFA" capacitor circled above. This capacitor is installed between "Hot" and "Neutral" on the 120V "Mains" side of the Power Supply and is in-circuit to reduce any switching noise from the power supply getting back into the Mains and interfering with the operation of other devices. These capacitors also help reducing the impact of any transients on the Mains from damaging the Power Supply and are designed to fail open. Somewhat like a Metal Oxide Varistor (MOV) they absorb some of a Transient's energy which blows-out some of the Metal inside of the capacitor. The X Capacitor in these Power Supplies is notorious for failure. The plastic cover on them cracks over time, this allows moisture to creep into the capacitor which accelerates it's failure. When these fail they boil off the dielectric releasing noxious "fishy" fumes. As they fail open the Power Supply will still work but with less Transient protection in-bound, and more switching noise out-bound. Sure enough a Visual inspection shows this 40 year old capacitor has the typical cracks in the case so it was removed. Amazingly enough an exact replacement is still available. Rifa got bought out by Kemet years ago who still manufacture these. Let's, however, use a more modern Metalized Polypropylene (MPP) "X" Capacitor. These have a much less smelly failure mode and though they have a lower slew rate are still rated for the same application. Fortunately, the Circuit Board has not only holes for the 0.8" spaced RIFA but additional holes for much more common 0.6" lead spacing: A Kemet 275V, 0.1uF X2 MPP Capacitor was procured as a replacement. In general, "X" Capacitors, which sit between Line and "Neutral", or "Line" to "Line" in 220/240V, can be replaced with "X2" or "Y" Capacitors. "Y" Capacitors, are designed to sit between "Line" and "Ground" and can ONLY be replaced with Y" Capacitors. This was ordered from DigiKey in Thief River Falls, MN and amazingly arrived in Eastern Ontario less than 24-hours later. Also ordered were some of the components for the Clone Motherboard and some spare 2.5A, 250V "Mini" fuses as replacements for the existing 2.75A, 125V Medium Blow Fuse in the Power Supply, just in case, as these are an unusual size and not commonly available. The circuit board and case were cleaned and special care was taken to ensure all the metal scraps from the rivet removals were removed. Keeping with the Retro theme I busted out the Weller WTCP-N Soldering Iron I inherited from my late father. Older than this Power Supply but it still works like a charm and the "PT" tips are still available. Shown here with the Desoldering attachment I gave Dad for Christmas one year. The replacement Capacitor was installed and the residual Rosin Flux removed with some Iso-Propyl Alcohol (IPA). Now to test the Power Supply. A quick test with the multi meter showed no shorts between the Common (Ground) and any of the Supply Voltages: +5, +12, -5 and -12V. Powering the unit on, however, showed no voltage on any of the leads with the Power Supply out of the system. These Astec Supplies are supposed to work with no Load so what's up? Connecting the unit to the motherboard and turning it on also resulted in no power light, no beep and no 5V on the TTL Logic chips. Houston, we have a problem. To keep the project moving forward in a timely fashion I applied $ and ordered a replacement Power Supply which should be here in a few days from Reactive Micro in the States. I was intending on doing this eventually anyway. I ordered the one with the "Universal DC Connector" which will allow its use in all of the Apple 2 Series and also provided a little more "Ummph" and a Molex and USB 5V for modding. I will put the original Power Supply aside for now as I need to get a better multi-meter and possibly an Oscilloscope to repair it.

Part 1: Disassembly & Cleaning The case is in great shape all things considered. The Cyan splotches on it are Cuperous Oxide from when my humidifier water feed's "Vamp Clamp" started leaking right above the shelves where the Apple ][ and a bunch of modern hardware lived. That was exciting! The case was cleaned with mild dish detergent and lots of elbow grease and came out nicely. After removing a bunch of screws from the base and disconnecting the Keyboard the guts are revealed. A little dusty but none the worse for the years. The Power Supply and motherboard were dismounted. You have to be very gentle releasing the nylon clips on the posts holding the Motherboard down as they can break if forced. A dust off with a brush and a wash with soapy water and the base came out nicely with the exception of a foam ring surrounding the speaker disintegrating when touched. The motherboard was gently brushed with a natural bristle brush (to avoid static from Nylon brushes) and then had the dust blown off it with Compressed Air. The motherboard (820-0044-C) appears to have been manufactured in mid 1981 but the chip date codes tell us that it likely sat in stock with a base 16k of memory installed until mid 1983 when an additional 32k of RAM was added and it was likely sold sometime after that date. It also has a factory "Bodge" (Pin3 NE555 @K13 <-> R15 (1M) & R29 (12K)) on the rear but shows no sign of rework: This "Bodge" will be examined later when I dive into the Schematics to install an improved Color-Killer mod. The Power Supply needs to be checked and will be dealt with separately. The Keyboard was very dusty and the key caps a little grungy but nothing a brush and blow won't take care of. It is the model with Alps key switches which still have good travel and bounce. The keyboard had all it's key caps removed and the bezel for the power lamp. Yes, a Lamp, a 5v, 115mA T-1 3/4 model 817 specifically. The multi-meter showed continuity on the lamp but I'm ordering spares as these "miniature" lamps are getting very rare. The keyboard was gently brushed up-side down over the sink to remove accumulated dust and then cleaned with Windex on a Shop Towel. Only one very tiny rust spot on the base which I'll deal with later with some black "Rust-Coat" paint once its burnished off. I'll likely hit all the switches with some DeOxit D5 later after I test the system for an initial boot. The key caps soaked overnight in a bowl with some dish detergent and then the caps were gently scrubbed with a micro-fibre cloth and then rinsed with clean water and left to dry overnight.

Part 0: Research Adrian's Digital Basement has a playlist on Apple 2 repairs and enhancements. Highly recommended. I've been playing with AppleWin which is an excellent Apple 8-bit emulator that can emulate all the 8-bit Apple Computers (Apple I/II/II+/IIe/IIc/IIgs ) I used the Eastern Software House Apple II Macro Assembler (MAE) back in the day. I found a copy of the Disk Image over on Asimov but no manual. So I scanned the manual and uploaded it to the Internet Archive and the Disk Image as well. I have 6 or 7 boxes of Floppies for the Apple ][ and another long-term goal is to extract the files from these ~40 year old disks and move them to newer media.

I built my first computer around 1975 using an Intel 4004 Processor, 2K of RAM and some logic on a perfboard. I wish I'd kept that Perfboard as it turns out there's not too many working 4004 based systems left in the wild. I had friends in High School that had computers, a couple with Apple's and one with a SWTPC. I never could afford a computer while in College but managed to do a lot of programming and hardware hacking on Apple ][/+/e clones. After graduating I picked up an Apple ][+ Clone Motherboard at Computer Parts Galore in Toronto round about 1985/6. It was a project that has sat unfinished since then. Back in the mid 90s I obtained an Apple ][+ from a friend as partial payment for moving his business from the Apple 2 to a PC. It was in good shape and came with a 10MB Sider ][. Somewhere along the way I misplaced the SASI cable so I removed the HDD Controller. I later found the SASI cable but the controller has since disappeared. The system was last powered on 10 years or so ago and my kids were greatly amused by Choplifter, Lode Runner and other vintage games. It then went back on the shelf. With retirement approaching I finally have the time to get this Apple ][+ Back up and running again. My long-term goal is to get back into Hardware Hacking and Assembly Language programming. My approach is to first do a gentle clean and restoration on the Apple ][+ then, with the help of an eBay find of a recreated HDD Controller Card, get the Sider working to pull the data off it and my Collection of 70 or so 140kB Floppies and, finally, finish the ][+ Clone board and retire the original motherboard to help preserve it.

I like p12266 better, around the same points, lower power consumption:

Thanks, good to know. I haven't dived down the Rabbit Hole of AMD GPUs lack of performance in OpenMM or HIP in Folding@Home in a while but it appears that progress is being made. But the "proof is in the pudding" as they say and looking at a large Atom WU with good results such as Alheimer's p12424 in @LAR_Systems database we still see the AMD GPUs vastly under performing compared to NVIDIA. Granted, a direct comparison is hard to determine due to AMD's annoying habit of using the same GPUID for cards in the similar class (7900XTX/7900XT; 7800XT/7700XT ...) This discussion of AMD versus NVIDIA for Folding should be moved to a separate thread.

ooooh a Torrent, Where's the Build Log?

Personally, I use Cisco SG-200 Small Business switches from many years ago. Inexpensive, only gigabit but manageable through a Web GUI and VLANs and lots of other options. These days I'd likely be looking at MicroTik.

You seem to have an issue with your config files and possibly a bad password or username. Unfortunately you'll have to dive down the black hole of NUT configuration. When I went down that rabbit hole under Ubuntu a couple of years back it took me many hours to get it working properly only to discover that the USB connection on the Cyber Power UPSes were flaky AF. Time is money so I gave up, re-purposed the CyberPower UPS to other uses and went back to using Eaton UPSes with Network Modules. You might try searching for recipes on FreeBSD forms for getting CyberPower UPSes working under NUT as the config should be identical to what TrueNAS Core uses for NUT.

Also, you can look at @LAR_Systems excellent GPU Folding Performance Database and /or use the FP32 Specifications for various GPUs from sites like TechPowerUP. Simply divide the Performance (PPD from LAR_Systems; OR FP32 from TechPowerUP) by the Retail Price in your area and that should give you a good idea of the Value of a specific model. Of course, if you want/need to use the GPU for other things such as gaming that can influence which GPU is the best for your specific use cases. It's not that I don't like AMD GPUs, I actually prefer them over NVIDIA, as they are, IMHO, much better value and my daily driver is a FirePro W4100, but for just Folding they just don't make sense. Some older Models, like the Radeon 7, are also really strong performers for other Distributed Computing applications (BOINC) as they have excellent FP64 performance, but as I mentioned AMD, Like NVIDIA, seems to be intentionally "nerfing" FP64 performance these days on their consumer GPUs, likely to drive users needing good compute performance to their much more expensive Enterprise Product Lines.

I don't use AMD for Folding as they refuse to port ROCm to their Consumer GPUs and Folding@Home and thus their GPUs are forced to to OpenCL so their Folding@Home performance is much lower than an Nvidia GPU with similar rasterization performance that can take advantage of the work Nvidia has put into the CUDA libraries. In general though the Folding performance will be directly proportional to the Single Precision Floating Point performance (FP32) so if you divide that by the Cost it should offer a good indication of the value of a particular model for just Folding use. Folding at Home does not use Ray Tracing nor does it use a significant amount of VRAM so VRAM Size nor the size of the GPU Memory Bus Width should have much impact, if any. Some AMD models, however, have severally limited amounts of Double Precision (FP64) cores which may impact their performance on Work Units that require these calculations as they will have to be off-loaded to the CPU with a significant performance penalty.

I use Pro Drivers (27.20.21026.2006)as well for my W4100 in my Daily Driver, which I don't fold on (It has a RTX 4070 for folding). I'm running Windows 10 Pro (22H2) and I also see un-signed OpenCL Drivers from the Kronos Group in my Pro Drivers but mine are 3.0.3.0. I suspect the culprit is either AMD or the FAHClient package. A slot getting into a "Failed" state can usually be fixed with a reboot in Linux but you'd have to wait for someone who folds on AMD GPUs to chime in

With a mid-series refresh under way for NVIDIA's Ada GPUs I started to wonder what the changes mean in terms of value for Folding. The short answer is that even with a $200US drop in price of the "mid range" GPUs it is a better value but still can't match the value of the new 4070 Super or even the 4060 in terms of Shaders/$ Price Shaders Clock TDP Value Value Model ($US) (#) (MHz) (W) (shClk/$) (sh/$) RTX 4090 $1,599 16,384 2,235 450 22,901 10.2 RTX 4080s $999 10,240 2,295 320 23,524 10.3 RTX 4080 $1,199 9,728 2,205 320 17,890 8.1 RTX 4070ts $799 8,448 2,340 285 24,741 10.6 RTX 4070t $799 7,680 2,310 285 22,204 9.6 RTX 4070s $599 7,168 1,980 220 23,694 12.0 RTX 4070 $599 5,888 1,920 200 18,873 9.8 RTX 4060t $399 4,352 2,310 160 25,196 10.9 RTX 4060 $299 3,840 1,830 115 23,502 12.8 My personal preference for a new build would now be the RTX 4070 Super as it is showing nearly as good value as the 4060 and is still a dual slot card making it much easier to implement dual GPUs in motherboards with 3-slot spacing for the PCIe expansion slots.

My guess would be to check your passkey, and/or userid it may have been mangled

Glad to hear you fixed the issue. Unless you have experience working with electronics DO NOT take apart your power supply. There are capacitors in there that will hold a charge, even when the unit is unplugged and can give you at the very least a nasty shock.

Correction: It apparently has an 8-pin PCIe PEG Power Connector which appears to be at the back of the card. Here's a comparison of the RTX 4000 SFF and earlier SFF Cards:

The 4060 (non ti) is bus powered (i.e. no PCIe PEG Power Required)

It won't be as power efficient or as performant (1MPPD vs. ~4MPPD) as the 4060 @RollinLower listed (Nice find, BTW - but I believe it is a just a 4060 NOT a Ti) ) but it does support NVENC. I've found if you clock limit the 4000-series cards to 2250MHz they sip power and are very quiet.

Good for Display Out but that's about it. For just Display Out a AMD WX3200 is, IMHO, the budget king. I've been running a FirePro AMD W4100 for years in my daily driver for Display Out with a RTX4070 for Folding. The W4100 can sit close to the Folding card and not impeded airflow too much. And, unlike their consumer variants, the Radeon Pro Drivers suck much less and you get 10-bit LUTs to boot for Color-Critical work and display calibration. If it's for a HTPC build the GT730 has no NVENC Support.

I also linked the A2000 which you might be able to find used. Any of the Turing cards don't have very many CUDA cores and are much less performant.

For Folding? RTX 4000 SFF or RTX A2000

Here is the Legacy Compatibility list at Corsair. So a Type 3 PCIe PEG 6+2 should work as it appear that just the 24-pin motherboard connectors are pinned differently between Type 3 & 4. You should have received 2 of these cables with your PSU so you would have to use both connectors on one cable and one connector on the other to power your GPU. If your missing a PCIe cable they can be bought directly from Corsair. The TBP of a 7900XTX is 355W with a recommended PSU of 800W so you should be fine.

Likely there is an issue with the Bridging configuration. It sounds like the LAN (wired network) can forward packets to both the 2.4 and 5GHz Bridge Groups but the Bridge Groups aren't configured to allow forwarding to each other or likely also to the LAN. Here is my router using FreshTomato: The above basically allows the LAN (bridge0) to talk to the WiFi (Bridge1). If I wanted the WiFi to talk to the LAN I'd have to add a policy allowing Src: br1 to Dest:br2. Under the hood, most routers use Linux and thus IPtables and all these rules do is allow Layer 2 forwarding of packets between Bridge Groups.