Turkey and Solder –or– Busy Busy!

After a whirlwind of activity, a lot of which I need to document here, I’m back online!

First of all, I wish you a Merry New Year and for those that took time off this season. Hopefully you’re back with renewed vigor to tackle that project you’ve had on your mind, or an old project that’s on your workbench.

Me? I couldn’t stop working away. Time off feels is nice; I put my toys aside while working on projects at work, and having time off gives me the time to work on my personal projects. I can only describe it as a runner’s high.

There’s a scene from Malcolm in the Middle where Hal needs to change a lightbulb. When Lois comes home, she finds Hal under the car! ‘I thought you were going to change that lightbulb?’ Lois inquires, and Hal exclaims ‘What does it look like I’m doing!!?’

I can’t explain how much I love it when this happens. I find that I need a tool, which I know I can make, however the thrill is more in the journey to get X working, as opposed to a simple fix.

One big project that I have been working on is getting an old Tektronix 7854 digitizing oscilloscope back up and running. If you haven’t read-up on this series of oscilloscopes, they’re fantastic–there are plug0ins for curve-tracers, spectrum analyzers, measuring signals down to  micro-volt-levels, single-shot captures in pico-seconds–really and truly, fantastic! However, early production models left the factory with MOSTEK ROM devices, and really, they’ve come to be known as F-ROMs (Forgetful Read-Only Memory). After a number of years, these ROMs, which contain all the information to tell the internal CPU what each button does, and how to act, has ‘forgotten’ its instructions.

Yes, this oh-so-awesome piece of tech that I bought, demonstrated as working, turned into a toaster when it was setup on my bench.

A few years earlier, after having discovered how older Tektronix equipment was built (and I do mean built), I joined a Yahoo Group devoted to these mammoths of design and engineering. They were my first stop.

Initial research found that there have been a few individuals who have successfully restored their devices to operation by pulling all four of these F-ROMs and replacing them with two larger EPROMs. This requires some code hacking and rewiring, but at least the firmware was out there and it’s a sort-of-well-documented issue. But first I had to get over one minor issue, I don’t own an EPROM burner.

I put out a few want-ads on some forums I’m a part of and had an offer of an XILINX Superpro, complete with it’s interface card. Sold! I collected the package some weeks later and lo, an interface card was included, as well as an industry-grade UV eraser box!

The interface card was really just a 25-pin parallel port. I keep some old laptops around for these situations. I pulled out an IBM 600E and it fires right up. I downloaded the software from the company’s site, plugged the burner in, and not a sign of life. I try different BIOS configurations, I try another computer, and I take another look at that ISA card. There’s a voltage regulator on the board.

I look some more and find that the old piece of tape stuck on the plate reads ‘SUPERPRO ONLY.’ This was no typical parallel port.

So now I needed this card. I panicked. Is that ISA? It’s ISA! It’s definitely not PCI. Current solutions out there for ISA-to-USB conversion is in the $250+ range, beyond that there are industrial-level computers boasting 20 ISA slots for $thousands. I checked my stash and while I found a PII 200MHz machine, alas no ISA slot.

I slept on the issue, and thank goodness.

The next morning I tore into my closet–ah-ha! My family’s second computer! I saved it from the curb after dad replaced it with a Dell over a decade ago, and my last hope at finding an ISA slot.

I was rewarded with three. Success! I put the card in and set to work. Running the software under (a fresh!?) installation of Windows 2000 that was already loaded onto the PC (the name of the computer hinted that apparently I tried to use it as a server at some point) was a futile task. Windows kept asking me about memory access and the addresses required were at least partly eclipsed by the ISA NIC.

I pulled the NIC, pulled the HDD, found an old IDE drive in the basement, and set to work installing DOS 7.1.

Loading the software up again and it just worked™. To my amazement the software was happy, the lights on the Superpro illuminated, and it was ready to work. I grabbed the lot of EPROMs which I had the forethought to order ahead of time, on eBay, and set them up in the UV eraser. Growing up with DOS, loading the software using only a keyboard was a breeze.

So now I had a couple 28-pin devices and I had to get them to work in a circuit board that has 24-pin IC sockets.

Oh, and the pin assignments don’t line up.

Right, and some chip-select signals need to be relocated.

Ah yes, the power pins are incorrect as well…

This involved developing a set of stacked-and-soldered IC sockets. Some pins are clipped and the signals are routed to the correct location via magnet wire, however this is a bit of a kluge and is causing some clearance issues with neighbouring cards.

Reworking the ROM card itself, I found three solutions, each involving different ways of reconnecting gates and adding different jumpers.

During one stint of testing I was rewarded with traces! The traces were gone with a power cycle, but came back after a few resets. Intermittent issue… great.

To-date I have tried all three and have exchanged some personal messages with those who have had success, baffled at why their method is not working in my device. One theory is that the CS signals are fighting, but with work starting back up again, I need to change my focus if only temporarily.

I wish I could end this on a lighter note, but after three weeks I still do not have a functional 7854. I’ve since ordered a paper-copy of the service manual so that I may further study the schematics. Ultimately I would love to design my own drop-in replacement PCB as at this stage I’ve begun to lift traces. I have another MOSTEK-filled ROM card on the way to continue my experiments with and when I get my replacement card designed, I will share that journey with you all!




Comments are closed.