VT102 something blown

[Xezlec]

Dear fellow retronauts,

I have quite a tale of woe.

I bought a VT102 advertised as "new in box" on eBay. It was very nearly that. I've never seen such an old device in such flawless condition. Baud rates above 4096 didn't seem to work for some reason (may have just been my setup) but otherwise 100%, with manuals and everything! I was very excited to have obtained possibly the last near-mint-condition VT102 on Earth.

After using it for a few days, there was a thunderstorm. One lightning bolt that night was so loud I thought it must have hit my apartment directly. Despite having everything in my apartment plugged into expensive surge protectors, and despite having everything but the cable modem turned off at the time, and despite there being fuses in the power supplies, I discovered the next morning that my cable modem, my PC, and by far worst of all, my VT102, lay in ruin.

Having been proud to be preserving such a lovely piece of computing history, you can imagine my agony at the realization that I had done precisely the opposite. It took me a while to go back to it.

When I turn it on, I smell the distinct aroma of roast resistor, but I don't see any components visibly blackened inside. The screen doesn't seem to be capable of forming a stable pattern anymore, though it tries. Based on keyboard lights, it appears possible that the logic board is OK. (What are the chances?)

My questions are two:

1. Do I have a snowball's chance in the core of a neutron star of getting it working again?
2. Where do I start?

Thanks for any help. I can assure you that from now on, everything I own gets unplugged whenever there is even a chance of a storm.

 

[Stone]

Sorry to hear this. FWIW, surge protectors are just that and they will not protect against lightning. They only protect againsst surges in the AC power. But I guess you know that, now. There is only *one* protection from lightning strikes that actually works. You must unplug the device you want to protect from all outside lines and that includes power, telephone and cable although i might be wrong on the phone/cable part because I don't know if fiber optic lines can dissipate lightning. But if they're still copper, definately a good conductor.

 

[Chuckster_in_Jax]

There was a thunderstorm. One lightning bolt that night was so loud I thought it must have hit my apartment directly.

You must live in Florida. I have been a victim of lightning damage 3 or 4 times. Ive lost a couple of TV's, a complete computer system including laser printer, a dryer and even the meter outside.

FWIW, I never leave any of my vintage equipment plugged in. The electronics are just too old and they can be a fire hazard.

 

[Stone]

Florida is the thunderstorm capital of the Northern Hemisphere. Buildings there need to come equipped with master disconnect bots for all electrical circuitry. It's just tempting fate to leave anything plugged in, especially when you're not at home. I guy I know that even had a tornado cellar built with his new house in central Florida. It wasn't even two years before that proved to be a life saving investment.

 

[Lou - N2MIY]

The VT102 has a baseband video output on the back through a BNC connector. Connect this to a video input on a television set and tell us what you see.

If the logic board works, then the power supply is likely OK. Do you have proper tools for diagnosing and repairing electronics?

Almost anything can be fixed on a VT102.

Lou

 

[Xezlec]

Sorry it took so long to reply. Too many projects.

Sorry to hear this. FWIW, surge protectors are just that and they will not protect against lightning. They only protect againsst surges in the AC power. But I guess you know that, now. There is only *one* protection from lightning strikes that actually works.

What about lightning arrestors?

You must live in Florida.

Austin, Texas.

FWIW, I never leave any of my vintage equipment plugged in. The electronics are just too old and they can be a fire hazard.

I don't suppose flipping the switch off on the surge protector is good enough either, because the lightning could just arc over the switch.

The VT102 has a baseband video output on the back through a BNC connector. Connect this to a video input on a television set and tell us what you see.

Zilch. I connected it using alligator clips and copper wire though since I don't have a BNC-to-RCA cable. Also, the LEDs on the keyboard no longer flash and I no longer see even a flicker on the screen when I turn it on (though I do still hear flickery monitor noises). I think it's getting deader every time I give it juice. Might be the power supply.

Do you have proper tools for diagnosing and repairing electronics?

I'm ashamed to say I don't own a scope at the moment. "Ashamed" because I'm supposedly an electrical engineer. I do have a multimeter, soldering iron, wire cutters, pliers etc. Maybe I should invest in a scope one of these days.

 

[Lou - N2MIY]

Yes, check the power supply voltages. (And buy a scope!)

Lou

 

[Xezlec]

Wow.

So I bought a scope, the venerable Tek 465B. Worked great for about 5 minutes. This time, I have no idea how I managed to destroy it. I was just touching the probe and looking at the noise it was picking up, and then it suddenly got super bright for a few seconds and then the traces disappeared forever.

I guess I'm going to stop touching anything older than me that uses electricity. This is just getting too expensive.

 

[k2x4b524[]

might also be the fact this was essentially new old stock, all the insides hadn't been throughly conditioned for electricity. Living in washington, we get our fair share of lightning and storms. Never had anything go poof through the surge protectors, however plugging into the wall is different, try to avoid doing that with vintage gears, i've also got an older digital "it was made by digital* line conditioner that has proven it's worth over and over again.

 

[Doug G]

One common failure after lightning was the rs232 line driver chips, common ones were mc1488 / mc1489 chips.

I had an old DEC 11/23 computer with a 25ft rs232 cable coiled and connected to the Decwriter next to it. Both units were unplugged from the wall sitting on the floor in my office room. Lightning hit the ponderosa pine in the back yard during the day and both the computer and the decwriter were fried, apparently just from the induction in the coil of serial cable connecting them.

For years we supported hundreds of DEC-based systems with thousands of rs232 terminals, and our experience was that the serial terminal wiring was responsible for probably 95% of the lightning-related failures, with very few problems coming in from the AC power side.

 

[Chuck(G)]

You know, the Tek 465 (my favorite old school portable scope) is darned near indestructible. There is a ton of information, repair tips, etc. on the beast, so you shouldn't have much of a problem. One big issue with these guys are failing tantalum caps (big surprise).

For the VT102 I'd start with the VT100 technical manual on bitsavers.org and work through the power supply section and make sure that it's supply what it's supposed to.

 

[Xezlec]

Lightning hit the ponderosa pine in the back yard during the day and both the computer and the decwriter were fried, apparently just from the induction in the coil of serial cable connecting them.

Crazy! Then presumably if half of the coil had been wound in the opposite direction, it never would have happened. Not something I would normally think to do.

Thanks for all the responses and encouragement. I guess I should just jump back in and keep trying. I'm starting on the scope now (maybe it's something I can track down just with a multimeter).

 

[Doug G]

Crazy! Then presumably if half of the coil had been wound in the opposite direction, it never would have happened. Not something I would normally think to do.

Actually, I don't think it would have mattered if the cable was coiled, straight, longer or shorter, lightning that close will probably hurt something. The strike hit near the top of a maybe 75-100' ponderosa pine with a trunk about 5' diameter at the base, and it blew giant hunks of bark off the tree, blew off a few limbs that were maybe 8-10" diameter, and came down the tree and jumped about 5 feet off the tree to the metal frame of my kid's swing set. As Huey Lewis might say, "Powerful Stuff"

I wasn't home at the time and my ex was over visiting her friend so we didn't get to see the actual strike, thank goodness. We moved out of that house a few months later, and some friends said the tree died within a year or two.

 

[Xezlec]

So I'm ignoring the scope problems for now. It now seems to work intermittently, so there must be either a loose wire or a scratchy intensity knob or something. After apparently bumping it the right way while messing with it, I seem able to use it for the time being.

Regarding the VT102, I'd like to debug it, but I don't have schematics or a service manual and can't find them online. I've found both for the VT100, but on close inspection, it is clear that they do not at all match the boards in my VT102. I can't even check that the power supply is working because I don't know where to check or what to check for. I poked around a little with a multimeter but couldn't really figure anything out.

Weirdly, I now find that if I wait about 30 seconds to 1 minute after turning it on, the VT102 DOES finally (suddenly) boot, with the usual beep and the appropriate keyboard lights, though still without video. And it still quickly starts to smell like fail (can't track the smell down though).

I guess the next step for me is to pull out the H7838-A power supply board and start following the traces to try and create my own schematic, which I can then study to try and understand its operation. Next would be the CRT board, I suppose.

 

[Chuck(G)]

Well, there are some things that you can observe without taking the thing to pieces.

For example, there are lots of TTL chips in the thing that one can use to observe the behavior of the +5V line.

 

[Xezlec]

Well, I was having trouble measuring those voltages due to my unsteady hands. I've always been better at theory than practice, so I went for a walk to try pondering it instead.

I realized that if it was taking that long to come up, there had to be something temperature-related going on. Since it wasn't working initially, something must have been shorting the power, and since it came up eventually, that must be because the dead component was heating up enough to start to have a significant resistance. I also noticed that the smell was stronger near the logic board. So, I figured, whatever was dead should be very hot when it finally boots. Being kind of desperate at this point, I decided to just sacrifice a finger and touch each component on the board until the problem revealed itself (because skin heals but chips don't). It worked! I burned myself on a little buffer chip right next to the RS232 port. So yes, it looks like induction in the serial cable may well have been where the problem happened. Good call, Doug!

In case anyone is on the edge of their seat, I'll report back after I find the time to replace it (I obviously suck at debugging, but I do know how to do this part, at least).

 

[RetroHacker_]

I have seen terminals zapped through the RS232 line where the line driver chips (1488/1489) were fried to the point of being shorted. This loads down the power supply and the whole thing won't work. Since the VT100 has a switchmode supply, it's shutting down.

That chip that's getting real hot, if it's a 1488 or 89, it shouldn't... You can clip the chip out carefully and not even replace it, and the terminal should now power up and work (it just won't communicate with anything). If that's the case, then simply desolder the legs and replace it (use a socket this time).

-Ian

 

[Xezlec]

So I replaced the bad chip, and now it boots fine, with video, and transmit is working fine, but I can't seem to receive.

Here's the weird thing. I check the open-circuit voltages on all the computer-to-terminal-direction lines on the cable from the computer (i.e. not plugged into the terminal) and they're all +/- 7 V. Perfectly reasonable for RS-232. Now when I plug the cable into the terminal and check the voltages at the traces/solder points on the logic board, they're all +/- 0.3 V, except the receive line, which is +/- 0.1 V. At first I thought maybe it was being shorted somehow, but all 4 of the lines I checked having that same problem and at about the same voltage? What are the odds? They don't all go through the same chip. So now I'm wondering if maybe the board is designed to work this way for some stupid reason, i.e. using fairly-low-impedance inputs or something, but when I look at the datasheet for the line receiver chips it uses, it looks like they are supposed to be ordinary high-impedance inputs.

I guess it is entirely possible that all the receiver chips are just busted in exactly the same way, but I wish there was some way I could double-check that before launching into the painful chore of replacing all these damn chips with my crappy soldering iron. I don't feel very confident that I'm on the right track now.

Also, if I'm going to replace them, really, I'd be tempted to just replace the chip that handles the receive signal, and leave the others broken since I don't really need them and am lazy, but I'd worry about whether that's going to cause gradual damage to something somewhere else. Any thoughts or experiences in that department?

 

[Xezlec]

Wow. Out of the 5 chips that talk to the serial port, each was causing a different problem. Only once I had replaced them all did I finally have a fully working VT102 again. I also ended up buying a fancy-schmancy desoldering gun because I got tired of desoldering the hard way.

So, in this journey, I started out with a mint condition VT102, and ended up with a working-fine-but-badly-soldered VT102, Tek 465B, Hakko 808, and an empty wallet. But it was all worth it to see that smiling face...

 

[MicrocomputerSolutions]

I've fixed many more bad boards over the years using my eyes and nose than I've ever fixed with a scope, logic probe, and DVM.

Did you take the time to install IC sockets instead of soldering the chips back onto the board so next time the chips will be easier to replace?

The 1488 and 1499 serial chips are pretty common failure parts, and I make a habit of inserting sockets when I cut out chips that I suspect have gone defective.

My goal when replacing chips is to save the circuit board traces, so I always sacrafice the suspected defective chips. First I use a pair of flush cutters to cut the legs off chips close to the body, leaving a good stub to grab with a pair a needle noses from the component side of the board, while applying a soldering iron to the solder side of the board to remove the legs, followed by a desoldering tool to clean out the holes.