• Please review our updated Terms and Rules here

Qume Terminal membrane keyboard repair

W

Worblehat

Member
Joined
Feb 28, 2017
Messages
36
Location
Victoria BC Canada
I recently acquired a Qume QVT-102 terminal. The display is in great shape, the keyboard ... well, not so much. Pretty much none of the keys work as they should. A few can mostly be counted on to produce a character on the screen if I press them firmly. A few more can be coaxed to produce a character if I repeatedly press them firmly from slightly different angles. And quite a few don't work at all. I've opened up the keyboard and it looks to be a fairly standard membrane type. I'd post pics but the ever-useful Bitsavers has already posted better pics than I could, as these show the membrane with the switches removed:


I don't have much experience with this kind of thing, but based on it seems likely to me the problem is with the membrane itself, rather than the keyboard circuitry. I'd really like to avoid removing the key switches if I can, and I've heard tell elsewhere on these forums of a fix that consists of heating the back of the backplate with a hot air gun, that has been known to produce good results in Osborne 1 keyboards that I believe are somewhat similar to this. Does that seem like a reasonable thing to try here?
 
So, I’ve been having some luck with repetitively - some might say obsessively- pressing the keys from slightly different angles. Several that weren’t working at all are now working reliably, and a few more are starting to come back. It’s a bit tedious but as it’s producing results I’ll keep going for now.
 
Hey - I just got a QVT-102 and have the same issue. Did you ever fix yours completely? The Bitsavers images are helpful, but I don't see any easy way to remove the keyswitch housing to get at the membrane to clean it?
 
guruchill101 said:
Hey - I just got a QVT-102 and have the same issue. Did you ever fix yours completely? The Bitsavers images are helpful, but I don't see any easy way to remove the keyswitch housing to get at the membrane to clean it?
Click to expand...
Unfortunately not completely or even close. It’s possible that if I had the patience of Job my approach might have worked in the end, because I did continue to see some improvement. But it was so slow and tedious I eventually gave up. I couldn’t see any way to get at the membrane either. I’ve wondered gently heating it with a hairdryer might have an effect but have not yet tried it.
 
I just received a QVT-102 from the Pacific NW via eBay. No keyboard. I'd like to reverse engineer the interface so we can build an adapter to us a PS/2 or USB keyboard. But neither of the Qume manuals for this terminal document the protocol. If anyone would be willing to lend me their non-working keyboard...as long as a small number of keys work...I should be able to figure out the protocol and return the keyboard. Then I'd design an adapter board to use a PS/2 or USB keyboard and send a working adapter board to whomever helped out. If anyone is interested in this endeavor, please PM me.
 
That was my fallback plan if unable to find a sample to examine. The main controller code is 8KB of 6800 so also manageable if all else fails. It's supposed to be a bidirectional one-wire protocol so it might be necessary to see both sides. Maybe not. I'll be poking through that. Thanks for the feedback.
 
Hey folks, I'm glad to see that this subject is getting some attention. Here is my experience as I have also recently acquired one of these terminals from ebay. Opening the keyboard is straightforward but once the assembly is removed from the case, the keyswitches are apparently ultrasonicly plastic welded with plastic pins. I thought about the best approach to getting the switches off, trying to cut the little plastic tabs with a hobby knife didn't work so I tried heat to kinda melt them, this was too slow and the plastic would cool too fast to smush two within a short period of time, finally I just broke out my Dremel with a small barrel sander on it and sort of erased the flared out plastic part on the back of the switches. It did cause some minor grind damage to the back of the PCB but since it's a single-sided PCB this is of no consequence. This finally allowed me to extract the darn keyswitches - some are rows of switches stuck together- very odd.

I peeled off the membrane and noticed that it is only one piece with the PCB directly behind it. The traces on the membrane are mostly behind a white backing on the rear of the membrane that cannot be removed. Almost like this white backing is painted on. I was thinking that I had a lot of trace repainting awaiting me since this could not be peeled back so I figured I'd check continuity. Get this, not one single trace on the PCB or membrane had a break anywhere. I could find nothing wrong with the traces but I still only have one single key that works - Caps Lock.

I checked all the diodes on the PCB and the resistor pack, and neither is the problem. So I'm thinking that for my board in particular it is either the 8048 (I hope not) or the 4051. I'll have to hook everything up and see what the oscilloscope shows.

IMG_2553.jpg

IMG_2554.jpg

IMG_2555.jpg
 
notneb82 said:
I checked all the diodes on the PCB and the resistor pack, and neither is the problem. So I'm thinking that for my board in particular it is either the 8048 (I hope not) or the 4051. I'll have to hook everything up and see what the oscilloscope shows.
Click to expand...

If you find the 8048 is bad, we can work out a mutually beneficial arrangement. Lend me the keyboard for a couple of weeks. I will program a replacement 8748, run some scans of the communication between it and the terminal, then return it to you with a new controller installed.

I originally had a QVT-102A back in the mid 80s. Typed millions of characters on the keyboard, I'm sure. My fingers still ache from the high spring force. I'm really looking forward to using this terminal once a mechanical PS/2 keyboard is interfaced to it. :)

First I need to replace the Rifas on the power supply board. The mainboard also had a leaked CR2032 but the damage doesn't look too bad.
 
Looks similar to some of the cheaper PC and laptop keyboards.

The white coating masks over/insulates the conductive black traces.
The conductive coating can deteriorate over time. Try metering from the edge connector to the pads and see what the resistance is.
When a key is pressed, there is conductivity between the pcb and black centre dot, sometimes in the kohms rather than ohms.
I have lightly cleaned these type of keyboards and restored functionality for a while.
 
mark0x01 said:
Looks similar to some of the cheaper PC and laptop keyboards.

The white coating masks over/insulates the conductive black traces.
The conductive coating can deteriorate over time. Try metering from the edge connector to the pads and see what the resistance is.
When a key is pressed, there is conductivity between the pcb and black centre dot, sometimes in the kohms rather than ohms.
I have lightly cleaned these type of keyboards and restored functionality for a while.
Click to expand...

Yeah, I don't have a lot of membrane keyboard experience except for a handful of 90's keyboards and they usually have a sheet of plastic taped to the back of the trace membrane - making it easy to repair breaks; so I was surprised and concerned with how to repair any extensive breaks since peeling this off is really out of the question. I suppose one would have to get some conductive paint and just paint each broken trace directly to the white insulation and then either cover the new traces with tape or put some solder mask / nail polish over the new paths.

I did check cont from the edge connectors to each pad, but I did re-meter them today as you suggested and the highest resistance I got was 15ohm on the longest trace with the others ranging from 5-10 ohms depending on how far away the keypad is. This is why I think there is some sort of logic issue. The terminal knows whether or not the keyboard is plugged in - it gives an error on the screen which shows as a cursor and an H (looks like this: _H) which according to the manual, means the keyboard was not found. With the keyboard plugged in, the speaker on the keyboard will give one audible beep and display a normal blinking _ as the cursor and no keys produce anything onscreen except for caps lock which shows "CAPS" at the bottom left of the screen and clears it when pressed again. At this point in time, I haven't ruled out the terminal's logic board, but I want to see what the oscilloscope shows of the keybaord's 8048 and the 4051 chips. I did check the keyboard cable's continuity to rule it out as a starting point prior to completely tearing down the keyboard - no problems from any of its wires either.
 
From memory, the 15ohms is good.

You could eliminate the keyboard by emulating a key - a substitute key + resistor in series across the right pins at the cpu maybe?
 
You must log in or register to reply here.
Back
Top