View Full Version : Sinclair QL - Where to start.

May 21st, 2015, 06:49 AM
This machine was given to me along with a copy of QL Quill on the condition that I repair it. I only very recently got myself a Spectrum and the QL is still a giant mystery to me. I wasn't even aware until now that Sinclair sold a north american AC adapter for it.

Anyways, the machine is as seen below. All the case screws are missing. The original ROMs are missing but a Minerva ROM kludge was included however it is also a bit of a mess with small reworks and missing pins. What does that switch even do? There's also evidence of work done to the logic board. The RF connector has been modified for composite (I would assume PAL but what about that 120v AC adapter?) and there's a circuit driven by a 74HCU04B1 glued on top of the ram. There's also a random green wire coming from somewhere under the PCB and all the wiring for the keyboard I have no idea where it goes. One thing I do know however is that the keyboard's membrane was supposedly replaced recently.


May 21st, 2015, 11:13 PM
Mmm, can't see the pics right now; but when back at home I'll give it a look.
Most of the mods you mention doesn't sound to me too standard. The keyboard is a pro one or is it still using a membrane?

Have you already powered on the machine or want to do first some basic checks?

May 22nd, 2015, 05:05 AM
It's still the classic membrane keyboard and no I have not powered it on yet due to the fact that I currently have no ROM installed in the machine (that and my only setup machine right now that can display PAL is having disk problems). You'll get a better idea once you see the pictures.

May 26th, 2015, 11:35 PM
Ok, from the top of my head:

- The Minerva kludge has probably a 27C512 that holds the Minerva (48K) and a Toolkit II ROM (16K). AFAIK the switch is used to enable/disable the "extra" 16K of the toolkit. Indeed there're two pins missing, you'll probably need to repair them or wire to the socket to make it work. The last Minerva ROM is version 1.98, you can burn it easily.

- You don't need to worry about PAL or NTSC as the composite mod seems the monochrome one. Maybe this QL was used in a professional environment with B&W monitors; really weird as it has an digital RGB connector that gives better quality and color no matters the color standard.

- No idea about the 74HCU04B1... and the green cable, is the one over 8049?. Never seen it, you'll need to check the pinout and figure out the possible uses, this controller is used for serial, sound and network...

- The cables are IIRC the led's connectors, you can leave unconnected. In my QL are wired from up (being up the side near the rear of the computer) to down: red, black, white, black, gray (green in your case?), black

May 27th, 2015, 02:19 AM
PAL vs. NTSC does present another issue, though, even in monochrome - 50 vs. 60 Hz timing.

May 27th, 2015, 02:44 AM
I've never found a monitor not capable of doing with 50/60Hz; but indeed, it can be an issue depending the TV set.

Anyway, the PSU seems to match a NTSC machine; so probably 1) the original user wasn't able to use the RGB (no idea why), 2) modded the horrible NTSC RF to be composite monochrome (pretty good quality in hires). I bet this QL works at 60Hz, if so, it's a rare beast :-)

May 27th, 2015, 08:07 PM
Okay then so the LED wiring, AC adapter and the video output questions have been answered (I got a PVM that can sync to just about anything). I'll pretty much have to strip down and rebuild the ROM board because it is a total mess but which socket should I plug it back into?

May 27th, 2015, 11:23 PM
You can use both; but I usually plug Minerva in the leftmost one (IC33).

May 29th, 2015, 07:16 PM
Well rebuilding the adapter did not go as planned at all. It's a very cheap PCB. Extremely thin, extremely thin traces and pads only on one side. No matter how careful I was the traces and pads pretty much fell off. About half of them in all were lost. The socket was toast and the socket pins were toast.

Talk about top-notch British engineering. ;)

I'll try to continue rebuilding it but the total lack of pads to solder to on the flipside doesn't make my life easier.

May 30th, 2015, 09:47 AM
Well it's no more pretty than it used ot be, but the board is rebuilt back to what it used to be before I touched it.

(The black areas are just marker so I knew where the traces and pads used to be)


Electrically it's intact now. Now to take a break form that I went to track down that mysterious green wire and repair the composite mod. The latter when I tested was shorting to ground because when the cover was snapped back on the RF module the cable was pinched. I fed a new yellow wire through the base of the enclosure and soldered it directly to the pin on the underside for monochrome output.


There are signs of modifications around the video port. Two traces were cut and a black wire patched between the cuts to ground, which made no sense as the traces on the other sides of the cuts went to ground as well... The brown wire took what the QL diagrams tell me is the "Composite PAL" and fed that to VCC via pin 40 of the 8049H. What the..... That mysterious green wire also tacked to Vcc, so it was to supply power to...something. Any possible explainations here?

May 31st, 2015, 11:49 PM
I fed a new yellow wire through the base of the enclosure and soldered it directly to the pin on the underside for monochrome output

Much better :-), the original hack was really dirty.

To what pin of the 8049 goes the green wire?

May 31st, 2015, 11:52 PM
Pin 40. Vcc.

Still baffled why it also links to PAL out. Wouldn't that nuke any video device attached to that pin?

June 1st, 2015, 11:18 PM
Seems extremely weird indeed, I haven't seen the QL's video circuit; but feeding +5V backwards can't be good, maybe it's modified in some other ways. The problem is to recover the RGB output if it's not working :-(

EDIT: Definitely +5V routed to the composite output of the MC1377 is not good...

http://i.imgur.com/xU5teQim.png (http://imgur.com/xU5teQi)

June 3rd, 2015, 04:18 PM
I really need to see the schematics or some sort of variant diagram for the NTSC version. God knows what Sinclair was doing.

Edited: Well, promising news at least. I verified everything, threw it back together, plugged it into a TV and crossed my fingers.
The monochrome output appears to be PAL so the image rolls but I can see the "garbage screen" initialization, followed by a black screen with MINERVA in the corner. :D

Thing is, that's as far as it goes. The function keys do nothing, typing on the keyboard does nothing and even trying to autoload a wafer does nothing that I can see. It just sits there.

June 4th, 2015, 11:19 PM
At least that's good news, the main ICs are working.

Without keyboard you can't go beyond the initial screens (Minerva in the corner and the following with info at the top of the screen), you need to press F1 or F2 to select between TV or monitor.
The keyboard membrane is 99% non functional, it's as bad as the one used in the Spectrum 48K, typical Sinclair quality. The same goes for the microdrives (the cartridges, not the drive), if you have some microdrives check if the small pad under the magnetic tape is worn out or it has even disappeared, without it, the microdrive won't work, The drives itself are not specially prone to fail, if their leds briefly lights at boot (at least the drive 0) and they make noise for a second, they're probably working.

June 5th, 2015, 06:05 AM
Thing is, I was assured the membrane has already been replaced, AND I repadded all my wafers last year. At this point I'm trying to make it do anything.

June 5th, 2015, 01:18 PM
Well, a new membrane is easy to spot as it's made of mylar & acetate (or something very close) and the old one is, as you must know, made of some cheap plastic/paper extremely weak. If it's really replaced and it's clearly a new one, there's obviously something wrong with the keyboard circuitry. You can always check the keyboard matrix and create a fake "F1" press jumping the appropiate pads in the keyboard connectors, this way you can check if the circuitry is working or not.
About the reppaded carts, have you seen them working?, never tried myself to repair a microdrive; but seems that 50% of them keep failing. The method is far from perfect...

June 5th, 2015, 02:32 PM
I have yet to test any of mine however that aside just the drive operating is an indication that the computer is functioning enough to communicate with the mass storage hardware.

June 23rd, 2015, 04:02 AM
Hello there :-)

Have you visited the QL Forum (http://www.qlforum.co.uk/)?

Lots of information and help available there.

And I'm sure the members there would love to know more about you modified QL.

Oh, the back wire jumpering between GND and GND. That may have been a modification to improve the performance of the data transfer of the microdrives. QL's are known to have rather a high level of electrical noise on the circuit board.

The Minerva board is a third party extra, so for once Sinclair are not to blame :lol: