The Sinclair QL Forum

Hi,

This topic I have wished to bring up for quite some time. Has anyone observed slight distortion of the case as QL warms up? I did. I was thinking what could cause such distortion. I tried to capture in the picture.
The warp is apparent on the right, in the worst case, causing top lid touching down on MDV2 head board. Sinclair rectified with a stand off spacer (plastic collar).

Has it not to do with two distinct materials (PCB, case), being firmly fastened with screws, due to some thermal dilation, leading to sort of `bimetallic effect'?

Provided above correct, which really is only my assumption, could it be rectified by making mounting PCB holes oval rather than strictly circular allowing for thermal dilation to compensate?

Tomas

I recall a discussion about this which happened at Sandy in mid-85.

Sinclair redesigned the base molding twice. Once because the mold was straight and when the parts cooled they had a slight bend. This changed how the screws tightened. The old tightening order (middle four then end four then snug) no longer worked as I was stripping out the front middle screw by the microdrive fairly consistently. I only had to do it three or four times to spot the change. Over time, the lower case was imposing its bend on the top case at the weak point where the microdrives started. The bend was about 3-4mm.

Sandy received a couple of batches of 1,000 brown box QLs. One batch was all new with linear serial numbers and Issue 7 PCBs. The other was made from whatever they had laying around in their junk pile (or at least it seemed that way) with a mix of Issue 5/6/7 boards, case moldings, microdrive types, etc etc.

I don't recall in their conversation why the other redesign happened, just that they said that.

Dave,

have not seen that in any of my QLs. Provided all the case screws are present and tightened, the QL case forms a torsion box that simply cannot warp much. If you operate the QL without a properly closed (just snapped-on) upper case, however, I can very well imagine some warp might occur. (Doing that is bad for the power connector as well - It locks with the upper body, and pulling and plugging the power plug puts a lot of strain on the PCB connections if the upper case is not properly closed and fastened - until they finally break)

Tobias

With the lower case having a slight warp, the front screws being short and having a higher load over a shorter thread distance, you'd have to do the screws up in order 1-4-2-3. I was doing 3-2-4-1 and the 3rd screw would sometimes strip its thread. I don't think it happened the first time the case was screwed up, but it did the 4th or 5th time.

It only happened on the pre-warped bottoms. Mostly early Issue 5s, though later when they were just chucking any recycled board in any recycled case it wasn't a hard and fast rule any more.

Hi,

Thanks for taking your interest.

All screws present and not over-fastened, no threads stripped. My QL is issue 6, with only 4 support rubber feet. It has no green serial, but had in place a yellowing label with some numbering, quite uncommon possibly.
When disassembled, bottom case shows no warp, perhaps I should take a second look for a pre-warp as per Dave's argument.

Regarding torsion, I can observe that as well. Both warp and torsion develop on the right side, where QL builds up some heat arround the heat sink, the bend starts from MDV1 unit rightwards. In cooller days smaller, in summer days of course bigger.
I was hoping to compensate by making two right PCB positioning holes, and one right mounting screw hole, bigger with some +0.5-1mm tolerance, to allow for PCB dilatation, if not too silly an idea.

This is only an assumtption, possibly entirely wrong.
Tomas

Hi,

With the lower case having a slight warp, you'd have to do the screws up in order 1-4-2-3.

I realised, I also need help with above screws numbering, simply I would not know which is what?

Many thanks so far.
Tomas

Does replacing the 7805 with a switching regulator to eliminate heat fix the problem?

With the QL sat in front of you, looked at from below, the screws are:

1---2---3---4
KKKKKKKKK MM
5---6---7---8

K is behind Keyboard, M is behind Microdrives.

Order was supposed to be:

5-8-1-4-7-2-6-3.

With this order and a bent lower, you never see stripped threads unless they're over-tightened.

Hi Dave,

Thank you, screw - unscrew order the same, or the reverse?

Does replacing the 7805 with a switching regulator to eliminate heat fix the problem?

Not yet, I somehow thought I have uncovered some common shortcoming typical for 5/6 Issue boards/mouldings. I am still interested in finding a suitable solution for what it had been back in the days.

I rember there was some sales statistics going on here in the forum, breaken down by serial numbers, and Issue models. I am curious what really was the mass produced QL model. Issue 5,6, or 7 (was that Samsung)?

Many thanks
Tomas

I can’t answer the question of overall sales, but of the QLs that I have (five cased, and some that are just PCBs) they are a mixture of issue 5 and issue 6 boards as far as I know. But then I only got interested in them In 2011...

I think somewhere on this forum I posted a picture of a QL circuit board which Sinclair never finished making.
Edit: here it is in this post It’s an issue 6 board.

Mark

Hi,

Cool, perfect, barely populated Iss#6 board. Now having close up in front of my eyes, simply proves making even slightly bigger mounting PCB holes, not brightest the idea, as there are some tracks going near by?

Tomas