At 18:26 28/09/2001, Tony Duell wrote:
>But I've
>never seen a PET fitted with one of those.
OK, I'll go with you on that one. I'm basing my dual-tap idea on the fact
there's three lugs on the input side; I'd guessed that one might deliver
1/2 the voltage. I've not tested it in any meaningful way.
>There are several possibilites.
>
>Firstly, we have 50Hz mains, the US has 60Hz mains. Therefore, for a
>given load current, smoothing capacitor value, and so on, there is more
>ripple on a PSU in the UK than in the States. I doubt it's that marginal
>('rock steady' -> 'wobbling' sounds a lot more than 'marginal'), but it
>can't hurt to check.
I'm guessing there must be a rectifier up in the monitor, since that only
takes 20vac. Last time I looked, there's no really big caps up there
(unlike the one in the main case); maybe the VDU supply is more plausibly
marginal than the mainboard?
To give you some idea of the wobble - it's oscillating through around 1mm,
possibly a shade less. It's quite a rapid wobble - not 50 times a second
though. It seems pretty consistent across the whole screen, but possibly
slightly worse on the side of the main transformer. There's also a
noticable downward bend on the top line which doesn't seem to be present on
the bottom line. However, I went crosseyed trying to look at it.... I'll
have to set my camera up to shoot a short movie of it, so I can do a
frame-by-frame analysis of the motion (assuming that's necessary).
>Secondly, stray magnetic fields from the transformer could be affecting
>the dispaly (by deflecting the electron beam -- electrons are not smart
>enough to realise they should only be deflected by the fields from the
>yoke :-)). Did you mount the new transformer in the same position _and
>orientation_ as the original one.
Yes - I've not unmounted the transformer from its base plate, so I've no
choice about orientation. However, it may be orientated incorrectly on the
base plate of course.... Also, it's worth noting that the US transformer is
a different layout anyway (primary inside secondary, I think).
>Thirdly, the machine might be designed to accommodate stray fields and/or
>ripple _at a particular frequency_. Typically the vertical deflection
>frequency of the display will be the same as the mains frequency, so
>although the ripple/stray fields will make the image slightly
>non-straight, it won't wobble.
It's definitely wobbling, although as noted above there's some distortion
as well (actually, there's slight pincushion on all four sides).
>Now, the 8032 uses a 6845 CRT controller chip IIRC. It sets the sync
>frequencies. Maybe it's configured differently on UK and US machines.
>Maybe, therefore, one of the ROMs is different between UK and US machines
>(I don't recall any setup links on the CPU board to configure this, as
>some other manufacturers used). Does anybody know if UK PETs have 50Hz
>vertical display systems? If they do, then you might need a dump of the
>appropriate ROMs.
Hmm. The wobble is no different on the UK & US machines; so I'd be tempted
to not believe this is the case. The voltage plates are on the chassis of
the machine, as opposed to the transformer carrier (which is where they'd
make more sense IMHO), so I know the UK machine really *is* a UK machine.
> > I genuinely don't know. That's the only part I've swapped in the new 8032.
> > Could a flakey transformer be making the screen wobble? How do
> transformers
> > go flakey anyway, if indeed they *can* co flakey? Or is the US chap
> telling
>
>Very unlikely. Transformers fail by open-circuit windings (no output) or
>shorts between turns (wrong output and the thing gets _very_ hot), or
>shorts from the windings to the core (ditto, sometimes blown fuses).
>Overloading a transformer might saturate the core and increase the stray
>field, but I think this is a very unlikely problem.
OK. There's no extraneous heat that I can detect, so presumably - unless
its in saturation - it's OK.
What I'll do next week is trace all the power wiring, so I've got some idea
of what pin (on the PSU connector) is doing what, which will give my
multimeter something to look at. And, I get an Oscilloscope on the 13th
October (huzzah!), so I'll point that at it as well; make sure the waveform
looks right.
Meantime, I've got a 3.5hr drive to do now, and it's late...
--
Cheers, Ade.
Be where it's at, B-Racing!
http://b-racing.co.uk
-------------- next part --------------
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.281 / Virus Database: 149 - Release Date: 18/09/2001
Received on Fri Sep 28 2001 - 16:51:31 BST