IBM 5155 analogue display fault

[Gordon JC Pearce]

Tony Duell wrote:
>>>That is not an excuse. Anyone working on IBM 5155s (or for that matter=20
>>>Friden Flexowriters, which are stuffed with setscrews needing Bristol=20
>>>Spline drivers) should have a set.=20
>>
>>Are they the ones that are the same shape as a tractor PTO shaft?
> 
> 
> Roughly, yes (although I've never worked on a Tractor...). I think of 
> them as : 
> 
> Allen : Triangular wave rapped round a circle
> Torx : Sine wave rapped round a circle
> Bristol : Square wave wrapped round a circle
> 
> Very approximate, of course (and certainly not enough to machine a tool 
> to fit them), but it gives the right idea.

I'd say the torx ones were more like a half-wave rectified sine wave, 
really - flat at the tips

>>>or had been moodifed to allow a normal driver to get them out (e.g.=20
>>>removing the cerntal pin or cutting a slot across them), said engineer=20
>>>was supposed to replace the entire PSU (and charge the customer for it!=
>>
>>).
>>
>>>=20
>>
>>This is still the case for a lot of "no user serviceable parts inside"=20
>>bits from IBM.  Billable call, and then some... 350 quid plus vat and=20
>>parts, or thereabouts.
> 
> 
> The comment on the PSU applied even if the service call was totally 
> unrelated to PSU problems, If an engineer came to, say, install some RAM 
> chips, he was supposed to check the PSU screws (and some other things) 
> and replace the PSU if necessary.
> 
> This was, apparently, for safety reasons. IF the PSU had been opened, 
> there was no way to know if any nasty bodges had been done internally...
> 
> (FWIW, I've opened IBM PSUs, I've repaired them, IMHO, properly and 
> safely, and I've put the tamperproof Torx screws back in again...)

There was a repair to a PSU which involved snipping a resistor off the 
board.  It was a ten second job, but needed to be done in the workshop 
for insurance reasons.  Oh, and so that the serial numbers all matched, 
it meant taking the PSU away, doing the mod, and bringing it back the 
next day...  all for the sake of a bean-counter's ego, I shouldn't wonder.

>>>>Having said all that, I tend to take the same view as you when it comes
>>>>to my classic car, and try and use the right bolts / screws etc.
>>>
>>>=20
>>>=20
>>>What have you got?
>>>=20
>>
>>Ah, see, I don't.  I want the damn thing to work, and reliably.  If I=20
>>can make it easy to fix when it goes wrong, even better.  That's why I=20
>>always replace the steel hydraulic pipes with kunifer, and use stainless=20
> 
> 
> Well, my Father's Citroen BX (touch wood...) has hard very few hydraulic 
> problems. I've had to replace just one high pressure pipe (to the back 
> brakes), which was damaged by an idiot garage mechanic when he refitted 
> the rear subframe after an accident repair. 

Easy enough.  You just need the funny pipe tool.  You do know that you 
do the brake and suspension pipe nuts not much more than finger tight? 
Too much and you deform the rubber seal, and it leaks.  Of course, this 
means that those not in the know horse the nuts up even tighter to seal 
it...

They're good cars, but they're starting to get expensive to buy unless 
they're utter sheds.  The early Xantias are starting to hit the "buy for 
hundred quid and spend that much again on the suspension" end of the 
market these days

>>>and not that many people have a reasonably complete set of Whitworth=20
>>>spanners and sockets any more.
>>
>>Uhm, I do.  And it's mostly old Citro=EBns I work on, all metric (but lot=
> 
> 
> So do I, but then I have a Myford lathe, which is assembled with an 
> interesting mix of BSW, BA, and metric bolts...

Argh.

>>s=20
>>of 7mm and 11mm bolts).
> 
> 
> 11mm is relatively common. It's 12mm that I find somewhat obscure, and 
> which Citroen seem to use to excess...

Hm, not run across many, with the notable exception of the hydraulic 
bleeder on the regulator.  They do, however, love their Torx screws for 
interior fittings, on the newer ones.

Gordon