Rebooting Racal 383-2 Oscilloscope

[knobtwiddler]

Hi,

A forum member has been extremely kind and let me take a classic 17” Racal 383-2 oscilloscope off his hands. The unit dates from 1972, and he informs me that it was last switched on 10 years ago.

It’s a rare unit, with orange phosphor display. With this in mind, I’m determined to take as much care as possible in getting the unit ‘rebooted’. At this point, I have not connected it to the mains, and have no intention of doing so until I’m convinced I’ve done every bit of necessary research. Of course, with units of this age, you can never be 100% sure something won’t play up when being returned to service, but I’d like to take every step that I can.

The most obvious first step should be to remove the external covers and look for any visual clues around electrolytic caps, i.e. leaking electrolyte or suspicious bulging. I will also look for any obvious humdingers such as RIFA caps (doubt it’ll have them).

What other checks would you do?

I have a large variac (about 7A output from memory), so I can ramp it up slowly – but if it contains switching regulation, I doubt that will help? I ought to add that my field is audio, and the HV / CRT area of electronics is well outside of my expertise.

Going ahead of myself and assuming all is ok: the max. BW of the Racal is 10KHz. I have other scopes and analysers with outputs. I was wondering if I'll be able to look at signals above this frequency, if I feed it from these... What about a Thurlby DSA524 ?

Any suggestions gladly received. TIA!

[dave cox]

Check it has the correct fuses!
Check it has the correct mains voltage selected.
Check it is correctly earthed, connect to an RCD and switch on but be ready to switch it off

dc

[chriswood1900]

It is certainly worth trying the Thurlby DS524, the large screen scope will need a horizontal sweep speed of 50usecs/div with a trigger input I did some experiments with a Leader large screen display and a DS524 but I needed a external horiz timebase using a function generator.
In the spec Thurlby says you need 5Mhz bandwidth on the scope but worth a try after all experimenting is fun!

[factory]

You may want something to monitor the current draw as you bring it up.
Also keep the intensity low in case the timebase decides not to run to avoid screen burn.

Found a older version of the 383 large screen oscilloscope from Airmec (taken over by Racal) here;
https://www.bidspotter.co.uk/en-gb/a...3-a8e9010e6f33

And another older Airmec large screen scope here;
https://www.***********/photos/anachr...7624562074349/

WME_Bill mentions he has a manual for the Airmec 279 here (post 6);
https://www.vintage-radio.net/forum/...ad.php?t=55842
No idea how much difference there will be in all these older models.

David

[knobtwiddler]

Thank you, all - much appreciated. I will be making a checklist and observing each point. The advice re: screen burn is particularly important, as it won't be easy to find a replacement 17" orange phosphor screen!

[knobtwiddler]

>You may want something to monitor the current draw as you bring it up

I have a current monitor. Are you advocating bringing it up on the variac, whilst watching draw? Does anyone think there could be any risk involved? With linear PSUs I think variacs are a great idea, but I am nervy with items that might have switching supplies. My understanding is that a lot of TVs from this period had switching PSUs (they are not a modern invention!). Is there a chance that the Racal from 1972 might?

[factory]

Probably not if it's a switching based PSU, what does it use? Most old oscilloscopes I've seen don't, but at least one tube based Tek scope does.
Monitoring the current should make it more obvious if something decides to short and you can switch it off before smoke escapes.

David

[Cobaltblue]

Thread re-opened for an update.

Cheers

Mike T

[knobtwiddler]

Update:

A couple of weeks ago, I managed to fire the scope up (with intensity set to low..). It became apparent that the TB wasn’t generating a waveform, so I patched in an external oscillator. As can be seen on this post here, it made some pretty Lissajous patterns: https://www.vintage-radio.net/forum/...5&d=1696544187

I looked away for a few minutes, looked back and was confronted with a dead screen…
One weird thing that I noticed: the TB rotary switch seemed to have an effect on operation, even when the jumper link was disconnected… Either the mechanical disturbance was doing something to a nearby circuit...or?? If I toggle the TB switch (with external oscillator into Y input), it seems to work for a few seconds.

I haven’t had a chance to take the lid off yet, so am posting here for a bit of preparation. The control panel looks absolutely identical to the one in Factory’s Link
The panel looks so similar (every nut'n'bolt is in same place), the only differences I can see are knobs / colour scheme. I am wondering if when it got rebranded / taken over by Racal, they simply gave it a facelift, changing the knobs and colour scheme, but leaving the rest as it was? The FlickR link suggests it’s a 19” scope, which would be the case if you measured the Perspex screen corner-to-corner (I am thinking that the tube itself is more like a 17”).

If it's got green Hunts caps in it, then they are known rogues to my knowledge. Does anyone see any other rogues in the FlickR link?

Anyway – I shall have some internal photos in the next few days, hopefully!

[factory]

The grey molded plastic capacitors are possibly paper dielectric too, I would check for electrolytics that might be leaky too.
Can't tell what is used in the monitor section due to the thick layer of dust.

David

[knobtwiddler]

Well, as the photos show, I was completely wrong about the circuitry being similar to the one in the FlickR link. It is solid-state throughout, with several T-03 BJTs and lots of smaller BJTs in the TB / amp section.

The HT PSU is encased in an enclosure of its own and I haven't opened it up and do not plan to unless I really have to.

The only non-electrolytic caps I am seeing are the gold-coloured WIMAs on the TB PCB. Unless anyone else can spot something that I haven't?

I plan to change all of the caps on the TB PCB, put some Deoxit in the rotary switches and take it from there. Any pointers gladly received, thank you!

edit - before I get ticked off, I will be treating it to a good old blast of air!

[knobtwiddler]

A couple of photos of the TB / input amp boards.

The large gold WIMAs appear to be film. Are they known failure points? I am concerned that removing them to test them could injure them if they are polystyrene (which I doubt for such a large cap)...

I suspect the smallest gold caps are polystyrene (from observing where leads come out). It is very unlikely that they have issues. Anyone?

The blue axial electro caps look similar to the Philips ones seen in much 70s electronics. It's easy to change these, so I will.

Any comments gratefully received!

[jjl]

The gold coloured Wima caps are early polyester film caps and tend to be reliable.
Paper dielectric Wima caps are just as bad as any other paper type.

John

[knobtwiddler]

Thank you, John.

Here is a view from the rear.

I pulled one of the electro caps out and measured ESR. For a 2u2 cap, it registered 47R, which doesn't strike me as particularly good! The electro caps are easily accessed, so I plan to order them from RS today, clean the rotary switches and try a reboot. I shall report back!

Note the large heatsinks (for T03 devices) in the top corners. You could use them for a decent sized power amp! The resistor in-between is curious. It seems that the v-shaped cowling is there to shield the massive reservoir cap in front of it from heat. I guess it also forms a heatsink as well.

The black crud that cakes the PCBs and components is stubborn to IPA. I am not entering it for a concours competition, so am not too worried.

edit - have not ordered the big electro caps... Plan to change the small axials and see how it goes. I don't blanket recap items as a rule, but the first cap I tested didn't look too clever (the others are giving inflated values, albeit still in circuit), so I think it is logical to change them.

[factory]

Don't forgot electrolytics were often specced at -20%/+50% tolerance or more, modern ones are precisely made, to be just below the printed value.

Originally Posted by jjl

The gold coloured Wima caps are early polyester film caps and tend to be reliable.
Paper dielectric Wima caps are just as bad as any other paper type.

John

WIMA TFM, they only gave trouble if badly cracked, or near very hot components, otherwise very reliable.

[knobtwiddler]

Most of the blue electro's that I've looked at have '2.5%' printed on them (although print is faded, so it's difficult to be sure about anything). 2.5% is a heck of a tolerance for a modern electro (good going for a film cap!), so I am wondering if it's inferring -20% / + 50% in shorthand?

[knobtwiddler]

Update: have changed the EL caps in the TB and signal path section, along with giving the rotary switches a good blast of Deoxit. I can now get lissajous patterns if I feed oscillators into the X and Y channels. However, I cannot seem to get the TB to perform any kind of horizontal scanning. I get a horizontal scroll for a couple of seconds when first switching on, but that is it. The manual trigger button seems to do something.

When feeding a sawtooth TB from a separate scope, I get the waveform shown in the attachment. It appear a little distorted and has the trace feeding back on itself in a loop. I am not a veteran of vintage scopes (my first scope was an '80s Kenwood) and have no idea if this is a hardware fault or operator error...

Without a service manual, I think that getting this scope tip-top will be a labour of love. I will have to read a book on scope design and sharpen my chops. The good news is that the tube appears to be in excellent nick, and anything beyond about 12 o'clock is too bright.

edit - while writing this, the trace has died and the fault that it originally had, whereby the trace only comes back temporarily when the TB velocity switch is turned has come back... It seems that changing the caps and cleaning switches has only worked for a couple of hours... I have basically wasted an afternoon...

[knobtwiddler]

One very obvious thing I ought to have stated: with an oscillator driving the X input and the jumper removed that connects it to the internal TB oscillator, I am using the scope as a simple X/Y display. Why on earth would the TB velocity swtch have any effect at all? It doesn't appear to be affecting the trace mechanically, i.e. from the clunk of the switch, as the trace comes back on exactly when the middle TB scan setting is selected...which makes me think the fault is electrical, and the TB veloocity is causing an electrical ripple effect.

[factory]

Using in X/Y mode you won't get any blanking of the trace during retrace. Normally the Z input would be used for controlling the brightness/blanking.

Is there nothing on the timebase output? Probably need to get this repaired first, the board is 6873-115.
The timebase not working might be causing the blanking fault too.

The other board with four identical sections, has the input amplifiers for each of the four channels.

Not related but I note in 1975 Racal had a model 9383 large screen oscilloscope, does anyone a manual for that one, or any of the older Airmec/Racal 383 versions?

David

[knobtwiddler]

Thank you, David.

Update as follows:

If I plug a pair of oscillators into X/Y it works for a while and TB generates sawtooth, then screen goes blank. TB velocity seems to have an effect, which is odd as it's disconnected.

TB oscillator only works if disconnected from X i/p. If I connect it, screen goes blank and oscillator gives -2.2VDC at its O/P.

If I switch TBV switch to off position for a while, it seems to recover when I turn back on.

Once it's been on for a while, the problem occurs more often (it'll work as X/Y monitor for hour or so from cold).

When TB is disconnected from X input, it puts out a healthy sawtooth. If I monitor the TB waveform on separate scope, it collapses into DC when connected to X...

Whatever the problem is, it seems to come on when TB connected to X input, but it also seems to come on after a while anyway. Both TB oscillator stops working, as well as general function as an X/Y monitor .

If I use a cap to couple TB into X input, it makes little difference. I get about half a screen of horizontal scroll, then TB waveform still collapses and display goes dim. If I disconnect cap and reconnect as X/Y display, still no display - I either have to turn off and leave for a minute, or put TBV knob to off for a minute, for it to recover.

Either TB waveform is weak and can't handle load, or X input is loading it too much.

I tried feeding a gate signal from one of my Tek scopes into the Y input, but that did not do anything. I've never used a scope as an X/Y monitor with external base before, so am not sure if the Airmec scope meets the same gate V spec as the Tek.

EDIT - once it's been on for hr or so and TB osc stops working, there doesn't seem to be a way of getting it to work again, even if I leave TB off for a while. Scope seems to work for a while as X/Y monitor from cold, but once fully warm the 2 problems stop it from working at all. You just get a brief flicker as you turn TBV knob - that's it.