Telequipment Serviscope s31 oscilloscope problems

[Bignumbas]

3 months ago I obtained an old Serviscope s31 in a non working condition. See photos.
There was also a Goldring G99 turntable and an AKAI 910 valve Reel to Reel tape deck. 
I have finally got it back to working condition.
But, the schematic available on the internet does not exactly match what is and I cannot find a revision.
Is there anyone who can show me the correct wiring of C27 trimmer cap in the time base & horizontal amp on page 26 of the manual.
I removed this early on in the restoration because it was wrong according to the schematic.
In the schematic it is wired with the other caps c27-c31 in parallel  but in reality it is not it is on the selector side of the switch in series going back to v7A anode.
I have left it out for the moment.
It can be seen in the circuit photo halfway up on the left, looks like a small can made of aluminium.

The previous owner had also removed other parts of the circuitry and miss-wired elsewhere. He was TV tech from the 60's and died about 5 years ago at 80.
So far I have replaced all the molded paper caps, v10, R107, R108, MR1-3, NT1-3, CR1, C43,C41,C41,C19, a multitude of resistors and some small caps.
v10 was acting like a Van der Graaf generator.
I replaced the Selenium rectifiers with a 1N4007 and a 100 ohm 1 watt resistor inside the original casing
The Germanium GD5 Crystal diode with a 1N5711 schottky diode, it was open circuit.
The Neon tubes with NE2 90v bulbs, all the leads were corroded through at the base

Other errors in the schematic include  NT3 being joined to junction of R127 & R128 in Power Supply, I have not done this.
CA = C1, CB = C2,  RA = R1  in Horizontal Amp.

Most frustratingly there are no voltages noted anywhere. This caused a small problem with replacing the Selenium recifiers.
By experiment with resistors and the 1N4007 diodes I now have 1250v after MR3 and the controls all work on the CRT. ie Astig, brightness, focus.
With the Selenium rectifiers I was only getting 900v.
Still waiting on some parts on the slow boat from China.
I am going to try some 1600v metalized film caps to replace the C49 and C48 0.5uf 1500v paper caps.
I ordered these https://www.aliexpress.com/item/Free-Shipping-5pcs-CBB-105J-1KV-1uF-P25mm-Metallized-Film-Capacitor/32381712753.html
and was going to put them in parallel to give me .5uf.
There seems to be some fluctuations, random spikes in the power supply.
So hard to find high voltage caps today.

If anyone can send me more specs/tips etc. it would be much appreciated.
I included the service manual pdf as an attachment.

[Cyberdragon]

The Neon tubes with NE2 90v bulbs, all the leads were corroded through at the base

I sincerely hope you meant you replaced the indicators and not some regulator tubes. If you did, you really need to get proper regulator tubes or you're asking for trouble (or at the minumum some zeners, they won't be as good a regulator tubes though). This would explain the voltage spikes. Indicator is to regulator as a normal diode is to a zener.

[Bignumbas]

This CRO was made in 1958.
NT1-3 are only to supply the calibration square wave which is working well with the replacement neons.
There are so many neon tubes and the manual does not specify anything except it being a NeonTube2 type which is what the NE2 is and 50V.  I measure 50v AC after NT3
As far as I can tell this is an equivalent part:
A1A (NE-2) NEON GLOW BULB WIRE TERMINAL BASE - 1/17 Watt 105/125 Volt T-2 Neon Glow Miniature Bulb Wire Terminal Base.
Breakdown Voltage Max (AC) 65, Breakdown Voltage Max (DC) 90, Series Resistance 150, 25,000 Average Rated Hours.
Have a look at http://www.tiffe.de/roehren/neon.pdf it has a large list of neon tube equivalents starting at page 100.

This has nothing to do with the power supply to the CRT to my knowledge.
I think the 60 year old C49-48, 0.5uf 1500v  paper caps are the culprit with the slight voltage fluctuations I see.
I live in fear of them going bang! 
I have no Signal  Generator as yet to calibrate the 1v square wave output but it definitely puts out a solid square wave as per the photo.
And I am not even going to go near replacing this neon tube circuit with silicon parts.

Cheers

[Bignumbas]

Has anyone got the actual specs for this scope so I know what it is capable of?
I have finally replaced everything to spec going by the schematic.
I had a whipping cat's tail at the start of the trace so I swapped v5 and v6  and that fixed that.
But I cannot get a reliable trace when I go up and past 500hz let alone 1mhz.
I bought a Hantek 20mhz usb Oscilloscope so I have now got exactly 1v pp for the calibration signal
The calibration square wave fits on screen exactly as described in the manual now.
There is however a sloping top and bottom on the square wave.
I have bought a proper probe but the trimmer cap builtin does nothing.
There seems to be a sync problem with the higher frequencies.

[rgarnett1923]

Hi,

Do you have the full manual for the S31 or just the schematic?   I have the full manual if you can't find it and also the valve and CRT data sheets although these are not hard to find.

The specs are on Page 1 of 26.  The scope has a bandwidth of dc to 6 MHz (-3 dB) and a rise time of 60 ns.  This is consistent with the use of 6BL8's (ECF80's) as I recall these were used a lot for TV video amps which needed at least  dc to 5.5 MHz bandwidth.

Old microwave oven's have good quality HV caps.  I pulled one that was 0.9 uF 2.2 kV with an inbuilt 10 Meg bleed resistor.   They may be a bit big, but the power supply tranny on my S31 has been moved outside of the scope as the original tranny's smoke escaped so I have a bit more room.

It's a pain that the manual doesn't have the power supply voltages.  My scope is in such a bad condition I have to replace all the caps and a lot of the resistors so I can't get a valid measurement of the voltages at the moment.  One of the caps (C41a and b)  had bubbles coming out of the rubber seal.  I pulled it out. Both halves gave 32.5 uF an the bridge, but sadly the "red" half gave a 250 volt megger reading of zero.  The "yellow" half was perfect.

What voltages are you getting for HT+1, HT+2 HT+3 etc?

I am going to pull all of the old power supplies caps, resistors and diodes out and replace them with switched mode, much lighter and better regulated.  So each voltage will have it's own regulator and not have to rely on resistors.  The original EZ81 rectifier and the seleniums were replaced about 30 years ago.

I am also thinking of rebuilding the circuit with a PCB instead of the point to point wiring, I know that the purists will think this sacrilegious, but I will still have the original valve and CRT design without the birds nest underneath.  I may even be able to scrape a few more kHz bandwidth out of it with good PCB layout and modern passive components. Who knows.

I am also going to put a small fan in the case  Cool electronics is happy electronics!

I think I will also get rid of the UHF connector and replace it with a good quality BNC as the old one is very corroded and you have to use an adapter for modern BNC probes.

You may have trouble compensating modern probes with this scope as the input capacitance is a bit high at around 30pf. Most modern scopes are around the  15 pF mark.   I notice they have a 15 pF cap C15 on the attenuator output.  Do you know why they would do this?   It seems to me that removing this would make newer probes work better as the input capacitance would be halved.

If you put the scope on AC, the high pass cutoff is 2 Hz (-3Db).  If you want to go lower put the scope on DC and use an external capacitor.

Of course as this is not a storage scope really slow traces are a bit pointless unless you set up a camera.  The ubiquitous mobile phone camera could be used and you could probably trigger the camera with the mic input or something. But let's not get carried away with all this.

It will be good to get it going as it's been sitting in my loft all neglected for the last 20 years.  In it's day it was a very good oscilloscope for TV and radio repairers.

[Cyberdragon]

SNIP

1. be careful with switch modes, LOTS of EMI noise! That's why modern scopes have sheilding. Not impossible, but might be more of a hassle than it's worth.

2. A PCB? REALLY?  No...just no. Not only is it a disgusting idea it's also totally pointless. No, you won't get more bandwidth, that would require a total redesign of the scope. If you want to design you're own tube scope then by all means get some junk and do it, but please don't tear up a working scope. Follow the rule "if it ain't broke don't fix it"!

[Bignumbas]

Hi there.
Yes I have a service manual.
My CRO seems to have had some revisions, its not exactly the same as the schematics. ie R107/108  were both 6k ohms, both had failed.
I have nearly replace all the resistors now. After running it for a few hours  after I got it going I had more fail.
I have finally replaced the last paper capacitor components C49/48, I had to use two 1uf/1600v metalised caps in series.
I have left all the old cans in place, all the new components are a fraction of the size and fit on the wiring side easily.
The 4 capacitors to replace C41/42 are only a bit bigger than a thimble each.
The best I can do is get a 2mhz square wave on screen. I have ordered a new signal generator from China to test further.
I measured the following  voltages while testing the calibration square wave.
The transformer puts out AC 365v to V10 which then converts to DC 410v
The voltages then output are HT+1=380v, HT+2=250v, HT+3=240v, HT-1=445v.
The voltage tripler  has 446v between MR1 and MR2, 830v between MR2 and MR3 and 1276v at C49, 1220v at C48 and 1100v going to #2 grid on the CRT.
I think replacing the selenium rectifiers was the major thing in fixing the HT volts. 
I put a 1N4007 and a 100 ohm resistor back inside the original casing to keep the vintage look. The do not look the part , they are so tiny.
The Germanium diode (CR1) also had to be replaced it was open circuit, that improved the signal. I used a 1N5711 Schottky diode after much research.
The Time/Cm switch  did not have C27 installed at all and had a trimmer cap installed on the V7A anode side?? why?. When I modded this back to the schematic the 1us works when I tweak the trimmer
The only problem I have now is C17. It was no good and I have replaced it but I can get no adjustment from it when calibrating the square waveform.
I think you are very brave if you are going to put in a component board I have seen this in later models.

[rgarnett1923]

Hi Cyberdragon,

The only bit of my s31 scope that is working is the Y Amp which I have spent a day fixing up. The rest is stuffed.  The power supply is all over the place and the timebase is like Monty Python's parrot, late, stuffed, dead, fin.

The switch modes I which have designed will be going into the tranny box which is attached to the back of the scope.  This is fully shielded.  The problem I have with the existing power supply is that the original transformer failed and my uncle who owned the scope put an oversize transformer in an aluminum box mounted on the back panel of the scope to replace it.  This box is the same cross section of the scope  with a depth of about 120 mm. 

The tranny is about double the linear dimension of the original and so weighs in at least four times as much.  The new tranny didn't  include a separate heater winding for the CRT cathode so my uncle put a small 240 to 6.3 tranny in place of the original, mounted on a piece of plastic to provide additional insulation.  So the whole thing weighs a ton and is very awkward to move around.  It's so far removed from the original construction that there doesn't seem to be much point in worrying about maintaining it's original form.  Using switched modes allows me to get rid of the big tranny making the thing a whole lot lighter.  I will probably keep the tranny for the CRT heater as this is not so big.


My uncle also got rid of the rectifier valve and seleniums replacing them with strings of 1N4007's.  The high voltage caps on the negative tube supply were replaced with strings of  polyester caps.  The whole thing is very ugly. 

He didn't much care what it looked like he just needed the scope working to repair TV sets.

The carbon comp resistors are falling apart. One of them , R26, a hundred ohm, going from the anode to one of the Y Amp triode grids had completely lost it's outer casing and all of the others are extremely fragile.  My uncle has replaced a few of the old carbon comp with carbon film resistors because of this. This scope is one of the earliest of this model made so its probably five years older than my 63 years. It has black and white rather than the red and black control knobs that the later scopes had.  The other problem the scope has is all of the valve socket pins are corroded with the result that moving the valves in the sockets a little bit affects the signals, so all of these need replacing. To fix it this scope I have to "tear it up"  anyway, or it will never work again.

In any event I would never remake the point-to-point wiring; it looks awful and usually is and I don't have the patience for it.

I have seen a number of modern valve amps that used PCB's and they are really neat, so I can't see much of an objection to them.  I don't want to change the basic electronic design as there is nothing much wrong with it and it's the electronic design that interests me, not the old fashioned construction of it. One has to ask if point-to-point was so good why did God give us PCB's.  Answer, they work and are cheap. I am not trying to restore it to factory I just want to get it working reliably.  Most of the problems we seem to get with valve equipment isn't the valves, it's the passives that have to work in all that heat. So in goes a small fan.  I would put it this way: I'm a modern, older man who likes valves!!  Most of my electronics has been in micro processors, c programming and DSP, but as a kid in the mid seventies I built a couple of 100 W guitar amps one with valves the other with transistors. I always liked the valve one the most. And it had a home made PCB for the valve pre-amps!

On the issue of bandwidth; long component leads that are used in point-to-point add inductance and capacitance just as the valve pins do. It's one of the reasons that valves went away from the old light bulb style glass "crimped" lead designs of big octal valves to the miniature seven and nine pin circular base.  This gets the leads shorter for extended RF performance.  Good layout and surface mount components could reduce these inductance and capacitances and maybe improve bandwidth a little bit. Not a lot. I'll bet a PCB will not make the bandwidth worse as long as I am careful about avoiding adding capacitance between signal tracks  and ground.  I will still have to use some leaded components for the higher power resistors and for higher voltage drops but will be able to keep leads nice and short.  Of course the high power resistors which run fairly hot can cause problems with PCB's, but good layout and the fan should avoid these issues.  I use Seeed Studio for all my PC boards.  They do a very nice job at a good price.

Possibly in the light of this additional information you may rethink your "disgusting" and "pointless" adjectives to something a little less judgemental.  I don't use sarcasm and negativity against those who want to restore equipment to it's original form.  I applaud them for their patience and ingenuity.

It's just not for me.

[rgarnett1923]

Hi Bignumbas

Thanks very much for that.  It's very helpful.

I made an LTSpice model of the power supply and got;

HT+1=400v, HT+2=260v, HT+3=220v, HT-1=420v. 

I was a bit out as I only estimated the power supply currents from the likely valve cathode currents for each part of the circuit so having the real values is a lot better as the power supply in my scope has been extensively modified and so even if I replaced all of the caps I still wouldn't be sure what the supply should be.

My scope is in such a bad way I'm not trying to restore it to original.  The main transformer has long gone and the valve sockets, carbon comp resistors and most of the electrolytics are kaput  so I  have to virtually strip it down to nothing.  Luckily and surprisingly all of the wafer switches seem to be in good repair.  I got the Y Amp going albeit with dodgy supply voltages and found the switchable attenuator was working well after spraying with a bit of contact cleaner.  The UHF connector was loose and corroded so I bypassed this when I tested the Y Amp.


But I retired from work last week and so hopefully I will have plenty of time to rebuild it.  I have done a lot of PCB work for solid state stuff using KiCad where a lot of the components are smaller than a black ant so using valves will be a delight as my tweezer and vacuum sucker skills are not the best.

Regards
Rob

[Cyberdragon]

 Oh, you didn't specify that your scope was totalled.

Still I counter that point wiring is not always spaghetti. If you use terminal strips like Tektronix it's nice and tidy.

[vk6zgo]

One has to ask if point-to-point was so good why did God give us PCB's.  Answer, they work and are cheap.

When PCBs first came out, some manufacturers leapt into using them, but ran into problems with heat from the valve heaters, & resistors in the circuit causing damage to the then phenolic PCB substrate.

The Marconi company used such PCBs in their  combined picture/waveform monitors.
Damage to the PCBs  was a continuing problem, even when they were only a few years old

Even FR4 substrate material  was not immune.
I remember fixing near new Sony BW valve TVs in 1974, where the PCB was showing signs of damage.

As you are not going to use the 'scope for long periods of time, It should be OK, but I'm still very dubious about the guitar amps using PCBs.

I am not trying to restore it to factory I just want to get it working reliably.  Most of the problems we seem to get with valve equipment isn't the valves, it's the passives that have to work in all that heat. So in goes a small fan.  I would put it this way: I'm a modern, older man who likes valves!!  Most of my electronics has been in micro processors, c programming and DSP, but as a kid in the mid seventies I built a couple of 100 W guitar amps one with valves the other with transistors. I always liked the valve one the most. And it had a home made PCB for the valve pre-amps!

As I said above, I'm still dubious about using PCBs with, especially, power valves.

Possibly in the light of this additional information you may rethink your "disgusting" and "pointless" adjectives to something a little less judgemental.  I don't use sarcasm and negativity against those who want to restore equipment to it's original form.  I applaud them for their patience and ingenuity.

It's just not for me.

I tend to be of two minds on this.
My feeling is that if the equipment has never had any obvious work done to it , try to go for a "showroom" look, but if it has had the usual amount of work, mods, etc done (& equipment that "has to work"will inevitably end up with "non-original" parts in it), this is part of the history of the device, & just as important as the original condition.

[fenland787]

Hi Folks,
My first post on this forum. I just won a very early S31 - pre-1958 'face lift' version anyway - and a 'C1' Scope calibrator. The latter is now (almost!) functioning but I've not started on the S31 yet.

Re the request for S31 voltages in the OP, I also found someone selling a partial photocopy of the S31 service manual which I got because it seemed to have a page I'd not found in any of the downloaded versions showing anode and grid voltages to be expected on each valve (tube) and CRT together with the HT supply levels.

I hope it is attached to this post.....

[Valveboy]

Hi All,

       I have been reading your posts and I also have a S31 in very good condition but the input is dead!
Before I start fault finding anyone have an idea.

Thanks

Mark

[Bignumbas]

The fault could be anywhere.
Do you have a flat scan line on the crt?
I would check if there is an open cct somewhere from the input through to the valves.
Are all the valves good, heaters working?
You cold try swapping the valves around and see if the fault shifts somewhere else. but remember where they came from because some of them are supposed to be matched pairs.
Very hard to diagnose without see it in the flesh 

Good luck!

[Valveboy]

Hi,

    Yes I have a flat scan line! All valves seem to be heating up fine but I will have ago at swapping them around.

I’ll keep you posted

Thanks

Mark