Scopex 4D-25 Oscilloscope.

[Brian of Romsey]

Hi, I am about to start looking into my 4D-25. It has a couple of problems.

1) The horizontal deflection has a problem. It only deflects right. The position control does work, and when well right the start of the trace is visible (and the end is off screen). As the position control moves the trace left there is no display on the left side of the screen. In the middle of the screen the left hand half of the display is compressed into a minimum width vertical line with all the vertical deflection information.

2) A lot of 'noise', on both channels, when changing vertical attenuation. This is hard to describe but changing the attenuation doesn't always seem to follow the 1-2-5-10 sequence. I shall start by cleaning the contacts of the attenuator switches.

I shall start by looking in detail at the power supplies. My actual question here is what level of ripple should I see on the regulator outputs? i.e. On the current measurement resistors R663, R683, and R643. I know that the voltages should be within about 1% of nominal.

Thanks, Brian.

[Diabolical Artificer]

No idea why post's in the "test gear" section attract no or very few answer's, (BTW, you tend to get more answers in a thread if you post schematics, pics etc. Gives folk something to go on) I'll have a shot at it, maybe this will cause other's more experienced to chip in.

1) I'd check paired up transistors here, EG in a P-P see if one side or one tranny has died. Probably talking out of my pipe.

2) sounds about right, dirty contacts.

Just looking at a generic regulator, the LM317 it says line regulation 0.01% ish, load reg 0.5% ripple rejection around 60dB, therefore if your scope uses a 3 pin regulator you should see roughly the same. Not sure your scope does though as it's the same AFAIK as a Farnell scope I had, which I can't remember seeing any 3 pin reg chips in.

Sorry can't be of more help, Andy.

[PJL]

The answer to why there are few responses is that they are very difficult to fault find without hands on.

I would focus on the deflection fault and leave the noise till last.
Check the 110V and +/-20V rails first.
Monitor the plate voltages on the collector of TR508 and TR509 differential drivers and see what happens when you adjust the position control. As one goes up the other should go down.
Has the scope ever worked? It's possible the tube is damaged.

[WME_bill]

Scopex 4-25 - General Fault finding on Oscilloscopes.
I endorse what "Diabolical Artificier" and "PJL" advise.
Perhaps I could add an initial guide:
1. Measure voltage on all power rails. Accept within 5% at this stage. Check EHT voltages. EHT power rail is a common fault area. (Notes on a high voltage probe posted some years ago by me and others.)
2. Confirm setting of all controls. Usually a section in the service manual gives you this.
3. Get the service manual, or at least the circuit with voltages, and try to understand how each section works.
4. What are apparent faults in each section: Display- size, completeness: X horizontal display : Y vertical display: Synchronisation/Trigger - a common fault area- (voltages here should be within 1V). Anything running very hot, any damaged areas.
5. Concentrate on getting it working more or less, then start refining. Secondary issues like noisy switches and pots and possible ripple on power rails can be solved later.
6. Resist removing components and transistors and testing them. This will often create more faults. Don't waste the effort in changing all electrolytic capacitors yet. Scopes are made to industrial standards and components last much better than in commercial radios.
7. Check voltages around transistors and integrated circuits. Remember emitter to base is about 0.6v. Collector much greater. So also diodes: 0.6v one way, infinity the other. You might have to unsolder one leg.
Don't confuse yourself and us by posting lists of voltages to 4 places of decimals from your digital meter. An Avo will indicate quite accurately enough to fault find.
And tell us what the transistor tested does if you can, so we don't have to find the circuit, find where the component is and work out what it does.
8. Check around the tube bright-up circuit. Another common problem area. (Transistors operating at higher voltages, even within ratings, seem to fail more frequently.)
9. Do NOT test a tunnel diode with an ohmmeter or similar. It will destroy it. Almost impossible to replace now. The Russian one are possible. 1L304 or GI304 =1N3716 or 3717 at 5ma. 1L or GI305 = 1N3718 or 3719 at 10ma.
I posted specifications and notes on testing Tunnel Diodes some years ago.
10. Forget the dramatic idea of a CRT failure. In 30 years I have had only two, and quite apparent from inspection - cracked glass and loose bits inside. Usual failure is for a dim low contrast picture at end of life. Get the rest working. I have posted data on comparable CRT types,so you can often get a substitute.

Your horizontal fault sounds like one of the X output transistors has failed. Test the voltages, should be symmetrical between the two sides, as PJL suggests.
wme_bill. m0wpn

[RogerEvans]

A very convenient way to check the X amplifier is to use the lowest sweep speed and preferably an analogue multimeter. If you measure corresponding points in both halves of the amplifier you should see one ramp up and snap down and the other ramp down and snap up between roughly the same end points. Could be s simple as one X plate disconnected.

Roger

[Brian of Romsey]

Hi,

Thanks to all who responded, very helpful. Here are the results from today's investigations:

1) Power supplies look pretty good. +112 V, +20.20 V, -20.18 V. The current sensing resistors R663 and R683 have 410 mV and 449 mV across them respectively. This means 124 mA on the +20 Volt supply and 136 mA on the -20 Volt supply

2) The -1,600 Volt supply looks a bit funny, the voltage is ~1,700. I 'measured' this by measuring the voltages across C603, C604, C605, and C606 which were 147 V, 520 V, 517 V, and 520 V respectively. I expected them to all be ~400 Volts based on the resistive divider network they are attached to. This leads me to suspect that C603 is a bit duff (leaky), I shall schedule it for replacement. It also means that the other three caps have been receiving some voltage stress. All four caps are blue and are marked:

DALY
16μF 500V
DI 8/28
XC
U.K. MADE

3) The horizontal section is mostly good. The voltage at the base of TR511 swings between -5 Volts and +5 Volts in response to the position control. A similar 'signal' is observed on the base of TR509. The signal on the Sweep Out socket at the back seems to correct (it is supposed to be '8V negative-going sawtooth symmetrical about earth' according the the spec) and is, as best I can work out, changing even when the trace is not moving in the centre of the screen. The signal on the collector of TR509 is swinging about, and the position control moves it up and down on the voltmeter. Sadly the collector of TR508 is stuck at about +110 Volts. This leads me to believe that TR508 (a BF337) has failed.

So, if I replace TR508, should I replace TR509 (also a BF337) at the same time? Anyone got any suggestions for a modern replacement for the BF337? (Vceo 200 Volts, Ic 100 mA, Hfe 20, 80 Mhz, 3 Watts.) There are two BF337s in the +115 Volt regulator, TR761 and TR763, these might be candidates to go to the horizontal deflection department as finding suitable replacements in the regulator circuit will be much easier.

An interesting note is that the power supply series pass transistors, TR663 & TR682, appear to be MJE340 devices rather than BD165 as expected from the schematic.

Cheers and thanks, Brian.

[Brian of Romsey]

Found some BF337s on eBay, £1.98 for a pair. Awaiting their arrival.

Brian.

[Brian of Romsey]

Hi,

I have found some potential replacements for the 1,600 Volt cap stack. Currently the stack is four 16μF 500V in series with 2M2 resistors across each to ensure the voltage is shared equally. This results in 4μF 2,000V overall.

I am considering replacing this with five 47μF 450V 105°C Radial Electrolytic Capacitors from Maplin, with five 2M2 resistors. This would result in 9.4μF 2,250V overall. Any thoughts on the suitability?

Thanks, Brian.

[Diabolical Artificer]

Good to see you're making progress, hope your BF337's are kosher.

These - https://www.watfordvalves.com/produc...il.asp?id=3295 are direct replacements Brian, if they fit. Why did you decide to go for 47u rather than 10u or 22u?

Andy.

[Brian of Romsey]

Andy,

I chose 47μF 450V based pretty much on price, after they met the original spec that is. At less than 50 pence a pop, they seemed like exceptionally good value. The obvious closer choice, 22μF 500V, is £2.10 from RS and £2.57 from Mouser.

I have been thinking about the the resistor chain. I might use pairs of 1M0 resistors rather than 2M0 resistors, as this will lower the voltage across the individual devices. Good idea, or not worth the effort?

Cheers, Brian.

[Neutrino]

Originally Posted by Brian of Romsey

I have been thinking about the the resistor chain. I might use pairs of 1M0 resistors rather than 2M0 resistors, as this will lower the voltage across the individual devices. Good idea, or not worth the effort?

All six 2.2 MΩ resistors in my SCOPEX 4D 10A were out of tolerance including one open circuit. I choose to purchase replacement resistors with adequate voltage rating instead of using pairs of resistors.

RS Components sell Vishay VR37 High Ohmic/High Voltage Metal Glaze Leaded Resistors in packs of 25. I purchased replacement 2.2 MΩ Vishay HVR37 High Ohmic/High Voltage Metal Film Leaded Resistors from Farnell.

For safety, I did not attempt to measure any voltages in the EHT circuits.

[Brian of Romsey]

Originally Posted by Diabolical Artificer

These - https://www.watfordvalves.com/produc...il.asp?id=3295 are direct replacements Brian, if they fit.

Ta for that, I checked it out. Almost £24 for a set of four caps (inc VAT and postage). Which is a bit more than I'm up for spending on this 'scope.

That said, that's a very useful site.

Cheers, Brian.

[Brian of Romsey]

Originally Posted by Neutrino

All six 2.2 MΩ resistors in my SCOPEX 4D 10A were out of tolerance including one open circuit. I choose to purchase replacement resistors with adequate voltage rating instead of using pairs of resistors.

RS Components sell Vishay VR37 High Ohmic/High Voltage Metal Glaze Leaded Resistors in packs of 25. I purchased replacement 2.2 MΩ Vishay HVR37 High Ohmic/High Voltage Metal Film Leaded Resistors from Farnell.

For safety, I did not attempt to measure any voltages in the EHT circuits.

I suppose I should check the voltages across C601 and C602 too.

Is it worthwhile lifting an end on the 2M2 resistors to check them all? Or is it sufficient to just establish that C601 and C602 have the same voltage - roughly 400 V - across them.

And yes, I was quite ginger when measuring the voltage across those high voltage caps.

Cheers, Brian.

[Brian of Romsey]

Hi,

Replaced TR508 and TR509 with new BF337s from eBay. Horizontal deflection now working properly.

Still waiting on the caps from Maplin.

Regards, Brian.

[Diabolical Artificer]

Glad to hear your making progress. Quite understand not wanting to spend all that cash and your replacements should do the job.

Andy.

[Brian of Romsey]

Hi,

I finally got some time to get back to my 'scope repair (and Maplin had come through with the capacitors at last). First off I removed all four capacitors, resistors, and links from the -1,600 Volt capacitor array on the PCB. A small point of interest here is that these capacitors are mounted on the back of the single sided PCB and so their holes are actually little eyelets, and this makes desoldering an bit of an effort. Lots of heat and wiggling is needed, but they do come eventually. The eyelets increase the robustness of the mounting because there is some mechanical strength in their attachment to the PCB (rather than relying on a track on the same side as the component).

I built the replacement array on a bit of fibreglass proto-board I had in the junk box. See the attached pictures for details. I added the 5 nF 2,500 Volt capacitor because I had loads lying about. For robustness I chose to mount the competed array to the PCB using two of the aforementioned eyelets. This necessitated two small insulated wire straps on the PCB to connect the actual endpoints on the PCB (ground and -1,660 Volts) to two of the now isolated eyelet / resistor / wire link nodes. You can see one end of the black wire in the third photo, the red wire to ground is completely obscured. I also added a polyethylene isolation barrier between the PCB traces and the capacitor array.

The four original capacitors and the five new capacitors were tested with my cheap Chinese component tester:

Part Value ESR Vloss
-----------------------------------
C603 18.55μF 1.5Ω 1.4%
C604 18.63μF 1.2Ω 1.3%
C605 18.68μF 1.0Ω 1.4%
C606 26.95μF 1.8Ω 5.0%

C60a 47.1μF 0.60Ω 0.8%
C60b 47.1μF 0.65Ω 0.9%
C60c 47.1μF 0.65Ω 1.0%
C60d 46.7μF 0.69Ω 1.0%
C60e 45.8μF 0.73Ω 0.9%

Interestingly trying to measure the voltage across the capacitor array varies depending on how I measure it. The voltages across the five capacitors were: 318, 362, 357, 355, and 332 which adds up to 1,724 Volts. Measuring (with the same DMM) across the ten 1.0 MΩ resistors gives; 156, 157, 169, 168, 168, 167, 165, 163, 158, and 158 which adds up to 1,629 Volts. Not 100% sure I have a clear idea of why, I'd appreciate any explanations.

So, I'm sure there was a small problem - as evidenced by the original imbalance in the voltages measured across the four original capacitors - and that has now been removed. Overall, I'm happier that the long term robustness and longevity have been improved at a modest cost.

Cheers, Brian.

[Brian of Romsey]

Hi,

Sorry, the posting software de-formatted my table in the previous post. It was clear and well spaced when I composed it, honest.

Brian.