Cossor 4000 oscilloscope

[RogerWalker]

At last! a precious hour or two to investigate the many excellent suggestions:
1. With PDA supply disconnected, there is no sign of a trace visible - even in total darkness.
2. I replaced all the high voltage 4.7nF and 500pF capacitors in the bright-up circuit with new ones. No difference.
3. I found one of a chain of 3 x 2.2Meg high stabs to be O/C. Replaced it with new. No difference. It was the one immediately after the 68/82V zeners, R36 on my circuit.
4. Checked the zeners - all seemed to have the right voltages acrosss them. I remembered the fact that some of them are at [what should be] -940V with respect to chassis, so avoided nasty shocks! Checked all the 1N916 diodes in situ using a DMM diode test. All gave seemingly OK results.
5. Noted that there's a 100V voltage drop across the 47k CRT supply smoothing capacitor, so something's still dragging the CRT supply down - though there's still adequate PDA EHT >10kV.
6. There is still nothing visible on the screen unless I press the beam-finder button.
7. Tried setting the beam current pot. There's supposed to be 0.4V across a 1k series resistor. The pot made little difference and the voltage across the 1k resistor was steady at 0.6V, 50% high.
8. Tried swapping round the three MM3001 transistors [plug in bases are handy sometimes] in the bright-up and amplifier circuits. No difference so put them back as they were.
I had to stop there - HID to arrive shortly expecting supper to be ready!
Any further input welcome - I'm a bit stuck now.

[RogerWalker]

As my last post would not make much sense to anyone not in posession of a circuit diagram, I have attached a .pdf scan of the EHT and bright-up circuitry in question. Apologies for the omission

The 1k resistor mentioned with regard to setting the beam current is R41.
The 47k CRT supply smoothing resistor mentioned is R42. Components I have replaced are: C11/12 [4.7nF 3kV], C2/C19/C23/C25 [500pF 10kV] R36 2.2Meg, EHT sticks D19 and D13. TR1/2/5 have been swapped around with no effect.

If my Ohms laws serves me correctly, 100Volts dropped across R42 suggests a current draw of 2mA, so not as heavy as I first thought.



Regards

Roger

[maninashed]

Hi Roger, 2ma is probably correct for the current draw on the crt. The fact that the 2.2M ohm resistor in the feedback chain was o/c would mean no regulation. Have you tried readjustment of the correct cathode voltage since? Try also adjusting back the grid bias as this will cut off the display. I'm surprised that you don't have something visible by now Roger. If the beam finder puts up a display, you are not far off, check the beam finder circuit and see where it connects in to the crt control circuit, and check all silicon and resistors thereabouts.
Bill

[RogerWalker]

Thanks Bill
The cathode voltage remains at -750V, the set EHT pot doesn't have any more range. I am not sure how to adjust the CRT grid bias as it seems fixed by the two diodes D20/21 and R23/24 in the bright-up circuit. I have scanned the beam finder circuit but have to confess I can't see how it works to limit X and Y deflection and centalise the traces. Copy attached for easy reference.

So far the 4000 circuit seems identical to the CDU120, but I think the delayed timebase and trigger arrangements are different.
Regards
Roger

[maninashed]

The cathode voltage is too low, but if you get a trace with beam finder pressed we can run with that for now. The problem could well be with the x amp, if the x amp is faulty the beam could be deflected to one side and you won't see it. Check that you have equal voltages on the outputs to each x plate, this is where another scope comes in handy. If you get a trace with beam-finder pressed, the fault will lie in the connected circuits. Long shot, but have you checked the brightness pot for continuity? Check the silicon in the output of the x amplifier. I take it you have good +50 and -50 supplies?
Don't give up now !
Bill
.

[Radio Wrangler]

One dirty trick is to disconnect systems one at a time from the beamfinder switch to see just what the beamfinder is defeating, putting eachbach after each test. A quick way to sort out brightness/Xamp/Yamp faults, if you have little other gear.

David

[WME_bill]

Cossor 4000 or CDU120. After your sterling work with the EHT supply, I wonder if your trouble lies in the Bright-up circuit. The transistors there operate at high voltages which seem to make them much less reliable. I usually look in that part of the circuit on any oscilloscope as soon as I have established that the EHT supplies are near enough right.
Have you checked the DC voltages at the output of the brightup.
Test 1 at emitter of TR1/collector TR2, with the timebase not operating- ie no trigger, no auto. The voltage should vary about 20-30v with the Brilliance control. = brilliance.
Test 2, same point. Switch from TB to X Input. This should bring up the brightness, so the same voltage change. = brightup from TB.
This circuit uses a multivibrator oscillator (TR5,6) with amplitude limited to the brightup levels required, which is then fed through capacitors (C2 and C23) to the tube grid, after rectification by D10,11. Never an easy circuit to understand. Philips were very fond it also. Look at PM3240, 3260 series for a good description.
I attach the description taken from the Cossor 3000, which is very similar, and may help. The circuit references are not quite the same, and I have noted the comparable circuit references from the Cossor 4000 as a guide. Ignore the references to TR6, an extra level transistor. I have not copied the circuits from the 3000 as they are on three different sheets and awkward to do. Send me a PM if you want them. wme_bill m0wpn

[WME_bill]

Cossor 4000 / CDU120 / CDU150. Bright up circuit explanation.
I have just posted circuit and a good explanation of this modulator circuit for the Enertec 5216 under the Enertec/ Solartron/ Schlumberger thread.The difference is that Enertec use a feed from the EHT oscillator, while Cossor use a separate multivibrator, but the explanation of the various brightness levels seems very helpful. wme_bill

[Radio Wrangler]

And pressing the beam finder button reduces the bias current in the emitter bias current sources, so the available current in the X amp into the load resistors of the amplifier cannot swing far enough to take the spot off the screen. It isn't so much a reducer of gain as a limiter of swing.

On a working scope Beam finder gives you a shrunken section of a picture, which spends a lot of its time against a pair of on-screen 'endstops'

It does the same thing to the Y amplifier.

It should bright-up the Z amplifier and in good scopes it puts the timebase into autotrigger so there should be some scanning activity.

David

[RogerWalker]

First thanks to both Bills and David for some insightful suggestions. I have read the blurb on bright-up and now have a basic understanding of how it works - but not of why it's there! I will copy the files to my tablet and re-read them under the bedclothes.

Now - some progress. There is a trace on the screen!

As suggested, I checked out the voltages on the X plates and the X amplifier in general and all seemed well. Turned to the Y amplifier and lo! there was a big discrepancy in the Y plate voltages, suggesting that one of the plate-driving 2N3119s was not drawing any current. First problem was to find the 2N3119s - they are buried in a plastic housing behind a couple of metal chassis parts, with only their wires projecting. I thought I measured an o/c 4.7k resistor in the common bias network, but when I replaced it, the old one measured OK. And so did the replacement in circuit. Seems there was a dry joint because, switching the set back on, the Y plate voltages were equal - and there was a single bright trace full width on the screen. Result! Unfortunately, the trace is immovable up, down or sideways and its intensity cannot be varied with the brightness control - and whereas before, in the fault condition, there were dual traces, albeit very narrow that could be moved around and were brightness variable, now there is just a single trace that can be focused - but little more.

I think I have to go back and check the CRT voltages again to see whether the cathode volts have come back into tolerance, and set up the beam current etc to see whether that allows the trace brightness to be varied. There are further issues to address in the Y amplifiers as there is no apparent activity when a square wave input is applied to either Y input. Still having a trace on the screen beats a blank screen anyday!
My resolve is rekindled. Wish there were more hours in the day - this retirement lark is just too exhausting!
Kind regards AND THANKS SO MUCH FOR ALL THE INPUT.
Roger Walker

[RogerWalker]

Here's a scan of the Y amplifier circuit. The 4.7k resistor referred to above is R977 and the 2N3119s are TR975 and 976. I have also scanned the layout of the components; if I can add it here I will - otherwise I'll add it in a separate post afterwards. Note the second part of the beam finder switch controlling the Y deflection [the ganging of which was not mentioned anywhere on the X amp circuit].
Regards
Roger

[maninashed]

OK Roger, we are getting somewhere. Personally, now you have a trace, I would concentrate on getting the cathode circuit functioning correctly. Otherwise you tend to get your attention diverted in different directions. I still feel that you have a bad power supply rail somewhere, check and double check the supplies to all the separate pcbs and don't forget a good ground connection to each pcb, corrosion under the mounting screws can cause all sorts of strange faults. Check the voltage rails where they enter a board. Check R21,22 and R6 and that the voltage at these points alters with the brightness control. The MM3001 is a NPN high voltage TO-39 transistor. These hv transistors can be problematic and go leaky, if possible remove them and test them on a dedicated tester.
Bill

[Radio Wrangler]

The X plates will be a lot less sensitive than the Y plates, and so will take a lot more swing of voltage to deflect the spot over the width of the tube, so it is highly likely that the X amp has a different (higher voltage) power rail.

THe Y amp will be a lot more complex to give greater and flatter bandwidth

You've only posted the middle section of the Y amplifier. Before this section there will be the input attenuators and differential amps with FET inputs. On this sheet we see the stage that drives the Y channel switch (all those diodes) along with the multivibrator at bottom left. This either switches triggered by the timebase so that alternate trace runs are from alternate inputs, or else it free runs quite fast to chop between channels. After the beam switch, there is an amplifier stage to drive the delay line.

The trigger for the timebase is taken off before the beam switch, so either channel can be chosen, the trigger sets the timebase going, and then the signalfinally reaches the CrT because of the delay down the delay line. This means that you get to just see the trigger point a little way into the scan.

Off the picture to the right are higher voltage amplifier stages to drive the deflection plates.

Find the missing parts and have a careful check around for every arked voltage.

David


David

[RogerWalker]

Thanks very much, Bill/Dave. Now that I've got a trace, I'll get back to basics and follow the maintenance procedure as far as I can. I need to address the trace intensity question sooner rather than later as I'm afraid of damaging the CRT as the trace at minimum intensity is still very bright.

I have scanned the other part of the Y amplifier [just one channel for now, as the other appears to be identical] and I have included a scan of the Low Voltage PSU circuit too. Apart from the CRT cathode supply, the basic top level voltages are correct, but I will methodically check all the iterations as there are any number of 'decoupled' sub-supplies to the various boards.

First I want to get the brightness under control . . .

Kind regards

Roger

[maninashed]

If the trace is very bright Roger, it may be lacking any grid volts. DON'T try and measure the grid voltage as you may inadvertently destroy the tube, these voltages are high impedance and your probe may send the grid +ve. Does the beam current control have any effect?
Bill

[RogerWalker]

I'll check the CRT cathode supply and try to set the beam current per the 'maintenance/setup' instructions and revert. There's a busy time coming, so I may not have (m)any opportunities to work on the Cossor for the next ten days. But I'll try to set the EHT and beam current tomorrow if I can.
Regards
Roger

[WME_bill]

cossor 4000 /CDU120
I should not worry too much about EHT supplies at the moment, as they are clearly correct enough to give a trace. But the excessive brightness is the problem.
ave you checked the bright-up voltages as I suggested on 20 May, and that it varies with the brightness control. Working on this doesn't appear in any maintenance procedures, as the makers assume the transistors are working properly.
While testing, de-focus as much as possible to give a huge blob, it will reduce the risk of damaging (burning) the screen.
I would endorse the advice from Maninashed about testing the grid voltages. Just don't probe that pin. wme_bill

[RogerWalker]

Back from the West country after my wife's 65th birthday bash. All went well but still have her sister and nephew over here from Canada, which cramps my style a bit as regards getting on with the Cossor 4000. I have read and re-read WME_Bill's piece on the 3000 bright-up circuit and I am pretty sure that that's where the problem lies. As soon as I can I will get my [working] scope out to examine this strange stepped waveform and see whether or not it is present. I must confess that I was rather hoping that there would be just one fault to fix - but this beastie seems to have a series of faults, all of which need to be fixed in sequence. I wish I lived somewhere near Exeter or Preston - or Fife! More later as soon as I can be excused 'tour manager' duties for the visiting Canadians [today Hever castle, tomorrow Scotney castle . . .]. I really appreciate all the wonderful input and will implement the advice given as soon as I can.
Regards
Roger

[377Ohms]

After checking the short circuit current, I used an old Robin 500/1000V Megger to test the stick rectifiers of a CDU 150 in situ. About 2mA.