Cossor 4000 oscilloscope

[RogerWalker]

I have recently acquired a 4000 scope that has no sign of traces on the screen. I am about to try to find out why. I have a manual.
Do any other members have any knowledge/experience of this machine please? Regards
Roger

[cmjones01]

I don't have much knowledge of these scopes, but your photo is the first time I've seen one since about 1987, when a whole load of them were sold off by the Pye TVT site in Cambridge. A friend of mine picked one up, but we couldn't get it working due to lack of information and skill at the time! I got an HP 5245L counter on the same day which I'm still using...

Chris

[Radio Wrangler]

It's got some features that make me suspect a relationship to the Cossor CDU150 military scope.

David

[WME_bill]

Cossor 4000. I saw this was on offer on ebay last month. Congratulations. I wondered about bidding, but as I am still repairing a CDU110, decided to leave it to a better man.
A bit of history may be helpful: The CDU110 introduced 1966 at £330, 20Mhz, dual trace plugin, delay TB, 6x10 tube. Almost the first fully transistorised scope.
This was repackaged for the Admiralty, with an 8x10 tube as the CT531, in 1969 (35 Mhz, dual trace, delay TB), sold commercially as CDU150 at £470. I have heard that the original intention was 50Mhz, but they couldn't get it, so settled on 35Mhz bandwidth.
This was later improved to 50Mhz and sold as the CDU120, in 1969 at £590. (50Mhz 5mv or 60Mhz 50mv, delay TB). Done I suspect by tweaking the Y Amplifier bandwidth, probably by removing the direct Y plate connections.
Reputed not to have been particularly successful.
The Cossor 4000 appears to be a CDU120 re-numbered, and still with the simpler delay TB.

Then in 1972 Cossor produced a new design, the CDU130, sold as Cossor 3000, dual trace 35Mhz/40Mhz, true dual TB, and plug in. They had been aiming for 25 Mhz, and were surprised when it turned out to be 40Mhz bandwidth.
The Cossor 4100 was 1972, portable, 75Mhz dual trace, true dual TB.
There was also a very pretty battery/mains single trace 15Mhz scope, the CDU130, in 1968.

I think that experience with the CDU110 and CDU150/CT531 will all read across to your 4000 /CDU120.
The troubles I have found with the CT531 are dry joints around the EHT tripler and the screening to some of the interconnection cables. The EHT feedback resistor chain amazingly has given no trouble. One STC stick rectifier faulty.
The usual problems around the Bright-up, with transistors or the two high voltage capacitors failing.
But the big trouble was the Low Voltage Power Supply. It uses dual capacitors in single cans on each rail. Each capacitor seems fine when individually tested, but there is an erratic short between the two positive plates within the can. This appears as though the series pass transistor is failing. I have spent hours on this. Yet reforming the capacitors seems to cure it.
Let us know how you get on.
wme_bill m0wpn

[RogerWalker]

Thanks for all the excellent responses, especially WME_Bill's product history. I have ordered two sets of 5 x ceramic disc and 1 x doorknob capacitors from Hong Kong [I have a second 4000 that might fail one day!] $55 including postage.

Initial tests indicate one of the EHT stick rectifiers may be faulty - anyone know a source or an equivalent please?

I can imagine that shorts across the two sections of the electrolytics could have some weird effects. I will reform the capacitors and have a good look around for dry joints per WME_Bill's advice while I am waiting for the capacitors to come from Hong Kong anyway.

Regards

Roger Walker

[WME_bill]

CDU120 or CDU150 or apparently Cossor 4000. The STC stick EHT rectifiers should not set a problem to find modern replacements. The main tube feed is K83/90D (4200v PIV, 1500v rms, 2.5ma) and the EHT multiplier four K83/150D (7000v PIV, 2500v rms, 2.5ma.)
Almost any TV EHT diode will do. I have raided scrap TV to get these, otherwise the usual sources: RS, Cricklewood, Farnell should produce something. Mullard/Philips BY409, BY476, BYX91 series. wme_bill m0wpn

[RogerWalker]

Thanks Bill. All the sticks read open circuit on a DMM 200 Mohm range regardless of the polarity of the prods, so I am assuming they are all duff. I couldn't find any of the diodes you mentioned for sale in UK [BY409, BY476, BYX91] but I did find BY140 at £6 a throw which seemed a bit overkill and expensive. A further trawl unearthed Vishay DO-204AL 3kV 250mA recovery diodes at Farnell. It wasn't clear whether the 3kV was RMS or 'repetitive reverse' rating whatever that is, but they were cheap so I was thinking of buying some and trying one for the K83/90D and two in series for each K83/150D. Has anyone ever tried series silicon diodes for EHT? I have a load of 1N4007 but seem to remember they need a chain of balancing resistors when used in series? Any comments gratefully received.
Roger

[Nicklyons2]

The e.h.t sticks may not be faulty, I suspect your DMM won't be able to produce forward voltage to 'turn on' the junctions. Many eht sticks of this period were tubes with 'pills' of small lower voltage rectifiers assenbled as a 'pile' so there may be 20 times V.fwds in series which your meter's supply won't be able to overcome.

[RogerWalker]

Thanks for the tip Nick - my Avo 8 has a 15V battery, maybe that'll kick 'em into life. Is there any sure way of testing eht sticks that determines good/bad?

I have checked the low voltage power supply adjustments. All to spec except "set EHT" that should return 960V dc but can only manage 750V dc. At least it indicates that the oscillator and transformer are working after a fashion. I am working on the theory that something's amiss in the quadrupler that's dragging down the set EHT voltage. If so, I guess if I isolate the quadrupler by disconnecting the first feed capacitor, the set EHT figure should rise to 960V. I'll set the EHT adjustment potentiometer low and give it a go. I just hope it doesn't turn out to be shorted turns in the EHT secondary winding . . .
Roger

[WME_bill]

EHT sticks. Following on from Nicklyons2 comment upon testing these things. I rummaged in my junk box, and found half a dozen or so, K83/15 or K83/25. All too low a peak voltage rating for you. But I though I'd test them.
On a digital. no reading both ways.
On Avo 8 Resistance X1. no reading both ways
On Avo 8 High Rest (=15V). Reverse no reading = infinity.
Forward 5 to 10 thousand X100 ohm
In series with 70v supply. Current through Avo: Reverse = nil, Forward was around 5 ma for lower voltage one, and 10ma for the higher ones.
So it suggests that an ohmmeter check is no indication as to whether these things are working. Try them in situ, and make up a high voltage probe for your Avo. I did post some ideas on this years ago, probably re Telequipment D83 or similar scope. Bill

[MotorBikeLes]

Way back when I was playing with TQ valved scopes, I had a load of these stick rectifiers of uncertain capability.
I made up a little test rig which quickly gave me a poor/fair/good selection.
I used an spare scope mains transformer with a 960v (I think) secondary winding. I fed that through the "rectifier under test" to a string of electrolytics. (3 off 2.2mFd @450v I think) to meet the full rectified voltage. A parallel string of resistors (today I would just use 2 VR37 types) to give a voltage division, with a neon indicator across the bottom resistor. Can't remember what values I used, but easy to find something suitable. I then switch on, allow a few seconds to stabilise, then count the number of flashes in a 10 second period. More flashes indicates better rectifier. I think I also used to check the voltage across the rectifier, but not certain. It is only in the last few months that I dismantled the bits and ditched the transformer, since I am unlikely to need it again. I have a good stock of EHT rectifiers, from those used in mono TVs of the early '70's, to various silicon types of more recent manufacture. I would likely use a BY187 to calibrate the resistor chain in my little tester if I was making one today. The resistor chain overall value would be chosen to put about 1mA load on the rectifier.
Les.

[KeithsTV]

I found a K8/30T rectifier amongst the diodes I have and measured it on the curve tracer at work. The plot below is 50v/division horizontally and shows it takes at least 20v before it starts conducting. It starts breaking down in the reverse direction at about 1400v.

Keith

[RogerWalker]

Thanks for those responses - I have good news to report! The final stick and its associated capacitor were duff. One end cap of the rectifier was loose and corroded. Replacing the rectifier with two 3kV silicon diodes in series helped but replacing the capacitor brought a crackle of static and oodles of EHT. I might fit capacitors across the diodes to even the load a bit, if anyone thinks it necessary.
The bad news is - still no proper traces.
The "set EHT" voltage is still only 750V versus the 960V specified so that's probably affecting the CRT operating conditions. If I press the 'beam finder' button, I can get tiny traces about 2" long which can be moved up, down and laterally and focussed, brightened and dimmed - which is good news - and the main timebase seems to be running as changing timebase speed has an effect on the traces. There seems to be sufficient Y displacement on both Y amplifiers just minimal deflection in the X department and the CRT remains stubbornly cut off unless the beam-seeker is activated.

Unfortunately the manual provides no circuit description whatsoever and no voltages. It assumes that anyone reading it is a fully trained Cossor test engineer familiar with transistorised scopes and expert on the model 4000. I wish!

I have to have a break tomorrow as the engineer is coming to install superfast cable broadband. Any input on where to look next would be appreciated.
Kind regards
Roger

[Radio Wrangler]

I don't know the innards of that machhine, but the usual ploy with a relatively modern scope is to have TWO EHT windings to make the cathode bias and grid bias voltages from two separate, floating, 2kV-ish supplies. One of these supplies works with respect to ground, and the other one works with respect to an amplifier. In this way the amplifier controls the grid-cathode voltage difference and the scheme works right down to DC, so you can have DC-coupled Z-modulation, very-slow-retrace blanking works perfectly and your brightness pot is at a reasonable potential. I gave my old CDU150 and manual away some years ago, so I don't have a schematic.

David

[maninashed]

The eht circuit of the CDU150 operates via T1 which is the eht transformer tr13 is the oscillator transistor bolted to the chassis. The emitter of which is connected to pin b on the transformer. Pin a and f are grounded, pin h is connected to D15 hv diode, this feeds the cathode at -1.2kv. Pin i is connected to hv cap 500pf and then to the hv multiplier chain, eht should be 11.25 kv. There are 2 zeners in this circuit one 68v, the other 82v connected to the set beam current pot, these zeners can cause problems. D16 and 17 connect cathode pin 4, to grid at pin 3, and then via C2 500pf, R46 150R, R8 15k, and R11 5.6k to the brilliance pot. Check all these components.
Bill

[RogerWalker]

Thanks David and Bill - the 4000 has different circuit references but I have located the components pointed to by Bill and will check them out.

Presumably using a DMM to measure the voltage across the two zeners should return 68V and 82V if the diodes are OK and some other value if they're not? The CRT cathode / grid diodes [D20/21 in the 4000] are 1N916s, so should be easy enough to test out of circuit.

I must say the guys who drew the circuit made it look easy to follow but, by concentrating on the physical locations of the components, managed to make it a "reverse tube map". I guess it's what one's used to in the end.

Presumably the advice relates to the cut-off CRT - any thoughts on why the traces are so narrow? BTW, the "x10" switch that should expand the traces, has no effect.

I am very grateful for all the help - as you can probably tell, I am substantially out of my depth with this beastie!
Kind regards
Roger

[maninashed]

Hi Roger, can you post a pic of the eht section please. You should have +50 and -50v to the eht oscillator circuit, make sure the lv supplies are ok. Your eht generator is definitely oscillating, but your cathode supply is low at 750v, so something is loading it. What are you measuring it with? These supplies are high impedance and if you are using an avo, that's no good as its 20,00 ohms /volt and it will load the circuit too heavily and give a low reading. Ideally, you need a hv probe. You can easily make one of these using a thick plastic tube containg a string of hv resistors, the end of the chain is grounded, you connect your digital meter across a 1 meg resistor at the ground end of the chain. You can calibrate it using say a 100 v source so you can calculate the multiplication factor. If the chain uses a total of say 100M ohms, this will not load the circuit too much and your readings will be more accurate.
If you disconnect the eht connection at pin i on the eht transformer, the voltage at the 47k connected to the hv rectifier at pin h(cathode supply) should be as close to the spec as possible (-1.2kv on my diagram), and yes that is a minus sign. With the eht disconnected, you 'may' see a dim trace at max brightness in dim lighting conditions. If this trace is visible and full width, report back.
Hope this helps.
Bill

[RogerWalker]

Thanks very much Bill - I am using a 10Mohm input DMM on 1kV range and getting -750 V at the 47k with set EHT control at max. I have a 25kV probe for an Avo 8 and that is showing 11kV EHT at the doorknob capacitor since I replaced the rectifier stick and capacitor.

I tried disconnecting the EHT feed to the multiplier at "i" in the tranny, but it didn't chage the volts at the 47k. As soon as the ADSL engineer has been/gone, I will disconnect it again and look for a dim trace without PDA.

Now that I know that the 3kV silicon diodes I bought are OK to use, I will try replacing the K83/90D cathode supply diode to see whether that brings up the volts - the manual says it should be -940V [not 960 as I previously stated] - lower than the CDU150's 1.2kV.

BTW: Further to my last post, I have examined the circuit further and found that the beam-finder switch and the "X10" deflection switch look to be inter-related so pressing the beam-finder may render the X10 switch inoperative. I will crack on with your suggestions and try to get traces on the screen without having to press the beam-finder.
Regards
Roger

[maninashed]

Roger, your measurements should be fine with that equipment. If the cathode supply diode proves good, with the pda supply disconnected, disconnect in turn each of the 4700pf caps to ground, as a leak in one of these would cause your problem.
Bill

[RogerWalker]

Thanks Bill - I'll check the 4700pFs as suggested.
Regards
Roger