Ok did some analysis on the flickering while it's open and safely connected to DC and I found that voltage at TP842 drops whenever it flickers. This is not on the driver side of things after hooking a scope to the collector of Q840, setting persistence to infinite and leaving it there for a few minutes so I assume the flickering is an HT issue. I will replace the known hooky HT resistors first and see if that solves it. If not, I will do all the capacitors. If not that I will cry into it and find a donor unit for multipler and transformer.
I will win if it kills me 
Never give up
The HV 0.01uF caps are pretty loud when they arc over inside. Looking at the pcb in the complete dark you can see any arcing. But the multiplier parts are encapsulated and hard to see/hear anything. It's possible to test the HV multiplier with a car ignition coil and HV probe, and CCFL tubes as a load.
Good thread on remaking a Tek HV multiplier:
https://www.eevblog.com/forum/repair/tektronix-sc504-hv-multiplier/I think the 2235 is well worth the trouble to repair. The scope is a great drive. So easy to see subtle signal qualities in the analog world, 2mV/DIV sensitivity you can actually use instead of SMPS hash all over the place like you see with modern DSO's.
The 2235A I have is a total dog in comparison. I hate that scope, it never triggers properly and seeing the plethora of Tektronix ECO's, it looks like the scope was a rough change when Tek went to CADD. I have to spend time and see if it needs repairs or ECO's done.
Yeah I’m really liking this scope. When it works
The multiplier doesn’t worry me too much. I built and played with lots of multipliers (usually resulting in swearing and zaps) as a kid. The transformer does worry me though!
My TDS210 is amazingly quiet. Another keeper that one
Thanks David for pointing out the two different grounds on C908. I'm not sure how T906 is wound but this extra winding surely is to lower EMI.
When I see care and attention like this, I wonder how using a 48VDC replacement SMPS can work without wrecking the scope's noise floor.
If the TL594 preregulator is not being used, then there is no interference from it to cancel.
Ok did some analysis on the flickering while it's open and safely connected to DC and I found that voltage at TP842 drops whenever it flickers. This is not on the driver side of things after hooking a scope to the collector of Q840, setting persistence to infinite and leaving it there for a few minutes so I assume the flickering is an HT issue.
Those points are the output of the z-axis amplifier which controls the trace brightness.
Are you saying that TP842 glitches but the collector of Q840 does not?
An intermittent failure on the high voltage side of the z-axis amplifier is possible but I would expect it to have the same effect on both side of R842 which is just there it isolate the capacitance from the amplifier output.
Before getting too involved with replacing parts, I would fix the problem with the focus resistor chain that you identified. The original Allen-Bradley carbon composition resistors R888 to R892 were replaced with Vishay VR37 resistors and R886 was replaced with a Vishay VR25 resistor. These high voltage film resistors are inexpensively available.
Yes that is correct. TP842 glitches but Q840 collector doesn’t. Looks like whatever is happening is only pulling the voltage down there by about 4 volts. It is weird as I’d expect Q840 to drop a bit there too but it doesn’t. I stuck the scope across both sides of it and did a differential measurement which matched. I have my suspicions the resistor isn’t healthy. It will be replaced with a nice fresh Vishay MRS25 when I do the dividers.
I have already ordered some fresh VR37 510k resistors. 50 of the things as it was cheaper than buying 6 from RS
... unfortunately back ordered due on 6th December.
Fuse is now sorted. I got some slow blow ones (arrived in a massive cardboard box big enough to get 2500 sheets of paper in covered in fragile tape. More
). Power up doesn’t pop the fuse now so the power supply is repaired and working. Runs pretty cool. After 30 minutes the bracket was only warm to the touch and nothing to worry about.
Can’t progress further yet so I am shelving this and fixing my 465 and 475A until the resistors arrive. Got a collection going now
Yes that is correct. TP842 glitches but Q840 collector doesn’t. Looks like whatever is happening is only pulling the voltage down there by about 4 volts. It is weird as I’d expect Q840 to drop a bit there too but it doesn’t. I stuck the scope across both sides of it and did a differential measurement which matched. I have my suspicions the resistor isn’t healthy. It will be replaced with a nice fresh Vishay MRS25 when I do the dividers.
R842 might be bad then but there is nothing special about it and any film resistor can be used.
I am not sure about the output impedance at the collector of Q840. That is the feedback point so up to a certain load, the impedance is effectively zero but the maximum load is pretty low. The quiescent current through Q845 is about 4 milliamps.
4 milliamps through R842 can only produce 1 volt but if TP842 is pulled low enough, the collector-base junction of Q840 will be forward biased and unlimited current will be provided through it and CR840 so that could explain the voltage drop across R842.
Ok have replaced the resistors in this and no change.
Did some more debugging and removed the anode supply entirely and the display is (as expected) dull and entirely stable.
That means it’s either the multiplier which is shot or the tube is. I can’t find any evidence of flash over on or around the anode cap so my current thought is the multiplier.
Fuck!
That means it’s either the multiplier which is shot or the tube is. I can’t find any evidence of flash over on or around the anode cap so my current thought is the multiplier.
You mentioned a crack when discharging the anode connection. Discharge then reconnect and repower. Shut down and check if the anode is recharged.
If the multiplier was causing a problem then I would expect more than just an intensity change; beam deflection also depends on the PDA (post deflection acceleration) voltage.
The CRT can be operated without the PDA attached. The deflection should be about half and the trace will be dim and fuzzy but it should still be visible.
The deflection changes as well. If you stick it on X-Y mode it loses focus and shifts slightly. In sweep mode, the sweep length drops. I made a very hooky looking high voltage AC probe with some left over VR37 resistors and 5x 1n4007's in series and stuffed it across the ground / T948 p23 junction and there was no significant amplitude change going into the multiplier when this event was occurring. I even disconnected the multiplier and checked the transformer unloaded and that was fine. The -2KV is rock solid out of the multiplier as well so I think this is the 6x chain in there. I have run it without the PDA and it's perfectly stable there but as you say half deflection and brightness.
Ergo I concluded the multiplier. It could be the CRT anode cap leaking but that usually leaves deposits around the cap and it's totally clean.
Good news however. I managed to get my hands on an entirely dead 2235. This one's mains filter exploded and the attenuators are shot but the thing comes up on DC apparently fine. Also the channel switch is good if I have to hook that out. The cost? One £12 crate of strongbow and waiting until mid Jan.
Seeing if I can get a multiplier before then as well. I really want this to work as for an analogue scope it weighs a lot less than a 465!
You bastard... you and your stick-to-it-iveness had me up at 2AM calculating the charge level I'd need to put to 3 LiPo batteries to get a stable 42-43V.
mnem
Where is the "Kick that guy in the junk!!!" button on my keyboard?!?
Trick is you don’t need exactly 43V. Anything around that but less than the trip voltage works nicely
In fact it seems to run around 35V fine.
Trick is you don’t need exactly 43V. Anything around that but less than the trip voltage works nicely
In fact it seems to run around 35V fine.
The inverter output transistors operate in their linear region when turned on to regulate the output voltage so the voltage from the switching preregulator needs to be high enough to prevent them from saturating or the voltages on the secondary will drop.
Later versions of this circuit in the 4-channel 22xx series and 24xx series did away with linear regulation in the inverter which may have been a reliability issue and relied instead on the preregulator to produce the exact voltage needed.
Yeah, well... my problem is not go/no-go, but one of noise, which I blame for a RAM error in the POST (Tek 2230, not 2235). I want to bypass the prereg entirely so I can determine if it is the source or just a victim of oscillation further down the line. So yeah, I wanna get pretty close to right on.
mnem
*poit!*
Trick is you don’t need exactly 43V. Anything around that but less than the trip voltage works nicely
In fact it seems to run around 35V fine.
I see what you guys are saying... at first I thought David's post was contradicting yours, now I realize he wasn't talking about the prereg and you're both coming to the same point by different routes. LiPo Storage charge is 3.7-3.8V/cell; With the three 4S packs I have handy that's 44-45.6V. Or I could do similar with 2 of my 24V drill packs. Should be good through a dim bulb tester I'd think. Now I need to drag it out and actually DO it instead of just thinking about it.
Yeah, yeah... I know. I should treat myself to a nice old Lambda CC/CV supply that can do it just by twisting a knob; what kind of TEA addict am I.
mnem
*ZZZzzzzZZZzzz...*
They are contradictory. The inverter will run on 35 volts but the outputs will not be regulated.
Well fine then... guess I need to read up on the "Principles of Operation" part of the service manual I bought so many years ago.
mnem
"Electrons may be very, very tiny; but when they gang up on you they ALWAYS win." ~me
On the 2213/2215 which have an earlier version of this power supply, operation is more explicit. There are two adjustments; one controls the regulated output voltage and the other controls the headroom which is the voltage across the inverter transistors which keeps them operating in their linear region. On the 2235, the headroom voltage is fixed by the switching preregulator.
The 4-channel 22xx oscilloscopes extended this design with a separate transistor used as the linear pass element and the inverter transistors run in saturation. I suspect matching of the two transistors operating in linear mode in the 2235 design caused problems with transformer balance although that does not seem possible with a two transformer saturable inverter; the parts list does seem to indicate that they were used in matched pairs.
Ok finally an update on this one. My projects take too long plus xmas went crazy wrong for me. Firstly the guy who was giving me the 2235 parts mule let me down. I don't have it. I managed to get the replacement HT transformer that arrived ages ago in this evening and it's working somewhat better than it did before. Off DC anyway. Can't hear any more ticking now which is good.
However it's still not totally healthy. It has still got somewhat jumpy focus (zips in and out) and shows some deflection changes when this happens. I'm going to measure, inspect and replace the remaining resistors in the HT section as I've done all the 510k ones now.
How's the focus pot looking?
My 2235 was a little jumpy when focusing and a NOS pot from ebay has it all nice and smooth again.
Seems ok. I can’t see any voltage jumps anywhere around it which I’d expect if focus was changing.
I’m wondering if this is leakage around the anode cap or in the tube base.
Seems ok. I can’t see any voltage jumps anywhere around it which I’d expect if focus was changing.
I’m wondering if this is leakage around the anode cap or in the tube base.
Anode cap you'd hear it and CRT base is only at EHT voltages.
Either Focus, Intensity or EHT components breaking down.
Yeah can't hear a thing. Anode cap is pretty noisy - been there
May replace all the HV capacitors in it and see if that helps as I have some of the right value. If that doesn't work I'll see if I can find a suitable focus pot. Seem to be unobtainum. It's a 5M AB modpot which has to withstand 2KV.
