EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: AaronB on October 15, 2020, 10:05:38 am
-
My 2235 scope has a problem with the power supply. It blew the fuse (1A slow blow) as soon as I turned it on. I have downloaded the service manual and it says the correct fuse should be stated on the rear sticker, that is no longer there, it's just a plain aluminium panel. Someone has pencilled in 2A FAST next to the fuse holder, but I'm pretty sure that is wrong for a UK machine at 240V, and 1A slow is correct.
Tracing back from the mains input, it looks like the A6 filter board is OK - measures 470K. The next section from C904, the 4 rectifier diodes, and C906 seems pretty suspect - C904 is a dead sort (obviously that's why the fuse blows). The diodes don't measure how I would expect, and C906 looks original and probably a good candidate for replacement. I will replace these components (suggestions welcome) but what more can I test to see if this is the only problem, and that there isn't something else further down the line that caused these failures?
I guessed on posting in the repair section, rather than testgear, I will repost if this incorrect.
Thanks in advance for any help.
-
Hello and welcome to the forum.
There have already been many topics on the power supply of the Tektronix 2235 oscilloscope,
Before opening a new topic, I advise you to use the search tool to check if the answer to your question is not found in another topic
A first tip: do not replace any part until you are certain that it is defective.
In your case, this is obviously a short-circuit and therefore easy to detect.
You need to check (in the order of the list) if the following components are not shorted (possibly unsolder one side of the component and unplug it from the PCB to check it)
- mains filter A6
- C904
- diodes CR901, CR902, CR903 and CR904
- -C906
- Q9070
- Q935
NB: fuse: 1A slow blow.
-
Thanks for the help, I did try searching but as you say, many topics, difficult to sift through.
I removed C904, it was out of spec at 3800pF instead of 2200pF, but that wasn't the short. Unplugging Q9070, allowed me to measure C906 - not bad at 77uF instead of 75uF, ESR 0.5. Diodes CR902 and CR904 measure OK on the board, It's the other two diodes in the bridge that are both shorted. After replacing them I need to check if the Q9070 FET and Q935 SCR are still OK.
-
Well done !
For capacitors, there is no need to worry if the measured value is higher than the nominal value.
The tolerance must be taken into account, for example, for an electrolytic it is + 50% and -20%
As for the ESR, there are many applications where it has little or no influence on the operation of the circuit.
In general, too high an ESR of a capacitor goes together with a much too low capacitance.
Preferably replace all 4 diodes, not just shorted diodes and check Q9070 FET and Q935 SCR
-
I have a stash of 1N4007 diodes that should do the trick for the rectifier circuit.
It's handy that it is so easy to unplug Q9070. My cheapo component tester will normally recognise a mosfet, but shows Q9070 as a couple of connected resistors - so I figure that is also bad as I believe they fail closed.
Urgh, Q935 is in the corner of the board, right underneath the edges of the chassis, but pretty sure it'll have to come out to be tested. I guess I could leave it and hope that it is OK after fixing the above.
Is there an average number of fuses that will get blown before people figure out all their PSU issues?
-
It's handy that it is so easy to unplug Q9070.
Later, after the fault is repaired, I advise you to eliminate this connector and directly solder the wires to the pins of the mosfet.
These crap molex connectors cause bad contacts and heating of the mosfet pins.
In the event of a fault, Q9070 short circuit or explode ..... easy to tell the difference between a faulty mosfet and a good one
Q935 will short circuit when it is defective, it is easy to check in circuit with ohmeter , no need to disassemble it.
-
The bits under the edges of the chassis are a pain but if you're careful, have a nice skinny tip (that's what she said), have just the right angle and enough light you can get to them.
If you scroll down you'll see I have a similar thread going on my own 2235 PSU. No luck yet.
I'm starting to think bd193's plan of action of sticking a 48V Meanwell unit in these and calling it a day is the way to go, which is what I'm planning on doing. The factory supplies are sketchy as hell anyway.
-
The bits under the edges of the chassis are a pain but if you're careful, have a nice skinny tip (that's what she said), have just the right angle and enough light you can get to them.
If you scroll down you'll see I have a similar thread going on my own 2235 PSU. No luck yet.
I'm starting to think bd193's plan of action of sticking a 48V Meanwell unit in these and calling it a day is the way to go, which is what I'm planning on doing. The factory supplies are sketchy as hell anyway.
The failure of your oscilloscope is different from that of AaronB and therefore requires a different diagnostic procedure.
I think we must therefore leave these two very distinct topics.
In fact, you have already been told what to do and you were asked a question that you did not answer:
"With an external 43V dc power supply, does it work or not?"
the crowbar protection trips at 51V, then 48V is a little too high .... to lower between 40 and 45V.
Why not just buy a 2x30V 2A laboratory power supply?
With the 2 adjustable 30V outputs in series, you can easily supply 43V to the oscilloscope.
-
Q935 will short circuit when it is defective, it is easy to check in circuit with ohmeter , no need to disassemble it.
It doesn't measure as a dead short, but the resistance values seem pretty low: 11, 120 & 131 Ohms - I think she'll have to come out. The parts reference says it is a C106B2, but not readily available here in the UK, C106BG looks like a good alternative. It's not expensive, so I'll add it to my parts order.
-
If this thyristor is not shorted, do not replace it.
In 90% of cases, a defective thyristor has a dead short circuit between anode and cathode following the melting of the semiconductor chip.
The resistances you are measuring are no doubt due to the fact that you are measuring it in the circuit.
In general, thyristors are very rugged semiconductors.
With a new fuse in place and the switch on, you should no longer measure a short circuit at the 230V input.
-
If there is no longer a short circuit, it is now time to try powering your oscilloscope with an AC voltage source with current limiting.
For this, take a 230V 100W incandescent bulb and connect it in series with the oscilloscope.
Supply the oscilloscope + 100W bulb assembly in series by the 230Vac
If the oscilloscope tries to start and the lamp only lights up halfway, all is well, you can then supply your oscilloscope with 230V without the bulb.
-
The parts reference says it is a C106B2
It's an internal over voltage protection and deliberately short circuiting thing.
Anything capable will do.
-
I did a video on restoring one of these scopes, you might find it useful.
I included the service upgrades too, so you can update your scope to improve performance and reliability.
https://www.youtube.com/watch?v=lrXCVg6T-ek (https://www.youtube.com/watch?v=lrXCVg6T-ek)
-
In fact, you have already been told what to do and you were asked a question that you did not answer:
"With an external 43V dc power supply, does it work or not?"
As I said in my thread, I do not yet have a suitable power supply. I've got one on the way now and will report back when it gets here.
-
The bits under the edges of the chassis are a pain but if you're careful, have a nice skinny tip (that's what she said), have just the right angle and enough light you can get to them.
And the CR903 diode is even more difficult to get to. But there is a solution.
(http://[attach=1])
-
Does anyone have a full size, single scan of the Power Supply, Z Axis + CRT fold out schematic from the Tektronix 2235 service manual? The pdf version I have has it split across two pages and there is a missing section between G/H.
-
https://www.eevblog.com/forum/testgear/tektronix-2235-repair-thread/ (https://www.eevblog.com/forum/testgear/tektronix-2235-repair-thread/)
-
Hopefully this helps. It's from the 2236, but the missing parts should line up. It's almost identical to the 2235 if you include the greyed parts. Plus the addition of the fan (a worthwhile upgrade to ensure longevity).
I also added some more useful info, I can provide specific details on the parts needed for the service update if you decide to do it.
Does anyone have a full size, single scan of the Power Supply, Z Axis + CRT fold out schematic from the Tektronix 2235 service manual? The pdf version I have has it split across two pages and there is a missing section between G/H.
-
I was thinking of adding a fan. I have a miniature 12V fan that runs acceptably, and only draws 35mA, from a 9V supply, so I was just going to hi-jack the +8.6 rail and mount it close to the heatsink for Q9070. But it may be overkill for something that I would rarely ever use for more than 1 hour at a time.
-
There's a place to mount a fan on the rear panel, it just screws right into place and blows directly onto the PSU heat sinks.
-
The fan I was considering is very small, maybe only 4cmx4cm. Kind of thing you would mount directly on a heatsink, like a CPU fan. I would stick some fins to the back of the existing, flat heatsink and then attach the fan to that. It may not be much, but better than nothing. Maybe even just the extra cooling fins on that heatsink would be enough. But that's in the future.
First off, I'll replace the broken bits I know of. If that doesn't fix it, then I'll consider the external 43V PSU option. If it is then working, then I'll think about the service updates and a fan. If I cannot get it working, then look out for an eBay listing for a broken 2235.
-
Well that was disappointing :-- It has stopped blowing fuses but is otherwise still dead. I get nothing between TP940-TP950, maybe it's time to try the external 43V supply test. Is there any reason to isolate any other parts of the board or is it just a case of attaching +43V to TP940 and 0V to TP950?
-
There is no reason to be disappointed, repair is that, step by step.
If you don't have patience, repair is not for you. :-DD
You have taken a step, you no longer burn a fuse, that's already very good.
Let's temporarily abandon the repair of the pre-regulator to check if the rest of the oscilloscope is working correctly.
I usually solder two wires directly on the printed circuit, on the terminals of the capacitor C940 to apply 43v there - pay attention to the polarity!
Give us the results of your test with an external 43V power supply
-
I tried a few more component extractions for testing that had been mentioned in other threads. The Q935 thyristor, C925, C943, and R898-R891 - no luck there. So I went for the addition of a 43V PSU across the two test points. The first PSU I tried didn't work so well and I switched it off as I thought things were getting hot. Second one I tried caused the classic tick, tick, tick and two flashing neons of a resetting power supply. Carefully touching the board for hot components, and visual inspection, showed a bit of shadow on the PCB around CR907. Sure enough the smoking diode - a dead short.
It is marked as UES1106, which is a fast recovery diode, not something I keep in stock. Not sure what part it has to play in the circuit when bypassing the original supply, but with it removed, and also Q935 and Q9070 out of circuit, I get the power LED on and no ticking :). I once briefly saw a fuzzy rectangle on the display when I pressed BEAM FIND :D, but it has stopped doing that now :-BROKE. The PSU is drawing about 700mA in this state. Do I need to replace that diode, or is it no longer part of the equation with the new 43V PSU in circuit?
Getting closer.
-
I'm pretty sure it will work without it (and I think you found your problem) but just to be sure just test all the power supply voltages to the right of the transformer except the -2kV high voltage, unless you can do that completely safely. If the scope seems powered up but you have no beam, check R886, R888, R889, R890, R891, R892 and R894 to see if they are high or open. If those are bad, let me know and I'll look up the p/n from Mouser, as there is a specific resistor that works well in that application.
-
Well done, you have discovered a new fault: cr907 is dead short.
This diode is essential for the operation of the pre-regulator, it will have to be replaced later but no need for it to do the tests with the 43V
No need either for the SCR Q935 (crowbar) or the Q9070.
The consumption of 700mA seems normal to me, the oscilloscope should work.
Check if all the settings on the front panel are correct to have a trace
-
The resistors you mention all look good. For the power rails I have:
TP961 -8.6V
W954 +100V
W956 +30V
W960 +8.6V
W968 +5.1V
I don't have any way of checking the 2KV supply. I started on the trouble shooting flow chart with the initial settings described, and got the above voltage readings, which look correct. Still no base line trace. But I'm a bit stuck going any further down the trouble shooting flow chart as I don't have another scope to check waveforms. Unless I can do it with one of the digital, ultra-cheap Chinesium, ones off eBay.
-
OK, if you turn the intensity all the way up and press beam find, do you get nothing? How about in a darkened room?
-
Hi
This scope has some 14KV as a Post Deflection Acceleration voltage, if you listen carefully at startup and shut down you should hear a rustling sound as the 14KV comes up and down.
George G6HIG Dover UK
-
I have played around with various front panel settings and Beam Find, and twice it displayed a brief, fuzzy smudge. But I cannot repeat it. Nothing obvious in the dark either, except a brief flash of the A Trigger LED.
Son of a Diddly. Just as I typed the above, it has sprung into life ;D ;D ;D
It's a little unstable, with some ripple on CH2 but I'm of to bed happy :-+
-
:-+
Excellent trouble shooting patience.
Recently I had a 2215 preregulator (A-19) with a very similar problem.
After verifying the rest of the scope was ok (with external 43 volts DC input) I eventually traced the problem to a bad solder joint.
AC line-powered the board externally into an incandescent 75 watt load while flexing the board and eventually found the problem
:-- I miss the days of no hf, just the line frequency.
Just restored (E-cap change) a Telequipment (Tektronix UK) D 54 with all DC voltages derived from an (impossible to replace) line frequency transformer with multiple input winding from 100 V to 120 V, 5 volt increments or 200 V to 240 V, 10 volt increments, 48-450 Hz.
Two vertical channels and no B sweep, etc.
One sided PC's, no IC's and all transistors in sockets.
Only 10 MHz bandwidth but good for any audio work.
Only draw back is variable/cal knobs don't have indicators or switch detents and change when changing V/div or time/div >:D
-
Although I know have a working scope, in that I can see a couple of beam traces, there are still issues that need fixing. Unfortunately they are intermittent and difficult to diagnose. At anything more than moderate intensity the trace becomes quite unstable, some kind of shearing of the beam pattern, and inability to focus. Also, sometimes there is a tinkling/ringing sound from the HT side of the board at high intensity. I went through the first steps in the adjustment procedure to try and set the grid bias, but it changes each time I turn it on. A couple of times I got a square wave trace from the probe adjust test. And other problems, nothing I adjust seems to stay consistent.
I think I will stick with the external 43V supply - not sure I have the stamina to debug something I can already bypass. I want to isolate the pre-regulator entirely so that I can put the PSU inside the case and use the normal power switch, also to remove any interaction that may be causing problems from the old, unused components. I'd like to do this without cutting tracks, but that may not be possible.
The problem that @jdragoset had with a dry joint may be something similar to my intermittent problems, that circuitry is very sensitive. Last time the scope was misbehaving I found a dead spider under the CRT, removing it fixed the problem.
-
Before making any changes, try to repair your oscilloscope pre-regulator.
When a device has been idle for a long time, it is normal to experience bad contacts issues.
On the other hand, cold solder/dry joint problems are quite rare with tin lead solder.
For bad contacts, start by moving all switches on the front panel a dozen times.
Mark the position of the trimpot sliders, move them around their position then return the sliders to their initial position.
Remove and reinsert all connectors, including the tube socket. (be careful, discharge the 1200V before)
Find out where the high brightness noise is coming from.
Then we will try to fix the pre-regulator
-
If you are triple-sure those resistors I mentioned are all good, then you will need to acquire the means to safely measure high voltage before you can do much else diagnosis wise. The only other way is to examine all of the solder joints and reflow them, measure all of the components (especially the resistors) in the HV section and so on--perhaps you'll get lucky. The noise sounds like something is breaking down or arcing, but it could be anywhere. Inspecting under power (carefully) in a dark room may help.
-
Before making any changes, try to repair your oscilloscope pre-regulator.
I appreciate your thoughts but I have to take a more pragmatic approach. The next thing I would need to do to fix the pre-regulator is to source a single diode to get to what may be a fix, or just another round of debugging to find another failed component. Rinse and repeat. Buying individual components piecemeal is slow, expensive and tiresome. I would probably just give up if another single component was the next blocker, without an end game in sight. The external PSU is an easier way forward, and keeps me interested in getting this fix completed.
On the other hand, cold solder/dry joint problems are quite rare with tin lead solder.
The intermittent failures I am experiencing can often be induced by just lifting up the chassis frame. To me, that indicates some physical PCB or connection issue is at least part of this.
For bad contacts, start by moving all switches on the front panel a dozen times.
Mark the position of the trimpot sliders, move them around their position then return the sliders to their initial position.
Remove and reinsert all connectors, including the tube socket. (be careful, discharge the 1200V before)
Find out where the high brightness noise is coming from.
I had been fiddling with the switches and pots and I think it was that which contributed to it suddenly springing into life. Also, the noise has now gone - don't you love intermittent faults :(. The most obvious strangeness now is with the trigger LED randomly on/off/flashing depending on the intensity setting.
Thanks for all your help to get me this far. :-+
-
To repair these oscilloscopes, you should ideally have:
- a good quality isolation transformer to be able to work and measure the circuits of the pre-regulator in complete safety
- a 0-50V 0- 2A laboratory power supply
- another oscilloscope
- a multimeter
- spare parts
- service manual
- a calibrated function generator
Since you don't have most of these, diagnosis and repair is much more difficult.
That is why we are looking for possible solutions for you.
Regarding the defective diode, you could try to replace it with a UF4007 (UF4007, not 1N4007 !!!) or BYV27 / 600.
For the pre-regulator, I use to supply C925 in 15V with an external power supply and I check the presence of pulses on the gate of Q9070 with an oscilloscope.
It is not possible for you to do this.
So it will take:
- ex officio replace Q9070 and check Q908, CR908, check the PCB tracks around these components, and we will hope U930 is not faulty.
After this, we will test the pre-regulator with an incandescent bulb 100W in serie with mains.
If it try to start, you can feed it directly with mains without the 100W bulb in serie.
NB: before testing the pre-regulator, Q935 MUST BE IN PLACE
-
To repair these oscilloscopes, you should ideally have:
- a good quality isolation transformer to be able to work and measure the circuits of the pre-regulator in complete safety
- a 0-50V 0- 2A laboratory power supply
- another oscilloscope
- a multimeter
- spare parts
- service manual
- a calibrated function generator
Sure, in an ideal world I would have all that stuff.
What I do have:
- a power supply that can do up to 48V at 1.6A
- a kit built digital scope that works OK up to about 50Khz
- regular and clamp multimeters, but only up to CAT II 1000V
- a small stock of resistors and capacitors, and other parts culled from dead equipment
- the service manual
- a couple of signal generators, one audio band, one RF
Regarding the defective diode, you could try to replace it with a UF4007 (UF4007, not 1N4007 !!!) or BYV27 / 600.
For the pre-regulator, I use to supply C925 in 15V with an external power supply and I check the presence of pulses on the gate of Q9070 with an oscilloscope.
It is not possible for you to do this.
So it will take:
- ex officio replace Q9070 and check Q908, CR908, check the PCB tracks around these components, and we will hope U930 is not faulty.
Spare parts are the main problem. I see that UF4007 is cheap enough on eBay, but I have been burned a few times recently from sellers claiming UK stock with 3 day delivery, when in fact it comes from China in 6 weeks, isn't the correct part, is fake, or doesn't come at all. I can't reliably test Q908/CR908 while on the board, I would have to remove them for testing. Not sure what they are until I pull them out and try to find a cross reference, and a reliable supplier if broken. While probing around that area I find that R909 is open circuit - but I don't see R909 on the schematic or parts list that I have???? So, I'm back to looking at the time consuming, and possibly fruitless, task of debugging and fixing this damn pre-regulator. Or, just plugging in the 43V PSU.
There is some good news: With some more time spent twiddling pots, flipping switches, and knob turning I have found that the intermittent problems I had before have now gone. With the 43v supply, it now works correctly. Or at least as well as it did before the fuse blew. It would be nice if Santa bought me some better test gear for Xmas and I could continue this adventure, but I'll leave things working as they are for now.
-
I did not find R909 on the diagram either, but it is probably a 10 or 15 ohm resistor in series with the Q9070 gate to avoid oscillations (to be checked :).
If this is the case, there are probably other defective components in the pre-regulator control circuit and in particular:
Q908, CR908 and U930.
U930 (TL594) will have to be replaced automatically.
Be careful, you must place a good quality socket if you replace U930 !!!!
But you may also use your oscilloscope with an external 43V power supply if you wish. :-+
-
I know there are a number of threads on the problems with this series of Tektronix power supplies, and it looks like there is a bunch of well known suspect components for the possible issues. Can we put together a list of cross referenced components that would cover the most common eventualities? I don't mind ordering a full kit of parts, at a reasonable price, in one shot to save on delivery and processing fees of individual items None of the proposed items in this thread are excessively expensive. That would solve the frustration of waiting for a single new component to be delivered at the high cost of a single item. No doubt you will end up with unused parts, but they can be added to your stash.
The goal would be to work methodically through the test/replace procedure without the frustration of waiting for components. I encountered this scenario when fixing my plasma TV - I found a kit of parts on eBay that contained all the usual failure suspects when that model of TV went bad, and yes, it was always the power supply. The TV broke at the beginning of the Covid lockdown in the UK - quite a lot of pressure from the family to fix it :box: ASAP. The single kit of parts saved me being lynched.
-
The deed is done. I have tried to post a picture of the mods I did but there is a limit on what I can post (apparently) . Maybe you can see something in this image.(http://)
-
You posted an image smaller than a thumbnail. :D
-
AaronB originally tried to post a 12 MPixel image, was told in the Moderation Reports thread (where he complained that this didn't work) that such big pictures are usually a bad idea, and now feels entitled to be peeved.
-
I did nothing intentionally to mess up this thread or annoy others, I got caught out by - let's be honest - some pretty antiquated software. Can I delete and then resize the previous image I had posted, then reattach a decent sized image of the final build. I don't know. There is not much guidance or obvious warnings for management of attachments - all I saw was a 5MB limit, I tried to upload 3.5Mb and got a security warning - huh?. All the other forum systems I post to are able to handle a phone camera picture, either as a raw image, resizing it on upload, or allowing the poster to choose an image resize. Now I know the limits of this system, I won't make the same mistake.
-
You just have to adapt to the forum ...
Anyone can make mistakes at the beginning, you should not make a drama ...
Forget all that and keep posting. :popcorn:
-
I did nothing intentionally to mess up this thread or annoy others, I got caught out by - let's be honest - some pretty antiquated software. Can I delete and then resize the previous image I had posted, then reattach a decent sized image of the final build. I don't know. There is not much guidance or obvious warnings for management of attachments - all I saw was a 5MB limit, I tried to upload 3.5Mb and got a security warning - huh?. All the other forum systems I post to are able to handle a phone camera picture, either as a raw image, resizing it on upload, or allowing the poster to choose an image resize. Now I know the limits of this system, I won't make the same mistake.
Attaching that 46*25 pixel image in response to the size limits you encountered came across as a bit juvenile...
You can, of course, change the picture later: Click the "modify" button to the right of your post's title; below the edit window which opens click "attachments and other options"; uncheck the old attachment and select a new picture to attach.
-
Attaching that 46*25 pixel image in response to the size limits you encountered came across as a bit juvenile...
Childish or not, that's your decision, but it was all I could do. I had already posted one large image, someone else posted 3 other images, which took the total thread size right up to the 5MB limit. What I posted then, was all that was allowed.
-
So now it should all be fixed. Two images of a size that fits the download limits of this forum.
-
it was all I could do. I had already posted one large image, someone else posted 3 other images, which took the total thread size right up to the 5MB limit. What I posted then, was all that was allowed.
No, you misunderstood. The 5MB total attachment size limit applies per post, not for the entire thread.
Edit: I just tried to attach two images here of 1.4 MB each to confirm -- works just fine. Removed the pictures now to conserve server space.
Yes, that means that a single 3.5 MB image attached to your earlier post should have been just fine, assuming that it was the only attachment to that post, and that it indeed had a size below the 4 MB limit for each individual attachment. Not sure what went wrong with your prior attempt.
-
No, you misunderstood. The 5MB total attachment size limit applies per post, not for the entire thread.
Yes, that means that a single 3.5 MB image attached to your earlier post should have been just fine, assuming that it was the only attachment to that post, and that it indeed had a size below the 4 MB limit for each individual attachment. Not sure what went wrong with your prior attempt.
The wording could be construed as a bit ambiguous, and I see now that the limit message has changed to be total size 5000KB, maximum individual size 4000KB. Not sure why my other image was rejected but hey, it's all good now. :-+ You could do with a :beer: icon
I know it can be difficult to interact with a forum when you are a noob and trying to find your feet. I did search around the internet for a couple of days researching my problem before choosing this site. And I did then search in the Test Gear and the Repair forums before posting here. I'm not sure if I have enough knowledge to contribute with answers to other people's problems, but sometimes it can be enough to further a discussion by just joining in and asking a question.
I very much appreciate the help you have all given, and now have a working 'scope, rather than some more "busted electrical tat" clogging up the place (she said).