Author Options:

That didn't just happen, did my oscilloscope die!?!? How??? Repairable???? Answered

I went to my workbench today to discover that my BK Precision 30MHz analog scope has stopped working. Symptoms:

❶  Still shows a trace

❷  The trace does draw across the screen* 

❸  The horizontal time base seems to work

❹  10x horizontal magnification seems to work

❻  XY mode only draws in the X direction

❼  Waveforms do not show up, for either chA or chB

❽  Triggering seems to work (it causes the trace to draw more/less frequently. No vertical deflection) 

❾   Vertical controls for both channels have stopped working

❿   Dual trace shows only one trace

⓫   I have not been able to do anything to make the trace move up and down, vertical deflection is definitely shot.

⓬   * When changing the timebase, I noticed that the length of the trace horizontally gets shorter, and when a signal is present, or the vertical position for the appropriate channel is twiddled, it becomes the right length, within certain ranges.

⓭    When I was reaching for a powerstrip to turn off power to the lab the previous night, I heard a loud buzz, similar to that of the degaussing on a old computer monitor, but louder. no sparks or anything were visible as I recall. scared the crap outta-me, and I could not figure out what had happened. First thought was that maybe somewhere there was a live mains exposed somewhere that had shorted to ground or neutral. That was quickly dismissed since nothing was exposed like that, and also it would have popped the circuit breaker, and instead would have sounded like a firecracker. This sounded more like a small transformer failing. Also the same day, I have been messing with a tesla coil circuit (my beefed up slayer exciter) and the probes were sitting here by not connected picking up interference. They were in 10X mode, I do not remember the volts per division, however. I do recall a trace being shown on the display the previous day.

With these symptoms, I think the issue has to do with the vertical control on the CRT, and perhaps another, separate issue with the horizontal time base control, or possibly triggering. I think I might do a teardown on it to see if there is anything obvious that has failed, such as a charred coil. Then I will need to figure out what had caused the issue.

It is obviously something gone horribly, horribly wrong with vertical deflection or something of the like. I will have a look at the schematic and see what I can learn about the operation of the scope. I am glad I do have the user manual and schematics, so it should be straightforward, it is a nearly 100% BJT design.

The scope was used and old, and lacked a lot of features, such as cursors and frequincy counters. Perhaps it is time for a digital upgrade anyway...



3 years ago

PICTURES!!! :D (note: the pictures were uploaded in random order. Stupid instructables will not allow me to move them around chronologically.)


I found inside what certainly appears like a "mod," it goes off to the 3 by 3 connector I have mentioned. Also I forgot to mention a small blueish purplish wire was next to it, similarly just sticking out the scope. I figure it is a MOD because it is A) precariously mounted to the transformer with only one small nut, B) Has so many bodge wires crawling around on it, that it looks more like a cob web, C) the sloppy soldering and flux residue left behind indicated it was hand soldered and D) the PCB has a very different appearance than the rest of the boards inside. There is no solder mask, greyish white, and there are no silkscreened values or numbers.

I traced both the wire and the connector and ribbon cable to the same mod board, which clearly got toasted. There was a small orange wire tangling off that board, it it broke off before I even realised it, luckily, the pictures show roughly where it is connected. Tracing the blue wire back to the board, I discovered it is connected directly to one of the cooked resistor, and I think the other one was in series with it. Certainly next to it, and cooked.


Perhaps I am a bit too optimistic, but since the only connection found to the main board is the orange wire, and there are ceramic capacitors in series on that crappy looking PCB, indicating it is capacitively coupled; Hopefully what had happened was the blue wire came into contact something it was not supposed to, and shorted the output of that chip causing it to catastrophically fail, knocking out power to that triggering board. I will need to plug it is and then follow Dave's (from EEVBlog)rule: "Thou shalt test voltages"!

Nothing appears obviously burned out on any of the other boards from what I can tell. I will do further examination, and plug it in and probe where there were some red and black wires tacked on to the connector for the mod board.


p.s. I LOVE the look of that old tube, looks so vintage and nostalgic look to it! Also, I for the first time in my life saw in person the technique of "wire wrapping." I have heard about it, but have never came across it myself. Really cool!

20150505_200413.jpg20150505_200417.jpg20150505_200434 (1).jpg20150505_200400.jpg20150505_195405.jpg20150505_195514.jpg20150505_200029.jpg20150505_195650.jpg20150505_195641.jpg20150505_200042.jpg20150505_200111.jpg

UGHH, I just tested the voltages on the PSU connector going to the "sweep board." (the one sticking that is vertical and clearly visible from the angle of the pictures, the one with the wire wrapping) And everything checks out. The 5V voltages were around 4.9V, the 10V rail and -8V rail was basically spot on, and the 120V rail was actually about 139V. (15%-ish out of spec.)


This sweep board does seem functional, as it governs the trace sweeping on normal modes of operation. This does seem to work (however, like stated, sometimes the trace is not as long as it should be.) With a bit less hope, I will proceed to test the "vertical output amp" board's voltage rails to see if it is something gone horribly wrong there and inspect it for dead parts.


I would not expect this board to have been affected at all, since there are no direct connections between it and the mod board that was so obviously obliterated. IDK how this board may have been destroyed, it guess it is possible for it to be an entirely separate failure, and maybe that is what went bang, while the mod board died years ago. Who knows?

Voltages on the other board also checked out, as expected. :(((

Of course it has to be some obscure transistor on the thing that did not appreciate all little bonus current. Any sugestions as to what to check futher?

UPDATE: Discovered today that some driver transistors, which are mounted to heat sinks, are overheating. I suspect they may be blown, by the blackening and discoloration around them on the PCB.


3 years ago

Could your probe be grounded ?

OK, I have got a chance to crack the scope open to have a look inside, and check out the pictures posted! Any tips on what to do next? Me as well as verence and steveastrouk All seem to be in agreement that it is in fact the main vertical amplifier, since the channel B does work in XY mode (it deflects horizontally) but no vertical Y deflection has been achieved through all my testing.

Simalarly, I connected both probes to the built in calibration thing to see if I could get any trace at all. Still, zipittee-doo-daa nothin' at all.

It seems to be something other than that. I expect that when turn down the volts/div to the lowest setting (I think it is 0.5v/div) and when I touch the tip of the probe, I will see a distorted 60Hz signal. This is not what happens. I am naturally floating relative to ground and picking up EMI. This is how I always have checked what channels are working.

If a flat line is all you get, I'd guess it is the Main Vertical Amplifier (top right hand side in the block diagram). Its task is to 'move' the electron beam up and down.

There is only one as there is only one electron beam (the two signal traces are shown alternating). And it works with high voltage, that would explain the loud noise. Of course, it could be that the amplifier is fine, but one of the coils driven by the amplifier died.

That sounds exactly what I suspected, though I am not sure. The CRT seems to be deflected using electrostatic charges rather than magnetic coils, so I may look into the HV power supply section, and hopefully I do not blow up my multimeter... or myself....

Thing is, I have a bit of a catch-22 situation on my hands... I think I am going to need a oscilloscope to repair this oscilloscope. :(( It certainly would be nice to have something to cross-reference to check that everything works properly.

Okay, so the coils are out. I'd still go for the amplifier, not the HV supply - as the horizontal part is working. Another scope would surely be beneficial. Whatever you do, play it safe. Use an isolating transformer and keep one hand in your pant pocket. Good luck.

Good point. I wonder if it could be another supply voltage that failed. Although I have a feeling the problem is not power.

There is a +5V, -8V, +120V, and +107V supply. Since the trace is still being drawn, I do not suspect it is any of these voltages. However, it is a good idea to check them anyway.

I do not have an isolating transformer sadly. But I do have a 100W modified sine wave 12V inverter that is really small, and a 12V 12A xbox power supply. The scope is rated at 25W, although I am unsure of the power factor or VA rating.

Traditional scopes always used electrostatic deflection, not magnetic, for reasons that are obvious when someone tells you, but not otherwise....

Let me guess, does it have to do with inductance? And the difficulty in changing the current instantaneously? and the power consumption of large, current-hungry coils?

Yep. The inductance seriously limits the current rise time in the coils. When the X can move across at 200Mhz....

That makes a lot of sense! Although it is thinking about it, wouldn't electrostatic plates might have the exact opposite problem (obviously to a lesser degree), leading to the exact same problem: bandwidth?

Since the plates work with charges, and charge is related to voltage & area, and consequently have some parasitic capacitance, wouldn't that also limit rise fall times of the voltage and the charge on the plates, Due to large currents needed?

I suppose it is is lot easier to drive a heavy current into them with beefy emitter followers or low impedance sources like that than it is with large voltage transients for inductors, but IDK. Is this correct? Very interesting, I learn something new every day!

The capacitance of the plates is very small. The inductance of the requisite coils would be very large. CRTs use magnetic deflection, because the scan is much slower than a scope, and the required deflections larger too.

It figures, considering the size of them:http://www.r-type.org/pics/aaa0283.jpg

However, for focusing and astigmatism, is a magnetic coil used? or is everything electrostatic based? (I can see how maintaining constant current through a coil would require more power than contant voltage on a plate)

No, AFAIR, its all done in electric land.

Cool, I have verified that too! (popped it open today, amazed at what I saw, and I think I found a total mod board that is not OEM, too! That explains the lack of documentation on the external connector.) I will be posting images shortly.

typo: remove 'it is' in first line.


By opposite, I mean the effects of capacitance vs. inductance, where inductors are reluctant to change in current and magnetic flux, while capacitors are reluctant to change in voltage and charge.

Looking at the block diagram, I think you're right.

Sounds like you have blown the Y amp on both channels.

How would both Y amps fail though? What would cause them to fail, I mean?

Also, I should note that the XY does work, only in the X direction. II would think this would indicate that at least chA stage is working, but the problem seems to be further up, I am not a scope expert though.

The timebases and X drive is a separate system from the Y amps. Its the fact that it IS working in XY that guides me to saying the amps are dead.

What are these Y amps you are referring to BTW? I assumed it was the amps right after the input stages. On the block diagram PDF, these are labeled as "channel [A or B] signal pre-amp"

It seems like according to the block diagram, that only after those 2 input amps that the signals are further processed for all the triggering, XY mode, adding/subtracting, etc. I could be wrong though.

I would think if both channel A and B signal pre-amp's were dead, that XY mode would show nothing more than than bright dot, but rather, I can make it work in the X direction with ch. B at least, what would this mean?

There is some weird connector coming out the side that I suspect has shorted to something, though I am not sure what. It is a 4 by 4 pin connector.

I think that is what "built in signal delay line" the manual refers to.


3 years ago

Here is the full-blown schematic. It is pretty intimidating, but I can spot some basic building blocks within it, I do recognize a lot of the power supply, and the astable oscillator used for calibration! :D


I noticed a lot of diodes in the BJT circuits, my guess is that they are using the 0.6v drop of them to offset voltages at the base, or something. No idea, really.


Perhaps later this week I will have the chance to pull the cover off the scope and have a look around for obvious issues.


3 years ago

The good news is that I have the original documentation for the scope, I will try to scan as much stuff in as possible. The scope is a BK Precision 1474 Dual Trace Oscilloscope. Below is the block diagram of the scope. I suspect something (main vertical amplifier, delay line, and/or delay line driver is at fault, judging by the symptoms.)