Oscilloscope Repair

[user_88]

I put together a HeathKit 4205 5 MHz scope several years ago ... 1980 something.
All the intermediate circuit board measurements checked out, but there was a screen anomaly .... something like yours .... can't remember exactly. For several weeks I puzzled over the problem. Then, by accident, I moved one of the harness wires running somewhere on the board, and the screen pattern also moved. Then I took that particular wire .... just a plain #22 insulated copper wire, and moved it around a little .... The result was that the problem was resolved ... The screen trace was perfect.
Apparently, that particular wire was picking up a magnetic flux coming from someplace, and causing the trace be interfered with somehow.

It seems likely to me that your particular problem might be caused by an electrolytic capacitor someplace .... maybe in the power supply section ... Not making any bets though.

 

[shimniok]

The problem is still, how do I know which side of the resistor represents which side on the scematic?

Good point. Measure both sides of each resistor. Which is what I marked for all but R82, so measure both sides of that one too.

If your meter doesn't have an ESR meter function we are kind of out of luck on that front.

Michael

 

[shimniok]

Maybe this spreadsheet will help...

View attachment scopemeasure.xls.zip

 

[pierreretief]

Ok, so my multimeter has a capacitance setting, and I measured C16 , C17 , C18 , C20 , C22 , C25 and they all checked out ok, better than the ones I actually wanted to replace them with.. maybe the guy I bought it from has done a recap already.. anyway, if my DMM says its ok, then its ok, right?

 

[Nigel Goodwin]

Ok, so my multimeter has a capacitance setting, and I measured C16 , C17 , C18 , C20 , C22 , C25 and they all checked out ok, better than the ones I actually wanted to replace them with.. maybe the guy I bought it from has done a recap already.. anyway, if my DMM says its ok, then its ok, right?

No - you haven't checked anything worth checking.

High ESR is the failure mode of almost all faulty electrolytics - you need an ESR meter to check them.

 

[pierreretief]

ok, so even if its got full capacitance, it still doesnt mean its ok?

 

[shimniok]

Correct. ESR will muck with circuit operation and typically start to go out of spec high long before capacitance value changes significantly. Or so I have been told.

 

[pierreretief]

mmm, so what am I to do? Replace them, even if the new caps have got a slightly lower capacitance, around 0.8 for a 1uF...?

 

[pierreretief]

So how do I measure it without a ESR tester?

 

[shimniok]

I mentioned earlier--you need a scope and a function generator. So here's what I'd suggest...

Take a look at the jacket labels on these caps -- or look in the service manual -- to find out the tolerances involved. Typically the tolerances are pretty loose, like +80%/-20% or +50%/-20% or ±20% or whatever. If the new caps measure within those specs, go ahead and switch out.

If you have enough caps to replace all the vertical amp electrolytics that is cool. Otherwise, verify first that we have the right "side" of the vertical amp. Do this by setting up to measure one resistor on each half of the circuit and measure AC and DC voltage for each of these two resistors when the trace is wonky. I would expect that wherever you see AC, that is the bad half of the circuit.

Once you get the caps replaced, if that doesn't fix it, we should try to get back to testing components.

Michael

 

[sPuDd]

Complete Hung Chang 3502 Service Manual

See attached. Mine is about 18 years old. I've had to repair it a few times already. Current fault is no control over brightness - trace at full bright. And the Calibration Osc is outputing something closer to a sine wave.

sPuDd..


View attachment Hung Chang 3502.pdf

 

[shimniok]

Cool thanks for posting the manual.

I am going to go out on a limb and guess that the brightness thing is either a problem with the power supply or possibly the Z amplifier -- but you might want to check the intensity pot to make sure it is ok too.

I'd start by verifying the output of the voltage supplies that feed the Z amp (or heck, all of them). There's an adjustment procedure for the +200V supply and the -1.9kV supply. Service manual, sections 5-2-1 / 5-2-2 "Power Supply Adjustments" If pwr supply can't be adjusted correctly then there must be a fault in there somewhere.

Under that same section, follow the procedure for adjusting the intensity limit and see if that helps. Section 5-2-2, "Adjustments of intensity limit" If intensity can't be adjusted properly then I'd suspect some component in the Z amp is going out of spec.

Michael

 

[pierreretief]

Oh, just so by the way, my intensity knob also doesnt change anything, I didnt really think much of it since it looks fine at what ever level it is fixed at, it might be full brightness, but the beam looks fine...

 

[unclejed613]

don't worry anout it... yet..... still have to find the big problem before chasing the smaller one if it exists.... if you have a function generator and your multimeter reads AC volts even reasonably well at 100khz (doesn't have to be accurate, just has to work linearly) you can make an ESR tester by putting your multimeter across the oscillator and dialing in 100khz and adjust the amplitude so that your multimeter reads 1V AC. connect test probes to this getup and just measure across capacitors. the output impedance of the function generator is usually 50 ohms, so if the ESR of the cap is less than 10 ohms, you should read almost nothing across most electrolytics larger than 10uf. for electrolytics smaller than 10uf you will need to compare to known good one. remember, the reading on the meter is proportional to the ESR plus the reactance of the capacitor. you can use the probes directly on caps in circuit as long as the circuit is unpowered and unplugged AND the caps are discharged and have no voltage on them. if you need to know what the reactance of the capacitor is, the formula is 1/(628000)*C at 100khz c is in farads. that's in series with 50 ohms, so you can figure what volage sould appear there.

 

[sPuDd]

All found wrong was one of the 22M resistors in the screen area was failing at high voltage. I put in a substitute but I suspect something else is faulty. Voltages all measure right and the pot is fine. There was a period of arcing somewhere, and I suspect its caused damage. I'm not really enthused to find it, its old. So that’s how its stayed.

sPuDd..

 

[shimniok]

So, back to the OP problem -- want to try and help you stay on track here. We need to eliminte the electrolytic caps next.

Do you want to go ahead and just replace caps on the "bad" side?

Or do you have a freq generator and want to test the caps first?

If you don't have a freq gen, the simplest route is replace. If the new caps measure within tolerance of the old caps, just replace 'em on the bad "half" of the circuit. You should verify that we have the right "half" of the circuit before doing this. Why?

The reason to only do half--don't fix what isn't broke or there's a chance of adding more problems rather than solving them.

It's easy to think "oh it shouldn't matter" only to screw yourself up royally in trying to troubleshoot. Preserve that parts that work at all costs!

Michael

 

[pierreretief]

Im just going to replace them, since I dont have a frequency generator. But I read through your description on how to find the "bad" side, and I dont really understand. And those 2 680ohm resistors you pointed out in the picture are not the ones on y- and y+, I checked, they are R97 and R102 so I am still on a mission to find them...

 

[shimniok]

Im just going to replace them, since I dont have a frequency generator. But I read through your description on how to find the "bad" side, and I dont really understand. And those 2 680ohm resistors you pointed out in the picture are not the ones on y- and y+, I checked, they are R97 and R102 so I am still on a mission to find them...

Let me know if you find them! I will go and look at my scope again too.

Meanwhile... let me explain what I mean about finding the 'bad' side ... let's take a step back.

We know there is a weird trace pattern when you position the trace below a certain point.

The way an o-scope works is that it sets voltages on the Y+ and Y- plates to "steer" electrons up or down (it also applies voltage signal -- a sawtooth in fact -- to X+ and X- to steer them left and right).

When a small, fluctuating, periodic signal comes in thru the probe, it is then amplified to a larger voltage range and sent to the Y+ and Y- plates, steering the electron beam up and down to form the trace on the screen (I am leaving out stuff, but you get the idea). You would see an alternating voltage on the Y+ and Y- plates if you were to measure.

When the scope has no signal or if it is GND coupled, then only a DC voltage appears on the plates, causing no deflection of the electron beam--it is a straight line.

When your scope is set to GND coupled and it shows the wonky trace, the beam is being deflected even though it shouldn't be. In some of the video you've posted, you can see that there is a pattern to the way in which the beam is deflected in this condition. That means that there is something other than a DC voltage signal appearing on the Y- plate -- in other words, there is an AC voltage signal appearing there.

What I theorize is that there is a DC voltage on the Y+ plate but a fluctuating (AC) voltage on the Y- plate.

Now let's talk about multimeters. When you measure DC voltage on your DMM, you are measuring ONLY the DC voltage; the AC voltage, if present, is filtered out (think capacitor in parallel). When you measure AC voltage, the DC offset (voltage) is filtered out (think capacitor in series).

So back to the plates. If your signal trace is flat, that means the signal at the plates is DC only. Otherwise the trace would be deflected up and down. If you measure AC volts with DMM it should read 0 because there is no AC component to the signal. But-- when the signal is wonky, I *think* the Y- plate signal has an AC component and so if you measure for AC volts, you should get a non-zero measurement.

In fact, one thing you could try to help me out bigtime is to measure at the Y+/Y- jumper --->

1. Power off/unplug
2. Unplug the jumper (black and white wires with white plug) that goes to the Y+/Y- plates.
3. Attach DMM using mini clip to ONE of the pins.
4. Power up and set to DC coupling
5. Measure AC volts reading as you move the position knob up then down. Record values for full up versus full down
6. Power off scope and connect mini clip to the OTHER pin
7. Power on and measure AC volts reading as you move position knob up and down -- record values for full up versus full down.

I am curious to know if you see an AC signal on one or both of those pins at any time.

I predict that you will see AC voltage only on one pin and only when the position knob is turned towards one direction and not the other (ie, when the trace position is downwards.

If only one pin shows an AC voltage, that pin is hooked up to the "bad" side and so you should replace the caps in that half of the circuit to see what happens.

If you were to replace all the caps on both halves as well as the 2 or 3 caps that are sort of "shared" or "in the middle" of the circuit, there is a chance of horking up the circuit even more which puts us farther away from getting to a solution because we now have to solve the new problem before we can get back to solving the original one.

I've done a ***lot*** of fixing of very complex, aggravating, bizarre, difficult, "why me, God" kinds of problems with cars, software, and some electronics and it is really, really important not to make things worse. So, it's important not to be really really careful not to muck with anything that already works.

Michael

 

[unclejed613]

this circuit is a lot like a "bridged" amplifier. the deflection voltage is the difference between the deflection plates, which is why the circuit can work with only positive supply voltages. one side of the bridge is flaking out on you, so the trick is to figure out which side of the bridge is bad. take a step back and look at the physics of it. you have a beam of negatively charged electrons passing by these plates, and they will be attracted to whichever plate is more positive. so you are either losing positive voltage on the Y+ side, or you are getting latching to the rail on the Y- side. so measuring with an AC meter will tell you whether the Y+ driver or the Y- driver is the one acting up, and this will give you better clues where to look for bad components. one of these two circuits is generating an AC voltage when it's output should be straight DC, so whichever of the plates has AC on it when the scope acts up is the side of the circuit you are looking for. the other thing that might be a clue is with the position controls centered, your trace is NOT centered. both plates SHOULD be at about +60V with the vertical position controls centered. my guess is that one of them is at +60V and the other one is at a vastly different voltage. the one not at +60V is probably the half of the circuit with the problem.

this is also a good reason to have the input switched to GND, so that when you have the position control centered, the bias on both Y plates is supposed to be half the rail voltage of +120V. changing the position control changes the balance of the bias between the plates.

 

[pierreretief]

Great! That was so informative... Now you got me thinking, and I love thinking, and not to just measure voltages without really knowing why.. A couple of questions again...

"(it also applies voltage signal -- a sawtooth in fact -- to X+ and X- to steer them left and right)."

For interest sake, why a saw tooth, how does that actually bring the trace back to start at the left hand side again? Was just wondering about that...


"What I theorize is that there is a DC voltage on the Y+ plate but a fluctuating (AC) voltage on the Y- plate."

Then, How is it possible that the trace will ever be straight, "like when the trace is quite high" if there is a AC on the y- side, or does the AC stop somewhere higher up and then both y+ and y- are DC?


"Unplug the jumper (black and white wires with white plug) that goes to the Y+/Y- plates."

Ok, trying to identify this plug. There are two white plugs, going to the CRT one is P3 and one is P203... Not quite sure which one is the Y+ Y- negative one...


Ok, lastly, you said if you measure AC on one of the resistors, there could be a problem.. If look back at one of my posts, I posted the voltages I measured on one of the resistors, It had DC voltages and AC voltages... Could that not have been the problem or part thereof?

Thanks once again for all the info, I really really appreciate it!