[FIXED] Analog Oscilloscope Repair Help

[emanuelemessina]

Hi,
i recently bought my first oscilloscope (second-hand, many years old).

I cant find the exact same model online, but using google image search i found two oscilloscopes (volcraft 2040, beckman industrial 9020) that look the same and have the same manuals (manual of beckman attached, the volcraft manual is in german but they are equivalent in content and schematics: https://pdfhost.io/v/9MZQ0cqfk_Beckman_Industrial_9020pdf.pdf). (i've also taken the scope apart and the circuitboards are the same)

1) The main problem is that using high time settings (from 20 ms up) i can clearly see that the sweep period is extremely long (https://imgur.com/a/9QHe852), so i can't visualize many waves that have low frequencies.
I viewed a video on youtube of a guy that have the same problem on a tektronix oscilloscope, and he commented that he resolved the issue by replacing a resistor that went bad in the sweep generator circuit.
I tried watching the sweep signal coming out the back of the scope and it does not look like a sawtooth, but a long oblique line (is it normal? i don't think so. Photo taken in search delay mode cutting the left part of the line and fit it all in the screen, to show that it's not a wave but just a line: https://imgur.com/a/B0L2CXl).

2) The ch2 variable uncalibrated deflection sensitivity (red VAR knob over ch2 bnc on the right) definitely has a problem, both touching it (https://imgur.com/a/YYkhFFJ) and not touching it (https://imgur.com/a/xRo45v7). (photos taken visualizing default cal square signal)
I don't know if it is a soldering problem or something bad happened in ch2 circuit.

3) Just.. i don't even know, really (https://imgur.com/a/FO066Xk). The probe is just attached to an alligator clip, attached to nothing, and it shows a kind-of-sine wave: what??? If i disconnect the alligator clip from the tip of the probe, the amplitude decreases, but the wave remains for quite some time (https://imgur.com/a/9vTjq2c), almost as if the circuit has some kind of hysteresis on the fact that i touched some piece of metal on my desk.
It happens all the time, it's really annoying, and i don't even have the minimum clue on what could be the problem.
It happens in both ac and dc coupling, it also happens when i touch the tip with my skin, it happens every time the tip touches something conductive really, the only case when this does not happen is when the probe is properly grounded (like + an - on a battery or transformer, L/R and GND on an audio jack ecc..)

4) The probe is correctly compensated (i hope) (https://imgur.com/a/JX0r8Js), but increasing volts/div shows lack of compensation (https://imgur.com/a/9TCi3IG).

Overall i could say the the scope needs internal calibration (i dont know what trimmers to touch for what purpose), and i think something needs to be replaced because it went bad (i dont know where to start looking).

I can't figure it out on my own, i have zero experience in oscilloscopes and repairing faulty circuits, i don't know how complex circuits work, and the schematics are rightly dense and complex, i don't even know where to start looking and what to troubleshoot, so i'm asking here, hoping someone with experience and knowledge may help me and guide me to repair this scope.

PS: I'm really sad because this is my first scope and i've been looking for one to buy for cheap because i'm just a student with really low budget, and i wanted to use it to learn new things and expand my knowledge and experience, but i can't do so with a broken scope, and i don't have the knowledge yet to repair the scope that would have helped me understand how circuits work and how to repair them... so, yeah, i need the help of someone, please..

[bob91343]

The oblique line is probably the beginning of a very slow sawtooth wave.  It does appear that you have an open in the sweep generator.

As for different compensation for different sensitivity settings, that is misalignment of the input attenuator.

Touching metal with the probe merely picks up radiated signal from the power line.  Try shorting it and the signal should disappear.

[Gyro]

PS: I'm really sad because this is my first scope and i've been looking for one to buy for cheap because i'm just a student with really low budget, and i wanted to use it to learn new things and expand my knowledge and experience, but i can't do so with a broken scope, and i don't have the knowledge yet to repair the scope that would have helped me understand how circuits work and how to repair them... so, yeah, i need the help of someone, please..

Welcome to the forum.

Firstly, don't be sad, it looks as if you have a scope that is mostly, maybe even completely, functional. You have also successfully tracked down a suitable manual, which is an essential thing for an old scope.

The X and Y circuits seem to be working fine, the trace is bright and well focussed which means that there is very little wrong with it - you don't need to go messing with the high voltage sections. It may even just be your lack of detailed understanding of how to use an analogue scope.

1) It is normal on slower timebase speeds to be able to spot moving across the screen rather than a line. I can't tell from your description whether this is happening too early (timebase above 20ms/div really slow too soon) or not. If so, it might just be something like switch contacts in the timebase circuit.

The sweep signal on the back is an output of what the timebase is feeding to the X axis. What you are seeing is a straight diagonal line, nice and linear, which is what you would expect - the voltage rises as the dot sweeps across the screen. You would only expect to see an actual sawtooth wave (multiple cycles) if you were looking at it with another scope, set to a slower timebase.

2) If I understand correctly, that just looks like noisy potentiometers. Exercising them backwards and forwards and maybe the appropriate cleaning spray will probably improve that.

3)  Yes, that's quite normal for an open scope probe hooked to an 'antenna'. It is picking up stray mains voltage fields, it would be very rare to see a perfect sinewave under these conditions. Please don't try to fix that by trying to scope the mains directly - if you have a low voltage mains transformer, try scoping that instead, or one of those really cheap function generator kits off ebay.

4) You should compensate the probe with the scope set to view a couple of cycles of the 1kHz compensation output, and with a nice big trace height on the screen so you can clearly see and adjust the square wave against the lines on the graticule. There may be a small issue with input capacitance on different ranges, but more likely it is just finger-trouble (until proven otherwise).


I would advise putting the lid back on for now, don't touch any trimmers yet. Find a good video on analogue scopes (there's probably one of Dave's on here somewhere, or on YT) and work through the operation section of the manual. Work through the scope operation slowly and become familiar with it. Work out exactly what doesn't behave as expected before attempting any repairs.

Be happy, your scope might have some minor issues (nothing that can't be fixed) but it appears to be producing clear and sensible waveforms. 

[emanuelemessina]

Thank you so much for the complete and reassuring answer,

It may even just be your lack of detailed understanding of how to use an analogue scope.

this is sure

I've made some videos to better show the problems

1) Here you can see that the sweep is reeeally slow https://imgur.com/a/zEhz7Tq

The sweep signal on the back is an output of what the timebase is feeding to the X axis. What you are seeing is a straight diagonal line, nice and linear, which is what you would expect - the voltage rises as the dot sweeps across the screen. You would only expect to see an actual sawtooth wave (multiple cycles) if you were looking at it with another scope, set to a slower timebase.

oh, ok.. on the pcb there is written sawtooth so i expected to see a sawtooth (https://imgur.com/a/YtRglgm), good to know.

Anyway something strange happens when i increase the time/div,  https://imgur.com/a/mmz1Wmh  , the line that should be a perfect line.. curves. what?? why?

2) Mmmh.. i don't know, watch this: https://imgur.com/a/gbA5HpD . The amplitude changes randomly by fixed values and not so fast, as if the potentiometer does not make contact completely, but there's this extremely amount of jitter and noise, are you still sure that's only a dirty pot?

3) I'm a good antenna i suppose    https://imgur.com/a/hfonpyJ 

It is picking up stray mains voltage fields, it would be very rare to see a perfect sinewave under these conditions.

 not a perfect sine but almost, sure this is just picked up from somewhere?

4) Here you can see that i cannot achieve perfect compensation by just adjusting the probe, https://imgur.com/a/41DvwUb , the square wave is never against the graticule, that's why i thought that some trimmer inside the scope needed to be adjusted.

I hope these videos clarify the issues or confirm what you already said

[Gyro]

You're welcome. I've got several windows open at the same time looking at the various videos so please forgive any confusion. It would be better to include any still images as attachments to your posts to save having to jump out to links, some members also don't like clicking on links in posts, so you might get more input. Videos are more difficult of course, there's no real option there.

1) Here you can see that the sweep is reeeally slow https://imgur.com/a/zEhz7Tq

Hmm. It is difficult to tell in the ms/div ranges due to the sweep rate beating with the video refresh rate but it does look slow, although each setting in the s/div range looks to be in correct proportion to the one above and below. I can't clearly see the knob positions (decimal points) where you are viewing the stray mains pickup but using this as a reference, how many ms is it reading for a full mains cycle (for 50Hz it should be 20ms of course). One thing that I immediately notice is that the 'uncal' light above the time/div knob is lit. This indicates that the small variable knob above is not at its click stop 'cal' position, where the main knob indications are valid. It might be as simple as that.

The sweep signal on the back is an output of what the timebase is feeding to the X axis. What you are seeing is a straight diagonal line, nice and linear, which is what you would expect - the voltage rises as the dot sweeps across the screen. You would only expect to see an actual sawtooth wave (multiple cycles) if you were looking at it with another scope, set to a slower timebase.

oh, ok.. on the pcb there is written sawtooth so i expected to see a sawtooth (https://imgur.com/a/YtRglgm), good to know.

That looks like a nice easy to work on scope, with simple single sided PCBs and nice clear markings (including adjustment trimmers). There won't be any difficult to source ICs etc. It looks nice and clean internally too.

Anyway something strange happens when i increase the time/div,  https://imgur.com/a/mmz1Wmh  , the line that should be a perfect line.. curves. what?? why?

I think this is because you have the channel input coupling set to 'AC', so the position is changing as the average DC level changes. Try it again with the input set to DC - It is normal to always leave the input coupling set to DC unless you are trying to look at a small AC signal superimposed on a higher DC voltage.

2) Mmmh.. i don't know, watch this: https://imgur.com/a/gbA5HpD . The amplitude changes randomly by fixed values and not so fast, as if the potentiometer does not make contact completely, but there's this extremely amount of jitter and noise, are you still sure that's only a dirty pot?

It does look as if you have a defective pot on Ch2. If I am seeing it correctly, this is the fine amplitude adjustment knob. These are very rarely used (hence they get noisy) - they are normally just left in their 'cal' position and the amplitude adjusted by the main V/div knobs. The fine adjustment pots (voltage and frequency ones) are very little use as they prevent you from being able to interpret voltages and time intervals directly off the CRT graticule.

3) I'm a good antenna i suppose    https://imgur.com/a/hfonpyJ

Yes you are. 

It is picking up stray mains voltage fields, it would be very rare to see a perfect sinewave under these conditions.

 not a perfect sine but almost, sure this is just picked up from somewhere?

Yes, it is all around you - at high impedance.

4) Here you can see that i cannot achieve perfect compensation by just adjusting the probe, https://imgur.com/a/41DvwUb , the square wave is never against the graticule, that's why i thought that some trimmer inside the scope needed to be adjusted.

Yes, it does look as if you may have some problem there, but again it may also be that you have the input coupling set to AC, so you may see a small amount of slope on the flat tops due to roll-off. Try the same thing with the input set to DC coupled and see if it improves. It looks as if the probes do have sufficient range to compensate the scope input as you are able to achieve both undershoot and overshoot.

[EDIT: Actually, after looking at the video again  the slope is on the wrong direction for it to be just due to AC coupling - It looks as if at one point you can almost get it flat with slight roll-off on the leading edge, this is preferable to overshoot for now. It may require internal adjustment to improve it later.]


I am not confident that I have definitely seen an actual hard fault at the moment (I may be wrong of course) other than noisy pots on some of the less used functions. What you are seeing may all be explainable by control setting issues at this stage. Try the things above and see what changes.

[PA0PBZ]

1) Here you can see that the sweep is reeeally slow https://imgur.com/a/zEhz7Tq

Set the HOLD-OFF knob to MINimum and the VAR to CAL, then see how it looks.

[emanuelemessina]

It would be better to include any still images as attachments to your posts to save having to jump out to links, some members also don't like clicking on links in posts, so you might get more input. Videos are more difficult of course, there's no real option there.

Sorry for the links, i tried several times to post with all kinds of attachments but it wouldn't let me post it, it just refreshed the page and erased all i wrote, without posting it, hence the links.
[EDIT: maybe it has to do with a filesize limit? now the attachments work, there are just two.]

1)

One thing that I immediately notice is that the 'uncal' light above the time/div knob is lit. This indicates that the small variable knob above is not at its click stop 'cal' position, where the main knob indications are valid. It might be as simple as that.

Set the HOLD-OFF knob to MINimum and the VAR to CAL, then see how it looks.

It looks exactly the same.. That's the first thing i noticed when i got the scope by the way, that playing with holdoff and var, the sweep in certain situations was just a little bit faster but.. nope.

Hmm. It is difficult to tell in the ms/div ranges due to the sweep rate beating with the video refresh rate but it does look slow, although each setting in the s/div range looks to be in correct proportion to the one above and below. I can't clearly see the knob positions (decimal points) where you are viewing the stray mains pickup but using this as a reference, how many ms is it reading for a full mains cycle (for 50Hz it should be 20ms of course).

Here i've made some videos (probing the cal square wave to better identify the slow beam) with the time/div settings in the description, starting from 50ms/div up to 5 sec/div, all with holdoff to minimum and var to cal: https://imgur.com/a/GSFaABe

I viewed a video on youtube of a guy that have the same problem on a tektronix oscilloscope, and he commented that he resolved the issue by replacing a resistor that went bad in the sweep generator circuit.

Here's the video, the first 25 seconds looks exactly my problem, except my scope doesn't stop the sweep, it just becomes slower and slower.

Anyway something strange happens when i increase the time/div,  https://imgur.com/a/mmz1Wmh  , the line that should be a perfect line.. curves. what?? why?

I think this is because you have the channel input coupling set to 'AC', so the position is changing as the average DC level changes. Try it again with the input set to DC - It is normal to always leave the input coupling set to DC unless you are trying to look at a small AC signal superimposed on a higher DC voltage.

That was it, now a straight line, thanks!.

2)

It does look as if you have a defective pot on Ch2. If I am seeing it correctly, this is the fine amplitude adjustment knob. These are very rarely used (hence they get noisy) - they are normally just left in their 'cal' position and the amplitude adjusted by the main V/div knobs. The fine adjustment pots (voltage and frequency ones) are very little use as they prevent you from being able to interpret voltages and time intervals directly off the CRT graticule.

Thanks for the info! In the meantime i found EEVblog #803 and #804 where Dave repairs an HP1740A that has, beyond other many problems, this same amplitude problem on ch2. I'll watch it and make an update in case.
[EDIT: it was a bad rectifier that cause all the problems, but not related to the var knob, no mention of that problem at the end actually, so yeah probably a noisy pot also in there]

3)

3) I'm a good antenna i suppose    https://imgur.com/a/hfonpyJ

Yes you are. 

It is picking up stray mains voltage fields, it would be very rare to see a perfect sinewave under these conditions.

 not a perfect sine but almost, sure this is just picked up from somewhere?

Yes, it is all around you - at high impedance.

Point 3 closed then 

4)

Yes, it does look as if you may have some problem there, but again it may also be that you have the input coupling set to AC, so you may see a small amount of slope on the flat tops due to roll-off. Try the same thing with the input set to DC coupled and see if it improves. It looks as if the probes do have sufficient range to compensate the scope input as you are able to achieve both undershoot and overshoot.

[EDIT: Actually, after looking at the video again  the slope is on the wrong direction for it to be just due to AC coupling - It looks as if at one point you can almost get it flat with slight roll-off on the leading edge, this is preferable to overshoot for now. It may require internal adjustment to improve it later.]

It's the same with ac and dc, there is something going on with volts/div:

10 v/div, undershoot:
5 v/div, overshoot:

[Gyro]

Ok, we've knocked a couple off the list then. 

1)

One thing that I immediately notice is that the 'uncal' light above the time/div knob is lit. This indicates that the small variable knob above is not at its click stop 'cal' position, where the main knob indications are valid. It might be as simple as that.

Set the HOLD-OFF knob to MINimum and the VAR to CAL, then see how it looks.

It looks exactly the same.. That's the first thing i noticed when i got the scope by the way, that playing with holdoff and var, the sweep in certain situations was just a little bit faster but.. nope.

No difference when the 'var' timebase adjustment is turned? That's probably a significant clue then.

It would be a good idea to first check the low voltage power supplies as detailed on Manual page 5-2.  Looking at the PSU PCB layout (page 5-3) it looks like the test points are all clustered around the TP2 end of the board. NOTE: Stay well away from the other end of the board, which has the -2kV and other high voltage supply components on it! Measure the low voltage (up to +24V). It is unlikely that there is a problem there (all fixed regulators) but it is sensible to just check. Measure them relative to the chassis.

Unfortunately the PSU board it the only one with a layout shown in the manual, so I can't see if the Var pot are PCB mounted or wired. Check carefully around the pot (VR936 on manual page 7-7) for poor connections, missing ground connection (particularly if the 'Holdoff pot next to it is also not behaving). Check the resistor values and joints in that area too (i.e. R935 and R937). There is a Timebase adjust preset (VR938) in that part too, but I doubt that it would have sufficient range to correct the issue that you're seeing (maybe though, particularly if it has failed open-circuit).

If you don't feel confident with the above then please ask somebody knowledgeable at your School / College / Uni (it is always difficult to judge age remotely) to assist. It is such an uncomplicated design (which is nice) that it ought to be an easy fix, but there ARE high voltages around so you do need caution. You ought to be able to do the checking around the Var pot with the power removed, but it would be useful to know if the voltage at Base of T902 (safer to pick it up at R933) varies with the VAR pot. If possible, follow it through to the Emitter of T901 - this is the common point feeding all of the Timebase range selection resistors, so is the last 'common' point before the range knob.

It is hard to see the connections back to the sweep generator circuit (page 7-5), but as it is horizontally sweeping, I think the problem is located on Page 7-7.

2)

It does look as if you have a defective pot on Ch2. If I am seeing it correctly, this is the fine amplitude adjustment knob. These are very rarely used (hence they get noisy) - they are normally just left in their 'cal' position and the amplitude adjusted by the main V/div knobs. The fine adjustment pots (voltage and frequency ones) are very little use as they prevent you from being able to interpret voltages and time intervals directly off the CRT graticule.

Thanks for the info! In the meantime i found EEVblog #803 and #804 where Dave repairs an HP1740A that has, beyond other many problems, this same amplitude problem on ch2. I'll watch it and make an update in case.
[EDIT: it was a bad rectifier that cause all the problems, but not related to the var knob, no mention of that problem at the end actually, so yeah probably a noisy pot also in there]

Yes, still looking like like a noisy pot then, ok that's probably improvable with just some cleaner.

4)

Yes, it does look as if you may have some problem there, but again it may also be that you have the input coupling set to AC, so you may see a small amount of slope on the flat tops due to roll-off. Try the same thing with the input set to DC coupled and see if it improves. It looks as if the probes do have sufficient range to compensate the scope input as you are able to achieve both undershoot and overshoot.

[EDIT: Actually, after looking at the video again  the slope is on the wrong direction for it to be just due to AC coupling - It looks as if at one point you can almost get it flat with slight roll-off on the leading edge, this is preferable to overshoot for now. It may require internal adjustment to improve it later.]

It's the same with ac and dc, there is something going on with volts/div:

10 v/div, undershoot: [ Attachment Invalid Or Does Not Exist ]
5 v/div, overshoot: [ Attachment Invalid Or Does Not Exist ]

Ok, maybe somebody has been tweaking the variable capacitors in the attenuator network before you got it then. Do your scope probes have a switchable X1 / X10 setting? If so, it would be useful to set them to the X1 position and repeat the test. The probe compensation trimmer will have no effect in the X1 position but it will enable us to tell whether it is the scope input capacitance is changing on different V/div settings or whether it is later in the attenuator (Manual page 7-2).

If the probes don't have a X1 setting, then a short length of single core wire from the probe cal terminal to the centre of the input BNC connector will be sufficient (be careful to make sure the wire is an easy fit in the BNC and doesn't stretch the contacts).

[emanuelemessina]

1)

It would be a good idea to first check the low voltage power supplies as detailed on Manual page 5-2.  Looking at the PSU PCB layout (page 5-3) it looks like the test points are all clustered around the TP2 end of the board. NOTE: Stay well away from the other end of the board, which has the -2kV and other high voltage supply components on it! Measure the low voltage (up to +24V). It is unlikely that there is a problem there (all fixed regulators) but it is sensible to just check. Measure them relative to the chassis.

Checked  : dc low voltage supplies are within tolerance.

Check carefully around the pot (VR936 on manual page 7-7) for poor connections, missing ground connection (particularly if the 'Holdoff pot next to it is also not behaving). Check the resistor values and joints in that area too (i.e. R935 and R937).

You ought to be able to do the checking around the Var pot with the power removed, but it would be useful to know if the voltage at Base of T902 (safer to pick it up at R933) varies with the VAR pot. If possible, follow it through to the Emitter of T901 - this is the common point feeding all of the Timebase range selection resistors, so is the last 'common' point before the range knob.

Hold-off and var do behave correctly, but the effect is subtle due to the fact that the timebase is so slow, but they seem do what they are supposed to do.
It is not practical to do visual and resistor values checking on the timebase board as it would require taking apart the front panel for me to access the board, so removing all the knobs and possibily the blue filter wich is very delicate as stated in the manual, plus a lot of screws, and other boards. So for the moment i'm not doing that since var and holdoff seems to do just fine.
Still, some portion of the board is exposed and i confirm that the voltages at B-T902 and E-T901 rise up with VAR.

There is a Timebase adjust preset (VR938) in that part too, but I doubt that it would have sufficient range to correct the issue that you're seeing (maybe though, particularly if it has failed open-circuit).

I turned timebase adjust all the way and now the sweep is a little bit faster but, as you guessed, not enough.

It is hard to see the connections back to the sweep generator circuit (page 7-5), but as it is horizontally sweeping, I think the problem is located on Page 7-7.

So.. i do have to remove all the knobs and screws.. 
2) Well, while i am at it i'll clean the noisy pot on ch2.

4)

Ok, maybe somebody has been tweaking the variable capacitors in the attenuator network before you got it then. Do your scope probes have a switchable X1 / X10 setting? If so, it would be useful to set them to the X1 position and repeat the test. The probe compensation trimmer will have no effect in the X1 position but it will enable us to tell whether it is the scope input capacitance is changing on different V/div settings or whether it is later in the attenuator (Manual page 7-2).

Yes they have X1, i checked at various v/div and there is undershoot both in ac and dc, and it becomes more marked with lower v/div settings of course, but it is always constant with respect to the wave amplitude on the display.

[Gyro]

Hmm, I was hoping that access would be much easier. Maybe a couple of photos of the inside would help.

I clearly misunderstood what you were saying about the Var and Holdoff pots, sorry. My best guess is that something in that area has shifted (solder joint?) to throw out all of the timebase speeds. Have you actually quantified the difference, I remember some videos, but it would be helpful if you could actually quantify it, ie. Using the graticule, what setting does is the timebase think it is on when it measures 20ms for a full mains cycle (ie. it ought to be 4 divisions wide at 5ms/div, 10 divisions at 2ms/div)). Same with the 1kHz cal (although this frequency is only approximate). How long does the spot take to traverse the full width (8 or 10 divisions?) at the slowest setting.

I wouldn't be too worried about the blue filter, they are designed to protect the front face of the CRT (in addition to colour filtering) so they are pretty tough. Obviously you don't want to scratch it though.

With regard to the input undershoot. It may be that somebody has tweaked some of the trimmer caps, but I'm reluctant to start tweaking them at random. Even though you have a X1 setting on the probe, it might still be worth trying a short length of wire as this would avoid the cable capacitance of the probe and might give a sharper result.

Any other suggestions welcome folks!

[shakalnokturn]

Not absolutely sure from the videos but it does look as if your sweep is non-linear at the end. It looks as if the sawtooth is good, although too slow, for 8 divisions then slows for the last 2 divisions.
You could check this with a variable frequency square wave input counting how many cycles you fit per div. at beginning and end of sweep in different timebase settings.
It would also give you a better idea if the timebase is defective on all settings.

The basics of the timebase sawtooth generation is charging of capacitor(s) at constant current (ramp) once a certain voltage is reached the capacitor is suddenly discharged (reset).
Frequency is adjusted by changing either the charge current (set by resistors), the value of the capacitor or a combination  of both.

A sweep that's too slow but linear throughout is going to be caused by a defect in the charge portion.
A non linear sweep is more likely to be caused by the discharge not happening soon enough because of bad threshold voltage or wrong delay.
The constant current charge has its limitations: Once the capacitor voltage approaches the power supply voltage the transistor used as a current source saturates, the charge finishes at constant voltage which is not linear.

[emanuelemessina]

I clearly misunderstood what you were saying about the Var and Holdoff pots, sorry. My best guess is that something in that area has shifted (solder joint?) to throw out all of the timebase speeds. Have you actually quantified the difference, I remember some videos, but it would be helpful if you could actually quantify it, ie. Using the graticule, what setting does is the timebase think it is on when it measures 20ms for a full mains cycle (ie. it ought to be 4 divisions wide at 5ms/div, 10 divisions at 2ms/div)). Same with the 1kHz cal (although this frequency is only approximate). How long does the spot take to traverse the full width (8 or 10 divisions?) at the slowest setting.

Don't worry, thanks always for the replies.
I've taken apart the front panel and visually inspected the timebase generator and the sweep board (pages 7-7, 7-5), the components look fine.
Just some components are boiling hot, like (page 7-7, sweep board) IC507 (voltage regulator, a little bit rusty but the voltages are right) and R971 and R972 (to CRT).
(PS: I've cleaned the noisy pot, now it works.)
There are so many components i don't know what to test first and how.
The solder joints are all on the other side of the boards, and they look fine too, i know that there are cases when you almost need a microscope to notice some cracks, but from what i've seen this kind of joints make the problem manifest itself after some minutes/hours of operation, when the scope gets hot, and that is not the case here.

The spot seems to take right about 10div*time/div seconds to traverse the 10 divisions on the screen (viewing it in the seconds per div range, in the ms/div range i cannot be precise but it looks right there too), wich is right, right? But then, why is it so slow? It cannot be right this way, it's so slow that one cannot even visualize the waveform past the 50 ms/div because it's just a point moving on the screen.
And worst of all, when i visualize the two channels at the same time, the beam has to halve the cycles, one cycle for each channel, so it is even worse..

Not absolutely sure from the videos but it does look as if your sweep is non-linear at the end. It looks as if the sawtooth is good, although too slow, for 8 divisions then slows for the last 2 divisions.
You could check this with a variable frequency square wave input counting how many cycles you fit per div. at beginning and end of sweep in different timebase settings.
It would also give you a better idea if the timebase is defective on all settings.

Thanks a lot for the info about the sweep generation process.
The sweep looks linear to me but maybe i don't have a trained eye, anyway i don't have a function generator to be sure.

I was going to measure the period of the cal square wave to give you more details but..

MAJOR PROBLEM (encountered while writing this post and doing the tests for you guys):

I would have done more tests BUT having taken the front panel apart to inspect the timebase board (wich is now floating mid air) and having it disconnected and reconnected to the sweep board, something has changed. Now channel 1 is funky and channel 2 beam becomes more curved with higher time/div settings... plus i don't know what other problems. In short, pretty much everything about visualizing the cal wave became screwed, the controls work, its just the channels and the timebase and the deflection and i don't know what..

So.. i guessed something snapped? Moved? Too much problems to be just gravity and interference. Ooooh god what just happened...
https://imgur.com/a/iOZlTCU

[PA0PBZ]

That just looks like the ground of your probe is not connected...

[emanuelemessina]

That just looks like the ground of your probe is not connected...

I don't think you actually read the discussion or even the last message carefully.

[Gyro]

The spot seems to take right about 10div*time/div seconds to traverse the 10 divisions on the screen (viewing it in the seconds per div range, in the ms/div range i cannot be precise but it looks right there too), wich is right, right? But then, why is it so slow? It cannot be right this way, it's so slow that one cannot even visualize the waveform past the 50 ms/div because it's just a point moving on the screen.

Now I'm really confused -  I thought you were saying in your description and videos that the sweep speed was too slow compared to the timebase knob setting (which is not really visible in the videos). It seems now that you are measuring them at the correct sweep rate. ie. the time for 10 divisions of sweep is 10 times the time/div setting. On a 50ms/div, the time to traverse 10 divisions is 0.5 seconds, so the spot will clearly be visible. To be viewing a waveform (fitting one complete cycle across the width of the screen) you would be looking at a 2Hz waveform at this setting, which is a very low frequency. To see one complete cycle of your picked up 50Hz mains waveform across the screen width the timebase would be 2ms/div and the 1kHz calibrator, 0.1ms/div. Is this what you are seeing? If so, then the horizontal sweep speeds are correct. 

Even on a digital scope, below around 50ms/div, they normally default to roll mode where the screen scrolls a little like a chart recorder with the new voltage coming in at the right and scrolling across to the left. Analogue scopes do not have this feature, so you see the dot on the CRT traverse the screen from left to right. For slowly changing voltages, you monitor the dot position against the graticule.

And worst of all, when i visualize the two channels at the same time, the beam has to halve the cycles, one cycle for each channel, so it is even worse..

The scope has two modes of display for dual channel. Alternate and Chop. Alternate is where it displays a sweep of each channel alternately, whereas in Chop mode, it switches quickly between displaying both traces at a very fast rate (500kHz according to your manual), which is too fast for your eye to see, so you see two continuous traces simultaneously. Normally you use chop mode unless the frequency you are monitoring is near the chop frequency, which can cause display artefacts, or sometimes at very high frequencies where it gives a better display. It sounds as if you have the switch set to alternate (button 13 on page 3-1).

You really do need to study the manual carefully and watch some instructional videos as I suggested.

MAJOR PROBLEM (encountered while writing this post and doing the tests for you guys):

I would have done more tests BUT having taken the front panel apart to inspect the timebase board (which is now floating mid air) and having it disconnected and reconnected to the sweep board, something has changed. Now channel 1 is funky and channel 2 beam becomes more curved with higher time/div settings... plus i don't know what other problems. In short, pretty much everything about visualizing the cal wave became screwed, the controls work, its just the channels and the timebase and the deflection and i don't know what..

I wasn't signed in last night when your posted. Hopefully you turned off the power and stabilized the board (it sounds as if it is more mechanically awkward that you would expect for a relatively simple scope - you didn't post the internal photos). It is not good to work on a board (potentially with high voltages) floating in free air, it is always best to secure it to something, even if you just use string to support it and take stress off the wiring connections.

So.. i guessed something snapped? Moved? Too much problems to be just gravity and interference. Ooooh god what just happened...
https://imgur.com/a/iOZlTCU

Very hard to say when all we can see is the screen and not what you are doing as the waveform changes.

It sounds as if you have hopefully just dislodged a connector or maybe broken an interconnect wire which can be easily fixed, again it's difficult to see without pictures of the internal construction. Sometimes (not often) boards have grounding points to the chassis through mechanical fixings. It looks as if you have maybe dislodged one of the Y axis (board?) connections. Hopefully you have tracked down the problem by now, but if not, leave the power off, take a deep breath, and methodically work through all interconnects, checking continuity where needed. Only apply power again once you have the board secured and have checked everything. You now have the CH2 fine gain adjustment pot clean and working at least.

Please proceed more slowly and with caution. Again internal pictures would help, it's very hard to see what you are doing at a distance.

[emanuelemessina]

First of all, sorry for the delay, but with all the boards floating one touched the chassis and shorted the 24v line i think.
A resistor in the power supply blew up, and a flip flop in the sweep generator said goodbye.

That taught me a lesson.

I insulated several parts of the chassis so i could work with the boards comfortably and understand the problem.
I replaced the resistor, and i had to understand how an analog oscilloscope works to come to the conclusion that the flip flop was burnt.
I checked, for what it was possible (xy mode helped), that all the other parts were working, the attenuators, the deflectors, ecc..
All lead to the sweep generator being faulty as there was constant blanking, i desoldered the monostable and the flip flop and tested them on a breadboard, only a flip flop on the right side of the IC701 was not working.

I ordered some chips online, i had to wait for them to arrive (a month), now i've replaced the flip flop and the scope is working better than ever.

As for the problem of everything being funky on the display, that was most certainly the sweep board touching the chassis randomly because i removed the screws that were securing it firmly. Insulating the chassis after the short circuit lesson fixed it.

And that's what happened behind the scenes, i wrote it so anyone reading this can learn something as i did.

Back to the main "problem":

Now I'm really confused -  I thought you were saying in your description and videos that the sweep speed was too slow compared to the timebase knob setting (which is not really visible in the videos). It seems now that you are measuring them at the correct sweep rate. ie. the time for 10 divisions of sweep is 10 times the time/div setting. On a 50ms/div, the time to traverse 10 divisions is 0.5 seconds, so the spot will clearly be visible. To be viewing a waveform (fitting one complete cycle across the width of the screen) you would be looking at a 2Hz waveform at this setting, which is a very low frequency. To see one complete cycle of your picked up 50Hz mains waveform across the screen width the timebase would be 2ms/div and the 1kHz calibrator, 0.1ms/div. Is this what you are seeing? If so, then the horizontal sweep speeds are correct. 

Even on a digital scope, below around 50ms/div, they normally default to roll mode where the screen scrolls a little like a chart recorder with the new voltage coming in at the right and scrolling across to the left. Analogue scopes do not have this feature, so you see the dot on the CRT traverse the screen from left to right. For slowly changing voltages, you monitor the dot position against the graticule.

Basically that was all of it, my misunderstanding, or i should say my "not knowing how an analog oscilloscope with no digital storage works".
Having to fix the short circuit i had to understand most of the boards functions in this scope, relate that to the schematics and vice versa.
Now it is clear to me, it's basically just an horizontal (sweep) and a vertical (probes) deflector of the beam, it cannot show the content of a wave past its current horizontal position, because it has no memory.
Very few (if any) videos show or tell this, even instructive videos for one who is touching an oscilloscope for the first time, well it is obvious now that i know that not all scopes have memory, but it was not obvious before and i bet it is not obvious to anyone approaching oscilloscopes for the first time.

So, in conclusion: everything is working fine now (maybe i have to adjust the attenuator caps), thanks for all the help and patience, i've learned a lot, i hope someone coming here will learn something as well. 

[Gyro]

Whew, that's a relief. I feared the result had been terminal!

Thanks for reporting back - we often don't get to hear how things turned out. 

Congratulations on the good fault analysis and repair too - You now have the very considerable advantage of thoroughly knowing your scope, inside and out. A very useful learning experience.