Oscilloscope Vertical?

[neglectedtamagotchi]

Well, hope it isn't rude for this to be my first post. I bought a Hitachi V-212 analogue scope off eBay a while back. It worked on arrival, but I figured it'd probably do well for a calibration, or at least verification against known goods. (My uni has some nice function generators and some good Tektronix DSOs, and got permission to use that.) Lo and behold, the day before, the vertical deflection gave out just a bit above the middle graticle, meaning that a sine wave would appear to be half-rectified, negative.

I never got a picture of this behavior before it failed entirely, but V-adjust worked, and that sine wave would appear as a full wave, if I just moved the beam down on the CRT. It refused to deflect the beam higher than halfway up. An hour later, it completely gave. The horizontal sweep seems to be fine, it responds to switching the time base, but instead of a point I just get a slightly broad, and somewhat faint vertical line from top to bottom. For twenty bucks it's not a big deal, but it was a nice scope and I'd like to fix it if possible.

Owner's Manual: http://dcfa.exa.unicen.edu.ar/lab/manuales/Osciloscopio_hitachi_v-212.pdf
Service Manual: http://oldradio.tesla.hu/rajzok/muszerek/hitachi%20v-211-212-222-422%20oscilloszkop.pdf

[FlyingHacker]

I am in the middle of another scope repair. I will take a guess.

This happens with both input channels, right?

Of course, check all the power supply rails before doing anything. Check for both voltage and ripple. If the ripple is high then this might indicate bad smoothing caps, which is highly likely, though not necessarily the cause of your issue (should be replaced though).

I would start by looking at all the electrolytic caps. Measure them with an ESR meter, ideally after lifting one leg. Once those have been replaced (good idea) then I would look at the vertical amplifier transistors, and vertical output amp transistors (block diagram shows both). If I had to guess I would say some cap(s) went bad and caused a transistor or two to fail. It looks like the schematic for V Out shows some voltages at various points. Check those to be within a margin of error. Usually the docs show what settings were used for those readings. Find that info. Also see if they say what type of meter was used for the readings. If it was a 20,000 ohm/volt analog meter then your readings with a DMM will be a little higher at lower voltages. Looks like it says DMM. So that should not be an issue. I would work backwards from Y+ to CRT and V- to CRT (top and bottom, towards the right side of the schematic for V OUT.


  Don't connect another scope to this scope without watching Dave's video on how to not blow up your scope. You may need an isolation transformer, or at the very least you need to be sure to ONLY connect the ground lead of your probes to case ground (or remove the ground lead from the probe - don't leave it dangling- and try that first, attaching it only if you need to).

If you are comfortable connecting a scope to this then by all means do, as it will be much easier to troubleshoot than with a DMM alone. Note that there are waveforms shown in the schematics which can be useful for troubleshooting.

Don't connect your meter to anything at high voltage. The CRT circuit is likely over 1500 volts, maybe double that. Most meters cannot go that high without an HV probe. Read the voltage rating on any caps you check before you check them. If they say something like 2000 volts then don't touch them at all. Remember CRT circuits can store huge voltages for DAYS depending on how they are designed.

Also note that there will be a lot of relatively high voltages (>30V, which is high enough to kill you) all over the scope. So be careful, and don't touch anything with it on! The transistor metal cans or metal tabs are frequently live. For measurements to ground clip one lead to the case so you only have to worry about one at a time, and can keep one hand behind your back whenever possible.

SOMETIMES (though with your symptoms I doubt it) simply doing the calibration procedure can get things going again.

[singapol]

If you see a vertical line means the horizontal deflection has failed or could just be cracked solder joints.
See schematic of horizontal circuit on page 81 of service manual. Or if it's a point then the vertical deflection has failed or cracked solder joints.

Edit: Sorry I meant if you see a horizontal line. A point means the sweep generator has a fault.

[FlyingHacker]

If you see a vertical line means the horizontal deflection has failed or could just be cracked solder joints.
See schematic of horizontal circuit on page 81 of service manual. Or if it's a point then the vertical deflection has failed or cracked solder joints.

Maybe I misunderstood the OP. What do you get on the screen? I thought your horizontal timebase was working OK. Is the beam sweeping across the screen at the slower settings? Is it a soft vertical line sweeping horizontally across the screen? Or what? Pictures? Videos?

Lo and behold, the day before, the vertical deflection gave out just a bit above the middle graticle, meaning that a sine wave would appear to be half-rectified, negative.

This comment makes me wonder if it isn't in the vertical preamp? Though I think that would only apply to a single channel. It could be a positive power rail is bad causing clipping? Definitely check the voltages first. If the power rails are bad or have crazy ripple then it could do lots of weird things.

[neglectedtamagotchi]

I checked the voltages already, the 2-odd kilovolt supply excluded. (Something feels stupidly dangerous about home-rigging anything HV, even a voltage divider.) The voltages were all well within spec, last I checked, and I replaced a bunch of dodgy capacitors already soon after buying it. The horizontal works just fine; the vertical line isn't stationary. It's blurry, but it makes sense in the context of no working vertical. In other words, yes, soft vertical line sweeping across, obviously only visible on low timebases due to phosphor persistence.

The issue of it clipping vertically above the midpoint occurred with both channels, yes.

Even if it's for horizontal, though, might as well take a good look at the circuit for page 81, maybe a solder joint went bad somewhere.

[tautech]

The DC voltage values in the schematics can be useful for finding the area of failure, in your case it would seem to be only the vertical system. There are only a few of these DC voltages listed and for them to be valid the scope must be set as outined on P11, Preliminary settings.
P75 has a DC value if 62.4V with the vertical position set midway and the range to which this test point can be adjusted to.(Vertical position)
It is not uncommon for the CRT plate output transistors to have failed and these DC values allow you to check correct operation.

[tautech]

After a further read of the posts you might have 2 problems. 

Have you access to another scope? As you see there are many waveforms in the schematics just for tracing faults.

But ensure the basics are correct first, PSU voltages and ripple within spec.
Then confirm you have a clean sweep.
Then the output stages.

This may assist:
http://www.sphere.bc.ca/test/tek-parts/troubleshooting-scopes.pdf

[singapol]

I checked the voltages already, the 2-odd kilovolt supply excluded. (Something feels stupidly dangerous about home-rigging anything HV, even a voltage divider.) The voltages were all well within spec, last I checked, and I replaced a bunch of dodgy capacitors already soon after buying it. The horizontal works just fine; the vertical line isn't stationary. It's blurry, but it makes sense in the context of no working vertical. In other words, yes, soft vertical line sweeping across, obviously only visible on low timebases due to phosphor persistence.

The issue of it clipping vertically above the midpoint occurred with both channels, yes.

Even if it's for horizontal, though, might as well take a good look at the circuit for page 81, maybe a solder joint went bad somewhere.

If the horizontal circuits are good then you may have a vertical problem. Check Vertical preamp trigger pickoff & channel switch on page 73.

Check Vertical out on page 75. Trigger generator on page 77 and sweep generator on page 79?

[FlyingHacker]

I know others mentioned it (and I did), but do be sure to check the ripple on the power supply rails, not just the voltage.

In trouble shooting a scope I was working on I found it helpful to work backwards from the deflection voltages going to the tube on back to the preamp. You can do a binary search method once you confirm the start and end points for what they look like on your scope (Check things in the in the middle and you know if it is in the first or second half. Then cut that in half, etc., etc.)

[neglectedtamagotchi]

Sorry for taking so long to reply. Power supply rails are actually just fine. One of the transistors in the vertical output completely shorted across, and took out some of the transistors with it, so it seems. One of the parts is already a Fairchild replacement, or so it seems, I wouldn't be surprised if that had something to do with it - mismatched to the rest of the original Hitachi and Sanyo 2SCun-obsolete-anium parts. So, now it's a matter of just trying to find either original replacements or substitute new parts. If I'm right, it's generally just a deal of the VBC/BE/CEO being higher or equal, and fT being around the same or higher. Some fudging room the bigger numbers. Yes?

[tautech]

They're readily available if you look harder.
With any substitution ensure you select a sufficiently high enough voltage rating for the output stages.
These transistors are normally of the video output deflection type, ensure replacements are too.

Put up links to any datasheets so guidance can be offered.

[FlyingHacker]

Probably not a bad plan to have both vertical deflection output transistors the same, but do post info for discussion (mainly so I can hear Tautech explain it, as he knows quite a lot more than I do).

[tautech]

Sorry for taking so long to reply. Power supply rails are actually just fine. One of the transistors in the vertical output completely shorted across, and took out some of the transistors with it, so it seems. One of the parts is already a Fairchild replacement, or so it seems, I wouldn't be surprised if that had something to do with it - mismatched to the rest of the original Hitachi and Sanyo 2SCun-obsolete-anium parts. So, now it's a matter of just trying to find either original replacements or substitute new parts. If I'm right, it's generally just a deal of the VBC/BE/CEO being higher or equal, and fT being around the same or higher. Some fudging room the bigger numbers. Yes?

You haven't said which ones, why, we do not have a crystal ball.

So here's the Vout schematic so we are all on the same page:



Links to all datsheets for the V ouput stage bipolars:
2SA1210-S
http://www.datasheets360.com/pdf/-4107135149606732446
2SC2901
http://www.bluecolour.net/data/personal/bluecolour/Transistors/NPN%20Small%20Signal%20Transistors/2SC2901%20TO92.pdf
2SA781K
http://pdf1.alldatasheet.com/datasheet-pdf/view/107602/HITACHI/2SA781K.html

Double check packages and pinouts match those presently installed.

Sources:
2SA1210-S
http://www.aliexpress.com/item/IC-2SA1210-2SA1210-10pcs/1949323771.html?spm=2114.01020208.3.85.LqalT7&ws_ab_test=searchweb201556_6_71_72_73_74_75,searchweb201527_5,searchweb201560_9
2SC2901
http://www.aliexpress.com/item/IC-2SC2901-2SC2901-10pcs/1944224208.html?spm=2114.01020208.3.11.vPEJ5Z&ws_ab_test=searchweb201556_6_71_72_73_74_75,searchweb201527_5,searchweb201560_9
2SA781K
http://www.aliexpress.com/item/4pcs-2SA781K-Original-New-Transistor/1431097357.html?spm=2114.01020208.3.18.UpbqGx&ws_ab_test=searchweb201556_6_71_72_73_74_75,searchweb201527_5,searchweb201560_9

The first two are available from UTSOURCE a well know seller also from eBay
The 2SA781K is not so common so I've picked what looks to be a quality seller. 


Note:
The S in 2SA1210-S refers to the Hfe range of this device, in this case it's noted on the datsheet as 140-280. That means all 2SA1210 devices that fall in that range.
When this parameter of a specific device is needed and not available it is normal to buy extras and attempt to find close matches when a pair is needed. In this case just finding a pair within the 140-280 Hfe range will be OK.

[neglectedtamagotchi]

Well, sorry for making you put all that up. I would've provided them and linked to the datasheets I dug up today if you hadn't.

Anyways, the aforementioned Fairchild part was a Fairchild Semiconductor KSA1381E, last letter specifying hFE class E, putting it at 100~200 according to the datasheet. It originally was a 2SA2912S, in the position of TR376. So it turns out, though, it wasn't alone. The pair, TR376 and TR375 were KSA1381s, leading me to think it might be fine to just buy a pair of new KSA1381E from Mouser.

In any case, I was more hoping to replace these with in-production substitutes where at all possible; the replacements, could total more than I paid to buy it, and put me instead in the market for a new scope. The unit cost isn't that bad with e-packet free shipping, but  I was thinking of replacing all ten transistors in the vertical section; just in case the failure might propagated or made any go marginal/intermittent; these are old parts, failure modes do start to get weird at this age. I might as well swap them all while I have it open, right?

The 2N3904 seems a close enough match to a 2SC641K (as installed, hFE class C, 80~160@Vce=0.5V and Ic=1mA, while the part here has a min hFE of 70 at Vce=1V and Ic=1mA).

It is worth noting that the schematic is actually wrong; specifically just for the parts specified. The second pair of 2901s probably got too much power pushed through them in development, so they replaced them with the second pair of TO-126 parts. Checking the helpfully included BOM (pg 40), the transistors are as specified, with links leading to the alldatasheet page:
TR331/322: 2SC641K
TR365/366: 2SC2901
TR371/372: 2SC2912S
TR375/376: 2SA1210S
TR377/378: 2SA781K

For the pair of 2SC2901s, they're installed for class L, 40-120 (Is there a standard for these or are they just part specific?/manufacturer's whim?), and it looks like the AliExpress you sent is for L anyways.

The 2SA2912Ses were replaced with KSA1381Es, as mentioned. They might be suitable for just buying a new pair and sticking those in. I'm still considering options for the other transistors, but I'm not pulling the trigger on any order until I'm sure what I'm buying from Mouser and what I'm buying from AliExpress.

[tautech]

To be honest (TBH) when the original parts ARE available I've always used them, not so much NOS but newly manufactured and only resorted to substitution when they have been unobtainium.

There are many parameters to consider when substituting transistors and when a match looks close, the thing that normally seals it for me is the datasheets listed normal use, be it a switch, GP amp, defection amp and so on, but especially when substitution involves CRT plate defection output transistors.
In every case of scopes where I've repaired output stages the OEM tansistors have been specified in the datasheet as a video deflection amp of words to that effect.

Now if you look at the fault finding oscillographs on scope schematics for plate outputs you will see quite large voltage swings and for this reason alone it is wise to stick with devices for which this is their intended use.
My favourite transistor for this use has been BF259, but alas they won't suit your needs, I've checked. 

Good luck.
 

[Kleinstein]

It's likely ok to use either the parts from the original plan or a direkt relacement of the parts installed. The installed parts may be more modern and easier to get. They may also be upgrades to withstand heat a little better.

With the transistors the DC current gain is usually not that important. The speed and base - collector capacitance can be the more important parameters.

The 2N3904 as a relacement for the 2SC641K seems acceptable.