Heathkit IO-12 recap and a question

[Rickenbackerman]

Hey guys,

I swear I posted this last night and today it's gone?  Moderators - did someone delete it?  Or did I not actually post it?

Been dealing with "scope wars" lately.  Had this old Heathkit IO-12 since college (I'm 45 now), it finally took a crap about 5 years ago so I bought a Tek 2235 off the site and shelved the Heath.  2235 took a crap last year and I spent a time recapping it and trying to bring it back to life to no avail.  There's a three page thread on that pig here somewhere - it's going back on the site for parts, I'm done with switch mode power supplies in "more complicated than I need" scopes is the short story.

So, back to the IO.  I figure it's all I need, and simple.  I just want to look at waves to see where tube guitar amps clip, I don't need stupid bandwidth and I don't even need it to be calibrated or even close to it - I just use an RMS DVM across the amp output to measure wattage at clipping.  When I shelved it, the issue it has was that the Y-axis would go spastic every minute or so and the screen would turn to noise - I'd flip the sync switch back and forth to get it to stabilize and then it would do it again in another minute.

So I just recapped it.  The big multisection can and both of the 1600V guys on the bottom were leaking goo everywhere.  I swapped all of the electrolytics and all of the film caps but I didn't check any resistor values yet.  Fired it up and I get a trace (yay!), and a sine with a sig gen hooked up.  The issue is the trace is only about an inch wide with the horizontal gain maxed.  I have a BK747B so I tested V6 and V7 and they're fine.   DeOxited all of the controls (they needed that badly!) and they all seem to do their job.   Any ideas? 

[WattsThat]

At the top of my priority list would be soldering. It probably has something to do with the fact that my first job out of high school was a bench tech at the local Heathkit store. Soldering, soldering, soldering. It’s all about the soldering.

Then be on the lookout for resistors that have drifted upward and well out of tolerance. It’s common with the carbon composition parts used back then. It seems to be more common with higher initial value parts from say 100K and up. Any one and two watters get checked first. But, they all are suspect for upward drift.

The film caps would have been at the bottom of my list.

PS: site was down for a while earlier today. Maybe the DB got rolled back and posts were lost.

[bob91343]

Recheck your work.  Measure the power supply voltages.

[james_s]

Hey guys,

I swear I posted this last night and today it's gone?  Moderators - did someone delete it?  Or did I not actually post it?

I noticed several of my replies to various threads vanished too and wondered the same thing, I don't believe there was anything offensive or forbidden in them so I suspect technical problems.

[floobydust]

I saw the post last night too, and it vanished today along with a thread about post counts dropping 1-5%.

+1 Check the power supply voltages, shorted caps usually take out the resistor. Is this the schematic to use?
Is your H pos'n in the middle or do you need the control cranked over, which indicates a lop-sided drive to the H-defl plates. How's the focus?
If you suspect a tube, you can also do swaps i.e. V5, V6, V7 are all 12AU7 I think.

[Quarlo Klobrigney]

Check the resistors, especially the wrinkled 1/2 resistors. They look rough on the outside compared to the standard AB resistors. They are notorious for being way off. I'm quite sure Heath used any surplus they could buy back in the day.

As far as posts disappearing, you are correct. I had 2 unread replies that vanished although they were in my email notifications.

[Rickenbackerman]

+1 Check the power supply voltages, shorted caps usually take out the resistor. Is this the schematic to use?
Is your H pos'n in the middle or do you need the control cranked over, which indicates a lop-sided drive to the H-defl plates. How's the focus?

I'm actually using an O-12 skit, the one you posted looks closer to the right one except my scope has a 12BH7 installed, not a 12BY7.  Speaking of the 12BH7, I tested ALL of the tubes just now and they all test good except for the 12BH7 which is weak.  I might have one in the tube stash...

H position is the middle, focus is perfect.  Every control seems to do what it's supposed to do.  I can get it to give me a really decent looking sine wave with my UNcalibrated IG-18 sig gen (I need the scope to work so I can calibrate it!), it's just that it's only an inch wide!  Couple of pics...

[Rickenbackerman]

Cap list in case anybody's interested:

Electrolytics:
Can G - four section - 40/450 20/450 20/450 50/300 - replaced with a CE manufacturing 80/525 40/525 40/525 20/525
Can F - three section - 20/20/20/250 - replaced with a CE manufacturing 20/20/20/450 - swapping this one out was tricky, it's PCB mounted and the tabs on the factory can are in a triangle and on the CE can they're in a | _ | shape, so it's kind of like stuffing a square peg in a round hole!
2) 40/150 axial - replaced with 47/160
1) 100/50 axial - replaced with 100/160

Film caps:
2) 0.1/1600 replaced with same
3) 0.25/400 replaced with 0.33/630
6) 0.1/200 replaced with 0.1/630
1) 0.05/200 replaced with 0.068/630
1) 0.01/400 replaced with 0.01/630
1) 0.05/400 replaced with 0.068/630
2) 0.1/600 replaced with 0.1/630
2) 0.2/200 didn't bother replacing these since they go to the horizontal presets which I never use and they look pretty hard to get at too!

I got everything from AES with the exception of the two 0.1/1600 caps which I had to get from mouser. Film caps are all the generic yellow polyester tubular guys except for the the two mouser ones.

[bob91343]

If the components associated with V7 are good and the supply voltage is proper, there must be insufficient input signal to the horizontal amplifier.  If you have another scope, measure the CRT drive to the horizontal plates.

[Rickenbackerman]

Unfortunately I don't have another scope...

I did some more poking around.  Checked the obvious stuff first like double checked my work (solder joints, make sure the 'lytics are pointing the right way, etc), everything looks good.

I rolled V5/6/7 around a couple times with no change.

So I poked around V7 a bit.  The two 100 ohm grid feeders - one has drifted to 110K, the other 130K.  The 12K 2-watter from pins 8/3 to ground is 13K.  The big 33K 3-watters on the plates are 35K.

Voltages on V7 (and what's on the skit I have in parentheses):
Pin 6 - 254 (260)
Pin 1 - 267 (260)
Pin 7 - 38 (43)
Pin 2 - 74 (74)

Those look pretty good, no?  I will poke around V6 next...

[Rickenbackerman]

Measured all of the resistors on V6, they all looked good. 

V6 Voltages (with skit value in parentheses again):
Pin 1 - 128 (120)
Pin 2 - 13 (14)
Pin 3 - 64 (62)
Pin 6 - 73 (68)
Pin 8 - 2.2 (2.4)

It's funny how this thing never gave me any horizontal gain issues when I shelved it, it was the vertical amp giving me problems.  Now the latter is fine and former is broken!   

[bob91343]

The horiz amplifier seems to be okay so it's apparently not got enough input signal.

[Rickenbackerman]

So that begs the question then, where does the input signal come from?  i.e. where do I look next?  Working on tube guitar amps I'm good at.  Tube scopes, apparently not! 

[Quarlo Klobrigney]

Which input signal? The vertical amplifier or the horizontal sweep?
There are 2 more. The Z axis and the sync.

[james_s]

The problem is with the horizontal sweep, so I think it's obvious that's the signal he's talking about.

[floobydust]

Where is the SYNC SELECTOR set to, does it have any effect?

Do the Heathkit test - switch HOR/FREQ over to EXT. INPUT and connect 1Vp-p OUT to the HORIZ INPUT banana jack there (keep intensity low). This bypasses the sweep oscillator and would indicate whether the H-amp V6 and H-defl V7 are working- although those stage's voltage readings look fine. Max. HOR. GAIN is supposed to give 1-1/4" wide trace.

I would measure sweep osc. V4 and V5 voltages.

[Rickenbackerman]

flooby - you read my mind - I was just looking at the skit thinking "I need to look at V4 and V5 next"...

If that all looks good I will do the Heathkit test...

Oh and the Sync Selector is set to +INT - I flip it to EXT the trace kind of "jumps" vertically and then settles back down.

[Quarlo Klobrigney]

I never assume.
Is there 370V at the junction of the 2 33K 1W resistors feeding the plates of V7?
Are the 2 33K resistors good?
So on and so on...

[Rickenbackerman]

V6 and V7 pin voltages posted above...

So...  I was looking at the schematic earlier today and thought I MUST have messed something up with the three-section 20/20/20 can so either V4, V5 or both weren't getting juice.  So I just poked at some more pins...

V4 Voltages (with skit value in parentheses again):
Pin 2 - 19 (14)
Pin 5 - 0.4 (1.4)
Pin 6 - -0.5 (0)
Pin 7 - 2.0 (2.2)

V5 Voltages (with skit value in parentheses again):
Pin 1 - 100 (110)
Pin 2 - 88 (95)
Pin 3 - 91 (95)
Pin 6 - 62 (52)
Pin 7 - -1.5 (-0.5)

Looks like I hooked that cap up correctly and those look pretty good too, right?  So what gives?  Try the Heathkit "adjustments" test next?

[Quarlo Klobrigney]

What are V7's cathode (s) at? I can't seem to read the schematic that was posted, but it also looks like the individual anodes should be at 230V not 260V indicating lack of plate current to pull them lower (drive) or their cathodes are higher off of ground than they should be (off value 12K 2W). Also what is point AA at in voltage?
Here's a later schematic of the H sweep area with reference numbers and a bit clearer.

[Quarlo Klobrigney]

Insomnia a wonderful thing I had time to separate just the schematic:

[floobydust]

V6 and V7 pin voltages posted above...

You're missing voltage measurements    there should be 3 per triode, 6 per tube. I phoned the Psychic Alliance and they want $4.99 a minute to do crystal ball readings.
I thought the sweep oscillator's amplitude is reasonable because blanking is working.  But the trace seems squashed on the left so the ramp linearity might be poor.

[Quarlo Klobrigney]

I was going to say that but I didn't want to be the probe nazi

[Rickenbackerman]

GUYS!  I know I missed a couple, my fault!  But we're all looking at different skits and there's not a voltage reading on every pin on every skit!  I'm not that much of a nutball, I swear!   

Hang on, lemme check V7 cathodes...  I "think" those are the only ones I missed according to the skit I'm looking at (O-12).  I checked the 12K  2watter and read 13K but let me check voltages quick...

V7 Pin 8/3 - 86V

"Psychic Alliance" and "Probe Nazi" - I'm cracking up!   

flooby when I had my sig gen plugged in and I was playing around with the settings on both the scope and the gen I could definitely see signs of the trace being MASSIVELY squashed on the left at certain settings.  Is that a clue?

Oh and THANKS for the help here guys!  MUCH appreciated!  I will not give up, just yet...

[floobydust]

I'm using the IO-12 schematic, same as IO-12U and looks similar to the predecessor O-12? Voltages are listed for every pin.

Usually you start a repair by first measuring all the power supply voltages. V10 is a master shunt regulator so CC needs to be 135V and not far off, EE GG for the sweep osc. and AA FF for the H-amp.

[Rickenbackerman]

I know but the goo leaking out everywhere made me want to just shotgun all the caps...

Does this mean anything to you?  I recreated the "squashed on the left" look.

edit:  I printed out the skit you posted earlier, maybe I'm just using the wrong darn one.  I thought they were all "pretty much the same" and I just realized the one I'm looking at doesn't even have a "CC" supply.  I'm gassed, maybe back at it tomorrow night...

[Quarlo Klobrigney]

Still waiting on those voltages. Come back when you're charged back up.

[Rickenbackerman]

Okay fellers, I got some fresh eyes on this tonight...

I printed out Quarlo's skit (thanks man!) since it shows a 12BH7 as V3 which is what's in my scope.  flooby yours shows a 12BY7 instead...  I've had this thing for ages so I'm almost sure nobody's ever been inside since it was built.  Quarlo's skit is quite a bit different than the 0-12 one I was using before and has a lot more pin voltage data on it.  Sorry for my "psychic hotline" moment the other night!

Anyway, VOLTAGES!  Get your fresh hot voltages here!

Measured first, then the skit value.  My line voltage here is a pretty stable 120V BTW...

Power supply:
AA 377/370
BB 347/340
CC 140/135
DD 110/100
EE 120/115
FF 125/120
GG 110/105

Toobz:

V4:
1  93/90
2  19/15
5  0.2/1
6  -0.7/1
7  1.8/2

V5:
1  106/105
2  92/90
3  98/90
6  77/45 (!)
7  -2.4/-0.5
8  0/0

V6:
1  125/120
2  12.5/12
3  62/56
6  72/70
7  0/0
8  2.1/2.2

V7:
1  241/230
2  79/75
3  86/80
6  279/230 (!)
7  38/35
8  86/80

Some have changed since I took measurements last night, could have been psychic nazi probe operator error though.  The two I put exclamation points after jump out to me.

Guess I should check R144 feeding pin 6 of V5 first.  Then, not sure why the voltages on pin 1 and pin 6 on V7 are so different when I already checked both of those 33K 1 watters and they both read 35K?

I'll be on here for another hour or so then leaving town tomorrow for about a week...

[Quarlo Klobrigney]

Unless you have access on the road, everything on V4-V7 looks good except pin 6 of V5.
That wouldn't effect the horizontal sweep width.
It does note that C124 may be leaking as the cathode of the CRT is at a negative potential.
Did you change the 1KV caps, and with what?

[Paul Moir]

I have my IO-12 on the bench with the cover off at the moment.  Let me know if you want anything measured.

(I'll turn notifications for this thread on.)

I had replaced the GG ballast resistor (1.5K) because it had drifted.  If I remember right, I was getting a narrow trace as well.  But that may be a false memory.  It's running 106V right now, and I think your GG looks good so probably not that. 
My V5 pin 6 is at 50.5V.  But that's the blanker so that seems unlikely.  C124 is the big original ceramic hiding behind V10 in Quarlo's picture and seems like an unlikely candidate for being leaky.  It seems like something funny is going on with V5.

The only other ones I found that had drifted too badly were R182 on the V10 grid (bleed resistor) and the power resistors around V3 (2.7K x 2, 1.2K)
V7 1 & 6 have an awful lot to do with the exact setting of your horizontal position pot.  If I adjust mine so the dot is about halfway between the edge and the centre of the screen, I get 279V on one and 241V on the other. 

[Quarlo Klobrigney]

What picture?

[Paul Moir]

Sorry, Rickenbacherman's.  The picture with V10 in the centre...

[floobydust]

I think we are as far as we can go with only DCV measurements 
Problem could only be a low value coupling capacitor e.g. C126 0.25uF 400V, C127 40uF 150V, C128 0.1uF 200V, C230 0.1uF 200V that you need to look at the signals with a scope, or test each part.
Paul Moir can you can look (with a scope) at the output of the sweep oscillator, for amplitude so we can know if it's weak here or not.  Maybe V5A cathode or V6B plate. You might be able to measure it with a RMS multimeter (using an extra coupling cap to get rid of DC) like at C126 or V6A grid.

There's a chance there is nothing wrong with the H-amp and H-deflection or sweep oscillator. Instead, it could be the blanking malfunctioning. The CRT-beam being enabled for too short a portion of the sweep ramp. It's sweeping full H width but beam current only on for the inch, so it looks like a squished display, only a snippet is lit. Just a possibility.
V5B the grid bias is formed by space-charge and a very high impedance node with 22MEG resistor R142, so measuring with a 10MEG resistance multimeter there, the value will be thrown off. I think V5B is pulsed off? only for the retrace (blanked) portion of the inverted sawtooth it sees. So I would also check:
C116 100pF, C124 0.02uF 1,600V, R142 22MEG.

There might be a way to disable blanking, lift a leg on HV cap C124 or R144 etc and see what the screen looks like.

[Quarlo Klobrigney]

I think I already mentioned the capacitor C124 from the CRT in the blanking circuit. We will have to wait until he returns for the answer.

[Rickenbackerman]

Posting from the beach trying to digest all that you guys said, and thanks for the help!

Quarlo you asked about the "two 1kV caps" - C34/C37?  I got a couple 0.1/1600 film caps from mouser.  Not sure what brand but I can find out.  The original ones were literally leaking goo out everywhere.

I'll be home tomorrow night but I might not get back to it for another day or two.  Sounds like C124 is first to test.  I don't have an ESR meter but I have a decent DVM (which can measure capacitance) and a heathkit cap checker which I redid and works well.

Funny how I shelved the thing due to vertical issues and now the horizontal amp is the issue. The "squashed on the left" trace pic isn't a dead giveaway to anybody?  I thought it would be.

[floobydust]

We don't know what signal you are putting into the scope and what the sweep speed is. Is it supposed to be displaying a single sine-wave cycle?
If you measure the sweep oscillator frequency (e.g. with DMM) and set it to say 1kHz and the vertical has a 1kHz sine-wave coming in, that would tell more about a non-linearity or squished sweep ramp. Or maybe try it with the 1Vpp reference and line trigger (all mains frequency).

[Rickenbackerman]

Sorry flooby, I should have said that it was just a simple sine wave...  I forget what frequency...  Anyway, got home tonight and played around for a little bit.

FYI it should be noted that ALL of the electrolytics and ALL of the black tubular film caps have been swapped out.  The only original caps remaining are the ceramics.  In case that wasn't clear.  Not that those don't ever go bad...

I measured (in-circuit) a few things you guys noted I should check:
C124 - 0.028uF but like I said I don't have an ESR meter so I dunno if it's leaky or not.
R142 - 33M - a little high versus 22M, no?  (my DVM has a 2000M setting)
C116 - seems like right around 100pF but my meter was jumping around on that one.

I hooked up the IV P-P to the vertical input and played around with it - I've never even tried that before.  I could NOT get a decent looking sine wave at any setting, and I noticed that playing with the frequency vernier wants to make the trace either jump off toward the left or disappear completely.  It definitely moves it around horizontally though and it affects the intensity as well.  And I still get the "squashed on the left" look at certain settings.

When you guys talk about "blanking" and "sweep oscillator", I'm in over my head.  If you tell me exactly what to measure to try next, though, I'm all ears.

edit - Quarlo - for C34/C37 I used these:
https://www.mouser.com/ProductDetail/75-MKP1839410134HQ

[Quarlo Klobrigney]

Expensive little buggers! "For C34/C37 I used these:"
For C124 - 0.028 µF (or 28 nF) take it out of circuit. Discharge it and use your DMM on the highest resistance range to test it. It should charge, then go infinite. Flip it around a few times and try it again. If it doesn't go infinite or has a resistive reading that stays, it's leaky.

I use the 630V yellow polyester or whatever they are made of today. 2 in series is plenty to replace those 1.2KV caps. Of course, you have to double the capacity in series .056 µF (56 nF). At 50 cents apiece they are affordable. https://images-na.ssl-images-amazon.com/images/I/41gxqCXq3UL._AC_SX425_.jpg

"Dioda Gona Wildaaa" has a video on the sweep 07:33 and and blanking at 11:00. Starting here, https://youtu.be/TPX9BMDK4w8?t=453 he explains the basic theory with his cat assistant.

That was the quickest version I could come up with before bedtime.

For the out of range DMM resistor, you could always measure it in parallel with a known resistance with an online parallel resistor calculator. Use the known resistor that fits in the range of your DMM (2.2 MΩ for example) for the first value and substitute the printed value of the unmeasurable resistor. Now calculate the parallel resistance. Compare this to the example value and unreadable resistor in parallel.
That will give you the differential tolerance.

For example should take the 2.2 MΩ measured, put it in parallel with the unknown resistor (say it is 33 MΩ) the result would be = 2.0625 MΩ proving that the unknown resistor is not equal to 22 MΩ like in the next equation.

It should, for the 2.2 MΩ measured and the 22 MΩ required in parallel = 2 MΩ.

Just run a few values that you can parallel, through the calculator to find out.

[Rickenbackerman]

I will check C124 doing exactly what you said to do.  And I will watch that video.

But I think you misunderstood - my DVM can read up to 2000 MΩ (not sure accurate it is on that setting though!).  R142 reads 33MΩ vs 22MΩ.  I'm an engineer, so I prefer to not have to do math if I can help it!   

[Paul Moir]

Sorry I'm so late in replying.  I keep forgetting my digital 'scope so we'll have to live with my old 465, so unless you really want pictures I'll do this verbally.

On Cathode of V5A we're seeing a nice sawtooth like you would expect; it gets a little rounded on the lowest range (10 CPS) but it has a nice sharp rise.  Vpp is about 8V with the vernier at 10% and goes to 14V with the vernier at 90%
V6A plate is practically DC.  1Vpp uuuu shaped wave at the lowest sweeps and diminishing as the sweep rate increases.

I've been trying to work out a system for measuring this amplitude but haven't been successful yet.  I couldn't get anywhere with AC even with a good blocking capacitance on a TRMS meter, I'm trying to get a good rectification signal so we can use (a reliable) DCV with a pretty simple setup.  I'll keep working on it.

Can I suggest a simple test?  If Rickenbackerman sets his Horizontal Selector to "Line SW" (60Hz Sine wave) he should get a line that can be expanded past the extents of the CRT.  This takes out V4, V5, and their associated circuitry as culprits. 

Quarlo:  are you thinking the blanker is overblanking the sweep?  If so I tried lifting a leg on C124 disables the blank just like you would expect without any obvious side effects.

[Rickenbackerman]

Paul, no worries - I was at the beach for a week!

Horizontal Selector to "Line SW" - check - with no input I get a single dot with the horizontal gain on all settings.  I hooked up the 1Vpp and I get a vertical line with plenty of vertical gain.  And here's a weird one - if I move the horizontal position pot back and forth real fast I can get a somewhat decent looking sine wave the full screen width, although still smashed on the left.  It only does that when I move it, if I stop it goes back to the vertical line...

Lift a leg on C124?

[Quarlo Klobrigney]

I was tired when I read the 33M - DMM part. In that case, it's out of tolerance. Change it to 22 MΩ.
Lifting a leg is good to check ones balance preferably the "Crane" method. It is also good to check components.

[Paul Moir]

OK, we may have the tiger by the tail now.  To validate my check could you please confirm with the Horizontal Selector set to "Line SW" that you measure 1.4-2VAC (meter set to AC volts) on the indicated side of C126?  This is the long capacitor you replaced next to V6, and the side we want to measure is the side near the front panel.
If we get that, it looks like something around V6 is bad.  Since the horizontal position control seems to work fine, and everything after V6 is affected by that control, all that is probably OK.
I'm glad to hear your visit to the beach was good.  Ours are still cold and miserable, but they might warm up by the end of July!

[Rickenbackerman]

Paul, I get 1.22VAC at that location with the horizontal selector set to "line SW".

[Paul Moir]

OK, looks like we're making progress then.  I get 1.44VAC here so close enough; something downstream is amiss.

Could you please check the two points A & B on volts AC, with the same setup as before but with the horizontal gain set about halfway.  I get 1.2V at point A and half that at point B.  They're both on the pot so pretty easy to get to from there if it's lying on its side.  That will confirm the V6A amplifier is working.

[Rickenbackerman]

Paul, while you were typing that I was testing every resistor associated with V6 and V7 - everything looks good.

I lifted a leg of C124 and tested it with my DVM as Quarlo suggested - even with the meter set to 2000Meg, the cap charges up then the resistance goes infinite.  So that's not the problem.

To clarify an earlier measurement - R142 - measures 33Meg, but with the leads on my DVM connected directly on the 2000Meg setting I read 9Meg...  so really R142 is about 24Meg.  I really need a new meter (and I've been through a few over the years!).

TEST 2:
Point A - 0.012VAC
Point B - 0.012VAC

It looks like we're on to something.  To check my meter I redid the first test you suggested and now I see about 1.8VAC vs 1.2VAC.  Not sure why it's moving around.  I've looked over every solder joint on this thing at least a dozen times and I don't see anything so I'm thinking I may have connected something incorrectly...

[Rickenbackerman]

Closeup of V6 socket and parts...

[Paul Moir]

Your recapping is a lot neater than mine.  I've got a two 22uF Nichicon radials in parallel there   I also have a very, very fine crack on a trace on V7 which is giving me an intermittent horizontal. 
I double checked those voltages with a 'scope.  There is no DC there and it's a normal sine wave, so if you're reading point A correctly, there is a problem with that amplifier.
Regarding the original voltage; it might be affected by the "Phase" control.  Mine is stuck and I've been trying to free it up.  But if you're seeing anything like 1V at that point you should have a nice wide horizontal trace.  So don't worry too much about that exact voltage.

[Rickenbackerman]

Ha!  (on the neat recapping - I try)

If you carefully scrape the busted trace with the point of an X-acto knife to scuff it up, lay a hunk of small solid core across the gap and solder it on that's the best way I've found to fix a busted trace.

Bedtime here.  Agreed, something is up with V6A.  Tomorrow I'll reflow all joints associated with it and see if that gets me anywhere.  The tube tests fine on my BK747B.

[Paul Moir]

Same here.

And you can play swaptronics with V5 if you want because it's not doing anything with the Horizontal Selector in "Line SW'.

I'll fix that trace just as soon as I remember to bring my soldering iron home.  I have a rather nasty Radio Shack firestick here but I would rather not use it!

[Rickenbackerman]

Swapped V6 and V5 before crashing just for grins.  Same dot on Line SW, same vertical line with 1Vpp hooked up.

I also have a few "nasty firesticks"!  Ha!  My dad has like a 200watter he uses to solder up the ends of model airplane fuel tanks he makes out of tin.  Thing is ancient and the tip is about 5/8" wide! 

[Rickenbackerman]

OK, I reflowed all joints associated with V6 and did some more testing and I am utterly stumped.  V6A simply refuses to amplify.  I get a good 1.8VAC at the grid (pin 2) on "Line SW" and the plate voltage is good (125VDC).  Something is up at the cathode (pin 3).  I lifted a leg of C127 (even though it's new), and tested it for capacitance and leakage per the DVM resistance check and it's fine.  On pin 3 with C127 disconnected I get like 5VAC on "Line SW".  I think the horizontal gain pot itself might be toast.  I will NOT give up.

[james_s]

Well pots do fail, they're easy to test though. You might also try cleaning the tube sockets and tube pins, I've run into situations with old radios where dirty tube sockets/pins caused a malfunction.

[floobydust]

V6a is a cathode-follower, so it will naturally have no voltage gain, that configuration has a voltage gain of 1. It's buffering the output of the sweep oscillator with bootstrapping to increase the input impedance (R148-R150) so it doesn't load anything down. The grid and cathode (AC) voltages I would expect to be the same.
So if you have 1.8VAC coming in at the grid, the most you'll get is 1.8VAC at the Hor. Gain pot when it's cranked to max.
V6b is the (voltage) gain stage. I would look around there.

edit: multimeter loading is likely making V6a grid (signal) voltage measurement appear lower than it is. So V6a cathode can have more amplitude than the grid, as an illusion.

[Rickenbackerman]

While you guys were posting I was poking and probing...   Thinking maybe I had a bad ground on the horizontal gain pot or bad pot itself or dirty tube pins or loose sockets, etc...

Stared at the skit and think I found a better way to test all the "junk" on pin 3 of V6:

Pin 3 to chassis ground (from the tube side of pin 3 to eliminate any bad solder joints) - 56K - which makes sense since that 47K resistor has drifted up to 54K.  Don't think that would affect much?

Pin 7 to chassis ground (tube side again) - Horizontal gain pot min - 0K.  Horizontal gain pot max - 12K.

Cleaned V6 pins (again) and retentioned sockets 1-3.  Still 1.8VAC at the grid on "Line SW" and basically nothing (0.01VAC) at the cathode with the gain maxed. 

THANKS fellas for the help!  I STILL will not give up.

[Quarlo Klobrigney]

R254 → 10MΩ, 257 → 100Ω, 259 100Ω?

Pin 1 & 6 going to the proper CRT pin without corrosion?

Disconnect C128 off of pin 6 of the V6 plate. Connect a 1KHz audio source thru C128 without anything that would cause a hot chassis. The Horizontal should be too wide depending on the Horizontal gain pot.

[Rickenbackerman]

R254, R257, R259, all good.  Squeezed those CRT connections with no affect.  Did find a sketchy V7 pin 6 to CTR wire rubbing on the chassis there and fixed that too, no affect.

I'm a scared of screwing around with 1300VDC after numerous beers so I tried something else - I should see increasing AC voltage on pins 2 and 7 of V7 as I turn up the Horizontal gain with the Horizontal selector set to "Line SW", no?  If so, I see nothing. 

So, it's either V6A or V6B. 




Or something else.

[james_s]

You shouldn't really be doing anything at all inside that scope "after numerous beers". Even the "low" voltages in vacuum tube devices can knock you dead, or at the very least cause you to jerk back and slice open your hand on something.

[Rickenbackerman]

James, WELL aware here.  I've got about 1800 all-tube watts worth of vintage bass amplification here in my basement, and have done work on all of it.  Plus work on dozens of tube guitar amps for friends.  Just got done redoing a 60's Fisher 400 stereo receiver and a couple of Roberts 6W single ended EL84 RTR monoblock amps for a buddy.  I know my limits, and I'm done for the night. 

The only time I've ever really been nailed is when calibrating my BK 7474B tube tester in my early 20's (I'm 45 now).  I ate about 350VDC in one arm and out the other, right through the ticker.  Low amps, but it still knocked me on my ass for a good 15 minutes and taught me a lesson I won't ever forget.

[Rickenbackerman]

Only had about 2 hours to "play around" tonight, the wife's got me on the honey-do list this weekend...

Reflowed a few somewhat sketchy looking joints on V7, rechecked some more components, tested a few voltages.

Thought maybe I hosed something up and the the power supply for V6/V7 had AC?  Checked that, nope.  Did notice the voltage on pin 2 of V7 has mysteriously jumped up to about 105VDC though.  Setting dependent maybe? 

So I tried Quarlo's test and yanked the V6 pin 6 wire and fed C128 to V7 with 1kHz next.  The horizontal gain pot has no affect since the signal is being injected after that point.  I do get a nice wide trace, and it is big time affected by the horizontal position pot.  The trace wants to be toward the left (and off the screen) more than it does center or right depending on the position of that pot.   And I need LOTS of voltage to feed the trace.  My sig gen is not calibrated, but I had to jack it up to the 10V setting to get a full-width trace.

Something is definitely up with V6/V7...

[Paul Moir]

I can confirm you need a good strong signal at C128 to get a wide trace; I'm using the 60Hz "Line SW" setting and I get 24Vpp needed to get a trace that reaches the limits of the CRT.

Let me know if there is something you would like checked.

Pin 2 of V7 is going to be dependent on the exact position of the Horizontal Position pot.  It's not super twitchy but that's only a 10-15 degree turn of  the pot.

Have you tried shorting out the horizontal gain pot?  Connect pins 1 & 2 together (the two lugs closest the edge of the chassis) with the control turned fully clockwise?  I would like to rule out that suspected pot.

[Quarlo Klobrigney]

So I take it you have the same working scope to take measurements? That's a good thing. At least you can break the bad circuit in half at least, and half again until as they say in France viola...

[Paul Moir]

Yes, have you not read any of my replies?  The problem is at V6A.

[Rickenbackerman]

Paul, thanks for confirming you DO need a nice strong signal to drive V7, I was unsure about that.  And thanks for having yours opened up and taking measurements for me!  A working test mule is the best mule!

In post #54 I said about V6 - "Pin 7 (tube side) to chassis ground - Horizontal gain pot min - 0K.  Horizontal gain pot max - 12K."

I would think that would rule out the pot, but I'm desperate so I went and tried your "shorting out the pot" test out just now anyway.  First I had to reconnect the gray wire coming from V6 pin 6 that I disconnected last night and had a little bit of an AHA! moment, as I had a bit of a tough time getting a good joint (dirty wire I reckon) and though maybe that joint was bad in the first place.

Powered on, same result.  Powered off again.  Pins 1 and 3 jumped, powered on again, same result.  Reflowed all joints on the horizontal gain and horizontal position pots and checked the connections on the gray, purple and brown wires (gray from V6 pin 6, purple and brown to coming off the horizontal position pot) from end to end and everything looks good there too.

V6 still simply refuses to conduct.  I'm out of ideas short of finding a nice tall building and gently heaving this thing off the top of it.

[floobydust]

Clearly you've offended the oscilloscope gods. Usually that tells me I need to change my technique lol.
If things aren't making sense then something got missed or a previous measurement is wrong.

In your pictures, what is that film cap 683K at V6 "Phase" going to the potentiometer and (-) of 47uF C127? There's no cap there on the schematic and none are that small 0.068uF?

edit: I found a pic of another IO-12 and it has a 0.05uF at V6 as well.
*Can you check your replacement capacitors are going to the right terminals on the two pots R52, R72? Looks different...

[Rickenbackerman]

CLEARLY I've offended the 'scope gods!

Hmm....  ok, on the skit, over by the PT, two lead go to "to all tube filaments", then below that two leads go to FIL (A) and FIL (B).  I just looked over the heater wiring and it looks nothing like the skit.  The heater wires go right to the heaters and that 0.068uF cap goes from pin 9 of V6 to the phase pot.  Is that cap C41?  No idea where R73 and C42 are or if they even are.  On the skit I don't see where FIL (A) and FIL (B) go...  you are getting tired... and sleepy....   

[Rickenbackerman]

"*Can you check your replacement capacitors are going to the right terminals on the two pots R52, R72? Looks different..."

Check and check.  R52 left lug, R72 right lug as looking at your pic.

[james_s]

Are we absolutely certain that the new capacitors are all the correct value? It's easy enough to select the wrong order of magnitude, especially when converting from vintage to modern nomenclature. It has happened to me more than once.

[floobydust]

Why is there no wire connected to H-OUT ?

At this point, I would re-check all the wiring and recap capacitors, there's a mistake somewhere.

[Rickenbackerman]

At this point I am certain of absolutely nothing.

flooby, there is a wire at H-out - it's a gray wire that goes to C128 and comes up from the bottom.  You just can't see it from the top.

Later tonight I'll recheck AC and DC voltages around V6.

[floobydust]

Once I found a component lead clipping floating around underneath a tube socket, you have to leave literally no stone unturned. It could be a problem from the original build, the solid hookup wires do break if flexed too much.
One approach that works for me is to trace through the schematic by using an ohmeter to verify each connection. Just red mark with a pen each connection. Old flux is not good for high voltages and high impedances, so don't leave chunks on the pcb bottom.

During re-cap projects I sometimes forget where I was at, distracted phone rings etc. and the wrong cap or wire happens. This is without any beer 

[Rickenbackerman]

Definitely found lead clippings floating around and lodged into places where they shouldn't be before!

The horizontal amp worked perfectly before, so unless I disturbed something we can be sure the original build was A-ok. 

Check on the old flux, too.  I use a tiny flat blade screwdriver and scrape it off between the traces and socket pins on the bottoms of the boards.  Did that a couple weeks ago all around V6/V7 already.

Paul if you get a chance can you take a close look at the pic I took that flooby reposted above and see if anything looks out of place if you haven't already?  The shot in post #46 is better, actually, but If you need any better shots just let me know.

[Quarlo Klobrigney]

1 question, in all these posts, have you cleaned all the pots with something like D5?
Also, check all pots in question with one lead of the ohmmeter on the wiper and the other to each extent. 2 tests.
I have seen pots go bad on one end of the rotation many times, believe it or not. The well traveled road and such, it wears out. Also the metal covers on those pots look like there is some corrosion.

[Rickenbackerman]

I never mentioned that I cleaned the pots?  Jeez...  Yeah, I D5'd 'em weeks ago, they all needed it badly.

Checked all connections and components around V6A for the 17th time.  Reflowed a few more joints and scraped a little bit more nasty old flux off the bottom of the board between traces.  All looks good.

V6A DCV:
Pin 1 - 126V
Pin 2 - 12.7V
Pin 3 - 62.2V

V6A ACV with the horizontal selector set to "Line SW":
Pin 2 - 1.7V
Pin 3 - 0V

The horizontal gain pot is fine, I'm AM certain of that.  I tested it about three different ways by now.

edit: I also retentioned V6 socket pins 1-3 and hit all of the tube pins with a spritz of G5 gold (the hi-temp stuff).

[Quarlo Klobrigney]

OK. If you wish to answer, what part of the US are you in?

[Rickenbackerman]

I'm in Maryland, why?

[Quarlo Klobrigney]

Just wondering if you were anywhere near. Nope

[Rickenbackerman]

Uuuuh, how is Portugal anywhere near the USA?  Do I need to buy a new map?

[floobydust]

V6a the only way I can see AC at the grid but not the cathode, would be C127 47uF (-) shorted to GND at the HOR. GAIN pot pin 1. The DC voltages are bang on for V6a. You checked for AC mains ripple at FF right, so the filter cap C133 20uF must be OK?

[Paul Moir]

Rickenbackerman:  Sorry about the pot, I don't know how I missed that.  I guess that's a rather dead horse at this stage.
I've been looking very closely at the pictures you posted for some time trying to find a problem, but I haven't seen it. 

Not relevant, but that 0.05uF Phase - capacitor floodbydust is on the bottom right hand side of the schematic, beneath the pilot light.  It looks like they're just picking up the filament voltage off V6 since it is a convenient spot.  It's kind of a silly thing.

The voltages match pretty well, at least the DC.  It took a while for pin 2 to drift down and a while for pin 3 to drift up while measuring, but they match your values.
For AC I couldn't get the cheap TRMS meter I have here to read properly due to the DC offset.  I had to resort to my (non TRMS) Fluke 8120A which read 1.3VAC on pin 2 and 1.1VAC on pin 3.  I confirmed their values with a 'scope so I don't think there's any funny business here.  I can bring my Fluke 175 or my EEVBlog 121GW(!) to see how they do if you like.
Dumb question:  both filaments are glowing right?  Sorry, I'm about out of ideas. 
Finally sharing my ugly IO-12:

[Rickenbackerman]

flooby - tested FF for AC, saw none.  C133 is one third of the 20/20/20 CE can which is brand new.  Lifted a leg of C127 and tested it, it's fine.  Thought maybe I got bad one from the factory there, too, but it tests good.  And not shorted anywhere.

Paul - I LOVE your "ugly" IO-12 since yours actually works!  The DC voltages take a couple minutes to settle down here, too.  I SINCERELY appreciate it, but don't worry about testing any further, we all know what the problem is now - "something" is going on at pin 3.  I get ZERO ACV there no matter what.  I've looked at it from every angle and scratched my head for weeks.  For the hell of it I even reflowed the V6 heater joints even though the heaters there work fine...

edit: Paul - NOT a dumb question about both filaments glowing which is why I just reflowed those joints!  I wonder if I should I look closer at that, though?  I have never even looked close enough at a 12a*7 to see if you can tell if both sides are glowing?  I always just assumed if it was orange it's good to go?


edit #2: what if I desolder the wire from pin 2 of the horizontal gain pot and then recheck the ACV at V6 pin 3?  That would at least separate V6A and V6B.  I wonder if something on V6B is dragging down V6A?  Pin 7 is the V6B grid, though, so that doesn't make much sense, but it's an idea, and I'm about out of those, too.

[Paul Moir]

I was thinking something similar but rather C127 from the pot:  Floodbydust thinks something is shorting this to ground which would affect the AC but not the DC.  I know you measured this and did not find it so it would have to be a magical tin whisker or something like that.  Actually let me test that and confirm that's what happens.
Yeah, it sort of duplicates your readings.  The DC volts go up a bit (62 to 67) on pin 3 but it completely zeros out the AC component.

[Quarlo Klobrigney]

Not only do I need sleep, I need to really sleep and not sleep post.
Not only the wrong forum, but the wrong website.

Quote from: Rickenbackerman on Yesterday at 14:50:15

Uuuuh, how is Portugal anywhere near the USA?  Do I need to buy a new map?

[floobydust]

I never mentioned that I cleaned the pots?  Jeez...  Yeah, I D5'd 'em weeks ago, they all needed it badly.

Checked all connections and components around V6A for the 17th time.  Reflowed a few more joints and scraped a little bit more nasty old flux off the bottom of the board between traces.  All looks good.

V6A DCV:
Pin 1 - 126V
Pin 2 - 12.7V
Pin 3 - 62.2V

V6A ACV with the horizontal selector set to "Line SW":
Pin 2 - 1.7V
Pin 3 - 0V

The horizontal gain pot is fine, I'm AM certain of that.  I tested it about three different ways by now.

edit: I also retentioned V6 socket pins 1-3 and hit all of the tube pins with a spritz of G5 gold (the hi-temp stuff).

Your measurements still do not make sense 
You say V6a pin 3 cathode 0VAC yet you have some horizontal sweep on the CRT which says that is not the case. What are you doing wrong here? Is there oxide on the points you are measuring, your DVM has beer spilt in it or what?

[Paul Moir]

Hi Floodbydust,

The small trace collapsed to a point when we started using the "Line SW" input.  I suggested this for two reasons, one of which was to put a 60Hz sine into the circuit so we could measure the voltages with a multimeter.  I wasn't having much luck producing a repeatable measurement test for the sawtooth.  At the time I presumed that the higher frequency of the sawtooth was leaking through somehow.  I also presumed we would find the problem quickly.

I've been concerned that the "Line SW" was somehow leading us down the garden path and so have been checking a few concerns I had behind the scenes.  I don't think I have but doubt is the name of this game.

[Rickenbackerman]

I did say I had the Horizontal selector on "Line SW" when I took those measurements, and get zero trace width on "Line SW" as I rotate the Horizontal gain pot.   I only see hints of width up on the other settings (10, 100, 1000, etc) but I didn't check AC at V6 pin 3 when set on those...

My meter IS flaky, though!  Sometimes I have to rotate the switch back and forth a few times to get it to settle down.

I had an idea earlier today for another test to try - I'll report back later...

[floobydust]

It looked like there's signal coming into a stage (V6a grid 1.7VAC), and nothing out (cathode 0VAC).  Either an open connection to V6a grid (at the tube) or a heater-cathode short, but the filament winding is supposed to be center-tap grounded and could not float up to 62.2VDC ... so the problem is likely voodoo.

For the sweep source, LINE SW is high impedance and lower amplitude than the sweep osc. due to the PHASE potentiometer, capacitive divider C41 etc. so FIL(B) might be really a too small signal and it's just hum pickup, so the 1.7VAC is a red herring. Many multimeters can't handle a large DC offset when measuring AC. Clipping at the A/D or just plain lie and give bogus values. Bad firmware as well. You have to know your test equipment so it doesn't lead you on a goose chase. Some multimeters can even get damaged because they auto-range down looking for AC yet DC is present, while the AC stage input cap gets charged up, which then discharges and kills the multimeter's front end when you next probe anything else. Might have to make an RC circuit prevent the drama. I think the ACV measurements are not working out as we expect.

[Paul Moir]

Hi Floodbydust,

If you look at the initial measurements I got Rickenbackerman to do on the other side of C126 and C127, it was to avoid the DC offsets.  They sort-of-suggest the measurements he made on pins 2 and 3 are possibly correct.

But the differences in impedance are something I wasn't thinking of.  I'll try loading down the stages at a couple places to see if I can get the results replicated.

[Rickenbackerman]

Here is the test I wanted to perform - some questions still as to the horizontal gain pot being the issue - I checked it three different ways and was sure it was good, then today I thought "you know what, I AM sure of nothing" and thought of a simple way to totally get rid of it as a culprit - remove it entirely from the circuit.

Disconnected C127 and the wire going to V6 pin 7 from the pot, connected them together and then add in a 10K resistor to ground to simulate the pot resistance.  This simulates the pot being in the circuit and turned up all the way without the pot actually being in the circuit.  Not a very complicated test, no, but now we can be 110% sure that pot is good, and I feel better about that now at least.  Pic attached of my temporary wiring...  V6 removed just to take the picture, obviously...

Same result - dot on Line SW.  Hints of horizontal gain up on 10, 100, 1000, etc...  Want any more voltages checked while I have it hooked up like this or should I put it back the way it was?  If you guys think the DC offset messing with my meter at pin 3 - or elsewhere - I have tons of coupling caps and jumper leads here so I can put a cap in series with the positive probe?  Am totally comfortable poking around, like I said, I got nailed once and I know my limits. 

[Rickenbackerman]

Anybody got any ideas?  I'm pretty much out of those...

[floobydust]

I would backup and try something different. Strong assumptions (if wrong) can really make a mess when troubleshooting.
I thought we were trying to see if the sweep-amp V6 was doing its job, and the multimeter seemed to be lying about things. On ACV there is DC present at most nodes, it's a sawtooth, high impedances at the grids, so a digital multimeter can give misleading results. Or the probes are pooched or something else is not working.

I would forget about the H-sweep, who needs it anyway lol and see just how the Vertical section is doing.
Is it reasonable height when the scope probe is connected to the 1Vpp ref? Although all you'll get seems to be mostly a line, it's good enough to use for checking signal amplitude as we probe the H-sweep section...
I'm saying use the scope to probe itself, inside the horizontal section. Unless you have other test equipment. A common aspect between the two sections (V, H) is the SYNC portion, so I'm not sure where you want that switch. But poke around and try see (again) if V6 is amplifying anything. You just need ballpark ACV measurements as we are looking for where all the sweep signal gets lost - if that's the problem.
When the HV caps C36, C37 0.1uF 1,600V died, did they short and damage R71 470k 1W or the 1V2? Just checking because low HV can give weak V and H deflection. Don't have a beer and try measure it.

[james_s]

I had a thought, are we sure the problem is with the H-sweep in the first place and not with the blanking? Does the trace look compressed or are the edges of it simply being cut off?

[floobydust]

C124 measured 0.028uF but did not hear back if OP ran the scope with the lifted leg (disconnect one end of the part) to see what happened with the blanking disabled post #45

[Rickenbackerman]

When the HV caps C36, C37 0.1uF 1,600V died, did they short and damage R71 470k 1W or the 1V2? Just checking because low HV can give weak V and H deflection.

C124 measured 0.028uF but did not hear back if OP ran the scope with the lifted leg (disconnect one end of the part) to see what happened with the blanking disabled post #45

Ok, I admit I must be getting frustrated and possibly leaving out critical info.  So, let me step back a few and give some more...   

BACK STORY!  When I shelved the 'scope, the horizontal amplifier worked fine.  It was the vertical amp giving me issues.  It would "freak out" every couple of minutes (screen would turn to a blur of mush (vertical amp) everywhere and if I'd flip the SYNC Selector back and forth it would stabilize...  for another minute or two.  Then freak out again.

Pulled said 'scope off the shelf (before recapping) and plugged it in.  No trace or even a hint.  Noticed the 1V2 wasn't lit, so ordered a couple used good ones and caps and then recapped the 'scope.  (C36 and C37 had puked their guts everywhere, and the cans didn't look much better).  Plugged in a "new" 1V2, was happy to see that I at least had a sign of a trace, and now four pages later on the EEVblog here we are...

I thought I tested R71 but let me test it again...  492K, so not THAT far off.  And the "new" 1V2 pegs the meter on my BK747B.

And I did run the scope with a leg of C124 disconnected, no change in horizontal deflection.  I thought I mentioned that but I looked back and realized I didn't say...

A little bit scared about using the 'scope to "probe itself" - I have Zero HV probes here, only a 1:1.  The vertical amp sure does seem to look like it's right on, now.  Connecting the 1V-PP gives me a nice controllable vertical line with the Horizontal selector on LINE, and barely a half inch wide "trace" on the other settings.

No other useful test gear here. 

Pic of scope with 1V-PP connected, ignore that the VERT input switch is on X1, I think pin 1 and 3 wires on that switch are swapped, been that way since I've owned it.  So X1 is X100 and X100 is X1.

[floobydust]

This stuff is supposed to be fun, it's an adventure repairing vintage gear and keeps your skills sharp 
There are many ways to fix equipment - keep replacing parts or troubleshoot each stage to the component level, start the end, or the middle or the beginning stage.
I was looking at the three sections responsible for the H-sweep: Osc. V4/V5, H-amp V6, H-CRT driver V7 and we need to have some idea what stage is not working. Your DCV measurements seem to show everything is reasonable.
So one approach is start in the middle, see if the sweep amplifier has good output voltage, comparable to Paul's scope.

OK, no probe but no problem to poke around in the horizontal section as long as what you are using is touch-safe. Or just power it down and connect the test lead, then power on and measure, every time.
Can you fix the V/div switch wiring with R1, R2, R3 to avoid confusion. Everyone note this scope vertical section is only AC-coupled, so DC offsets do not get in the way.

I would do a measurement at the top leg of the Horiz-Gain pot R52, we need to know how many ACV is coming out of the sweep amp buffer V6A .
When you have the HOR/FREQ selector in LINE SW position, you are not even using the sweep oscillator. It's just a sine-wave 3Vpp FIL-B coming into the horizontal amp, so completely bypassing the sweep oscillator V4, V5. This leaves only V6 and V7 stages to measure.

Using the Vertical section, can you look at:
V6A output pin 3 or at the pot at R52. Should be around 3Vpp?
V6B pin 8 plate H-amp output. Should be around 30-60Vpp?
Biggies are V7 plates pin 1, pin 8 should have lots of voltage swing there but we don't seem to have it.

With blanking disabled, C124 leg lifted I would have expected to see a retrace line, the time when the beam zips back from right to left in readiness to do the next sweep.

If you're not wanting to make measurements, then carefully retest the all parts in the section - R260, R261, C230, R151, R254 etc. without assuming any of them are good.

[Paul Moir]

I don't want to interrupt this path since I think it is a good one, so this is directed to floodbydust:  If you're worried something is broken in the LINE-SW setting, I was going to propose connecting 1Vpp to the horizontal input and using the sweep selector on EXT. INPUT.  It gives us a lower impedance source, though that shouldn't matter. 

[Rickenbackerman]

Ok I'm back after another relaxing week at the beach away from this confounded scope that's making me pull my hair out.

Just kidding! Like flooby said this stuff is supposed to be fun.  Step away if you get irritated or confused.  Rethink things.  Double check previous work.

Didn't have much time tonight to poke around but I'd like to fix the 1X/10X/100X vertical V/Div selector switch first if we're going to use the scope to probe itself so nobody get confused.  Good idea and I like it, too.

Paul can you flip yours upside down on the bench and take a looksee if you still have the cover off and it's not too much of a pain?  I think I need to swap wires 1 and 3 on the switch (counting lugs 1, 2, 3, 4 up from the bottom of my picture)?  I'm seeing conflicting things on the various skits I have printed out here, not all things are labeled, etc... and figured it'd be easier to just ask you.  And those wires are solid core, not too easy to get to, wrapped around the lugs and don't look like much fun to swap if that's not it.  Pretty sure it is, though.

[Rickenbackerman]

One last quick test before I hit the sack...

V6 Pin 2 to vertical input.  Sync selector to -INT, Horizontal selector to Line SW, I still have the horizontal gain pot bypassed so that's maxed out.  Pic below of this test.

Same settings - V6 Pin 3 to vertical input only gives me a dot!  So V6A is clearly not working.  Why???  I don't get it...  At least we've proved what Paul said back in post #62.  Although he already knew it!  Ha!

[MathWizard]

The schematic is quite similar to an old scope I have. It's transformer is fried and I never got around to getting a suitable replacement. Maybe this year when i have some extra money.

[Paul Moir]

No problem at all.  It's good to take breaks from this kind of work when you can, especially if you're not having fun.  I've got email notifications on so it's not any burden.

Unfortunately it looks like another mystery, since the wiring looks correct.  And the labels match so that's not it.  I think you're trimmers are installed the same way around, with the trimmer with the white plastic cylinder beneath the adjustment screw towards the front & bottom of the 'scope. 

And another horrendous shot from my 'scope.  If I remember right, R1 had drifted really badly.  Also if I remember right, the trimmers took some serious exercising to get working cleanly so there might be an open there.

EDIT:  Oh, I think we just have a terminology problem here.  They work identical - X1 really means divide-by-1 while X100 means divide-by-100.  Not multiply-by.  So a 1Vpp signal fed to the vertical on my 'scope is biggest on X1, smaller on X10 and tiny on X100.

I also replicated the V6 pin 2 setup and I'm getting a diagonal line, due to the functional horizontal amplifier.  I get a higher amplitude (10% on the vernier is about the same as Rickenbackerman's 30%) and no opening into a circle, except briefly during warmup.  I'll try a few things later tonight.

[Rickenbackerman]

"X1 really means divide-by-1 while X100 means divide-by-100.  Not multiply-by.  So a 1Vpp signal fed to the vertical on my 'scope is biggest on X1, smaller on X10 and tiny on X100."

JEEZ, that's not confusing at ALL!  Good to know yours and mine work the same, though.

I was wondering about the "opening into a circle" too.  When I touch the probe to V6 pin 2 with the scope powered on, the trace jumps off the screen for a second or so due to the DC present there, comes back as a vertical line and then slowly widens into the diagonal circle in the image I posted above.  Not sure if that info is relevant, but thought I'd add it here.

Paul can you probe V6 pin 3 and see what you get? 

[Paul Moir]

It's funny how when you're young, your critical thinking is underdeveloped but your naivete kind of makes up for it.  The x as a divisor rather than a multiplier got locked into my young mind as "that's just the way things are."!

Picture is V6 pin 3 with the horizontal gain turned down all the way.  Increasing it to about 30% produces a 45 degree angle line since my horizontal amplifier is working.  I also tried this by routing the 1Vpp test signal into the horizontal input and setting to sweep to EXT. INPUT.  Results were pretty much identical.

[floobydust]

So OP has no signal present at V6A pin 3, but it's present at V6A pin 2?
I'd think that is an open connection to the triode at its grid. If you're measuring at a nearby part there might be a busted connection or dud tube socket pin.
The plate and cathode must be connected to give the cathode the OK DC voltage seen.
The V6A grid circuit is really high impedance, if not it would squish/distort the sweep waveform.
I would look really hard at the 2,200 ohm grid resistor if cracked or open.

[Paul Moir]

Floodbydust,
I've been trying to replicate Rickenbackerman's results on my own scope by adjusting resistors, opening things, etc, so far without much success.  I have lots of experience with electronics but very little with tubes; this has been an education for me.  Let me know if you have any specific failures you would like to see tested.   

I tried opening the grid resistor and probing, and I get a diagonal about the same size with the horizontal maxed out, which is very sensitive to how close I place my hand to the probing wire.  The thing is, it's a diagonal rather than an open circle, which tells me that the signal getting to the horizontal circuit is *different* than the one getting to the vertical circuit.  Can it be possible that the other side of the filament is somehow leaking through?

[james_s]

"X1 really means divide-by-1 while X100 means divide-by-100.  Not multiply-by.  So a 1Vpp signal fed to the vertical on my 'scope is biggest on X1, smaller on X10 and tiny on X100."

JEEZ, that's not confusing at ALL!  Good to know yours and mine work the same, though.

I was wondering about the "opening into a circle" too.  When I touch the probe to V6 pin 2 with the scope powered on, the trace jumps off the screen for a second or so due to the DC present there, comes back as a vertical line and then slowly widens into the diagonal circle in the image I posted above.  Not sure if that info is relevant, but thought I'd add it here.

Paul can you probe V6 pin 3 and see what you get?

All scope probes are like that. If you have a little more background then it makes perfect sense and is a lot more intuitive. Many old scopes don't have a separate set of numbers on the V/div dial for different probe types, they just have the X1 value. If you connect a X10 probe you have to multiply the voltage you read on the scope by 10 to get the true voltage of the signal you are measuring. It's pretty much as the tachometer on a car, it will typically say X1000, you take the displayed reading and multiply it by 1000 to get the actual RPM. The X setting isn't multiplying the signal, it's telling you what YOU need to multiply the reading by to get the true value.

[Rickenbackerman]

James that makes PERFECT sense.  I never thought of it that way!

Just now I soldered up the horizontal gain pot back to stock (I had it totally bypassed for testing), but I left the center tap wire going to pin 7 disconnected to separate V6A from V6B.  I thought maybe something was wrong with B and dragging A down...   I'll keep it this way until I can get V6A working, and so far no luck. 

flooby - that 2.2K resistor is fine, I posted that pic I took of pin 2 which is downstream of that resistor.  There's AC voltage there but none on pin 3.

Paul - HUGE thanks for helping me out with this, again!

This "opening into a circle" thing is curious.  Is something wrong upstream of the feed to V6A? 

[floobydust]

When the Horiz. Amplifier is switched to LINE SW. as a (sweep) source, it gets mains reference AC signal from FIL. (B) which is a 180° plus delay or phase-shifted (by the PHASE potentiometer), compared to FIL. (A) which is the 1Vpp ref.
Said another way, Line SW feeds in a phase-shifted sine-wave to the Horizonal (sweep) amplifier. So, compared to the 1Vpp ref, you'd see an ellipse or circle.

It's puzzling there is no AC coming out of the cathode follower, yet the grid has signal and the triode's three DC voltages are OK. When things don't make sense, it's usually a flawed measurement or something unexpected/overlooked. You might have a bunch of bad 12AU7's with a common cooked grid, or there is some bad connection or short to the triode on the PC board, wiring/recap mistake made there etc.

[Rickenbackerman]

OK fellas, some MAJOR progress tonight!  I swapped V6 and V7 for grins and rechecked some voltages and...  no AC at V6 pin 2 all of a sudden.  Huh? is what I said!  I think my DVM was lying all along.  AC reading are sketchy with this thing, which is why I only really look at DC with it.  I need to buy a Fluke.

Traced the FIL (B) wiring to the Horizontal/Freq Selector switch and eventually found a cold joint (my fault of course) between the wire coming off pin 10 of the switch and C126? (I think?  my skit is blurry - the 2.2K R147 that feeds the V6A grid.  Reflowed that and we have horizontal gain again!  WOOO!!!

I was able to recreate the 45 degree line that Paul talked about by probing V6 pin 3 while on "Line SW".  No "opening into a circle" now, however with the probe not touching anything it tends to want to do that still.

However, not totally out of the woods just yet.  I hooked up my signal generator and played around for a bit.  Everything looks good on preset 2, on other settings the trace is smashed to heck on the right side.  Here's a pic of 1KHz sine wave. 

I think there are still bad caps around the Horizontal/Freq Selector switch possibly?  There's still a couple of stock films there...  I'm wiped but will sleep happy knowing I made progress, finally!

[floobydust]

I did feel hell freeze over, this scope needs to be on a reality TV show lol.
Next would be figuring out if the horiz. sweep trace being squished or non-linear is the sweep oscillator, or stuff downstream of it. Not easy to do without a scope to look at the ramp waveform at V4 or V5A. If you've got a different 12AU7 in there then recheck the DC voltages and also see what the SYNC selector does as far if EXT has the same squished look.

With your multimeter not able to read ACV with DC present, you can add a blocking capacitor plus resistor in series with the probe.

[Rickenbackerman]

This scope doesn't need to be on a reality TV show, it deserves its own show.  This thing is possessed. 

Fired it up, still doing the "smashed on the right" thing.  So for the heck of it and because it took two seconds I swapped V5 out with a used good one from my stash and fired it back up.  1KHz sine pic below.  It was working perfectly.  Then put the original V5 back in.  Still working perfectly.  I was stoked.  Messed around with some other stuff, took out the trash, watched some TV, went to the bathroom, etc...   Came back to the scope, fired it up, smashed on the right again.  Pic of that here as well.  There's gotta be a bad V4/V5 connection somewhere, right?

[james_s]

You already found one bad solder joint, this looks an awful lot like another bad solder joint. I've also seen flaky tube sockets, usually cleaning the pins will do the trick.

[Paul Moir]

It's a crapshoot, but check the ground tab next to the frequency selector, the one that screws into the preset pot support and gets grounded through the frame via the horizontal position pot and the horizontal gain pot.  I think I could see a bad connection there doing that, and you would have been into it replacing your caps.
(For those following this is just below R84 150K in the schematic.  C21 and maybe C23 got replaced.  I ran out of time trying to replicate the problem.  I also am very slightly squished to the right like in the "working good" picture.)

[med6753]

(For those following this is just below R84 150K in the schematic.  C21 and maybe C23 got replaced.  I ran out of time trying to replicate the problem.  I also am very slightly squished to the right like in the "working good" picture.)

That design of Heath scope at certain sweep settings will have a slightly non-linear trace. That's normal. But not as bad as op's issue. 

[Rickenbackerman]

Just reflowed a bunch of joints around V4/V5, no change.  Still smashed on the right.

James I hit ALL of the tube pins with G5 Gold awhile ago, that doesn't mean they're clean though!  I've wiggled them all around in their sockets a bunch too.

Paul - good call on that ground - looks like a trouble spot for sure but mine is fine.  I checked it and it was good, then tightened it up anyway.  Can you check something else for me? 
SYNC Selector set to -INT
HOR/FREQ Select set to LINE SW
I get about an inch wide trace, HOR gain maxed, with nothing connected to the vertical input.  With 1VPP hooked up to the vertical input, I get a 45 degree line (it still opens up a bit) top right to bottom left with the Phase pot full CCW.  As I rotate the Phase pot clockwise, it opens into a circle and then goes top left to bottom right (still slightly open) as I reach the end of full CW rotation.

Maybe I haven't played around with the Phase pot on these settings before, but I don't recall the phase pot ever doing a dang thing.

edit: Crap, I think you mentioned earlier in this thread that your phase pot was froze up solid.

[Paul Moir]

Actually with all the messing around I got it freed up.

Definitely a difference there:
Same setup, I get a nice big horizontal line the full width of the CRT.  To get it down to 1" I need to set the horizontal gain to about 30%.  (for comparison I get full width sweeps with the gain set to about 70%).
With the 1Vpp hooked up, I get an open circle which collapses to a 45 degree line with the phase control half-way, which again opens "the other way" as I advance past half way to full.  I get roundish circles with the phase control at both the maximum and minimum.
Changing the sync selector does nothing, except a momentary jump in the trace when it goes through the line trigger setting.

The phase pot only affects the phase of the horizontal sine wave going into the horizontal amplifier in the "Line SW" setting, so it doesn't do much.  There's a reason mine was frozen I think! 

[Rickenbackerman]

IT WORKS!  (for now)...

Did some more measurements, things weren't making sense around V6A again - no gain.  Swapped it out with a used good one, working again.  Maybe it's eating tubes?  Scope-O-Saurus?  Although this 12AU7 and the 1V2 that was dead when I started are the only tube swaps so far - the rest all work fine even though the 12BH7 tests weak but seems to run the scope OK.

[Quarlo Klobrigney]

Bad tube sockets. DeOxit or change them to ceramic. "Troubles-Be-Gone"
You can find them everywhere online.

[Rickenbackerman]

Quarlo I hit all the pins with G5 Gold weeks ago.  And I've tried ceramic sockets, I don't like them because they don't play well with the high voltages in tube guitar and bass amps (which is what I work on).  They arc.  Doesn't make sense, but I swapped out the power tube sockets in a 1970 Orange 120 with ceramics once.  Two months later they were scorched and the tubes were still fine.  For something like this they'd be okay - but - I think it was the tube.  I'd test it (again) but I'm going to bed victorious.

[james_s]

It could be a mechanical fault in the tube I guess, but sockets or solder joints at the socket would be what I'd suspect first. I've seen tube sockets a couple of times that had bad connections not resolved by a squirt of DeOxit.

[Quarlo Klobrigney]

So a new socket wouldn't fix all of that then?

[Rickenbackerman]

Quarlo I don't think you read the whole thread.  And I don't blame you.  It's long and it's a whole damn saga.

I've seen some wasted sockets.  Melted, and I mean seriously melted sockets.  Like turned into molten lava and then rehardened - phenolic, bakelite, "plastic" or whatever you want to call it.  Black arcing lightning tracks, power tubes glowing bright orange...  10A fuses swapped in for 3A fuses, etc...  guitar players are rough on amps!  I also had an old stock GE6550A short out plate to screen and smoke the power transformer in one of my 1973 Ampeg SVT's once.  Of course it happened while I let a "friend" borrow it.  I had Heyboer rewind it.  Works great again.

[Quarlo Klobrigney]

Oh but I have been reading all along.
I'm just to busy to reply, however I just had to.

As far as the ceramic sockets, i use them all the time up to 500V and have had no problem.
I don't know where you are getting them from, but you should change sources. Even a 9 pin socket made of ceramic should be able to withstand 1KV.
I replace a lot of sockets from 50's vintage radios both PC and PCB mount. No problems so far.
Just saying, that scope is low voltage in the signal paths.

[Rickenbackerman]

I agree on all points.  Ceramic sockets would have been fine in this application, but a simple tube swap fixed the problem at the end problem (for now), so no use swapping sockets.  I just don't like ceramic ones, I had that bad experience with them once.  Left a bad taste in my mouth...

[Rickenbackerman]

Update - couple days later and...  IT STILL WORKS! 

HUGE thanks to Paul Moir for helping me out by using his scope as a test mule, and thanks to everyone else that chimed in on this.

So glad to have a working scope again.  It's not calibrated?  It's low bandwidth?  It's perfect for what I need it to do.  It works again, and I fixed it myself (with help from you guys).  And the trace is actually sharper now than I ever remember seeing it.  Sweet!

[james_s]

Good job, glad to see an old instrument like that still getting some use. It's every bit as capable as it was back when someone paid what was likely several weeks wages for it. People managed to accomplish all sorts of things with test equipment that is primitive by modern standards.

[med6753]

With a few minor tweaks and upgrades the IO-12 is essentially an O-10 first introduced by Heath in 1955. Back then most low cost scopes had bandwidths well under 1MHz. The O-10 was specifically designed with a 5MHz B/W so it could display the NTSC 3.58MHz color burst signal for the newly introduced color TV standard. But there were trade offs in the design such as a non-linear sweep as shown in this video.