Tek 575 Curve Tracer - HV Collapsing - Fixed!

[Marc M.]

I picked up a Tek 575 curve tracer off Fleabay.  Big surprise, it was poorly packed and took a bit of a beating during shipping.  I've already found over a half dozen tubes that had inter-electrode shorts (majority were 12AU7's and 6AU6's).  All have been replaced except for 2 6AN8's (and stupid money on Fleabay - damn AudioPhools  ) that I don't have any replacements for but they are located in the step generator therefore shouldn't affect the HV.

When I turn it on, I'll get a very ugly trace, for a short period of time (around 1 minute or so).  While it is visible, the Intensity, Focus, and Astigmatism controls all work (to some extent).  Monitoring the CRT anode V directly, it peaks at around +1600V (+2300 is normal) before slowly dropping down to under +100V which takes another minute or so to decay.  Checking the -1700V point, it peaks at ?-1000V before dropping down to near nothing.  The main supply voltages (-150V, +100V, and +300V) are all reasonable and stable. Both the 12AU7 and 6AQ5 have been replaced with a couple of different NOS tubes with no change.  Checked all the associated resistors and all are within tolerance. I also checked the caps with an LCR meter and all are also within tolerance.  I pulled a couple of the caps and checked them for leakage at 450V (the max. the Heathkit Condenser Checker goes) and they were good.  The Service Manual says to check the 6AQ5 oscillator stage by holding a neon lamp near the secondary's of T801 and it should glow, which it does.  There are no signs of distress on the transformer.

I've attached a .PDF of the HV/CRT schematic which I've annotated with voltages.  Those with a green background are measured after it warms up and all are pretty much stable.  If I turn it off and back on within a short period, the voltages never hit their max.  It needs to sit for 5-10 minutes before they will reach their maximum values.  It's a pretty simple circuit, but my gas-state troubleshooting is marginal at best so I thought I'd ask here.   

FWIW, I much prefer Dave's forum over the Yahoo groups simply because you guys are friendlier and generally speaking more knowledgeable than the Tek group so naturally I came here first.  I've also attached a few obligatory photos 

[PaulAm]

Might be the transformer.  I don't pay a lot of attention to 575 issues, but I seem to recall seeing a few things go by on the tekscopes list.  It would probably be worth checking the archives (assuming yahoo actually works).  Somebody on the list just recently rewound a transformer for a 576.

Might also be leaky caps in the HV side; they do go bad and it can make the HV drop.  This is probably more likely than a bad transformer.

Here is a link with some photos to inspire you
 
http://antiqueradios.com/forums/viewtopic.php?f=8&t=264438

[VK5RC]

I am no CRT expert,  what is the voltage drop across R812,  the 27k,  does the manual tell you what current the 'focus etc'  circuit should draw?

[Alex Eisenhut]

I picked up a Tek 575 curve tracer off Fleabay.  Big surprise, it was poorly packed and took a bit of a beating during shipping.

It's hard to understand why in 2016 people still aren't clued in about this. My much smaller NTSC test generator arrived dented, it could have been avoided by simply scrunching up some cardboard to bulk up the front panel so the knobs don't get the brunt of the impact. 30 seconds extra packing time.

 

[dave_k]

Under the fault condition, are both the HV rectifier tubes glowing bright?

[Cyberdragon]

He is getting some voltage, so the rectifiers aren't blown. They still may be bad though, so definitely check them.

[drjprv2]

This is my first post. I know it's customary or by protocol to first post in the introduction thread but I thought I'd jump right in and post a reply to this thread.

I'd suggest that the problem is due to leakage from bad capacitors in the HV section, specifically C809, even if the capacitors check okay for value or leakage. At least that was my experience.

I picked up a Tek 575 at a hamfest about a year ago. It had the same problem with the HV collapsing after a few minutes of operation. I replaced all of the capacitors in the HV section except for C809. I initially skipped replacing C809 because I was fearful of damaging the extremely fine transformer lead wires on the connectors where C809 is soldered to T801. Unfortunately the problem with the collapsing HV remained after replacing all of the other capacitors in the HV section. At that point I decided that I had no choice but to replace C809. That solved the problem with the HV and I’ve had no issues with HV on the 575 since. I replaced C809 with a ceramic capacitor of the same value with a 3kV voltage rating.

Hope this helps and good luck.

[Marc M.]

I've been working 10-14 hr days during the week leaving me very little time in the lab.  I'm also trying to locate the proper silver solder for use on the ceramic strips.  Out of the half-dozen Teks from this era I've had over the years, not a single one still had the Tek provided solder despite very few repairs if any done on any of them  .

 rant: Fleabay is full of 'silver' solder aimed at Audiophools.  I contacted a seller that has silver solder for 'high end audio  ' that replied to my query of composition that it was a secret formula  .  /rant

Until I can locate credible silver solder, I can't easily pull components which is why I just pulled a few of the tubular brown caps to check for leakage.
 

Might be the transformer...

  That is also mentioned in the SM.  Unfortunately, Tek didn't provide nominal resistance figures for it.  Visually inspecting it, I don't see any signs of distress and it's not getting appreciably warm in operation. The fact that I get some HV before it decays and it looked fine, I initially discounted that as the issue.
 

Might also be leaky caps in the HV side...

It's a big ceramic cap which I only tested with the LCR meter.  My old Heathkit Cap checker only goes up to 450v for leakage tests.  I do have a 6KV/10ma power supply I can use to check it but need to source the correct solder first.
 
Thanks for the ARF link.  I did a bit a sniffing over there via Google but oddly that post didn't appear in the top few results.
 

It's hard to understand why in 2016 people still aren't clued in about this...

In the words of Zeppelin: Nobody's fault but mine.
John Q. Public isn't very bright (spending too much worrying about the Kardashian's -whoever the heck they are).  I do know better and should have offered them an extra 20 bucks to pull each tube and wrap it in newspaper with a piece of paper identifying its location.  It's always best to double box gear but due to it's size that probably would have meant freight.
 

what is the voltage drop across R812,  the 27k...

I have to clear the bench to get my 3D printer set up to print out some prototype parts for a product I'm developing (not electronics related), as well as a bushing for the rear of the CRT on the 575 which was brittle and basically disintegrated in shipping.  I won't get this back on the bench until the weekend when I can check that.  Since the HV divider chain seemed to be functional (all related controls did 'stuff') and the resistors all measured up, I had discounted there being an issue there or within the CRT but maybe I shouldn't  .

Under the fault condition, are both the HV rectifier tubes glowing bright?

No, they visually appear to be fine, no signs of arcing or corona.  One had been changed so I was a little suspect there could be a problem there.  I had a Tek 551 Dual Beam scope with a lot of HV issues some of which were caused by a bad rectifier.  I picked up 2 on Fleabay for $5 shipped just in case (thank goodness no Hi-Fi sets used these or they'd be an order of magnitude more expensive). I've got a pair of x100 scope probes I bought for use at work (I repair CNC machining centers) that I'll try to remember to bring home this weekend to check them.

This is my first post. I know it's customary or by protocol to first post in the introduction thread but I thought I'd jump right in and post a reply to this thread.

In honor of Dave - No Worries!  Ultimately, this forum is all about helping people, not talking about ourselves.  Great 1st post.  Nothing is as valuable as direct experience.

I picked up a Tek 575 at a hamfest about a year ago. It had the same problem with the HV collapsing after a few minutes of operation...  I replaced C809 with a ceramic capacitor of the same value with a 3kV voltage rating.

Excellent, like yourself, that would have been the last one I'd of pulled due to it's problematic location. As you experienced a similar fault, there's a high probability that is going to end up being my problem as well.  In the thread on ARF, a member posted a cap list I think with DigiKey part #'s.  It's pretty evident that I should just recap the whole HV section.

Thanks to everybody who's replied.  I apologize for the long reply.  I've been attempting to reply since Monday but life keeps getting in the way.

[Cubdriver]

Marc -

I bought a spool of Kester '44' 62/36/02 Sn/Pb/Ag solder a year or two ago to have on hand for Tek gear as I've always found the stuff included from the factory to have gone missing somewhere along the line as well.  It's 2% silver rather than the 3% that I believe they originally used, but I think it should work just fine, and it has thus far in my admittedly limited experience.

If you'd like a bit, PM me your address and I'll pop some in the mail to you.

-Pat

[Cyberdragon]

Probably not the tranny, either the rectifiers or the CRT itself. Transformers are usually go or blow. Might be the cap, but usually it would start smoking or getting hot pretty quickly.

Wish you could be here, I can get tons of tubes for cheap. We have truckfulls at the radio club.

In fact, I used a 12AX7 to drive some LEDs. I have plenty, they mean nothing to me.

How many audiophools did I just melt?

[Cubdriver]

I'm not ambitious enough at 3:30 AM (when I really should be sleeping) to flip my 575 on its side to look at the HV section, but I'm assuming that the problem cap being discussed is the one soldered directly to the transformer shown in the first posted image.  Since I haven't seen it other than in the attached pic I'm pretty much talking out of my backside here, but to reduce the chances of damage to the transformer winding would it be practical to cut the existing cap out by nipping the leads as close to its body as possible, and then soldiering a replacement cap to the remaining lead stubs? 

I realize this will prevent properly testing the old cap, but in this case based on someone's previous experience and the symptoms, it seems to me it's a good bet that it's breaking down and dragging the HV with it, so replacing it is one of those 'you might as well' things.

-Pat

[Cubdriver]

Probably not the tranny, either the rectifiers or the CRT itself. Transformers are usually go or blow. Might be the cap, but usually it would start smoking or getting hot pretty quickly.

Wish you could be here, I can get tons of tubes for cheap. We have truckfulls at the radio club.

In fact, I used a 12AX7 to drive some LEDs. I have plenty, they mean nothing to me.

How many audiophools did I just melt?

You melted me because I see Tek carcasses that have been stripped of their tubes all to often, and it pisses me off (as anyone who remembers my 6DJ8 rant from when I got the 515A a few months ago well knows).

-Pat

[PaulAm]

Since I haven't seen it other than in the attached pic I'm pretty much talking out of my backside here, but to reduce the chances of damage to the transformer winding would it be practical to cut the existing cap out by nipping the leads as close to its body as possible, and then soldiering a replacement cap to the remaining lead stubs?

Should be fine.  Just don't leave any sharp edges to encourage corona.

[Marc M.]

...replace C809. That solved the problem with the HV...
Hope this helps and good luck.

Had a some time last night to work on the 575.  I pulled C809 and checked it for value/leakage.  According to the Heathkit Condenser Checker (not calibrated in years), it was around .005uF and was leaking at 25v.  Great first post, you nailed it  .  I had discounted that cap since the neon bulb test showed the 6AQ5 was still oscillating after the collapse.  It's on the low side of the transformer so the original cap was only rated to 600vdc and no worries about corona.  I had a NOS .001/600v Mylar cap from the 70's on hand.  Checked it with the Heathkit, close in value and no leakage at 450v so I put it in.  Since this didn't involve the ceramic strips I just used Kester 44 63/37 solder. I cut one side, pulled the cap, then the left over stub.  Having high quality soldering equipment (Metcal) made this process trivial. Fired it up, adjusted the -1700v and had a fuzzy, green haze I could move around the screen and highlight the burn spot near the center of the CRT  , but the HV held .  I had the collector voltage turned fully CCW.  Once I turned that up a bit, I got a pretty nice trace after setting focus/astig.  Popped a transistor in, fiddled with the controls and ended up with something that resembles a curve.  There's obviously still issues, but at least I now have basic function. 

Thanks to all the contributors to the thread.  I attached a pic of the new cap and what I'm currently getting for a display.  We're going to call the HV fixed.

[VintageNut]

Very nice rig. The old technology to make this measurement happen is intriguing. Thanks for sharing the result of the repair.

[drjprv2]

Excellent! Looks like you're well on your way to success. Glad to see another one of these fine instruments being restored instead of being cannibalized.

I don't know if you've seen this but the calibration procedure is available here: http://w140.com/tek_fcp/tek_575_fac_cal.pdf.

[Marc M.]

...I don't know if you've seen this but the calibration procedure is available here: http://w140.com/tek_fcp/tek_575_fac_cal.pdf.

No, I wasn't aware of it, just had the Tek SM for it.   

[Cubdriver]

Excellent!  I'm very happy to hear that the old beast is up and running again!   

-Pat