Tektronix 468 repair/refurb

[tautech]

Another thing that seems strange to me is that now that the trigger board is the only thing removed, all the loading on all the regulated rails is gone except for the -8.  hmm

Don't forget the "trap for newbies", installing polarised components incorrectly on a negative rail.
Very easily done, especially without an overlay.
Pay attention!

[Addicted2AnalogTek]

Another thing that seems strange to me is that now that the trigger board is the only thing removed, all the loading on all the regulated rails is gone except for the -8.  hmm

Don't forget the "trap for newbies", installing polarised components incorrectly on a negative rail.
Very easily done, especially without an overlay.
Pay attention!

I havent yet removed or installed anything on the -8 rail.  I just got done checking all the components in the 110 reg, and VR825 is bad.  It is conducting like a regular diode instead of a zener.   there is no way I would have known that without this octo setup I made last night.  See any reason why I cant power up to check the other rails with the 110 reg out of comission??
Also, that dim bulb tester you mentioned, that is for short circuit testing only, correct?  What I was getting at is actually controlling the input voltage to the scope. Plus I will need one when I start doing the cal on these scopes.

***If It's not a prob to power up without 110, which I don't think it is, I know I need to disconnect CR840 sp no power feeds those transistors***

[Addicted2AnalogTek]

Forgot to mention, I did use the octo to check parts I have, and checked some 100V zeners in scope and out to make sure that my rig is able to show me the output to the scope correctly. I did manage to figure out that the voltage and current to the CUT needs to be raised sometimes. Such as Q828 - I had to turn the transfo up to about 12V to get it to show me the "Z" and "long L" shapes.  (my transfo does 0-21V, and I have a 15K R + 10K pot , so I have a pretty good V and I range.    )

[Addicted2AnalogTek]

Well, I feel like a complete noob idiot.  I did not have enough voltage going into the zener to properly test it, and then said that I knew I had to crank up the voltage for some components!  shortly after I realized that the zener I had checked that was supposed to be 100V but was actually a 22V. So I wired up a transfo out of a high end stereo amplifier, and now I have 56V, 98V, and 115V.  I also realized that I can use it a 2 output isolation transo of sorts, and use it to give me a steady 115V.

But anyway, I did find a bad part in the 110V regulated section.... a resistor... R741.  a lead fell off when I it to isolate Q828.  I replaced it with 2 10ohm comp resistors for now.  Hasn't made all that much difference, though.  Only thing that has improved is the amplitude of the waveforms on the unregs.  everything else is the same. 

[Addicted2AnalogTek]

I found some bad CRT caps. and a bad cap on the trigger board. Of course there is a whole lot more checking to do...   I get my 6kv probe tomorrow, too!

[SoundTech-LG]

Let me know if there are any particular voltages, or resistances you would like to confirm. I have a running 468 I can check.

[Addicted2AnalogTek]

Ok, thank you.  I am just now getting ready to do a little more checking around.  There are some caps in the CRT, Trigger, Sweep, and vertical section that are on their way out, plus at least one diode somewhere that's clipping my waveforms.  Once I have a few things sorted out on it I will definitely be taking you up on that offer.   

[Addicted2AnalogTek]

What do metallized poly film capacitors look like on the scope when measuring the ESR (w2aew style)  ?  all 3 of these MKT1816's look like this.  I don't know if this is some sort of artifact of being low capacitance/high voltage, or if they are bad.  According to my scope and the ESR calculation, these capacitors have an ESR of 0 because I am putting in 1V and getting back 1V. 

in the pics, the scope is on DC input, .2V/div.  In AC mode, it's centered but the same. My input from FG is 200khz 1V positive going sq wave.

I don't have any similar caps to compare to.  I have another MKT but it is 1.5uf 100V.
These caps are C218, C220, C225 of the CRT circuit. They are 6.8nF 5kv.

[Addicted2AnalogTek]

and I have no idea where to find out what the ESR of the caps in this scope is supposed to be. none of the datasheets I have found specify. very irritating.

[Addicted2AnalogTek]

ended up finally finding some info about ESR:

http://patriotgaming.com/manuals/gaming_monitors/sencore_instruments/LC103/TT104-4416.pdf
https://en.wikipedia.org/wiki/Equivalent_series_resistance#Typical_values_of_ESR_for_capacitors_.5B3.5D
http://preher-tv.blogspot.com/2011/03/typical-electrolytic-capacitor-esr.html

[tautech]

I doubt the values are critical, you could just replace for piece of mind.
http://www.aliexpress.com/store/product/10pcs-20KV-10nF-0-01uF-103-High-Voltage-Ceramic-Capacitor/929745_1382467275.html

[Addicted2AnalogTek]

yeah they just go to the Z axis ground plane.  but replace metal-poly with ceramic?  I know that those are what is used in the 465's for the same purpose, or at least in mine.

[Addicted2AnalogTek]

I also forgot to mention that I figured out why I am seeing what I am on the scope for ESR.   Under about .5uf, the readings using this method become unreliable (on the equation side).  I have probed a bunch of various low value caps, and it seems if they are good, there will be a smooth curve on the + and - charge cycles of the waveform. at those values, it pretty much acts as a cap in circuit and smooths the charge cycles of the waveform.  So if the cap is bad, and it would probably have to be pretty bad to do this, but they are either choppy or just square. I guess indicating a short or near-short.  So more or less, under .5uf or so, my ESR contraption is useless.

Oh yeah, and the lower the value, the smaller the curved portion of the +charge and "- charge" cycles are.  It probably doesn't look the same for polarized caps, though.

[Addicted2AnalogTek]

Persistance pays off!!! I'm starting to find bad components in the 110V circuitry.  The astigmatism pot is bad.  It's a 100k pot and it goes from .6ohm to 30kohm and back down to 15k ohm. C142 was also dead. I was only able to tell by comparing the ESR trace to other .01uf ceramic caps to figure out it had a high ESR. (Also, increasing the frequency helps with seeing if very small value capacitors are bad.  I doubt the measurement is accurate, but comparing them to new ones of the same value is what I did. )

C303 of the CRT circuit measured 16ohms ESR.

Another thing is that there is a dead short between the Z axis ground plane and the main ground plane. 

The resistance measured between the +110V (at TP642) and ground is 19.56Kohm - which doesn't seem like anything, but.. I measure the exact same resistance between the +110V and -8V, and also the +110V and +5V.  but not between the +5 and -8.

-8V reg and -8V unreg are a dead short from the filter cap + to to TP740.
110V reg to 110V unreg reads 1.46Mohm and climbing, like my meter is charging a cap.
+15V reg to unreg does the same thing, except it starts at 0 and rapidly climbs. It's up to 5kohm within 4 or 5 seconds.
the other rails seem fine when measured from reg to unreg. stable readings, no complete shorts.

I've already checked every component in the 110 V reg section, so there are no issues there.  I'm about to check Q401 and it's buddies in the -8, and all the diodes in the +55.   here very shortly, whether I find issues with it and do some temp fixes til I get proper parts - or not, I am going to stick it all back together and power up to see where I'm at.   I've got my HV probe now, so I can measure the CRT, but my Fluke is wayy off and needs to be cal'd so unfortunately I will need to use the RS.

One thing I thought of that I find strange, is that when I had the Vertical Amp section isolated, the only voltages that weren't affected are the ones that go into it - 55, 15, and -8.   I thought that meant that there is nothing wrong there, but it can't be coincidence that all the voltages but those 3 changed.. 

[tautech]

  I've got my HV probe now, so I can measure the CRT, but my Fluke is wayy off and needs to be cal'd so unfortunately I will need to use the RS.

The only thing you need to check is the ~2kV CRT HV.
You don't really even need to bother until you're ready to go through the cal process.
You have a display so the PDA multiplier is working, no need to measure that.

[Addicted2AnalogTek]

Welll I found at least one part that is causing my power supply mahem.     I had tracked my problem down to the region on the output pin of U712A and determined that either U712, C609, R608, or VR604 was bad.  The 15V rail is fine, so that made me think that U712 is likely good. C609 didn't seem like it was a likely candidate, since the voltage at pin 1 of U712 is low, not high. R608 would have been hot or noticeably cracked (when pushing on it with pencil eraser).. so I decided to pull the zener and test it with my PSU and a resistor. It doesn't start clipping until 48V, so there is the 5V that I am low on pin 1. 

I noticed from some pics of a 468 main board from ebay that the CRT caps I have are not original.  The only one that is in there and still original is the .0047uf 6kv, which I found an exact NOS replacement for for $4. However I'm sure that is not the main issue in the CRT circuit that is dragging down my 110V unreg. 

I decided to map out what components get unusually warm, hot, and scalding on the main board, trigger, and vertical preamp.  Mapping that out on the circuits is where I am now, but one thing I noticed is that Q530 is likely bad.   It appears to have been replaced previously, also, since it is the 2n2484 and not the ST1046.  When I touch the can, all hell breaks loose with the CRT.  That transistor is right there smack in the first part of the CRT circuit at the 110V unreg feed, and the 5Vreg feeds into it also.   So now I am going to put the bad zener back in so I can keep going until the replacement gets here, and check the voltages around Q530 to confirm. Fortunately enough, the voltage on all 3 pins is labeled on the schematic! 

As for the other issues... I noticed that quite a number of components get really hot in the Horizontal Amp circuit as well.  I wish I had a 43V zener on hand to stick in there for now so I could figure out what issues that solves, and work on the ones that aren't caused by the bad waveforms that the zener issue caused.  I'm thinking that fix will cool down all the bridge rectifiers and power supply rail components - and numerous other things, too. 

I modified my UPS so that I can run my 465 off the battery, so now I can reference on stuff other than ground!!! YAAYYYY!  That should help quite a bit! 

I have a decent list of parts to get, and some I already ordered (109D's, 30D's, 500D's, Z5U ceramics..). So far my list contains solid tantalums, some of the common diodes, zeners, 2n5551's... does anyone have any suggestions for spare parts I should have on hand for this scope and whatever one I end up with next?

[tautech]

Other than getting a few more of each component because they are usually cheaper in a lot of 5 or so and never too expensive to have anyway, the one transistor that seems to be common for CRT output stages is BF259. If not spec'ed for it they are generally an excellent substitution, often better than the OEM device.

[Addicted2AnalogTek]

yeah it specifies ST1046 and the recommended substitution is 2n2484. so you're saying I should replace the pair of them with BF259's?

I'm sorting thru my zeners and I have a 36V that clips at 41.   That 43V specifies 5%.... 5% under 43 is 40.85V.   do you think I should try it if I can't find a closer one? or just put the old one back and wait?

[tautech]

yeah it specifies ST1046 and the recommended substitution is 2n2484. so you're saying I should replace the pair of them with BF259's?

Carefull study of datasheets for any substitution requirements is the only way.
I'm not saying BF259 is a "general" output stage replacement, it's the highest voltage device in that family and IME has been a good substitute IF REQUIRED.
These metal can transistors are quite common in output stages in many CRO's and BF259 is often a perfect "drop in".

I generally just check a BJT with a cheap DMM Hfe tester and ensure it falls within typical gain specs for that device. Sometimes they are used as push-pull, so check both devices have guite similar gain.

[Addicted2AnalogTek]

Now that I understand opamps a bit better, I understand what that bad diode has been doing to my power supplies... the output voltage of U712A was getting too close to the 9V on the input pins, so it was clipping the waveform. That replacement should help clear up the waveform itself, but I'm not so sure it will stabilize the jitter.  My guess is that is being caused by the excess current that is being drawn which is destabilizing the opamps and transistors. (am I close, or correct?) 
Until my zeners and 2N2484's arrive, all I can do is try to figure out what's causing the channel 2 vertical instability and continue mapping out the overheating components in the schematics to figure out where the caps that are trying to fry my scope are.  I'm considering just replacing all the tants just to be sure. I already listed all that are in the 468. 
(((I am just going to stick with the original match components for now [2N2484] until everything's fixed, then I may try the BF's.  It was cheaper for me to get 10 mil spec 2N's than it was to get 5 ST BF's anyway.)))

This may be a little premature, but I'm wondering what a good alternative to the type 106 would be.  Basically I am wondering if I could get a FG with near 1ns rise-time and high amplitude that would be more worth the money than just buying the 106 for $120.00.  My Wavetek 166 only does around 7ns.  The PG506's are insanely expensive so I bought the 067-0502-01, and a type 191.  still need the TMG and a fast rise/high amplitude sq gen.

[tautech]

For plate output transistors, I found devices used in OEM sometime have little voltage headroom, more so in designs that operate at ~150V. It seems the 468 has only 55V for the output stages so to use a 60V device should be OK but it is too close to it's Vce max for my likeing.

For Calibration checks you can get away with quite little gear, Dave's got a good vid on a Cal check:
http://www.eevblog.com/2013/08/03/eevblog-502-19-hameg-analog-oscilloscope/

Something like that can be quite adequate for other than precise use when you would buy a new unit or get it professionally Cal'ed, wouldn't you?

[Addicted2AnalogTek]

Good point.  I'll order a few of the BF529's to try.  they may be a better choice for while I am troubleshooting and getting everything back in order, and then when everything is back where it should be I could put the 2N's in if I want.  I'll just need to double check the circuit and make sure that allowing more voltage thru those won't allow something else that isn't so easily replaced (or multiple parts) to be blown. 

Of course a professional NIST traceable cal is best for precision.  I do want to get my scopes back into calibration fairly well, but mainly I want to get the gear and do the cals myself for the knowledge and experience.  Also, I could offset the cost later on by calibrating other peoples gear or by in-house renting them.  I'm lucky enough to know some people in the engineering dept of the local college where I could go and check the cal on all my gear for free.
I bought an STP2145A double OCXO that I am starting to design a board/LiIon constant supply/enclosure for soon, which I will use as my freq standard. I'll probably try to score an old 11 or 12 digit counter to mod for it.  Something like that I could easily take there and calibrate.

Other than all that, I just want the old Tek analog gear!      I intend for my hobby to be buying, restoring, and calibrating old analog scopes and gear, mainly Tek. Once I get a good bit better at it, I may even do some to sell to offset the cost - If I can convince myself to part with them! haha

I'm thinking the filter cap on my -8V is bad. I need to pull it and check. When I have several boards isolated from it, the reg output is OK, but the unreg jumps to over 200%. When nothing is isolated, the unreg is 1V high.  Also, Q401's emitter is 1.7V instead of 1V with nothing isolated (didn't measure it with boards isolated, unfortunately.) Either Q401, Q402, or CR304 is my guess for the culprit. I'll be powering it up with the bad zener and doing a bit of digging to see what I can come up with soon.  I have 30 values of zener diodes, but nothing close enough to 43V.  Nothing at all between 36V and 50V. figures, right?

[Addicted2AnalogTek]

I may have just found the source of the excessive current problem...... I really got to slowly and closely inspecting parts... mainly looking for zeners with burnt leads, and boy did I find a bunch.   
So I decided to look at capacitor ESR... a few of the tantalums are.. ehhhhh...  which could be part of the problem. But those little bullet "electrolytic" capacitors that are mostly on the trigger board... there are several of the same value that do not have anywhere near the same ESR.  From what I gather, 1uf 35V capacitors should have an ESR of about 4ohms.  most have 5ohms (acc to my device) and some have about 1ohm. I'm about to pull some leads and take some better measurements of ESR, capacitance, and check resistance across them to see if any are shorting. 

However there are zeners on several boards with burnt leads, so I'll be doing a lot of capacitor checking for now. Some of the zeners must be ok or I don't think the scope would be working, but the longer I have it on, the worse they are going to get, so I'm not going to power it back on until I've found some caps to replace and get the current draw down a bit.

Have any other ideas for why the zeners are getting so hot that their leads turn black?

[Addicted2AnalogTek]

I was mistaken about some of the capacitor values, which makes the dramatic difference in ESR understandable since some are 2.7uf and some are 1uf, but I am still fairly certain that several of them are bad.  I actually think that there are only two good ones on the whole board. They are the only ones that give me a "flat" response on the peaks of the waveforms. the rest of them are ramps.  I think that these ones are originals. There is no variation in any of them at all, not even in the text. I know that several tantalums have been replaced, a couple of the main filter caps, and all the electrolytics were replaced with ITT's at some point. Makes me kinda wonder why the one in the CRT circuit was soo bad (15ohms). That area shouldn't have been getting so hot that it killed that one cap..

[Addicted2AnalogTek]

I accidentally shorted the B-E of 740, so now I'm scrooged until Mouser gets me my new ones. 
Installing that new 43V zener diode worked wonders for the 110V and 55V waveforms, and of course brought the voltage on TP158 up 5V.  Also allowed the 5V and -8V rectifiers to run a little cooler, but I'm still waiting on the new caps so I can lower the current draw. It's so high that it's killing the 43V zener again. The new one started at 14.8V and dropped to 13.5V fairly quickly.  I didn't get all that far testing stuff once I killed Q740 when I was checking it's Vbe.   Oh well...

The most expensive capacitor in the entire scope needs to be replaced! but fortunately it is in the storage PSU so It's not crucial to get the main fixes done.   It's C162, and it's a 3 lead VPR cap.  It's just a filter cap, so when the time comes I,m just going to stick in a 390uf 16v radial, and I guess I could just cold solder a lead to the top of the can?  Unless I find out that it's a bad idea before hand.

at this point, I'm just worried about getting it running properly. Im not going to buy top notch parts, or the manual specified -expensive- ones anymore, unless I can get them cheap.

I also noticed one more BIG problem.......   Once the scope warms up, once every several minutes, it will "almost" power down.  Everything goes to shut off, and then it goes back to normal. The trace never disappears or anything, but the rail voltages drop by a volt or so and then jump back up.

Overcurrent protection of the 55V rail, maybe?   (that's why I was checking Vbe)  Q740/Q828 provide 110Vreg, but also provides 55V for the non-inverting input to U712 for the 55V rail.  So if it is overcurrent protection, it would have to be affecting one of those two rails in order to cause a "shutdown" like that, right?