Tektronix 2213A Project Thread <Novice, first timer>

[guy232]

Goal: As a novice, handle this encounter in a way that uses as many "best practices" in regards to both troubleshooting, reworking and selecting replacement components that may be needed. Wanting to try avoiding the typical new guy approach of shotgunning components before completing tests and getting numbers etc.
-------------------------

I recently purchased a used tektronix 2213a from an auction site. Tried it once, displayed both traces and seemed responsive to input. I powered down and let it sit while I spent quite a few days researching old posts around the web to get an idea of what I need to do.

Total novice but have bought some equipment to hopefully even the playing field a bit. I will be taking my time and researching things as I go. I will also be posting links to imgur albums with captions to try and keep any main points streamlined. Completely welcome all criticisms/suggestions if anyone wishes to point something out to me.

I'm aware of HV present in the machine when on/off and know about discharging CRT to chasis & manually resistor bleeding main caps.

Tools/Materials:

• Hi-res scan of 2213A Service Manual courtesy of Artek Media
• Aoyue 469 variable watt soldering iron w/ Hakko T18-D16 Chisel Tip
• Hakko FR-301 w/ n61 1mm nozzle
• Kester "44" 63/37 .020"
• MG Chem RA Flux Paste
• Digital multimeter, standard probes & 6" extended reach clip test leads

Objective 1: Safety/Reliability:

Pics for this section w/ captions for clarity

My work done: https://imgur.com/a/oDoDqyv

The components I have questions about: https://imgur.com/a/8NsWGsh


Before testing, based on what I've read online, there are a couple hazardous components that may explode when EOL present on the 22xx series scopes (2213a, 2215a, 2235 etc). I read that it is normally bad to change components prior to taking measurements because it can introduce many issues. I may be wrong, but In this case it is advised to replace the known hazardous components prior to doing any readings.

2 of these caps are located the small filter board on rear of chasis, other is located on mainboard HV.

They are RIFA 2200pF plastic safety capacitors that craze (surface cracking) over time. I inspected and found that each cap exhibited this.

Researched and believe I found proper replacements RIFA PME 271 Y 40/100/56/B (datasheet)

Applied RA flux, desoldered, cleaned, inserted replacements, RA flux, soldered, cleaned.

With those taken care of the only other possible explosive component that I'm aware of is the IEC AC power cord inlet. It is a schaffner branded one that apparently can blow up when near EOL. I believe the proper part to replace it with is a "Qualtek 858-03/007" I do not have this part yet but I am using 120V USA mains and have read that they are less likely to pop when not using 220V. I presume it may be safe to reassemble and take voltage readings of the LVPS before continuing my diagnostics on this device?

Lastly, I inspected some red disk shaped components in HVPS that appeared to have cracking at where the leads meet the component's coating. Is it unsafe to continue with powering on & testing until these are replaced?

[floobydust]

I start by replacing all the electrolytic capacitors in the power supply. Only a few dollars and gets rid of many oddball problems that can waste time troubleshooting.
Excel spreadsheet with (same) Tek 2235, and 2235a electrolytic capacitor sizes and modern part numbers: Badcaps Re: Recapping Tektronix 2235 PSU saved time measuring the sizes and getting part numbers.
The 840uF Sprague caps are surely getting too old. Don't forget the Rifa cap hidden behind the on/off switch.

The red disc caps are cracked on the legs-not the body which I think is OK. It's best if the HV parts are not touching each other with some air space between. I would not worry about the label peeling.

[guy232]

update: After looking over the cap list I question if the 3 caps I replaced need to be Y2 X1, because my current selection is Y2. Will report back if I find the info.

update2: after double checking it appears the replacement cap I used is the exact same as original (PME271Y)which is only rated Y2, I suppose the selection feaured on that 2235 cap list possibly works just as well, it just happens to also have an x1 rating.
------

@floobydust
Cheers for the info, will do. Was working on cap list but will definitely be checking into that one linked, searched quite a bit on the web for cap lists and never found that link, thanks!

[floobydust]

These are great scopes and a lot of fun to add life to. I think you have a good tool list.

I found it hardest to desolder the big primary (can) capacitor. It takes a lot of heat and you will have to add new solder to the pad, let the heat flow through to the other side of the PCB and then hit the suction. If your desoldering skills are new, maybe practice on some old boards. Being too quick or too slow with suction causes things to go wrong.

[David Hess]

Leave the red high voltage disk ceramic capacitors alone.  That cracking where the leads are covered with the encapsulation is completely normal for disk ceramic capacitors.  Preemptively changing them risks damage to the high voltage cathode circuits.

The old RIFA X and Y capacitors should of course be replaced.  I would not bother with exact replacements but it is not critical.

Any original aluminum electrolytic capacitors in the power supply will be nearing the end of their life.

The only other thing I might preemptively replace is the focus resistor chain which I believe is made from Allen-Bradley carbon composition resistors on the 2213A.  Later models in this series of oscilloscopes used high voltage film resistors like the Vishay VR25 and VR35 series.

[guy232]

Thanks for the info, also wanted to say thanks to David Hess for all of the old posts regarding the 22xx series that I have found while using search engines.

In regards to the focus resistor chain; I actually was about to change that next!

https://imgur.com/a/TWawyse

I desoldered the six 510k ohm allen bradley carbon comps. Testing out of circuit, the lowest resistance was 620k, the highest was 780k.

I replaced them with Vishay HVR37 510k ohm 0.6W Metal Film Resistors, I believe I actually selected these based on one of your old posts, David. I believe you also mentioned in the post that using the HVR37 could potentially be a bad selection because they tend to fail in short circuit, as opposed to carbon comps failing open?

Forgive me if I am misquoting.

My parts arrive tomorrow for the electrolytic re-cap. I used the list provided by floobydust's link, all of the 2235 capacitors in the list did appear to be proper choices.

As always, if anyone happens to notice something I did poorly or a poorly selected part please let me know; I have no issues about going back to do things right when needed. Cheers!

[David Hess]

David. I believe you also mentioned in the post that using the HVR37 could potentially be a bad selection because they tend to fail in short circuit, as opposed to carbon comps failing open?

I do not remember saying that about the HVR series in connection with the VR series but it is a problem with standard film resistors.  I do not remember the details but I think the HVR and VR series are suppose to be "fusible" meaning they fail open.

[particleman]

I do not remember saying that about the HVR series in connection with the VR series but it is a problem with standard film resistors.  I do not remember the details but I think the HVR and VR series are suppose to be "fusible" meaning they fail open.

David, if you ever write a book I'll buy it.

[guy232]

Update: answered my original question below while searching. Looks like it had already been answered and the hvr37 is an acceptable replacement unless anything has changed.

I am probably the one who mentioned them in that other thread.  The Vishay HVR37 series is essentially what Tektronix used although I am fuzzy on the differences between the HVR37 and VR37.

If your 2235 has the original carbon composition resistors in the focus chain, then I would replace them.

Quoted from this older thread: https://www.eevblog.com/forum/repair/tektronix-2235-upgrades/

I'm comparing the vr37 to the hvr37 datasheet, after reading the posts a couple times maybe I'm misreading things but:

is the VR37 series better suited than the HVR37 for the focus resistor chain? If so, what are the reasons for that choice?

Feel free to be brief if this question has a complicated answer, I can do some searching around on the bullet points if it helps save time!

Just trying to try to get an eye / understanding on component selection in case I am overlooking something.

[David Hess]

I remember looking through various sources for the differences between the HVR and VR series of high voltage film resistors and not finding anything conclusive.

When I researched the changes over time between various 22xx model oscilloscopes, I found that Tektronix initially used Allen-Bradley CB1845 and EB5145 carbon composition resistors and then replaced these with high voltage film resistors made by "Mepco/Centralab".  Based on tracking some part numbers through non-Tektronix sources, I think Vishay ended up owning the "Mepco/Centralab" resistor production lines and the specific resistor series became the VR25 and VR37.

[guy232]

Great info, thanks for the input!

Was delayed by 2 day migraine. The PS capacitors are all desoldered & the schaffner IEC AC outlet was snipped off. I'm replacing now and will update when complete.

Once complete, I can then begin to test and gather numbers I think.
---------------------

The only other area that didn't pass visual inspection was around the Vertical Output / Load Resistor area. Aside from the obvious darkened PCB, I also spot some orange residue around transistor Q254 (SRF3188) Old flux? Evidence of past rework?

High res & captioned pics:
https://imgur.com/a/Jv8drGC

[David Hess]

Oh, I forgot about the vertical output load resistors.  They run hot and especially so on the earlier 2213/2215 where only two were used in series.  The 2213A/2215A are lower cost 60MHz alternatives of the 100MHz 2235 are replaced the 2213/2215.  (1) One of the improvements was to use three resistors in series to distribute the heat better.

On a 2212/2215 I would recommend changing them but on a 2213A/2215A, I would not bother.  They are 1% 1/2 watt metal film resistors which are not common but Mouser carries them for a reasonable price.  If you replace them, remember to space them above the printed circuit board.  I used the shaft from a cotton swab as a temporary spacer when I did it last.

(1) The 2213 and 2215 are really a different (and short lived) series of oscilloscopes with just those two models.  They were followed by the 2235 series which includes the 2236 (counter/timer), 2213A/2215A (60MHz low cost), 2230/2232 (DSO), and some other variations which are all based on the 2235 design.  The 2225 is another separate series which was produced later.

[guy232]

Aye, added them along with the HVPS components order. Good advice, going to only replace if they end up being found faulty during testing. Since one was kinda discolored I thought I may as well have them on hand.

Some novice project drama finally cropped up. I have all of the SMPS work done without any snags except for C944. The positive lead was crimped down on the pad quite well and my ignorance caused the solder pad to become damaged on the bottom side of the board.

I first looked up C944 on the circuit diagram to hopefully get some clues about the positive pad. I wasn't sure what the empty triangle symbol meant so I browsed through the SM trying to find a symbol legend, the SM said to refer to ANSI 32.2-1975 for symbol definitions. Tracked down the .pdf and read through it to find out the symbol indicates a circuit return/signal ground.

Continuity checked, the C944 solder nub poking out near the lifted pad seems to share continuity with all other "signal ground" pads in the HV side of the SMPS. I believe that the bottom trace in this situation was just there for strength and not a connection? I.e. the top pad is what is connected to this signal ground whose plane is on the top/component side of the board.

Going to do a little more reading/probing to see if I can verify my assumption that this particular lifted pad won't impact the circuit. If that is the case I will continue with the project and practice patching up the damage when my trace repair materials arrive.

[David Hess]

If it matters, the ground return for C944 is the ground side of R949 so if the trace is missing, those two points can be bridged with a wire but I suspect it is part of the ground flood fill.

[guy232]

Powered up the scope, set to "Baseline Trace." Idles fine for 3-10 minutes before trace shrinks and blinks. Power indicator LED starts dimming in/out before everything goes dark. Toggling On/off doesn't turn the device back on. When left unpowered for 30min-1hr the scope turns on again before repeating the above.

Feel free to skip forward to 2:20 where the anomaly begins.

https://youtu.be/rC6rcQxdX1w


I disassembled and inspected. No visible issues. The only damage caused during repair was C944 positive bottom pad lifting (Mentioned in reply #12.)

I believe my best bet is to attempt to take LVPS here shortly after the machine sits unpowered for a bit. Will update.

[David Hess]

Does the inverter transformer get hot?  It might be damaged and failing at high temperature.

[guy232]

Is there a safe way to check the temperature of the inverter transformer without an isolation transformer? The only non-contact option I have is a k style thermocouple.

Just to double check, the inverter transformer is T944? Circled in light blue in below pic. Or is it one of the other 2 near by transformers? Reading in the manual they are labeled by different names.

Will update if I answer these questions, running searches

[David Hess]

No, I mean T948, the big one which includes the high voltage outputs.

T906 is usually the one to fail but since you are using a separate 42 volt power supply and not the built in off-line buck switching regulator, it is not a factor.

[guy232]

Alright, rigged up a stick and thermocouple to use multimeter. Ran scope for ~4 min and it experienced the same problem. I then quickly probed for temperature.

Ambient, Room, No fans/AC: 22-23 C

t906: 25 C

t948: 25 C

u975: 24 C

Chasis near heatsink: 27 C

T944: 30 C


I'm hoping this crude method of temperature testing offers viable data.

I don't think any of those temps are too insane? After measuring I attempted to power back on with no success so I doubt things cooled too much.

Will replace shield and probe LVPS points once the device allows itself to power up.

[David Hess]

I don't think any of those temps are too insane? After measuring I attempted to power back on with no success so I doubt things cooled too much.

No, those measurements are fine.  It was worth checking because it is so easy to do.  If a transformer was failing due to overheating, it would have been very hot.

[guy232]

Just wanted to update, have recieved a few suggestions from tekscopes and have done the following:

- Unplugged the CRT PDA (HV Anode connector) and powered on. Anomaly still occured after 3-5 minutes as before. Reconnected PDA and continued on.

- Checked LVPS voltages, all in spec but did notice some fluctuations on the 100V rail when anomaly was causing horizontal distortion/LED dimming etc. Unknown if the other LV rails fluctuated, will have to retest that later.

- Noticed the exterior chasis near Q9070 heatsink screws were warm to the touch. Tried hitting that area with upturned can of air to freeze chasis, the anomaly instantly stopped. The trace intensity got so dim that I thought it disappeared at first but then I realized that this was because the intensity knob was set low. In other words, I believe the anomaly is possibly causing the intensity to be much higher than what the knob is actually set to?

After spraying, the trace & intensity remained stable for 5-7 minutes before starting to show signs of the anomaly starting up again. I sprayed it a second time in the same spot, instantly returned trace intensity and stability again.

I edited up a compact video with captions, 57sec in length that document the above mentioned testing (I know how painful most videos can be so I did my best to make this one deliver as much info about the problem as fast as possible)

https://youtu.be/c0Ht9Crb6h4

I have just recieved the only E.C. that hasn't been changed, C956 (Nichicon UHW 330uF 50V) but I don't believe this cap would be causing this issue.

Someone suggested I solder 2 test wires to TP950 and TP940 and clip on multimeter to check voltage for stability. I imagine this will be safe if using insulated wires and clipping up the meter before powering the 2213a up.

Beyond this I believe I may be running out of options available based on having no isolation transformer and no extra scope. I have read some of David Hess's older posts on the 22xx and understand it takes quite a bit of gear and skill to properly probe around in this SMPS. Also is a task that isn't reasonable to walk someone through.

Either way, will see what I find and will report back!

[pbarton]

Just seen your YouTube symptoms. I would speculate that you have an intermittent high voltage discharge in the vicinity of the HV multiplier that produces the CRT anode supply. You need to remove that large black plastic cover which protects you from hazardous voltages on the non-component side of the main board. Switch the scope ON and get it into the failing state, turn the room lights down and look for pin-prick arcing between PCB tracks in the vicinity of that HV multiplier. You need to view around the multiplier on both sides of the main board. However, I've been wrong before!

[guy232]

Will definitely try that next time I power on for testing. I read a few old posts talking about that exact issue. The posts never mentioned what the actual causes of the arcing is though. In those cases, what would potentially be the cause of HV arcing?

[pbarton]

The HV multiplier has a high voltage and high frequency AC input, on one of the four PCB mounting pins. High frequency arcing can occur in the vicinity, if the motherboard is contaminated or otherwise compromised. Arcing produces more carbon deposits which assists the breakdown. You may detect a pin-point (high frequency) white light between adjacent PCB tracks. I'm going to be so embarrassed when you prove me wrong.

[tautech]

Just read this thread again and I can't stop thinking you might have a thermal issue in that components that are continually thermal cycled place stress on the solder joints and their own leads internal connections also.

IMO you should reflow all joints around the area you're hitting with the freeze spray as it's easy to miss fine solder cracks with a visual inspection.

If that fails clean that area with IPA too just in case you have some intermittent leakage issue.

[guy232]

Finally feeling a bit better and had time to open it up. Turned the lights off and turned on the scope. There are indeed very tiny and very faint white pin prick lights visible on the bottom of the board in the HV section. Could barely see it even in the dark, magnification helped.

I notice small traces of flux and very light contamination. Going to hit it hard with IPA, brush and kimwipes. Will retest and report back.

[guy232]

Reflowed some questionable (to my novice eye) joints using amtech flux and kester '44. Did a thorough scrub with 99% IPA/brush/kimwipes

Plugged in, fault happened within 3-5 min. Fault stopped after 3 minutes and then both traces remained steady for 10+ minutes until I powered down.

I let the unit sit an hour or longer then repowered up and experienced the same results as mentioned above. Did this 3-4 times while looking for any signs of arcing.

Would arcing be very obvious? Or could it be so dim that it is barely visible even in a pitch black room?

Going to try desoldering the Hv multi lead as suggested and see what happens.

[David Hess]

Sometimes the encapsulation around high voltage parts breaks down with only a tiny hole showing on the outside.  This also applies to potted assemblies like high voltage multipliers.

[guy232]

Made some progress!

Failing state no longer causes shrinking horiz/vert. After powered on for ~5min the traces just barely pulse intensity (as if you were barely turning the intensity knob back and forth by 2-4 degrees.) After ~5 more minutes the traces stop pulsing and remain steady until I power off the scope.

Used cheap usb microscope on PCB. Are the pics below tiny fractures that were imperceptible to my magnifying glass as @Tautech mentioned?

I reflowed Q946, Q947 and C907 with flux and new solder.

I changed out C956 (270uf 40v 10+100%) with a new nichicon UHW 330uf 50v. (Had to use the original cap during the previous tests because I initially ordered the wrong replacement. The original cap tested good but I know there are multiple ways ecaps can fail so unsure if it had any effect.)

link to pics:
https://imgur.com/a/VtTQGXD

[David Hess]

That is a nice find.  I keep an insulated tool like a bamboo skewer or diddle-stick around to tap on boards to see if it is a mechanical problem.

[jdragoset]

Back side of Harbor Freight brush gives a wide area (3/8" x 3/8") to push and distort in all directions for mechanical test.
I am changing (in succession) PS electrolytics and those insipid RIFA's in several Tek 2215's and one 2235
except the Sprague can (75 uf, 450 volt), as they test very well, likely because these have always been operated at 120 vac input.
If they fail, will blow the line fuse and will then be replaced.
I've many, very much older instruments, where those old 60 Hz DC filter caps still work fine, particularly if the voltage rating is much higher than the operating voltage.
e.g. Tek TMG 184, SN: 0412 powers up well with less than 40W consumption on the Weston 310 Watt meter, PS ripple less than spec', so why change electrolytic caps. 

[sprit]

I have one similar to yours.  Looks like it was repaired by someone (some parts were lost) below are their names.  Currently in Vietnam, these components cannot be ordered (some components have strange names, I don't know what to replace them with.) Can anyone here tell me the names of the transistors.  this?  i have some transistor i don't know if its suitable for it anymore.  They packed them in a lovely bag and of course no label on it.  so unlucky!  Thank you very much.  Wishing everyone all the best.

[David Hess]

Those are 2N3904s and 2N3906s, so not critically performance parts.