Tektronix 465 Oscilloscope Troubleshooting

[DanielSpokane]

Ok I'm looking over the schematic and I'm thinking I want to add the capacitors in parallel to C1542 not C1552. Right? C1542 looks like it is in power supply for the 15V rail. For some reason I was thinking C1552... but that is on the 5V PS. Correct?

Also, after I connect this... I assume I'm still taking the scope reading at TP 1548 and compairing to the funny looking output I posted earlier to see if the ripple dispoates which maybe points to C1542 as as the cuplrit (in part at least)?

[tautech]

Ok I'm looking over the schematic and I'm thinking I want to add the capacitors in parallel to C1542 not C1552. Right? C1542 looks like it is in power supply for the 15V rail. For some reason I was thinking C1552... but that is on the 5V PS. Correct?

Also, after I connect this... I assume I'm still taking the scope reading at TP 1548 and compairing to the funny looking output I posted earlier to see if the ripple dispoates which maybe points to C1542 as as the cuplrit (in part at least)?

You're onto it. 

Now a few clues about that screenshot.
Frequency = 2x mains
From Trigger level box; trig level plus ~2 div = ~15V where the flat top was and where the regulator was holding it to 15V. The sawtooth was typical 'pre' regulation and the degree of ripple tells us the main filter cap is not doing its job.

The blue trace is what it should look like after the cap and the min amplitude = Vreg + 0.7 to allow for the diode drop in the pass transistor.

[DanielSpokane]

Wow, that graph is very helpful. Thank you!

I paralleled 600uF to C1542 (using (3) 200uF caps I found stashed away). I fired up the scope and the voltage at TP 1548 now reads 15.4V. I posted the scope output and a picture of the test setup (for fun). It looks like DC voltage is right and the rectified power looks correct (best i can tell).

I also took a video of the scope screen while this is occurring. The trace is now linear and looks very nice (I did a little adjusting after the video and it looks even better). A video of it is at:



Looks like I need to replace C1542... and then recheck everything? Any suggestions on replacement technique and cap to use?

[tautech]

Great ! 

Now we back up some and recheck all the other rails for accuracy and ripple and decide how many other main smoothing caps need attention. All rails must meet the Tek spec for ripple. Levels we can adjust later.

The manual BOM will have the specs for each cap and when making a selection for this older gear be sure not to pick anything with too low ESR.....it only adds to the diode surge current at power ON.
Chris (bd139) will know where he gets them from but you're on the other side of the pond so I guess your best bet is Digikey or maybe check Tedds:
https://www.tedss.com/

You'll probably need to pull C1542 to measure the pinout and mounting lug layout but these 'cans' can be difficult to get so you might have to 'shoehorn' a radial in there instead.
Again, bd139 will know as he's done plenty of these over the years.

So recheck the others, build a list and park it until you get replacement/s sorted.

[DanielSpokane]

Awesome! I will wait and see what bd139 has to say regarding suggested cap.

I checked voltages again w/ the 600uF in parallel with C1542:

TP1568 -- DC 8.11V -- AC 1.4mV
TP1548 -- DC 15.28V -- AC 1.2mV
TP1536 -- DC 56.0V -- AC 1.5mV
TP1558 -- DC 5.09V -- AC 1.5mV
TP1518 -- DC 110.07V -- AC 2.5mV

All appear to be within spec.

[bd139]

Looking good.

Two options for replacing that cap. One nice and one not so nice.

First step is to get the thing out which is a nightmare. I use desolder braid and a 50W Weller and it’s hard work.

There are three terminals and two are shorted. They need to be shorted on the scope as well as I think if I remember correctly it bridges two parts of the circuit. New capacitors are tiny compared to the originals so your options are usually getting one in as best as possible via lead bending and then jumping the other two pins on the bottom of the board.

The last one I replaced however i made a carrier board. I took a piece of scrap FR4 PCB offcut, cut it into a circle with aviation shears, marked the original pins on, gouged out a channel to separate negative and positive, drilled holes for the new cap and old pins and the used snipped off 2W resistor leads to jump this board. Didn’t take any photos however

As for the cap, I tend to use Vishay, rubycon or Epcos ones for this job but to be honest any brand cap will do. Forget looking for an original can - they’re stupid expensive ($50+ each!). Make sure it’s 105oC rated, voltage is higher than the peak voltage and that’s about it. If you’re feeling sneaky, measure the lead pitch of two of the holes on the board and order one that fits. Don’t get low ESR ones.

To quote Wes Hayward, it doesn’t need to look good to work good.

[tautech]

A few possible options I found while having a look around but most seem to be low ESR 
http://nz.element14.com/vishay/mal205016472e3/cap-alu-elec-4700uf-25v-solder/dp/1166572
https://nz.rs-online.com/web/p/aluminium-capacitors/4340699/
https://www.tedss.com/2020038380
https://www.digikey.co.nz/product-detail/en/vishay-bc-components/MAL205156472E3/4778PHBK-ND/5632831

[bd139]

That last one is interesting the MAL205156472E3. Worth checking the pin pitch on the data sheet.

Still 4x the price of the ones I use though

[tautech]

That last one is interesting the MAL205156472E3. Worth checking the pin pitch on the data sheet.

Still 4x the price of the ones I use though

http://www.vishay.com/docs/28346/051053pe.pdf

Remember your Pound is ~1/2 our NZD

[bd139]

Yeah true. The one is pick is about £2 including delivery from RS though

[tautech]

Yeah true. The one is pick is about £2 including delivery from RS though

Here's a cheaper selection but few of which I'd consider worthy of putting in a Tek PSU. 

https://nz.rs-online.com/web/c/passive-components/capacitors/aluminium-capacitors/?applied-dimensions=4294826632,4294518902,4292338771,4294885341&pn=1

[Addicted2AnalogTek]

You could go fancy with it and order some cap adapter boards (courtesy of Dave K). I ordered some just to have, and recently replaced a couple of caps in my 465B using them.   "CAPBOARD_S_v2.zip" are the files for the caps with 3 ground pins (the ones you would use) , and the "L" is for the larger caps that have 4 ground pins.

https://www.eevblog.com/forum/repair/making-dave_k's-capacitor-adapters-at-osh-park/

I went a bit overkill on the caps and used these:

https://www.mouser.com/ProductDetail/EPCOS-TDK/B41252B7568M000?qs=sGAEpiMZZMtZ1n0r9vR22flQyDdOCPao7Ig0dxc2YlI%3d

[bd139]

Those are really neat but I couldn’t be bothered with that hence why I snipped my own out of FR4 stock ... I might be tempted if I ever get my hands on a high end 7000 series.

Good choice on capacitor. Epcos ones are really nice.

[Addicted2AnalogTek]

I can see why you wouldn't bother to go that route on a 400 series... one small board costs a little over 2 USD shipped from OSHPARK, and you have to buy at least 3. Next, wait about two weeks.   I bought the recommended pins from Mouser at $0.24 each (x4 necessary for each board) IIRC.  Then you still need to pay for the capacitor and shipping.   It just doesn't make sense unless you're placing large orders and don't mind buying boards for possible future use, not to mention waiting weeks for the boards.

Buying all of that to restore a high end 7xx4 scope... yes, definitely!  I wish I would have grabbed that 7844 option 4/78.... oh well.

[bd139]

OSHpark always takes over a month for me. I’ve moved to JLCPCB. You can get ten 4x4” boards for $15 shipped! Reckon your could get 4 per board so 40 for $15

[DanielSpokane]

Thanks for all the help everyone. I'm at work now, but tonight I will pull the capacitor and will review dimensions. I like the idea of making a carrier board. I will give that some thought tonight as well. Regarding the cap size... looks like several of the caps suggested are 4700uF. Since the original is 5000uF... I assume that is close enough? Also, what value would you consider low ESR?

Regarding the cap adapters... what file format are those files in? I have AutoCAD but it doesn't look like a typical AutoCAD format.

Another idea I was thinking about is connecting wire leads to each of the 3 cap pins (using a crimp or friction type connection) and then working the wires through the holes pulling tight and clipping. Easier said than done? 

Thanks again!

[DanielSpokane]

I went a bit overkill on the caps and used these:

https://www.mouser.com/ProductDetail/EPCOS-TDK/B41252B7568M000?qs=sGAEpiMZZMtZ1n0r9vR22flQyDdOCPao7Ig0dxc2YlI%3d

Do you know off hand if the diam. is ok for this cap? I haven't pulled mine yet and am at work so can't measure right now. But... I'm order some mouser parts and was thinking about including this.

[bd139]

4700 is probably ok. The capacitor usually specifies a tolerance. If it’s -20 to +80% on the service manual, make sure the new capacitor matches or exceeds that and it will be ok.

Size, no idea. The new ones are usually only 18mm which is much smaller than the originals.

[Addicted2AnalogTek]

@ bd139: Thanks!  I may place an order from JCLPCB for some 7K adapters in anticipation of the need to change some caps in my 7854.  The ones I got from OSHPARK are A+ quality, so I'd like to compare them, anyway.


@ Daniel: OD of the cap can be no more than 30mm.  Even that is a squeeze, so I would go 28mm or less diameter. Obviously, you'll never find longer caps than the OEM, so any length will do.   

 The manual states 5000uf -0%/+100% 25V  (C1552) and 5500uf -10%/+100% 30V  (C1542), so I chose 5600uF to replace both the 5000uF and 5500uF, which worked fine for me.   I tend to aim for slightly higher voltage rated caps, so I originally chose the following 63V rated, which are 30mm diameter and barely fit:
https://www.mouser.com/ProductDetail/871-B41252A8568M000

The ones I posted above previously seem ideal for replacing either, and are 5600uF 35V and 25.4mm diameter.   Just purchase ones that aren't stated in the datasheet to be "low ESR".  You also want ones that are designed for SMPS or line filtering, which you'll be able to tell easily by the stated "test frequency" on the datasheet.  If you click the link and scroll down, you'll notice that it states "Test Frequency: 120Hz".  Another indicator for this is that the datasheet will state under "Applications" that they are for ideal for power supplies.

If you want to be quick about it, you could just solder some insulated wires to the capacitor terminals (with some heat shrink tubing over the exposed conductor), feed those through their respective holes, and then bridge the 3 ground holes together with more wire.  The original caps are metal cans, so all three of the grounding holes are connected in the original design.  You could just zip-tie the new cap to one of the original ones if you so choose, and it will work fine.  Myself, I prefer to have the new cap firmly attached to the main board, though.

[MarkL]

...
Regarding the cap adapters... what file format are those files in? I have AutoCAD but it doesn't look like a typical AutoCAD format.
...

If you mean CAPBOARD_L_v2.zip and CAPBOARD_S_v2.zip from the other thread, those are Gerber files plus the TXT file is a drill file, ready for submission to your favorite PCB manufacturer.  You can use any Gerber viewer to look at them.

[Addicted2AnalogTek]

Ah yes, thanks MarkL, I forgot to address that.   

I'm familiar with OSHPARK so I'll explain it for that website.

After you download the capboard gerber files, all you need to do is open OSHPARK website, click "browse for files". Next, access the folder you downloaded the files to and select "CAPBOARD_S_v2.zip".  That's it!   You don't need to worry about the warning it will give for the lack of a bottom silk screen.

It's probably very similar on the other PCB manufacturer's sites.

The mounting pins that I used, as recommended, are "Harwin (p/n H2101-01)", which are available from Mouser.  Those fit nicely into the smaller holes on the 465's main board.  However, I noticed that in order to have proper clearance between the capacitor leads and the main PCB, I had to solder them to the adapter board without them being fully inserted into the adapter board.

[DanielSpokane]

Thanks for all the explanation. Very helpful. I ordered the EPCOS 5600uF 35V Ecap... actually I ordered 2 just incase the other goes bad at some point... that and I love to stockpile haha

I'm getting ready to take this cap out and the best i can tell there are 4 connection points.  See picture below. Looks like 1, 2, and 3 are all grounded and 4 is the +.  Does that sound right to you? I assume on the new cap I could jumper the ground legs on the cap... but would need to make sure they aren't bridge some part of the circut. If they are then I need to bridge it on the circuit board or provide a leg to that point. Am I saying that right? 

Regarding mounting the new cap... I'm thinking i'm going to try the insulated wire, heat shrink, and zip tie method. If I feel like it isn't sound enough then i will use the adapter board. If I use the adpater board I will buy some of the "Harwin (p/n H2101-01)". I assume those would only be needed if I used the adapter board?

Thanks again.

[Addicted2AnalogTek]

Yes, as per your picture, 1, 2, and 3 are ground, 4 being the positive.   When you install, just insert the ground wire from the capacitor into any of 1, 2, or 3.  Assuming you choose 1, you'll want to use some wire to connect 1 to 2 and another wire from 2 to 3.   

You are correct about the Harwin pins, you don't need those unless you're using an adapter board.

Have you formulated a game plan for removing that cap?  Those things are tricky to get out without lifting the pads under the board.

[DanielSpokane]

I was planning on just heating each connection point up and removing solder (using a sucker first and then desolder wick). After I get as much off as I can I was just going to heat and push, heat and push, etc. Think that will work?

When you say "... lifting the pads under the board" what of you mean by that? Do you mean watch out for traces lifting because of heat?... or physically removing the board(s)?

[Addicted2AnalogTek]

What I meant by "lifting the pads under the board" is that when you apply mechanical force on the capacitor pins without them being free from the board, there's a chance the trace around the hole will be separated from the board.

I've been rather lucky when I've had to swap a cap and haven't had this happen, but I've seen a few posts where others have had that happen.

I don't have a vacuum desolering iron, so I use pretty much the same method you intend to use.  I try to get enough solder removed from the large ground pin (pin 2 in your picture) and positive pin to be able to free them entirely from the board.  Next, I grab the top of the capacitor, heat the solder on one of the remaining two pins and tilt the capacitor to pull the lead through just a little.  Then I heat the other remaining pin and do the same, and switch back and forth until it's completely out.  After the cap is out of the way, it's easy cleaning.     

If you remove the trigger board, you might be able to get some side cutters in there and cut one of the pins off to make it a lot easier.  It's not really necessary to remove the trigger board though. All a matter of preference.