What's the likelihood of virtually all the caps being dead on a Tektronix 2235?

[requim]

I sat down to test the capacitors on my Tektronix 2235 since I bought it as a repair unit.  Virtually all the caps showed up as 0L that I tested and I tested 20-30 caps.  A few caps gave values but they were way out of spec.  I know I switched my meter into capacitance and not into ohms.  The resistors I tested all seemed within spec, though I tested far fewer of them.

How likely is this?  Am I doing something wrong?  What would cause this?

[hacklordsniper]

Possibility that all caps are dead on such equipment is not very possible. However you do not know where unit has been, its storage and operating conditions.

Capitance meter is not a way to test capacitors

[requim]

What's the right way?

I'm using my Fluke 87 V to read the capacitance.   From what I understand I need to know the capacitance as well as the ESR.  But when the caps are showing as being shorted, what's the point of getting the ESR?

[amspire]

I sat down to test the capacitors on my Tektronix 2235 since I bought it as a repair unit.  Virtually all the caps showed up as 0L that I tested and I tested 20-30 caps.  A few caps gave values but they were way out of spec.  I know I switched my meter into capacitance and not into ohms.  The resistors I tested all seemed within spec, though I tested far fewer of them.

How likely is this?  Am I doing something wrong?  What would cause this?

The capacitance range of a multimeter is not designed for in-circuit tests. If you try testing another board, you will probably find the same thing happens. The capacitance test in multimeters requires the capacitor to be part of an oscillator circuit, with the period proportional to the capacitance.  For the meter to oscillate, the capacitance has to be able to freely charge and discharge  between the two preset  voltage levels in the meter. 

If the circuit does not let the capacitor freely charge or discharge to these voltages due to a parallel resistor or other components, the meter cannot oscillate.  OL means that either the capacitance is too big for the meter, or the circuitry on the board is preventing the oscillator circuit in the meter from oscillating.

It is hard to think of a capacitor failure that would show up as an OL reading - so basically OL is not telling you a thing except that the meter is not able to measure the capacitor in the circuit.

To test a capacitor, you will probably have to remove at least one lead of the capacitor from the board.

Richard.

[hacklordsniper]

Nicely said. However even out of circuit capitance measurment will not tell you if a cap is bad or not most of the time, you need an ESR meter, possibly one that can measure in circuit.

I use one from GME tecnology and it has proven itself very good

[mikeselectricstuff]

Nicely said. However even out of circuit capitance measurment will not tell you if a cap is bad or not most of the time, you need an ESR meter, possibly one that can measure in circuit.

I use one from GME tecnology and it has proven itself very good

Although by the time you've taken it out you may as well replace it anyway...

[hacklordsniper]

Although by the time you've taken it out you may as well replace it anyway...

You are right, but it depends how something is old and is it necessary to replace capacitors or its enough to find just one that died

[requim]

Looks like it's time to pull out the soldering iron and desolder a few ends on caps on a couple different boards to see whether I get different results. I'll post the results.

[tekfan]

The likelihood of virtually all the caps being dead on a Tektronix 2235 is virtually 0. Tektronix used high quality caps in every piece of test equipment they manufactured. The really problematic caps are tantalums from the 70's. They usually go short circuit. The 2235 is from the 80's so as far as I know there are no tantalum capacitors in it.

Checking a component in circuit is always tricky. I usually just use an ohmmeter to see if the cap is shorted or not and an ESR meter to see if the impedance is low enough.

A complete recap shouldn't be necessary. Recaping the switching power supply is always a good idea.

The only capacitors you should in no way touch are the timing capacitors near the time/div switch. They are high precision film capacitors.

[requim]

I think I'll plan on recapping the power supply and crossing my fingers then.  I pulled one cap out from the power supply, but I can't seem to find an equivalent replacement.  Here are the specs:

Mallory
840MF (though the service guide says 840uF) 12V  105C
290-0945-00   (Tek Part #)
-10+100%
235-8443K

I've been looking on Digikey w/o any success.  This site has them for 3.87 a piece, but I'll need around 10 or so of them.

http://store01.prostores.com/servlet/ucite1/Detail?no=18721

Surely there has to be a better price/source.

[tekfan]

I think I'll plan on recapping the power supply and crossing my fingers then.  I pulled one cap out from the power supply, but I can't seem to find an equivalent replacement.  Here are the specs:

Mallory
840MF (though the service guide says 840uF) 12V  105C
290-0945-00   (Tek Part #)
-10+100%
235-8443K

I've been looking on Digikey w/o any success.  This site has them for 3.87 a piece, but I'll need around 10 or so of them.

http://store01.prostores.com/servlet/ucite1/Detail?no=18721

Surely there has to be a better price/source.

well the tolerance is -10 to +100% so it won't bother the scope to have double the capacitance in there. You might try a 1000uF 12 volt low ESR cap rated for switching power supplies.

If the scope doesn't want to turn on at all you should check other things in the power supply aswell (transistors, diodes, rectifiers, open or burned resistors, etc.).
You will not believe how many times it was just a fuse in many of my used test equipment.

Anyway if you've got the time to repair it, don't give up on such a nice scope.

[requim]

I bought the scope for parts/repair because I wanted to repair it.  So I'm in it for the long haul (within reason of course).  The scope blows the fuse immediately after turning it on.  I bought a replacement HV transformer, but have yet to swap out the old one.   I bought the replacement because I had read it was a common failure point in these scopes.  Speaking of which, how can I verify whether the old one is working or not? And do you have a particular cap you'd recommend? I'm not finding much in the way of 12V caps.

[tekfan]

Well if it blows the fuse on turnon it's probably not a capacitor. It's almos definitely on the primary side of the transformer. It shouldn't blow a fuse if there is a short on the secondary since the supply is current limited.

The cap doesn't need to be a 12V one. It can have a higher voltage rating eg. 25, 35V... as long it's roughly the same capacitance and the voltage rating either the same or larger you can put there anything you want (as long as it physically fits in there).

The instant fuse blow can be caused by one or more of the rectifier diodes being shorted (CR901-CR904), the fet Q9070 could also be shorted and in the worst case the transformer can be shorted.

Try checking the transistors in the power supply with a multimeter in diode check function. You should get a drop of about 0.5 to 0.8V between base-collector and base-emitter.

[SgtRock]

Dear Requim:

--Keep in mind if your have a cap with a tolerance of -10 + 100, you indeed could double the capacitance, no problem.  But, you would have to measure the capacitance you install to make sure it in less than double of the original. For instance if you have a 1uF cap with a tolerance of -10 +100; you could replace it with a cap with a measured 2uF capacitance, but not necessarily with a cap marked 2uF with the same tolerance, because its capacitance might be as high a 4 uF. So unless you are measuring as you go along, you should try to stick to exact replacements, and in any case, stay as close to the specified values as possible. As the unit ages, parts will change in value. If they already are right at the line, they do not have any room for variance without possibly going out of spec.

--Also, you should work from the schematic and the original specifications whenever possible, as the cap or resistor may have already been replaced using the "old time radio wanker, anything close will do" method.

"I happen to have discovered a direct relation between magnetism and light, also electricity and light, and the field it opens is so large and I think rich." Michael Faraday 1791 1867

Best Regards
Clear Ether

[BravoV]

--Also, you should work from the schematic and the original specifications whenever possible, as the cap or resistor may have already be replaced using the "old time radio wanker, anything close will do" method.

This one is a really good advise.

Get your self the original schematic from the service manual, it will pay off especially you're troubleshooting a dead unit.

[requim]

@SgtRock - Thanks for the advice.  When available I always use a schematic.  In this particular case the parts and the schematic are in sync, at least so far as the devices were originally spec'd as.  In fact the caps I pulled have all been within spec.  Though I haven't been able to check the ESR on them.

I went ahead and measured the diodes CR901 - CF904 and all of them were about 0.650V.  I haven't been able to identify the FET Q9070.  I see printing on the circuit board and the circuitboard diagram in the schematic for P9070, but when I look at the parts list P9070 doesn't appear but Q9070 does.  I assume the PCB print and the schematic have the FET mislabeled.

I went ahead and tested other diodes in the PS but I don't know whether values were within spec or not.  A few of the diodes measured OL when I had the probes one way and some other value when I changed the probe/lead pairing.  Not sure what that means.  I tested all the diodes using both methods. The rest of them seemed to behave as expected (ie one way was higher than the other way).  I was thinking about recording all the probed values and posting them here in some hope that someone would see what was off at a glance.

I also probed the HV Transformer and the values on the primary side were around 1.5 ohms and the values on the secondary side were around 2.5 - 5 ohms.  I assume that's probably within spec, but having never dealt with transformers I don't know.

Any help would be appreciated.  Thanks!

[amspire]

I haven't been able to identify the FET Q9070.

If you do a Google search on Tektronix Q9070, you get all these posts of people fixing the power supply in Tektronix 22xx series oscilloscopes. Might be worth a look to see what problems other people have had.

One thing I noticed - there was a problem with a crowbar circuit triggering to early, and Tektronix changed zener VR935 from a 51V zener to a 56V zener.

Richard

[vk6zgo]

Well,if the thing blows the main fuse as soon as you turn it on,the approach of "an old time radio wanker"would be to disconnect anything from the secondary of the mains transformer (I assume from Tekfan's comment,that it does have a mains transformer).

Once he had done this,the "old wanker" would turn the 'scope on again & see if the fuse blows.
If all is good,he would then reconnect circuits to the secondary one at a time until the fuse blew again.
This process can be done again with circuitry connected to whatever  seems to be blowing the fuse,to see if it is a further circuit doing this.

As far as component substitution is concerned,this requires the ability to analyse  the circuit the component is used in,which ability you  may not possess at this point in time.
If you don't,Sgt Rock/Clear Ether's advice is excellent.

VK6ZGO

[requim]

Previously when I searched for problems with the 2235 the two primary issues people had were that the caps in the power supply needed to be replaced and the High Voltage Transformer needed to be replaced.  As I said earlier I've pulled most of the caps off the power supply and capacitance wise they all check out fine.  No ESR testing has been done on them.  The HV transformer also seems fine, but I really don't know much about how to test them beyond what I saw in a youtube video and a document on some website.

I'm 99% certain the problem is restricted to the power supply because when I received the unit it had a blown fuse.  I put in the specified 1 amp 250 V slow fuse and the first time I turned it on it put on a pretty good spark show.  It was in the general vicinity of the HV transformer, but I honestly couldn't tell you what it was.  I haven't been able to find any burn marks on any of the components.  Ever since that first time I haven't seen anything other than two lights on the board.  I haven't found anything to tell me what the lights represent.

[vk6zgo]

Maybe you should have disconnected the HV circuit then, & checked if it still blew fuses,or at least given us the information so we could perhaps help more.

If you know your way around a particular circuit,"scattergun" replacement of capacitors can be a useful time saving technique.

In this case,I'm very doubtful that it will help.

Try the stepped approach,you have very little to lose!

VK6ZGO

[requim]

I don't expect replacing the caps will solve the problem, however I do think it's not a bad idea considering how common a problem it is on the units. I've already pulled them.  At this point the only difficult thing is finding the parts and parting with the money to acquire them.  I'll pull the HV Transformer, but I'm not sure it will tell me anything new since it seems to be ohming out just fine.

[vk6zgo]

I Googled for a bit of information,& found the power supply for the 2235 is a switchmode type without an 50/60 Hz power transformer.

All my references to  a transformer in my first posting should be ignored.

The "divide & conquer" method is still viable.

If you disconnect the mains input to the switchmode the main fuse should not blow.

If it does,your SMPS is OK.

One problem with some SMPS is that they won't work correctly without their normal load--others will.

One person on one of the other sites removed the normal loads & replaced them with resistors.

Re the switchmode transformer--It is very hard to test these things with a DMM.A shorted turn will cause it to fail,but you won't see it with a resistance reading.

VK6ZGO

[SgtRock]

Dear VK6ZG0:

--When I referred "old time radio wanker, anything close will do" method, I guess I should have said "present company excepted" I was specifically talking about some of the curmudgeons at the "AntiqueRadioForums" who routinely advocate some rather bizarre substitutions for resistors, caps and even tubes. I should add that most of the substitutions they recommend will work after a fashion, for an All American Five, but who knows if they provide peak performance or not. I have repeatedly seen them advise substitution of caps based on the label and not measurement. I certainly did not mean that Ham Radio Enthusiasts are "Old Time Radio Wankers". Folks that build and repair their own Xmitters and Receivers, are a lot closer to engineers than part swappers. I sincerely regret any distress my overly broad characterization may have inadvertently caused. I hope that is 5X5.

"Marconi invented the radio but he had to wait years and years till anything decent was on." Johnny Carson

Best Regards
Clear Ether

[vk6zgo]

That's cool,Sarge,no offence taken!

I would say,though,that most times I've had to make component substitutions have been when I was being paid for my time,rather than in my hobby.
In many cases in TV & Broadcasting,the important thing is to get the thing back into service,& with the manufacturers &  original parts
supply on the other side of the world,intelligent substitutions become essential.

Another thing is that OEMs sometimes make horrific misjudgements,particularly when dealing with power circuits,so mods become necessary,in order to increase reliability in service.

By the way,sorry for my snide comments in the Solar/alternative power thread.

VK6ZGO

[requim]

@vk6zgo - would you mind telling me how to disconnect the mains input from the SMPS?  I'm new to this whole game and from my perspective the mains input connects directly into the SMPS (or is a part of the SMPS).

Just for kicks I put a new fuse in and turned it on to see what would happen since I've removed a bunch of caps.  Unexpectedly it still blew a fuse, which I wouldn't have thought would happen since I assumed with the capacitors missing it would just go open..  Not sure what that tells me.  A short prior to any of the capacitors being removed?

[requim]

Looks like I tracked it down.  Well credit goes to an earlier poster.  It appears the mosfet Q9070 is toast.  With any luck that's the only thing wrong with this.  I just placed an order w/ Digi-Key so with luck I'll have the parts by Tuesday or Wednesday.

I'd still like to know how to disconnect the mains for future reference.

Oh I do have one question.  So when I was putting all the caps back on the board I realized something was amiss.  In the schematic and the parts list it lists C956 as a 270uF Electrolytic Cap +100-10%, 40V.  However I didn't have that cap in my bucket of parts from when I stripped the caps off the power supply.  I skipped it and when I was finished with the rest of the caps I still had a single 840uF 10+1000%, 12V Electrolytic cap.  I rechecked the board and where I stored the caps several times and couldn't find any empty spots on the board so I assume the 840uF was in C956.  Since I did not have a 270uF cap I put a  330uF 50V cap in its place.  My question is this -- Can someone please look at the schematic and tell me whether the schematic is wrong or whether I should double/triple check the caps on the board again?  I googled around and didn't find anything on it.  I do know that there are several misprints on the schematic from previous searches but this doesn't appear to be one of them.

Thanks!

[SgtRock]

Dear Requim:

--Doing the math, any cap that ranges between 243 uF and 540 uF will meet the specifications. So if you measure your 330 uF cap, and it measures 541 uF, is does not qualify.

--If you cannot measure the 330 uF cap, you should by a 270 uF cap. I like to measure everything before installation, as I have, from time to time, had new components that were out of specification. Also, you might want to review my earlier remarks. I hope someone can verify the value from the schematic for you.

--I hope this helps. That is a great scope.

"My favorite things in life don't cost any money. It's really clear that the most precious resource we all have is time." Steve Jobs
 
Best Regards
Clear Ether

[requim]

@SgtRock - The Panasonic 330uF 50V cap I put in is spec'd +/- 20% so assuming Panasonic doesn't have any major issues with quality control the cap should be well within spec.  I'm just new to this whole game so I'm extra paranoid about substitutions. In the future I'll verify the values of the caps before soldering them.

[SgtRock]

Dear Requim:

--The Panasonic 330uF 50V cap +/- 20% should be just fine, barring the one in a thousand or so chance of a bad one.

--I am sure everyone would like to see some pictures, if you
have the time and a camera.

"I have had my results for a long time: but I do not yet now how I am to arrive at them."
Carl Friedrich Gauss 1777 1855
 
Best Regards
Clear Ether

[requim]

I have both and will oblige.  Didn't seem to be much interest in my posting of the Lambda LC-512 I tore down and restored so didn't see much point posting photos of the scope. 

I also tore down and fixed a Harman/Kardon receiver and HP e3610a DC Power Supply but didn't take photos because I didn't think anyone would be interested.  I am currently working on two HP 204C Sine Oscillators and will soon be working on a Agilent 6611C DC Power Supply and HP 34401a DMM.  If you want pictures posted of any of my current or future projects just let me know which ones and I will oblige.  Oh and I will probably be taking apart my old Yamaha receiver to figure out why the headphone jack suddenly went down a large number of decibels.

[FenderBender]

Yeah as others have said, some of the old style wet tantalums tend to go bad. I've seen nothing but Japanese or American caps in Tek stuff, so shouldn't be an issue of the electrolytics.

[requim]

Here are some pictures as requested:
















The culprit:




If you want to see them in full-resolution you can find them here:

http://www.flickr.com/photos/54406519@N08/sets/72157627955284523/

[hacklordsniper]

I have both and will oblige.  Didn't seem to be much interest in my posting of the Lambda LC-512 I tore down and restored so didn't see much point posting photos of the scope. 

I also tore down and fixed a Harman/Kardon receiver and HP e3610a DC Power Supply but didn't take photos because I didn't think anyone would be interested.  I am currently working on two HP 204C Sine Oscillators and will soon be working on a Agilent 6611C DC Power Supply and HP 34401a DMM.  If you want pictures posted of any of my current or future projects just let me know which ones and I will oblige.  Oh and I will probably be taking apart my old Yamaha receiver to figure out why the headphone jack suddenly went down a large number of decibels.

A teardown which will not be interesting does not exsist 

[SgtRock]

Dear Requim:

--I agree with Hacklordsniper, and uninteresting tear-down does not exist. And in the case of very popular pieces, like this one, people may even, go back some time later to view the pictures, so they can confirm things about their own equipment. Hence, it is always nice to have:

(1 Exact name of the unit in the title line, as you did here. So people can easily find it, sometimes much later, maybe even yourself.
 
(2 Lot of pictures. You could increase the resolution a little bit on yours, if you wish. DJ does not stint on bandwidth, and I think, he likes to look at the pictures himself. Also a bit more light, next time would be nice.

(3 Also, a complete list of all parts, including fasteners, along with multiple sources. Had you didn't I. Snork!

--I sincerely appreciate the trouble you have taken, and I look forward to seeing her up and running.

"You know that I write slowly. This is chiefly because I am never satisfied until I have said as much as possible in a few words, and writing briefly takes far more time than writing at length."
Carl Friedrich Gauss 1777 1855
 
Best Regards
Clear Ether

[requim]

@SgtRock - I posted the full resolution images to my flickr account.  I didn't think most people would like it if I linked in to 10 2 mb photos into the page.  Not everyone has uberfast bandwidth.  This site restricts images to 1 megabyte in size and 10 attached images per post.  So in some ways it was easier to just post it to flickr and link to it.

As for lighting, I can't do much about it for the time being based on where I'm staying/living.  But I agree.

[SgtRock]

Dear Requim:

--I stand corrected. Very satisfactory.

"Reality is merely an illusion, albeit a very persistent one."
Albert Einstein 1879 1936

Best Regards
Clear Ether

[lowimpedance]

Hi Requim
 When servicing any gear where removing a number of components or connectors /wires etc I either draw a rough pictorial style diagram to show where the connections/components go with colors or values etc, even when I have a service manual. Or these days a number of photo's is allways useful.
 One other thing I have one of these scopes and have on two occasions had the IEC inlet filter smokem up. The original spec'd tektronix component was not rated properly for the mains here in Australia, (240V, component rated at 220 or 230 from memory so marginal), and was a known problem. I have replaced it with my 'chosen' substitute with no further problems.
 The thing to note with service manuals even ones still on the manufactures websites for obsolete gear is changes may not have been reflected in those versions of the manual.
Case in point I have a HP1741 CRO for which I also have the manual that came with it when new and there are many changes to the circuits either in values or complete new designs which for some have been supplied with manual changes sheets however the description of one key cct that's been completely revised has not been changed to reflect the new cct design.
And yes downloading the manual of the Agilent website does not even show some of the component value changes let alone give any details to the newer designed cct.
 Sorry to ramble but the point is whats to say that the cap you had left over was not the correct one for the position you indicated in the supply.

Incidentally in my 2235a manual C956 is a 10uF cap, C957 is a 270uF cap on the schematic, both on the 30v line forming a PI filter with the 33uH L956.
(If I am looking at the right part of the cct that is!!!!)
I also looked through the manual change info at the back of my manual but did not see any reference to changes of this component number.

Anyway was it only the mosfet switch?? , nice CRO which you will enjoy using for years to come.
regards
john

[requim]

So I put the new mosfet in, braced myself for it, and had a big letdown when the unit, while it didn't short out anything, didn't turn on either.  No lights, nothing.  Could the mystery capacitor be to blame?  I probed a few points on the board and saw current is flowing.  I plan on looking in the service guide and trying to probe various points where they give known ranges/values and seeing if any of them match.

As I said, no apparent short circuits and no blown fuses.  One step at a time I guess. What I am surprised by is that before, when it shorted instantly a couple lights on the board light up ( I was able to capture this using high speed cameras) but now no lights on the board turn on.  I suppose I should check my solder joints for the caps I desoldered/resoldered.

[lowimpedance]

I have not read all the replies, but has anyone suggested to feed 40v dc from the connector T948-3.4 from the pre reg. output which goes to the main switch mode supply (that actually powers all the rails of the cro. That will tell you if the cro itself is working including the main supply and the problem is with the pre reg.
I repaired a 2215 which is similar and I was able to isolate the problem to the pre reg supply in this way (proving the cro worked). So if you have no 40v (or so) at tp940 or at R907, and you have a bench supply that can be adusted to simulate the output of the pre reg you should be able to test whether or not the main supply is okay and indeed the cro is working.
Have you measured that test point and have the 40 or so volts?.

The problem with the pre reg in the 2215 was the transformer t906.

I guess you know to be very careful when working around these area's

John
 

[spec]

I know it's been around 3 years since the last post on this topic but I am intrigued to know if the fault was located and fixed. It seems to me from the posts that the blown power FET may not have been the cause but a symptom.  By the way, Tekfan informed me that, "The difference between the [Tektronix 2235] A and non A versions is an improved power supply in the A version. There's a factory mod for the non A version which involves replacing a transistor with one with better specs and adding a few diodes and resistors. ". This may have some bearing on the matter.

[M0BSW]

Nicely said. However even out of circuit capitance measurment will not tell you if a cap is bad or not most of the time, you need an ESR meter, possibly one that can measure in circuit.

I use one from GME technology and it has proven itself very good

Although by the time you've taken it out you may as well replace it anyway...

Well said, If it was a  refurbish job I would anyway, at least that way you shouldn't have to do a revisit for a long time, if you replace with decent caps.