Tektronix 2221A focus problem

[wolfalex]

Hello,

this otherwise wonderful Saturday did not start 100% well. After some weird behaviour of the Focus setting on my Tektronix2221A yesterday evening today the Focus decided to go
bad completely Yesterday I was able to get a wonderful focused CRT picture somewhere in the middle of the focus pot range..The only weird behaviour was that the Focus setting did not stay, and it was wandering around. I mean for example I set the Focus perfectly, and 10 seconds later the screen was out of Focus again. But with constantly tweaking the Focus Pot on the Front of the scope I was able to regain a nice picture until the focus drifted again.

But today after 1 min of operation of the scope the Focus did go completely. Now the Focus pot has no influence anymore, and I get only a 2.5 inch wide band of light on the CRT screen

I read in the internet about a famous Focus fault, regarding some 510k resistors in series which go bad over time, (the go open) and then the Focus has problems. Could this be the problem? Is this also the case in the 2221A?

It would be so wonderful to get the scope working again, I really like it. Thanks in advance for every information that I get,

 Alex

[wolfalex]

I am currently on my way to double check the resistors for the Focus. Unfortunately there is no Service manual for the 2221A in the internet, but somewhere I read that the 2230 should
be from the same family. Therefore I am using the 2230 service manual as a reference. First I had to flip the Storage Board into the service position. Because I have never repaired a device complicated like this in construction I'm a little bit nervous But after flipping the storage board out, and opening the protective covers from the power supply unit section I am now ready to measure these resistors.

I added some pictures from my scope if somebody is interested in it. I am sorry if the picture quality is not that great, but I am not sure how big the uploaded pictures are allowed to be in size. Therefore I decided to start with a not so good quality.

Next step will be to discharge all the caps inside the unit to be sure that I don't get a lethal shock, or my fluke 87 of course I can already see that somebody did the same repair before me in the past, because the resistors pads are full of old solder flux remainings

Does somebody know what type of resistors are used for the Focus? I mean the 510kohms ones which I have to change maybe? Are these normal standard troughhole resistors, or are these special ones? If I solder around in the scope, I read somewhere a special silver solder wire must be used to prevent some kind of damage in the scope? Is that true or can I use just a standard solder wire?

Thanks,
Alex

 

[SeanB]

If you are going to do one do all 6, as if one is drifting all will be close to it now. That can be done with regular solder, the silver loaded solder was reserved for some really hot running resistors that would with time melt regular solder.

[wolfalex]

Hi SeanB,

 thank you very much for your information regarding the solder wire. I am really happy that I do not have to buy a special one. Do you maybe also know more about the type
 of replacement resistor I have to use? Are these just standard ones, or special ones? Yes, I will change all of them, only to be sure

 Thanks,
 Alex

[SeanB]

I would suggest using 0.6W metal film resistors there, and space them slightly off the board so they do not touch each other. Buy some RD  506-5636, 683-3809 which will fit direct, or buy some 683-5534 which are 1W resistors that will need a slight adjustment on the spacing so they do not touch, but which will run cooler.

[wolfalex]

I measured all the resistors which could be the culprint, here the list:

R886 = 176.6 k measured (nominal 180k)
R888 = 512k measured (nominal 510k)
R889 = open (nominal 510k)
R890 = 566k (nominal 510k)
R891 = 572k (nominal 510k)
R892 = 5.95 Mohms measured (pot?, therefore R892 open), (nominal 510k)
R894 = 523k measured (nominal 510k)
R893 = pot endpoints = 5.95 Mohms measured (nominal 5 Mohms)

I am really happs now, because it seems I am on the right track to get this wonderful scope back. It seems at least 2 resistors are completely open, 2 others have at least drifted. Also it seems the pot is not 100% happy anymore, it is already 1Mohms higher than nominal. But I guess it is not easy to get an replacement, and therefore I guess I wont swap it. 

Do you think I should also swap the 180k resistor? Value wise its ok, but if there will be also more drift in the future I want to change it now. I am sure that I will swap all of this 510k resistors.

@SeanB: Thanks for this really helpful info on the resistors. I think with a probability of 80% I will use the 0.6W resistors, but at least I will try to do  the 1W option, because otherwise I probabliy have to change the resistors again in future.

[David Hess]

I have always wanted to see the 2221 and 2221A service manuals but I have found no free scans online and I try not to buy manuals for test equipment that I do not have.  The Tektronix 2232 is closer to the 2221A than the 2230 and there is a free 2232 service manual online.  All of these oscilloscopes use the same basic design though so any of the manuals will work for a focus problem including the 2235.

The older oscilloscopes in this series include an automatic focus circuit at the -8.6 volt side of the focus divider but it can be ignored.  I assume the later ones used a better CRT which did not require this.

The resistors are 5% 1/2 watt film resistors but the voltage across each 510k resistor is about 320 volts which is marginal for most 1/2 watt film resistors explaining why the focus resistor string is not as reliable as it should be.  Carefully check the voltage rating of the replacement resistors or use something like the Vishay HVR37 series which are high voltage.

[wolfalex]

Hello David,

 I really appreciate your information about the resistors and about the scope, thank you very much. I already ordered 2 sets of resistors. When the arrive I will decide which set
 I will use. One set are the HVR37 resistors which you have mentioned (a lot of voltage, but only 0.5W), and the other set has more power capability, but only a maximum pemissible voltage of 500V.

If I take your figure of approx. 320V over a 510k resistor I get approx a power dissipation of 200mW, which with the 0.5W resistors give a factor of approx. 2. This also seems a little be close, or? Or is a factor of 2 more then enough to get a longer lifetime? The same problem with the other set, very high power (2W), but voltage safety factor slightly under 2.

So I think it will be difficult for me to decide which ones to take. What do you think? I have the feeling you have much more experience with this kind of resistor application than I have. I never repaired something near a CRT

Thanks,
Alex

[David Hess]

I would usually use the highest power part that will fit but in this case there are voltage derating and safety considerations which are uncommon in most resistor applications.  It is not immediately clear from the part numbers given in the service manual what kind of resistors Tektronix used or why.  The failure mode of going open is consistent with long term high temperature operation however.

I did some more research and as far as I can tell, Vishay ended up with the product lines of Mepco/Centralab and I did find evidence that the 5053CX series of resistors become the Vishay CR37 series of high voltage resistors.  Older oscilloscopes in the same series like the 2215A and 2235 used Allen-Bradley CB and EB series carbon composition resistors rated at 250 and 350 volts respectively.  The implication is that the resistor was selected based on failing in an open state instead of shorting like film resistors tend to do.  Based on that, I would use the CR37 resistors and *not* a general purpose film resistor even if it is a higher power replacement.

The three common CRT circuit failures are focus resistor chain, vertical output load resistors which fail open do to extended high temperature operation, and the DC restorer circuits which have to hold off the high cathode voltage of about -2000 volts but that is a capacitor and/or diode problem.

[22swg]

I had to replace all the same ladder of 510k resistors in a Tektronix  2235 , I just used 1 W Metal film , focus is fine  but tends to drift if on a long time. Old resistors seem to be carbon , but were not as per colors, crazy values.

[David Hess]

High temperature over time could have discolored the resistor markings.

None of my oscilloscope have their focus drift so as I would notice so I might suspect there is some other problem with either a supply voltage or the astigmatism.  Their brightness does change a little bit until they warm up though.

[22swg]

Point of interest , I see some Tektronix... scopes have a fan , should i fit one to my 2235 ?

[David Hess]

The fan was needed for the 2235 models like the 2230, 2232, and 2236 because they draw more power to support their extra circuitry.  You can certainly add one and it will keep it cooler probably making it more reliable but it was designed to operate without it.

There were a couple of different fan implementations.  The 2230 used a custom Tektronix 12 volt 2.4 watt 6500 RPM 37 CFM tubeaxial fan running on 8.6 volts.  The 2232 started out with a similar custom Tektronix 1.5 watt 4600 RPM 19 CFM tubeaxial fan which was replaced by a practically identical model at least as far as specifications which I think was made by Sunonwealth which I expect became Sunon.  The 2236 used the +5 and -8.6 secondary winding to generate 13.6 volts and a 12 volt 2.4 watt 6500 RPM 31 CFM tubeaxial fan made by Buehler which was later replaced by a 12 volt 0.95 watt model apparently made by Tektronix.

I think the implication is that if you add a fan, you should copy the 2236 implementation and use a 12 volt fan that draws roughly 1 watt or 83 milliamps.

The 2232 Sunon fan looks to be 60mm frame size and had a support bracket.  I do not know the old Sunon parts numbering scheme to be more detailed:

http://www.qservice.tv/vpasp/shopexd.asp?id=9698

Note that I just looked up all of the above in the respective service manuals and only because I have several 22xx oscilloscopes and was interested in the answer.  The same goes for the focus resistor answer.  I do not have any special knowledge about the 22xx series.

[wolfalex]

Thank you very much for your research. Maybe you deem yourself not as an specialist in the 22xx series, but I have already learned a lot from you about this kind of Tektronix scopes:)
 
 I also thought power and voltage rating is all which goes into the choice of the resistor in this case, and I again was surprised: I would have never thought that the kind of failure
 mechanism differences between the different parts could influence the selection which part to use. But it makes sense, I guess a shorted resistor in this circuit would lead to a lot
 more damage than something simply going open.

 Should I also double check other resistors if the scope is already open anyway? Because you told me there seem to be 3 famous failure modes of this scopes at least...




 

 

[David Hess]

Thank you very much for your research. Maybe you deem yourself not as an specialist in the 22xx series, but I have already learned a lot from you about this kind of Tektronix scopes:)

I just wanted to be clear that I claim no special knowledge.

Reverse engineering such well designed test instruments is fun and I keep several 3 ring binders with documentation and notes.  None of my 22xx oscilloscopes have suffered any of the common problems so in a way I have missed out.  I did repair a 2213 for a friend that suffered from the vertical CRT load resistor problem.  My first 2230 had a very unusual problem that took weeks to track down.

I also thought power and voltage rating is all which goes into the choice of the resistor in this case, and I again was surprised: I would have never thought that the kind of failure mechanism differences between the different parts could influence the selection which part to use. But it makes sense, I guess a shorted resistor in this circuit would lead to a lot more damage than something simply going open.

Carbon composition resistors were often used for their high peak power capability and safe failure mode.  Film resistors tend to short out.

The working and peak voltage rating for resistors is often overlooked just like ripple current rating is overlooked for capacitors.

Should I also double check other resistors if the scope is already open anyway? Because you told me there seem to be 3 famous failure modes of this scopes at least...

I might do a visual inspection and in circuit test if I saw something suspicious but I would not preemptively change anything unless there was evidence of a problem.  The vertical CRT amplifier load resistor issue is more common on the early models which used 2 resistors in series instead of 3 to distribute the heat.  DC restorer problems are difficult to diagnose so there is really nothing to look for.

[wolfalex]

It was some months ago, but maybe you can still remember..one of my first posts in this Forum was a question in which test gear I have bought at that time on Ebay I can find lithium batteries, and which of them have to be changed as soon as possible. It was a very important question for me at that time because of my 3478a where I also had to swap the Bat.

And at that time you have given me information about a possible Lithium battery in the 2221A in combination with the GPIB option and the waveform storage memory. I was not able to find a battery in the schematic in the past, and so I thought there is maybe nothing in there, but now during the search for the focus fault in the scope I was able to find it

It is very well hidden below the storage board. I mean as long as you dont lift the storage board out. If you lift the storage board into the service position the battery can be easily seen. But I dont think I will change it, because what I can remember you have told me in the past that it is only for the waveform storage function, and not for calibration itself. And for the calibration I have seen a lot of pots on the PCBs itself. And If the waveform storage does not work anymore one day it is not a big deal for me.

I really like the look of the storage board, I think this is one of the most complex boards I have seen for a long time. At least it looks complex because of the high amount of logic devices. Todays stuff is way more complex I am sure, but now all is hidden in FPGAs, ASICs and so on Especially nice are the hybrid ADC modules. I only hope the do not break, because if the do I guess it is game over.

My scope has also the GPIB option, and I hope one day I will be able to connect to it. GPIB is the last untested part of this scope, and I do not know if it works or not. I added some more pictures of my unit, maybe it is interesting for someone

[David Hess]

It was some months ago, but maybe you can still remember..one of my first posts in this Forum was a question in which test gear I have bought at that time on Ebay I can find lithium batteries, and which of them have to be changed as soon as possible. It was a very important question for me at that time because of my 3478a where I also had to swap the Bat.

I remember that discussion.

And at that time you have given me information about a possible Lithium battery in the 2221A in combination with the GPIB option and the waveform storage memory. I was not able to find a battery in the schematic in the past, and so I thought there is maybe nothing in there, but now during the search for the focus fault in the scope I was able to find it

It is very well hidden below the storage board. I mean as long as you dont lift the storage board out. If you lift the storage board into the service position the battery can be easily seen. But I dont think I will change it, because what I can remember you have told me in the past that it is only for the waveform storage function, and not for calibration itself. And for the calibration I have seen a lot of pots on the PCBs itself. And If the waveform storage does not work anymore one day it is not a big deal for me.

I think the 2232 battery is on the underside of the storage board which is why I originally missed it as well.  There is no calibration data to be lost but there are some configuration settings like the acquisition mode.

I really like the look of the storage board, I think this is one of the most complex boards I have seen for a long time. At least it looks complex because of the high amount of logic devices. Todays stuff is way more complex I am sure, but now all is hidden in FPGAs, ASICs and so on Especially nice are the hybrid ADC modules. I only hope the do not break, because if the do I guess it is game over.

The storage board looks identical to the 2232 storage board.  The 2221A came out one year later than the 2232.  As far as I know, the only differences are the lack of the delayed sweep timebase and the REF button under SETUP, and of course different firmware.

My scope has also the GPIB option, and I hope one day I will be able to connect to it. GPIB is the last untested part of this scope, and I do not know if it works or not. I added some more pictures of my unit, maybe it is interesting for someone

My 2232 has GPIB as well and I have not used it either.

[wolfalex]

I would really like to use GPIB in the future, but I dont have an interface yet. Maybe I find a cheaper one in future, because the average price for them is slightly too high for me.

Another thought about the scope... should I make a backup of the ROM chips in the scope, or in my other older test equipment? Is a ROM problem/defect a really uncommom
 problem, or is the probability high enough to better do something about it? When I read about old test equipment  in the internet from time to time I run into stories about "ROM got defect", or I read very often that people did a backup of their ROM chips.

Problem is I do not have the necessary equipment yet, and I also fear that maybe with the "take device out, make backup, and put it back in" something can maybe break during the process. And I am not only talking about putting the device wrongly into the programmer :)This question is not only interesting for the 2221A, but also for the HP 3478A, 5316A, and the 3325A I have now.

[David Hess]

The digital display image can also be transferred over the 2221A serial interface option in the form of HPGL or digital data.  I do not know if the 2232 and 2221A communication option boards are compatible.

The ROMs which seem to be unreliable are the mask ROMs from Mostek and PROMs *not* based on anti-fuse technology.  Floating gate memory like EPROMs seem to do much better.  Backing them up though is a pretty trivial exercise.  Maybe find some random EPROMs to test your procedure with first.

[wolfalex]

Hello,

 after I got the resistors and I had some time left on saturday, I was finally able to fix the scope again. I had two sets of resistors here, and after a difficult decision I decided to use the
 HVR37 series resistors for the fix. After swapping all 6 resistors, and putting everything back together everything works now again I am really happy about the outcome, and I
 hope I will never have to change this resistors again.

 I want to thank all the people again which gave me information here in this forum, it really helped a lot, and without the tipps it would have taken so much longer to bring the scope
 back to life.

[elliott]

My 2221A just did this today, R889 open.

The other day I got a hold of the programming manual for the GPIB/RS-232 interface and decided I'd get that working, which I did after replacing the level shift IC, with the old WaveStar for Windows 3.1. Today I was going to play around with GNUPlot and see if I could get it to plot the binary waveform data and when I turned it on the focus was gone.

Have the HVR37s been working out well for you?

[elliott]

They are a bit fatter than the originals and I wasn't happy about them touching, so I staggered them to give them a better chance of staying cool.

[wolfalex]

Hello elliott,

 I am very very sorry for my late answer. I only saw your post today by accident, because I had not much time lately for reading on the internet. Yes, until now the HVR37 are holding
 nicely. But I also must say, that the scope was only turned on for some hours until now since my repair job..But no smoke has until now escaped Yes, the new resistors are
 somehow bigger than the old ones, but I tried to keep some distance. But I am not sure if I did a great job on that.

 I am very careful with the scope anyway, I will only connect it to the power grid and especially power it on when I am next to it. I would never leave it alone, especially because of
 older tantal caps

 I think I read once on the internet that somebody mentioned that someones place almost burned down because of an older Tek measurement device/scope and an older tantal cap 
 which died.

 How did your repair go? Did you get GPIB working? Sorry again for my late answer and therefore not really helping you

 I'm happy now, because now I know that there is at least a second person here with a Tek 2221a

 

 

[elliott]

How did your repair go? Did you get GPIB working? Sorry again for my late answer and therefore not really helping you

The focus has been great, seems to stay stable with temperature as well.

Mine is RS-232, not GPIB, it is working sometimes. I haven't checked too deeply to find out what is causing it to fail. Sometimes I'll send it a command and everything works great, sometimes it responds with garbage and then ignores further commands. They are fine up to the UART chip, beyond that I don't know what is happening. This might be motivation to pick up a logic analyzer.

When it works, the WaveStar software is usable. I want something Linux native if I get it reliable though.

I'm happy now, because now I know that there is at least a second person here with a Tek 2221a

The 2221A isn't the fanciest thing around, but it is a solid scope. I've had it for a few years, but never dove too deep into the digital stuff until the last 6 months or so.

[David Hess]

I am very careful with the scope anyway, I will only connect it to the power grid and especially power it on when I am next to it. I would never leave it alone, especially because of older tantal caps

I think I read once on the internet that somebody mentioned that someones place almost burned down because of an older Tek measurement device/scope and an older tantal cap which died.

Tektronix was always good about including power supply protection in the form of fold-back current limiting or an over-current shutdown so that in the event of failure, no extra damage would occur.  I am not aware of any tantalum capacitor failures in the 22xx series.