Tektronix 2232 scope.... in need of a doctor..

[wasyoungonce]

wasn't there a guy over at Yahoo Tek scopes re-winding these?

[Vince]

Is there really ? Taht would great !

Not sure how active this old Yahoo group might be today, given the popularity of the EEVBLOG forum ? Will have to join to find out (can't even read past messages un til I subscribe).  Unfortunately for me, I keep clicking on the "Join" button, but nothing happens at all, the web browser does not even seem to be trying to reach any server, it just does nothing at all... not good 

Anyway, been thinking of rewinding the coil myself actually (nothing to lose after all), if just for the sake of seeing if I could even manage to do it. Sadty I fear there is not much point in doing so, given that the case is split and cracked... unless this outer/visible part is only there for show, doesn't not really matter, and the actual ferrite core is inside, and might still be intact...

But first I must get when of these "Dremel" miniature tools, if I want to have a remote chance of opening it up without destroying everything... all I have is an angle grinder and hacksaw, not really the most precise tools for that job !

Other than this, last night I ordered all the (other) parts I need to fix the regulator. Figuring out replacement for the MPS2907A trannies and FDH9427 diode was easier than I feared. Trannies were just regular 2907A which I should have guessed from part number... the PNP companion of the famous 2222 NPN. Now they make a PN2907 version with just the same specs.
FDH9427 diode, marked as "super fast" in the service manual, believe it or not is still super fast by modern standards. I did a parametric search on Farnell's site and 4ns is about as fast it gets even today. The 1N4151 has spot on the same specs : 4ns, 2pF, 150mA continuous, and a tad more head room with regards to reverse voltage, 50V instead of the original spec of 40V. Can't be a bad thing.

The few resistors I needed will get an upgrade.... just out parts availability, the 5% will get to be 1%, and the 1% jobbies of the precision voltage divider setting the current limit threshold, will be 0.1% for the lower resistor, and 0.5% for the upper one, from memory. Can't be a bad thing.

Even the mundane 39ohm resistor driving the gate of the FET will get an upgrade from 5% to 1% again because no choice and not much cost penalty anyway. Won't make any difference whatsoever to that part of the circuit though, obviously...

[wasyoungonce]

Vince the Tek group is very active.  You will even find guys and articles by members who worked for Tek in that era.  Also a Google search for "rewinding Tektronix transformer" brings quite a few hits as well as "rewinding Tektronix pre-regulator transformer".

Most are hits for HV transformers but you get many for the pre-reg txfmr.  Of course this PSU is similar on all 22XX scopes.  I think its the 120-1439-00 or 120-1469-01 transformer.  Would be nice to find a blank bobbin and make a new one...except these are probably all gone as well.


Edit:

Ummm Sphere have qty 2 used 120-1439-00 transformers avail.  I think these are pretty much the same???.  Mine is a Tek 2236 and this part is on my parts list...120-1439-00.  Even if it isn't compatible...and I'll eat mu hat if it isn't...then you can use its bobbin for a re-wind!

http://www.sphere.bc.ca/test/tek-transformers.html

[Vince]

OK then, will try and get my computer to cooperate and subscribe me to this Tek Yahhoo group...

I am wary of getting a different part-numbered coil from another 22xx series scope because... there must be a reason why they don't carry the same number, and unless I know what it is exactly, I can't make an informed decision... Sure it looks the same from the outside  (I guess so anyway), but are the inductances exactly the same, same number or turns for all 4 windings ? Are all 4 wires the same gauge ? A bit risky/hasardous at the least.
Since I don't have reference measurements for the resistance and inductance and wire gauge of all 4 windings of the 2232 coil, I can't compare to other 22xx coils in any meaning/useful way...

But I will definitely join this Yahoo group then, since it's still active and there are knowledgeable people there, and since this coil issue is apaprently very common. People there might have data on the various coils and know if I can actually use a 22XYZ coil on my 2232, and maybe some people already have done it and can report...

Problem as always is that buying from the US is prohibitive for me being in Europe. Probalby economical to actually wait to find a donor 2232 locally and salvage the parts I need.

As you said though, if I can find a coil for a reasonable amount of money once shipped, then yes even it's not  the correct part number I could still use it as a blank, then I can do the autopsy of my broken so I can replicate the exact same windings.

Anyway, great to see that it's actually technically possible to rewind them (I guess the two halves of the body are just glued together then, easy to put back together ?). Will practice on my broken unit and join that Tek Yahoo group to benefit from their expertise on the subject...

In the meantime I will replace all the other damaged components, and then after that, fixing or replacing the coil might take several months maybe, or more who knows, depends how lucky I get. But I thought what the hell... I can still get the scope going in the interim, by putting together a 40V power supply and strapping it to the rear panel of the scope...

[Vince]

OK done, subscription to the Yahoo Tek group is being processed by the moderator, waiting for the green flag...

[David Hess]

Rewinding transformers from tube era oscilloscopes is common because fewer spares are available.  There are lots of 22xx parts donors available and as far as I know, the -00 and -01 transformers are interchangeable.  Incidentally, the 4th winding between chassis and earth ground was a clever last minute addition for EMI suppression.

QService is a good place for Tektronix parts if you are located in Europe but they do not have any of these transformers at the moment.  They have a 2213A main board available but given the cost, I would look for a cheap parts donor.

TekScopes@yahoogroups.com is very active and is completely accessible via email which may be preferable to the web interface.  TekScopes2@groups.io made the jump away from Yahoo not long ago and is worth checking as well.

The small signal transistor types are not critical.  2N4401s and 2N4403s are suitable replacements.

I know of at least one person who just replaced the preregulator with a 48 volt off-line power supply module which could be adjusted to 42 volts.

[Vince]

Thanks for the pointers Dave, will check that Qservice site regularly. They say you can e-mail them your request because not everything is listed on their site... will do !

Rewinding has two advantages for me : not being in the US with plenty of Tek parts available for cheap (no duty taxes no VAT no outrageous shipping fees), rewinding is a way to get my scope fixed in a reasonable time frame, rather than waiting a couple years hoping the right par comes along in my neck of the woods. Second advantage being... I have never rewound a coil/transformer and find it cool... resurrecting a transformer, I like the idea. My broken Tek just gives me a good "excuse" to get some hands on experience on the subject, and satisfy my curiosity...

Most probably what will happen is that I will eventually a coil at a decent price, within a few months, which I will fit, but whilst waiting for this to happen I will "play" with my broken coil 

[Neomys Sapiens]

Just noticed that no one pointed you to THIS document:

[Vince]

Thanks for that.  It helped me a bit with regards to the way the PWM choice is powered. Looking at schematics, I could not make heads or tails of it.  I was assuming there would be some kind of crude voltage divider of some sort to lower the voltage for the sole purpose of powering the PWM chip, but  with this assumption then I can male no sense of the schematics. The ground pin of the chip is being connected to the 40V rail... and a diode feeding that back upstream humm... was starting to think of some kind of boot-strapping mechanism but hardly familiar with this kind of stuff... At least this document confirms that indeed I was on the right track. Still need to figure out/analyze exactly how it all works in practice...

[Vince]

Dave, I am checking Qservice's site and unless there are several similarly named companies bearing the same name... then it looks like they are not operating in Europe sadly for me. Their "About us" page tells their story but doesn't even say where they are located ! However their "Contact us" page mentions a street address which is clearly in the US... Los Angeles California...
Still, with Sphere and Qservice, I now have two major parts suppliers that I can check regularly, that's good... gives me hope, in the long term at least (go figure when they will have the magic part available again...).

[David Hess]

When I ordered parts from QService, they shipped from Greece.  The padded envelope they used for packaging had a very strong smell of tobacco which reminded me of my grandfather.  Their EU web page shows a Greek address.

Skevou Zervou 10
Rhodes Island 85132 Greece

[Vince]

Oh, thanks for that ! 

Glad they operated in EU too !   

The site I found a minute ago was this one, probably their headquarters :

http://www.qservice.tv/vpasp/shopcustcontact.asp

I have already contact them, so I guess they will redirect me to their US site.

Greece what a good idea... that country is collapsing completely, so probably the cheapest/most profitable place in Europe to locate their EU operation !

Greece hum... make sense now : a few days ago when I searched Ebay to replace the little plastic knobs that fell apart on the 2232, the only place selling some in EU was... from Greece !  And the pictures and everything were clearly very professional, it was clear someone was doing that for a living.

So that means they sell their parts on ebay, which I check for Tek stuff every single day, so hopefully when they eventually get the coil I need in stock, they will advertise it on Ebay as well, increasing my chances to catch it...
Obviously it would be daunting to create Ebay ads for every single part they have for sale, but since it seems this coil is a common failure, hence many people looking for one, it's worth taking the time to advertise it as it would sell quickly I guess.

[tautech]

@ Vince
If you do go down the transformer resurrection path this site might help with tips and techniques:
http://ludens.cl/Electron/trafos/trafos.html

[Vince]

Thanks, will have a read ! :-)

Will also see on Youtube if I can find something specific to the Tek coil.. everybody uploads to youtube these days, maybe some Tek aficionado recorded his coil rewinding, who knows...

I just posted on the Tek yahoo forum, the mandatory presentation. They required one, understandable of course, we are civilized being aren't we. So, a presentation they got ! Vince-style.... 9 pages long.   
 They will probably hate me and not be willing to help me as a result ! LOL  But well, a presentation they wanted, a presentation they received !     

[oldway]

You will need a new pot core.

It is a gapped 30x19 pot core.
I don't know how much is the gap but it should be easily be calculated from the data of primary inductance and primary number of turns.

[Vince]

Thanks for that !  So these things are actually standard parts one can use to do their own coils... how great !  Just checked Mouser, indeed these pot cores can be bought, and for cheap ! Brings a smile to my face...

Gap needs to be known but alos maybe also the specs of the material used for the core ? Looking at data sheets on Mouser, it seems to be an important information ?! 

I couldn't help, last night I tore the coil apart ! Sorry I am so impatient. No need for a Dremel in the end.... what looked like a plastic body was obviously as you know, the pot core itself, argh. So, obviously as it's brittle, no other choice but break it into bits in order to get to the windings. Was cracked in the first place any way !  It looked like it was held in place with some transparent orange glue of some sort... which too was very hard and brittle. Tried to soften it with my hot air station, hoping the core halves would come off nicely... but no joy what so ever.

Underneath the Tek sticker, bearing the part number(s), I noticed markings printed on the core itself, so most likely OEM information about the core... which would be useful to figure out it's specs I would hope !  But sadly didn't think of writing it down before "disassembly"/destruction... I just tried to reconstruct the core from the bits laying around on the bench... could only do partial job, but luckily looks like it's enough to read the markings, see below.

It reads:

GF 43019
16   00
 F   170

As for the gap, I can't measure inductance as I don't have an RLC meter and the core was broken to start with, but I guess that's something that someone with a 2232 could fairly easily measre on this own scope, and report. Will have to ask nicely on this forum and the Tek Group...

As for number of turns, that would require destruction so can't ask anybody for that... never mind since I did it on my own coil as I said. Data follows further down.

So, looks like there is hope... maybe I can actually get the parts to make my own coil, for a reasonable amount of money. Will keep digging the subject while searching in parallel for a used coil on Sphere Qservice and Ebay. I might get lucky you never know.  Would still find it fun to do my own coil though.

I would also need a new coil "former" (looks like that's how it's called, if Mouser is anything to go by ?), because mine as you see, is... not in good shape ! It's "tilted" like the old Pisa tower, probably due to overheating ?!... overheating which might also explain why the inner winding managed to sculpt grooves into the plastic ! 

So, I need to find a new coil former with the appropriate dimensions...

Then of course I need enameled wires (of 3 different gauges), but I guess that should not be too difficult to find... might be expensive though, copepr not being cheap, but I only need 4 meters wires total, after all... I don't need a kilogram of the stuff...

So, here are summarized my findings about the windings, and some pics of course, I love pics... reworked to be light weight of course...

Winding (pins)     | Wire      | Résistance | Number of turns       
Outer listed first | Diameter  |            | and orientation/phase, viewed from the top
----------------------------------------------------------------------------------
9 - 10             | 0.20mm    | 0.26 ohm   | 5, CCW from 10 to 9
6 - 7              | 0.20mm    | 2.18 ohms  | 50 (x1 layer), CCW from 7 to 6
4 - 5              | 0.45mm    | 1,65 ohm   | 159 (x6 layers), CCW from 5 to 4, green colored
1 - 2 (primary)    | 0.30mm    | Open/cut   | 151 (x5 layers), CCW from 1 to 2


In the end it was nothing extra-ordinary, I don't think it would be a problem to rewind it, even though I have never done it before. Obviously that's assuming I can find the appropriate pot core and former, but as we have seen, there is hope...

The primary was indeed open circuit/cut. I found that the cut/burn in the wire was located 15 turns in.

Don't why, but one of the 4 windings is colored green, go figure. Maybe because it's bigger than the other wires, some sort of wire gauge color coding ? Or maybe it's a standard way of saying "this winding serves this particular purpose". No idea... any suggestion ?

Anyway, what I found interesting (because I know nothing about SMPS and how all these secondary windings are used exactly...), is that unlike what intuition would suggest, the primary is actually NOT the winding using the thicker wire, as you can see from my table above.  The green secondary winding is by far the thicker, at 0.45mm diameter, and the primary winding is "only" 0.30mm.

Anyway, just thought the curious out there might be interested in this mini tear down 

[Vince]

Google just served me this interesting catalog from pot core manufacturer :

http://www.mmgca.com/catalogue/MMG-Ferrite-PT2.pdf

Indeed all these pot cores have standard sizes, how great. 30x19 is in there no problem.

One interesting thing is that the cores they sell, can be adjusted with a screw, to adjust the inductance. I assume they mean that it adjusts the gap.. adjustable gap, how practical  eh ? Well that's assuming that once adjusted, you Loctite it firmly into place so that the gap stays put ! ....

So that means if someone can just measure the inductance of their primary coil for me, I don't would not have to worry about the gap.. I would just adjust it until I the primary reads the same inductance. No ?

That being said, make me feel a bit nervous.. if the screw were to move at all, might make the coil overheat and the scope blow again ?! Noooooooo... please no.

That said... thinking about it... although the Tek coil had no screw in it, it did have the inside of the core (where a screw would go) filled with some white glue/gunk... maybe Tek too adjusted the gap "by hand', then locked the two halves of the core into place with that white gunk... so the idea might not be as silly as it sounds ?

Any way, food for thought at the least !

[Vince]

Hmmm, another interesting link, "Magnetics" web site, a core manufacturer :

https://www.mag-inc.com/Products/Ferrite-Cores/Ferrite-Pot-Cores

Looks like the " AL " value , the "permeability", actually defines what material the ferrite is made of. So, if there was way to figure that permeability  ? That would solve the problem and I could order the pot core... or,since they appear to be fairly cheap, I guess I could order one of each material/AL value, and try them all until the coil measures at the expected inductance ?

I mean, most of the work is wiring all the windings on the coil "former". But once this is done, you don't have to touch it ever again, you can fit whatever pot core on top of that, try 20 different core models if need be, not a problem I think ?!

[Vince]

Hmmm... guess work :

See below the page from the PDF catalog from 'MMG' which I linked to earlier.

Their table for the 30x19 core, lists a model with an AL value  of 1600... could it have something to do, with luck, with the "16 00" marking on the Tek pot core humm....   If it does then problem solved for the gap : table specifies it, it's 0.08.

[Vince]

Making some progress on the gap side... what do you know, the coil former just snapped !  No I swear... I did not brutalize it !

Was hardly my intention, but I guess it was destiny : that allowed me to see for myself the gap between the two halves of the ferrite core.

See picture below. They is a partial bead of orange glue between the two parts. That bead was actually squeezed, therefore we know for a fact that the thickness of that bead is an exact image of the gap that used to be in there. Unfortunately the bead is split, but there is one spot where 100% of the bead is on one core only. So, I proceeded to measure the bead at this spot, as accurately as humanly possible using calipers. Did all I could to get a decent accuracy, though obviously you can't expect much in these conditions. But still. Even without measuring anything, the first thing that catches the eyes is how large the gap is, many tenth of a millimeter, at least 0.5mm and under 1mm.   MY measurement falls in the range so I would say it's plausible.. though I still find it very high...

Here goes : I estimate it at 0.80mm tops, with a likely-hood of being a tad under that. 0.75 or maybe 0.70, but not much less...

I could maybe make a more accurate and reliable measurement at work, using a dial gauge, will see what I can do. But 0.75mm is the ball park figure any how.

Hey hold on a minute... looking at catalog page I uploaded just above, 0.70mm actually does exist ! Material code " F5A " and AL = 250.

What do you know...

Does any of what I said in the 5 previous messages make any sense to any of you ?? Or am I just talking BS again... God I need to learn the basics of coils... we didn't learn about this at school, and it's far away... I need to get hands on experience on real/practical stuff... like I am doing right now... learning along the way, thanks for your patience...

[Vince]

Just got a reply from Qservice about the coil.  I didn't get lucky, their physical stock does match their on-line stock : no 2232 coil available, all sold. Apparently the things sell like hot cakes...

That makes me want to really go ahead and rewind my coil.  Once we have figured all the practical details, will be nice to have a turn-key solution for the numerous people who will need that coil in the future. A "howto" guide with all the technical information needed (like my table above, a good start), what core to buy exactly, wires etc... so they can feel more comfortable doing it themselves.

[oldway]

It is very interesting to see how you spent your time to know in detail how this inductor was made.

I have always repaired this kind of failure by recovering the inductance on other non-repairable oscilloscopes.

There are two versions of this inductance, the 00 and the 01 .... I do not know what the differences are, but I think they are interchangeable.

These are the same inductors on the 2235, 2236, 2232, ....

Your findings confirm what I have always written in different forums: this inductor works  too hot, that's why it often ends up losing the insulation between turns or layers.

I always advise to install a fan in the 2235, as I did in mine.

From a ferrite quality point of view, I would choose a ferrite with the least possible losses at about 40Khz due to the overheating problem of this inductor.

It is also necessary to check whether this ferrite has a high induction saturation level.
But in general, ferrites saturate at about 2000 Gauss (0.2T) and this does not vary much.

You should also care about insulations between coils....Are there any ?

[Vince]

I have always repaired this kind of failure by recovering the inductance on other non-repairable oscilloscopes.
These are the same inductors on the 2235, 2236, 2232, .... 

Yeah, I check the local ads for Tek gear every single day.
Sorted by price, here are the Tek 22XX I can find right now :

https://www.leboncoin.fr/annonces/offres/pays_de_la_loire/occasions/?q=tektronix%2022*&sp=1

The cheapest one at 60 Euros, looking very sorry but sellers says it's working:

https://www.leboncoin.fr/informatique/1133453935.htm?ca=18_s

Next cheapest one is at 100 Euros, starting to get pricey for a donor scope...

https://www.leboncoin.fr/equipements_industriels/1148674764.htm?ca=18_s

But these are 221X and 222X... looking at your list above, looks like only the 223X are compatible between one another... so the 221X, 222X and 223X are technically distinct families with different designs ? Too bad... a 223X surely is never going to be as cheap as this 221X...

OHHHH !!!!!  What do you know !!!!

I just ran a similar search on Ebay and looooook at that !

Just the part number I need ! 

I am contacting the guy RIGHT NOW !!!!


http://www.ebay.fr/itm/Tektronix-120-1439-00-01-SwitchmodeTransformer-for-22xx-Oscilloscope-/282530665982?hash=item41c823d5fe:g:VFwAAOSw-0xYf5hj

In the US but item is small and light so he states he can ship it abraod for only 13+ bucks !  Taht will be under 30 bucks delivered, I'm in ! 

Might buy two of them just in case I get unlucky and one of them is "noisy", as he states.

Says it's compatible with other 22XX scopes as well, not just the 223X series.. that answers my question.


There are two versions of this inductance, the 00 and the 01 .... I do not know what the differences are, but I think they are interchangeable.

That chap on Ebay seems to know what he is talking about, and agrees !


I always advise to install a fan in the 2235, as I did in mine.

Luckily the 2232 had one, phew ! 

From a ferrite quality point of view, I would choose a ferrite with the least possible losses at about 40Khz due to the overheating problem of this inductor.

You mean 60kHz ? That's the frequency that regulator operates at, if I am to believe the waveforms in the service manual. Or were you talking about something else entirely ?!

It is also necessary to check whether this ferrite has a high induction saturation level.

Wow, I don't master this aspect at all    .. .and don't know how we could figure it out.

I guess I can only do so much in trying to come up with an adequate core. Then I would test it in situ to make sure it operates properly, and if it does then I guess that means I did an OK job ?!   First, before even whacking the coil back in to the scope, I can check that teh inductance of the primary is correct, compared to another working scope. The I will solder it and check waveforms to make sure it oscillates properly at 609kHz, with adequate waveforms. Then check for any funny noises, then check for the 40V rails accuracy, the put the cover back onto the PSCU section, fan spinning, and let it warm-um, then check temperature of the coil and check all the rails again, especially ripple. If all is fine then.. I think we can call it done 

You should also care about insulations between coils....Are there any ?

Yep sure, didn't mention it but yes of course the windings are insulated from one another. A couple turns of transparent tape, not more.  The outer tape is orange but I guess that's just for show...

[Neomys Sapiens]

Maybe you should look at this
http://www.hakanh.com/dl/surplus.htm
or talk to that guy. That's where I got the document from.

[Vince]

Thanks for that.

guy on Ebay already replied to me, how quick. Said he would (at my request) send two coils with the white gunk in the center hole as usually more "quiet". I had that type in my scope and it was quiet. Said he would mail them tomorrow !