Just got a Tektronix 2235 (Opt 1)

[belzrebuth]

I've been searching for a scope for the past months and after a bad experience with a Tek 475 (there's a thread about it)https://www.eevblog.com/forum/beginners/about-to-buy-tektronix-475-first-oscilloscope-so-need-price-advice!/msg1011159/#msg1011159 ,
I finally got myself a 2235 USM version.

It's in pretty decent condition but unfortunately the Volts/Div knob is missing 1/3 of its backplate.(see attachement)
If someone knows where I could find a replacement knob please give me a clue.
There's a service center for old Teks in my country but doesn't have one in stock.

The other issue I'm having is that when I turn the scope on the traces seem to have a bit of noise (if I change the timebase value I can see that they are displaying something like a saw wave).
Traces slowly straight up afterwards.
What could be the issue here?
Does the scope just need a little bit of time to warm up?

[belzrebuth]

Here are some pictures with various settings on the scope.
Hopefully if someone has the same or similar scope can confirm that mine is fine.
I'm more concerned on what I see on the first picture.
Obviously I'm a noob regarding oscilloscopes but I intend to study the basics and learn how to use one (just need to ensure my scope is fine and it's me who doesn't have a clue about its operation)

If someone needs more info I can take additional photos with other settings and post them here.

[MosherIV]

Hopefully if someone has the same or similar scope can confirm that mine is fine.
I'm more concerned on what I see on the first picture.

Hi, I do not have that particular scope but it looks fine to me.
The first picture is just showing noise, not surprising since you have selected Channel2 and Chan2 is on 50mV/Div or 5mV/Div and it is AC coupled. I would expect some random noise pick up at that low volts per division.

Try changing the timebase to see what the period is, try turning electrical stuff off around the room/house and see if the noise goes away - you will figure out what is radiating lots of noise this way.

[belzrebuth]

At first I wondered whether this is DC ripple (from where you may ask and I don't know) or just plain noise .

Changing between AC/DC coupling (while the scope is set like the first picture) the wave remains the same.

So it must be mains noise?
I have a lamp connected to the same power strip.
I'll try without the lamp.

The trace should remain flat when nothing connected in all timebases and volts divisions?
Right?!
Not necessarily a flat line of course but flat nonetheless?!

[JPortici]

if i'm reading the panel correctly it looks like a 22kHz ripple, could very well be a residue from the power supply
could also come from the lamp. those are nasty

[belzrebuth]

It's not the lamp:(

So it's probably from the power supply,maybe I can move the scope to another power outlet (other room) and check if it behaves the same.

Channel 1 is slightly more noisy for some reason.

From where would this ripple come from ?Bad filtering?
Should I recap the power supply,maybe?

[JPortici]

i certainly would

[Tom45]

The first channel with the noise has the input set to 2 and input coupling set to AC. What does it look like if the input coupling is set to GND?

Before tearing into it I would suggest leaving it on for a day or so. Perhaps that might more fully form the capacitors in its power supply and also drive out any moisture that has settled in from not being used.

[belzrebuth]

Here's both channels with their Volts/DIV turned fully clockwise.(2m or 20m(for x10probe)).
Timebase is 20?s per division.
GND.

I don't very much like what I see.
I'll turn the intensity almost to zero and leave the scope on for a day like suggested.
Maybe remove the cover so that it doesn't get too hot or it's ok?

[KerryW]

I have a 2236, which is the same but with a computer module (freq counter, voltmeter, etc).
There is a similar signal on my display at those settings, but only 1/5 of a division high.

Try some experiments.  Short out the input BNC.  Still there?  Invert channel 2 and put it in ADD mode.  Still there?

[belzrebuth]

Same settings with ADD (first picture)
and channel 2 inverted (second picture,it's not straight it's still noisy)
ALT mode with channel 2 inverted(third picture)
.


@Kerry
maybe later you could try to apply these settings in your 2236 and post the results?!

[belzrebuth]

Forgot the shorted BNC input
(same settings as the first picture but with the bnc of channel 1 input shorted)
When shorted the amplitude of the signal is reduced.

[NexusKoolaid]

It's in pretty decent condition but unfortunately the Volts/Div knob is missing 1/3 of its backplate.(see attachement)
If someone knows where I could find a replacement knob please give me a clue.
There's a service center for old Teks in my country but doesn't have one in stock.

If you have access to a 3D printer or know someone who does, fabricating a replacement could be unsightly yet serviceable solution while you search for an official part. 

[KerryW]

Mine has a few idiosyncrasies of its own.  Channel 2 only shows the noise if inverted, non-inverted is a flat clean trace.  The GND position doesn't work on channel 1.  The noise is at the limit of syncing, so most tests (like ADD) just result in a slightly fuzzy trace.

[tautech]

@ belzrebuth
Good source for Tek bits and bobs is:
http://www.sphere.bc.ca/

[belzrebuth]

Mine has a few idiosyncrasies of its own.  Channel 2 only shows the noise if inverted, non-inverted is a flat clean trace.  The GND position doesn't work on channel 1.  The noise is at the limit of syncing, so most tests (like ADD) just result in a slightly fuzzy trace.

So if your scope is set like mine in the photo (both volts/div fully clockwise and timebase at 20?s ,only ch1 active ;you're getting some kind of noise or just a slightly fuzzy trace?


@tautech
Thanks for the link.
I checked but I didn't find my part:(

[tautech]

@tautech
Thanks for the link.
I checked but I didn't find my part:(

There's a few listings for 22** knobs, some of which are out of stock however you should email them.
http://www.sphere.bc.ca/test/tek-parts/tek-knobs.html

Edit
This one?

2mV -50V/div
366-2148-01

If you search Google there's plenty of hits using the part #

[KerryW]

At those settings, I can trigger on the noise.  It is about 0.4mV P-P and shows 6 or 7 cycles.  It is not big enough to trigger the freq counter.  Shorting either BNC reduces the signal but does not eliminate it.

I suspect it is from a switching power supply in the scope.  Yours may have a bad cap.

[belzrebuth]

I think I'll make a short video showing other settings and how the scope behaves in general in order to give you a better idea.

I may have to open the scope up and visually check for bad caps.
That makes me a little uncomfortable because from what I've seen the power supply is not easily accessible.
If I start desoldering caps just to test them I could just as well replace them anyway;first step should be just to open it up and write down what caps are already there and order replacements.
I think Nichicon pw series might be a good candidate since they're known for their long life inside switching PSUs.

I'll make a video as soon as I find some time and we'll go from there.

[David Hess]

That noise has got to be from the switching power supply which uses an off-line buck preregualtor followed by a saturable core inverter.  The output capacitors are probably worn out which is a common problem with 22xx series oscilloscopes simply do to their age.

I would not remove any capacitor for testing without replacing it with a new one.

A replacement knob should not be difficult to find.

[belzrebuth]

I opened it up and looked thru the power supply grill vent.
There are five or six caps that are slightly leaking on top.Tops are slightly bulged too.
The scope definitely needs a recap.I hope that will fix my noisy trace problem.

I was a bit lazy and didn't remove the vent but I intend to do this tomorrow.
Anyone has any suggestions on what replacement caps I should buy?

I will probably order from TME.eu

[JPortici]

check the requirements for ESR on the service manual, if you can keep a simillar ESR try to increase voltage and temperature ratings (don't bother if it costs too much though, those specified were good for about 25 years)

[belzrebuth]

I tried to remove the PSU shield by unscrewing all the torx screws but I can't.
Something is holding it back!
Anyways I peeked inside and the bulging caps are branded "IC" (Illinois) and are 1000uf at 25V.
I've never encountered/used illinois caps before so I can't recognize what series they are.
They're brown and don't seem very specialised judging from their bad lettering.
I could very well be wrong on this but they seem generic.

I'm not sure about the c968 but appears to have similar size.

These are the caps that appear to be problematic.
C960
C961
C962
C963
C968
C970

I can't find any info on these other than their values on the service manual.No voltage ratings and no esr.

[belzrebuth]

Service manual states these values for the 960-63 and 68/70.

C960 840uf
C961 840uf
C962 840uf
C963 840uf
C968 840uf
C970 840uf

[David Hess]

check the requirements for ESR on the service manual, if you can keep a simillar ESR try to increase voltage and temperature ratings (don't bother if it costs too much though, those specified were good for about 25 years)

There are no requirements listed.  Finding the specifications will require the original datasheets and the part numbers may not even be available; I have a whole set I removed made by Mallory which use Tektronix house numbers.

As a practical matter, the capacitors at the output of the saturable inverter do not have stringent ripple current requirements because the inverter output is a square wave.  The capacitors at the input and output of the buck preregulator are another matter of course.

Something else which can be done is to use the capacitor list from a later 22xx oscilloscope that has a higher power power supply.  The 2235 was the base model and the 2236, 2230, and 2232 used slightly larger capacitors.

[vulturebetrayer]

I found the original datasheet and original manuals online for you.
I have linked them here:

https://vulturebetrayer.com/tektronix2235.html

I have hotlinks disabled on my site which is why I had to make a page for you.

edit:
spelling fail

[David Hess]

I am sorry if I was not clear and just to avoid confusion, I was referring to the datasheets for the capacitors.  The 2235 service manual has the capacitor part numbers but to find their ESR and ripple current ratings, the datasheets for the capacitors will be necessary.

But again, they are not that critical; capacitors intended for low ESR high frequency switching regulator applications will work fine.  I would probably use the Nichicon PW series for the output capacitors and maybe for all of them but there are other higher performance choices.  Increasing the voltage derating to get larger case sizes will help with meeting any ripple current requirements.

Also in 1990 when the 2232 was released, the 2235 and 2236 became the 2235A and 2236A with uprated parts lists so the 2235 and 2235A are not quite the same.

[vulturebetrayer]

I am sorry if I was not clear and just to avoid confusion, I was referring to the datasheets for the capacitors.  The 2235 service manual has the capacitor part numbers but to find their ESR and ripple current ratings, the datasheets for the capacitors will be necessary.

But again, they are not that critical; capacitors intended for low ESR high frequency switching regulator applications will work fine.  I would probably use the Nichicon PW series for the output capacitors and maybe for all of them but there are other higher performance choices.  Increasing the voltage derating to get larger case sizes will help with meeting any ripple current requirements.

Also in 1990 when the 2232 was released, the 2235 and 2236 became the 2235A and 2236A with uprated parts lists so the 2235 and 2235A are not quite the same.

ahhh so we would need to know the exact cap model for that device and what year and model it was specifically. . .
 oh well, worth a shot I guess

[R005T3r]

It's in pretty decent condition but unfortunately the Volts/Div knob is missing 1/3 of its backplate.(see attachement)
If someone knows where I could find a replacement knob please give me a clue.
There's a service center for old Teks in my country but doesn't have one in stock.

If you have access to a 3D printer or know someone who does, fabricating a replacement could be unsightly yet serviceable solution while you search for an official part.

Actually there's a simple way to fabricate your own parts: silicon molds and resin. Also, you will need someone who make the design for the printer and it's not that simple... Since the object is not too small, and considered the knobs are identical, I think it's the easiest way to have the replacement. Also you will have:
1. a greater level of detail than 3D replicated stuff (to help you understand how good it is, you can clone miniature soldier out of that with the finest details printed...)
2. You can also load the resin with colors if you want or other agents to make it heavier, shinier and glittered...
3. there are many types of resins with various characteristics and proprieties.

[belzrebuth]

I'm waiting for a friend to decide what he also wants from TME and then we will place an order.

I think PWs should be fine for the task .Their specs are decent and they are rated specifically for switching psus.
PWs are also much cheaper than Panasonic FCs (not that cost matters for this kind of repair)
Here's what I plan to order:
http://www.tme.eu/gr/details/upw1e102mhd/puknote_-ilektrolut-tit-chamil-antist/nichicon/

Would it be a wise choice to replace every electrolytic in that section or just try my luck with these six?

Also,is there any trick to remove the psu shield?!
I've removed the screws but it's somehow stuck in there.
Maybe remove the little pcb in front of it?The one with the soldered leg-pins.

@vulturebetrayer Thanks a lot for these files and for your time iploading them!

@David Hess You've been a great help so far.Thanks!

While the soldering looks factory I think the caps on mine have been changed..
I saw pictures of another 2235 supply and the originals should be mallory brand and 840uf at 12V.


@R005t3r
I think I'll just wait for a replacement to appear.
I don't have a 3d printer or access to one but even if I did I wouldn't bother since ,as you say, it might be more difficult than one would think.

[David Hess]

Would it be a wise choice to replace every electrolytic in that section or just try my luck with these six?

Replace them all since you will have access.

Also,is there any trick to remove the psu shield?!
I've removed the screws but it's somehow stuck in there.
Maybe remove the little pcb in front of it?The one with the soldered leg-pins.

I cannot say since I have never had mine that far apart or it was so long ago that I do not remember.

[oldway]

Power supply of 2235 is working very hot and I recommend to install a fan.
Look at the schematics of the 2236 to see how to do it.
This is still more important because you live in a place where the weather is quite warm.

[belzrebuth]

Yes I read somewhere that there's even a connector for a 12V fan on the board.(for the 2235A at least).
Mine has the fitting ready in the chassis but I don't know where the connector is.
I'll try to find where I read about it and maybe I have the same connector.
I think I'll probably do that just after I install the new capacitors and check for good operation.

If only I could take that PSU shield off!
I need to write down the values of the other caps in the board but the shield just won't let me see.

[belzrebuth]

Okay I got the shield off.
The pcb that connects to the board in front of the power supply is almost impossible to remove (gently at least) so I tried to remove the shield without disconnecting the pcb.

I eventually did it but I think there should be an easier way.

Anyways the six caps I mentioned earlier are definitely looking bad.

I've also got a 330uf 50V cap (C956) that I'm going to replace because it's "IC" brand ,same with the other bad caps so I doubt it's ok.
There is a nippon 1000uf 50V cap that looks ok.(c940 or 941)

And there are three other capacitors at the right back that their voltage values I can't see.(c942 / c943).These look okay too.

There's one (c954) right next to the 330uf cap that I think it's 33uf but with high voltage rating.(200V maybe)

And the last electrolytic in this section is a C925 whose voltage value also I can't see.I think it is 100uf but can't see a thing the way the chassis gets in the way.

[belzrebuth]

Anyone's got experience with Chemi-con KYA series?

A local store has them in stock so I thought I could replace the 1000uf 25V with these instead of PW because I can get them today and finally enjoy my scope!
Should I wait a week or so for PWs or these could also do the job?
I'm thinking out loud here but I could get those and check if the problem is indeed the blown caps and then order PWs for the whole supply board..

[belzrebuth]

Okay!
Scope's okay now.
Replaced the 1000uf caps and the 330uf cap and the trace is now crisp with no noise even at the mV range.

Surprisingly putting the power supply shield back was far easier than removing it!
Thanks everyone:)

(if anyone need any pictures let me know)

[David Hess]

The Chemicon KYA series will work fine except for the input capacitor to the switching preregulator which needs to be higher voltage.

[tautech]

Well done. 

2 pics of the display showing before and after would be good for anybody finding this thread in the future.

[belzrebuth]

Which capacitor is the input cap to the preregulator?
The 33uf 200V?(I have one 33uf next to the 330 which might be 200V but I couldn't see well,I need to desolder it to check its ratings )


I intent to order chemicon PWs just in case I go paranoid and decide to re-recap everything but unfortunately when I opened the unit I couldn't see the values of some of the caps the way they are put in place.
I wouldn't want to desolder them and put them back so I could just wait for something to fail again before messing again with the insides of the scope.

I have one question though.
I see the trigger led randomly lighting up with nothing connected.It does that every 40s or so.
The traces dissappear for maybe hald of a second and then come back.
What could be causing this?

@tautech:pics are on the way..

[David Hess]

Which capacitor is the input cap to the preregulator?
The 33uf 200V?(I have one 33uf next to the 330 which might be 200V but I couldn't see well,I need to desolder it to check its ratings )

It is C906 which is marked 75uF on the schematic and 450V in the parts list; it is the big one on the right side of the power switch from the front.

I have one question though.
I see the trigger led randomly lighting up with nothing connected.It does that every 40s or so.
The traces dissappear for maybe hald of a second and then come back.
What could be causing this?

That is normal when there is no trigger available.  What happens is that occasionally when no trigger is available in AUTO mode, there will be enough noise to cause a trigger but this resets the trigger timeout and without further triggers, the display will be blanked until either a valid trigger occurs or the trigger timeout finishes.

[belzrebuth]

Thanks again David:)
I'm a total noob regarding oscilloscopes but that's the reason I got one (one has to start from somewhere!)
One thing I noticed is that I left it on for sometime and got hotter than before;maybe it's just me or just the ambient temperature is higher than yesterday.
Here are some pics with the same settings (volts/div fully clockwise and 20us timebase)
The difference is obvious.
And bonus audacity generated 1khz sine

[David Hess]

I'm a total noob regarding oscilloscopes but that's the reason I got one (one has to start from somewhere!)

The 2235 is a great oscilloscope to start with and actually at the higher end of the features curve with delayed sweep capability.

One thing I noticed is that I left it on for sometime and got hotter than before;maybe it's just me or just the ambient temperature is higher than yesterday.

The original design lacked a fan and I am not aware of any heat related issues.

[oldway]

I am not aware of any heat related issues.

In fact, there is a big one: internal isolation failure between layers in the pre regulator transformer (in reality, an inductor) T906 (120.1439-00) for working a long time at to high temperature.
As this transformer is no more available, the oscilloscope with this fault is not repairable

[belzrebuth]

I am not aware of any heat related issues.

In fact, there is a big one: internal isolation failure between layers in the pre regulator transformer (in reality, an inductor) T906 (120.1439-00) for working a long time at to high temperature.
As this transformer is no more available, the oscilloscope with this fault is not repairable

I'd be interested in installing a fan on mine as I indeed live in a quite hot area and that can't be good for my scope.
Some info that I got online  states that I need a 12V 60x60mm maybe SUNON brand like the ones found in servers.
There is also a blog post about it but the author doesn't describe the whole process well enough imo.
http://blog.cluepusher.dk/2011/02/16/simple-fan-mod-for-tektronix-2235/

If you have more information (a short guide maybe?!) please share..

[oldway]

You must download the service manual of the 2236 and look at the schematics.

http://elektrotanya.com/tektronix_2236.pdf/download.html

[belzrebuth]

So I check the 2236 schematics and "copy" its fan circuit to the 965 area of my 2235.
Correct?

[David Hess]

I am not aware of any heat related issues.

In fact, there is a big one: internal isolation failure between layers in the pre regulator transformer (in reality, an inductor) T906 (120.1439-00) for working a long time at to high temperature.
As this transformer is no more available, the oscilloscope with this fault is not repairable

I am aware of problems with the preregulator transformer/inductor from discussions over on tekscopes@yahoo.com but have never tracked down the actual failure mode.  They do not develop open or shorted windings but appear to have low inductance so my hypothesis has been either a cracked core or a material change do to overheating.  I myself have never gotten ahold of a bad 22xx to test; all of mine work fine.

[David Hess]

So I check the 2236 schematics and "copy" its fan circuit to the 965 area of my 2235.
Correct?

That is basically it.  There is not much to it.

There were some variations in the series in how the fan was setup but any of them will work.  The 2230, 2232, 2236, and 2236A schematics cover most if not all of what is needed.  The air volume is pretty low so a slow fan will be more than adequate.

[belzrebuth]

Okay.I'll see what I can do.
I'll probably order a SUNON fan like the ones used in servers and check the schematics in the meantime.

[David Hess]

I found my old tekscopes@yahoo.com post about the 2-channel 22xx fans:

1. The 2213A, 2215A, 2235, and similar models which lack a storage or
counter/timer/multimeter (2236 and 2236A) board lack the fan power
circuits and fan.  From what I remember at least on the 2235, the main
printed circuit board (A1) is the same but lacks the parts so the fan
is easy enough to add if desired.

2. The 2230 and early 2232 and similar power the fan from the +8.6
volt supply with a parallel set of rectifiers and an RC decoupling
network.

3. The 2236 which has the counter/timer/multimeter in place of the
digital storage board powers the fan from the difference in the +5 and
-8.6 volt supplies with a parallel set of rectifiers (3/4 wave?) and
an RC decoupling network.  Maybe the 2236 ran hotter than the early
2232?

3. The late 2232 (and 2235A and 2236A?) powers the fan directly from
the difference in the +5 and -5 volt supplies and includes the same RC
decoupling network (2.7 ohms and 4700 microfarads) as the early 2232.

That makes for 8.6, 10 and 13.6 volt supplies for the fan depending on
model.  I have never encountered a 12 volt DC brushless fan which
would not operate over that range of voltages.

[belzrebuth]

I ordered the diode and the capacitor needed for the fan circuit.
Here's the original part for the fan used in 22xx series.
It's a motor blower fan.
PAPST 812 12VDC 2.5W
http://www.qservice.eu/shopexd.asp?id=4356&bc=no

I can't find a blower avalaible so should I replace it with an axial fan?
The original part is overpriced and old so I'd prefer a new SUNON fan but right now I can only find axial type.
I guess from nothing even an axial would be better but I'd like some opinion..

I attached the schematic for the fan in case anyone need it.
For the diode I ordered a BYV 26C as stated in this blog post I found here http://blog.cluepusher.dk/2011/02/16/simple-fan-mod-for-tektronix-2235/ because the service manual for the 2236 doesn't have a part list.