Teardown B&K 2120 Scope

[med6753]

This is my first teardown thread. A B&K 2120 20Mhz Scope I purchased new in 1991 to replace a Heathkit IO-4540 5Mhz Scope I built in the 1970's. Back then going from 5Mhz B/W to 20Mhz B/W was quite a leap in performance which looks rather trivial today. The B&K has served me well over these many years and for a while was the only scope on my bench. Lately I've noticed intermittent and noisy contacts on both channel's attenuator switches and high frequency roll off. So it's time for a good cleaning. But getting to the switches involves major teardown and shield removal. So let's get started....

With the case removed there's the upper board consisting mainly of the trigger circuits, a portion of the linear LV supply, and the high voltage supply. The CRT is to the left. Both the board and CRT will be removed.

Close up of the upper board.


The CRT is from Japan by Toshiba. The scope itself was made in Taiwan.


Neck and gun assembly of the CRT. Since the HV is only -2KV there is no anode cap on the bell of the CRT. The HV enters via one of the base pins.


Upper board and CRT removed. No unsoldering was required to remove them. I was surprised that the CRT had an internal graticule just like the big boys. Having it removed gave me an opportunity to clean the face plate for the first time.


The Input Board with the shields covering the input attenuator switches. They are soldered to the foil side of the board.


I got the shields unsoldered but things did not go as planned from that point. They would not clear the vertical position control assembly immediately above it. That assembly is attached to the front panel with a set a screws requiring a sub-minature allen key in a size I don't have. Plus I discovered that the shields themselves have a front lip that wouldn't clear the attenuator switch decks if I tried to pull them straight back. I looked at removing the entire input board and that required a lot of dis assembly and unsoldering. Not something I really wanted to get into. Since the shields were now loose from the board I was able to pull them up slightly and I could see all 3 decks of the switches. So I carefully sprayed the deoxit with the extension tube on each deck. Not the best solution but hopefully I would get lucky. Resoldered the shields and reassembled the scope.     


Success! The Volts/cm switches for both channels are now rock solid. Channel 1 shows some peaking on the square wave. Channel 2 looks much better. I also noticed that the timebase is slightly off. The source waveforms are from a homebuilt function generator utilizing the 8038 chip. The outputs are fixed at 0.5V p-p and 5Khz. I use it specifically to quickly check basic scope functions. So a project for after the holidays is a complete alignment/calibration.   

[med6753]

Update to this teardown. Finally got to perform a complete calibration and everything came within spec.

Source: 5Khz, 0.5 v p-p Square from 8038 function generator chip. No more peaking on Channel 1. 50uS sweep on scope and it's right on the money.


Pushing the envelope for a 20Mhz scope. 1 Mhz Square from a Heathkit IG-4244 Oscilloscope Calibrator. Special cable with built in 50 ohm termination. Scope sweep set to 2uS with X10 engaged. Looks reasonable.


Ready for another 25 years of service 

[alank2]

Great teardown.  I get a little noise when switching the vertical - where exactly did you spray deoxit?  I've got some radio shack contact cleaner - would it also work?  I'm not sure I want to tear it down as much as you did!

ALSO - one thing I didn't realize until a few days ago was that you can pull the trig level switch to only have it display when the trigger is satisfied.  It probably seems basic for an analog scope, but I didn't know it was there.  Your mention of the X10 tipped me to look for the X10 on the timebase too!  I wondered how it could display anything close to 20 MHz until I found it!

[med6753]

Great teardown.  I get a little noise when switching the vertical - where exactly did you spray deoxit?  I've got some radio shack contact cleaner - would it also work?  I'm not sure I want to tear it down as much as you did!

The RS contact cleaner isn't going to cut it. Those deck switches seem to get a lot of corrosion. Look at the Time/Div switch deck and you'll see what I mean. You'll need this...


Unfortunately you can't do an adequate job of cleaning the volts/div switch decks unless you pull the shields as I did. The culprit for most of the noise is the deck closest to the front panel. You can't get the deoxit on it unless you lift the shield.

And if you are going to pull the top board I recommend changing out all the resistors in the focus divider circuit. I've had several of them go open over the years. I don't have the schematic in front of me at the moment. I'll post it later and identify the resistors you should change. 

[med6753]

Here's the HV and CRT schematic. The resistors that you should consider changing are R526, R527, R528, and R529.

[alank2]

I appreciate the help - I'll order some of the deoxit and see about getting it apart!  I may have more questions when I get there!!

[alank2]

Is deoxit D5 the same stuff?  You get 2.5x as much for 2/3 the price...

[med6753]

Is deoxit D5 the same stuff?  You get 2.5x as much for 2/3 the price...

I don't know. Perhaps someone else can answer that.

[Fsck]

Is deoxit D5 the same stuff?  You get 2.5x as much for 2/3 the price...

it isn't. D5 is a 5% solution, D100 is 100%.

[alank2]

it isn't. D5 is a 5% solution, D100 is 100%.

That answers it allright!

[alank2]

Those service manual says the focus resistors are carbon - should I order that or would metal film be an improvement?

[med6753]

I used carbon because that's what I had on hand. But I think metal film would be a better choice. Just make sure they are .5 watt or better.

[alank2]

I'm waiting for the Deoxit to arrive.  Will order the resistors Monday.

Did you have to remove the tube for some reason, or just the top board?  Do I need to be concerned about emptying any capacitors?

Is it easy to remove those black metal bands for the controls?  where they connect to the pot.

[med6753]

Did you have to remove the tube for some reason, or just the top board?  Do I need to be concerned about emptying any capacitors?

In order to gain access to the shields from all angles it's best to remove the CRT. It comes out fairly easily. Yes, good idea to make sure the caps in the HV generator circuit are fully discharged.

Is it easy to remove those black metal bands for the controls?  where they connect to the pot.

If you are referring to the extension shafts such as on the Intensity and Focus controls all you have to do is pull forward carefully and they come off.

[alank2]

It is coming along great.  I got it all apart and it did take some work to get those shields off.  I cleaned up the switch decks with an eraser and then sprayed them with the deoxit afterwards.  I went ahead and sprayed the other pots that had access as well.  I went through the calibration steps and all is going well so far except:

The timebase talks about VR801/VC801, but my PCB as three adjustments VR803, VR804, VR805 instead:

VR803 is for the fast timebases
VR804 begins at 50us
VR805 begins at 5ms

These were not too hard to calibrate.

I had a problem on step 13 of the trigger adjustment - is talks about adjusting VR309 to get 0V on the collector of Q610, but my meter shows +23VDC.  Changing VR309 does not do anything.  Then you repeat it for CH2, changing VR310.  Still not working - still +23VDC.  Any ideas?

ALSO - the corner where the transformer is gets quite warm without the case on - I wonder if it gets as warm with the case in place.

The scope is working so much better now...

[med6753]

It is coming along great.  I got it all apart and it did take some work to get those shields off.  I cleaned up the switch decks with an eraser and then sprayed them with the deoxit afterwards.  I went ahead and sprayed the other pots that had access as well.  I went through the calibration steps and all is going well so far except:

The timebase talks about VR801/VC801, but my PCB as three adjustments VR803, VR804, VR805 instead:

VR803 is for the fast timebases
VR804 begins at 50us
VR805 begins at 5ms

These were not too hard to calibrate.

I had the same issue. The service manual and the scope did not match. There must have been a revision but the documentation was not updated. I was able to tweak the 3 controls and get the timebase reasonable throughout it's range but it did take some time.

I had a problem on step 13 of the trigger adjustment - is talks about adjusting VR309 to get 0V on the collector of Q610, but my meter shows +23VDC.  Changing VR309 does not do anything.  Then you repeat it for CH2, changing VR310.  Still not working - still +23VDC.  Any ideas?

My trigger circuits were in spec so I did not attempt any further calibration. So I don't know what the issue might be.

ALSO - the corner where the transformer is gets quite warm without the case on - I wonder if it gets as warm with the case in place.

The scope is working so much better now...

Yes, the transformer get quite warm even with the case on and it's normal.

[alank2]

Does your CAL output go to 0.2V p-p?  Mine won't adjust that large, I ended up going with 0.1V p-p instead...

[med6753]

Does your CAL output go to 0.2V p-p?  Mine won't adjust that large, I ended up going with 0.1V p-p instead...

Mine is exactly 0.2V p-p so I would say there's an issue with your calibrator circuit. There's not much to it so it should be easy to identify the problem.

[alank2]

4.92V is going into the calibrator circuit.  The resistors check out as does the pot.  I tried cleaning up some of the rosin junk around it but that did not change it.  Bad IC?  If I changed R901 to a larger value I could maybe get it to go higher...

[med6753]

Check R904 (6.8 ohm) to make sure it hasn't increased in value. If it has it will lower the Vcc to U901 which will reduce it's output. If all the other passive components check out then I would agree that U901 is bad.

[alank2]

It was like 7.02 ohms.  I sure wish it was socketed!!!

[alank2]

I wonder if the heat producer near the transformer is actually the BUX86 (Q503) transistor bolted to the back of the chassis.  It seems especially warm.  It is part of the +140V circuit.

[alank2]

I took a look at the scope running with my Flir E4 and, wow, it has some resistors (R838,R839) that run 100 deg C !  Would this unit benefit from forced air?  That seems pretty warm but perhaps that is just the way it is made.

[med6753]

Yes, the scope does run very warm and I'm sure could benefit from some air circulation. Problem is there's no room for a fan.

All I can say is that I've had mine since 1991 and it's never had a heat related failure and it's sometimes been on for hours.

[alank2]

I agree, I don't see anyplace inside where a fan could be fit.  I'm tempted to connect a 60mm square type fan to the back cover with some zip ties and at least try to pull some air out at that point so air will enter all the case vent holes.

[Performa01]

I took a look at the scope running with my Flir E4 and, wow, it has some resistors (R838,R839) that run 100 deg C !  Would this unit benefit from forced air?  That seems pretty warm but perhaps that is just the way it is made.

While usually not really a problem for a resistor, 100°C isn't particular sexy either. But the same goes for forced air cooling, doesn't it?

Why not just replace these hot resistors with bigger ones (rated for 2 or 3 times the power)?

This will provide a lower thermal resistance, thus lowering the temperature on the surface.

If the footprint doesn't fit, then just leave the resistor leads a little longer and bend them in a nice curve so they can fit. This way you also increase the distance between resistor and PCB, which isn't a bad idea for hot components anyway.

[alank2]

The two resistors in question are already 3W !!  I think mouser did show some 5W ones were available...  They do have a hole in the pcb underneath them probably to provide for some air flow...

[Performa01]

The two resistors in question are already 3W !!  I think mouser did show some 5W ones were available...  They do have a hole in the pcb underneath them probably to provide for some air flow...

What resistors are R838, 839 exactly? Carbon film? Carbon composite? Wire wound? And what values?
Maybe you could also show the relevant part of the circuit diagram, so we could decide which kind of replacement would be appropriate...

...because if inductance isn't a problem and ww. is an option, you can easily get cemented resistors with very good properties, e.g. KH214 series from Vitrohm.

[alank2]

They are for "Horizontal Final Amp" from XY mode. and go to the CRT.  15K.  Page 18 of the 2120 schematic.

[Performa01]

Okay, had a look at the service manual and found the answer: R838,839 are metal oxide resistors, hence have a rather low inductance. But then, there are two inductors L801,802 in series ("peaking coils")...

If your adventurous and like experimenting, you could get two 15k Vitrohm series 214-8 resistors. These are 9W and will have some 250µH inductance (measured at 100kHz). When replacing R838,839, the total inductance will rise, but that probably won't hurt. But these are rather big: 9 x 9 x 38mm - and Mouser doesn't seem to have them.

The 5W metal oxide resistors Xicon 286-15K-RC (Mouser) would be the safer bet.

And then, you could not change anything and just ignore these two little toasters in your scope ...

[alank2]

What about these:

MOS5C153J

Mouser has them in stock...

[Performa01]

Yes, they would be fine too of course.

The Xicon would just have been a bit cheaper, sorry, didn't realize there is no stock...

[elex_enthusiast]

Great teardonwn. I tooo have the same scope given by a friend here on eevblog with the same peaking issue on both channels and definitely needs calibration.. Here's the story: http://everist.org/NobLog/20150731_simple_scope_funny_fault.htm
It has been repaired by him but he didn't had time to calibrate. Its my first scope and its was a great opportunity to own one to learn from it. Sadly I dont have any function gen to use for calibration.

[alank2]

Thanks everyone for the help given in this thread - my 2120 is working so great it is like a different piece of equipment!!!

elex - calibration was not too difficult - the service manual covers it in very good detail.  Just be careful with the high voltages.  I picked up (any excuse!) a high voltage probe so I could check the -2kV test point as well.

[elex_enthusiast]

I already had the service manual. Unfortunately acquiring a descent function gen is way far beyond my financial means,  another thing is, I haven't figured out yet if there's somebody near my place has a function gen that I could borrow or rent for calibrating my BK scope. I'm leaving in an urban place and there's no such place like a cal lab here..

[alank2]

Some of the calibration steps are visual, "adjust this until this is visible on the display", and some are voltage only, "adjust this until this test point is 95VDC +/-5V", so you might be able to calibrate some of them.  I didn't end up having to change any of the trimmers for the various vertical gain settings which was nice, but it certainly helped having a frequency generator.  I picked up a SDG2082X last year and it is now one of my favorite tools.

[med6753]

Thanks everyone for the help given in this thread - my 2120 is working so great it is like a different piece of equipment!!!

elex - calibration was not too difficult - the service manual covers it in very good detail.  Just be careful with the high voltages.  I picked up (any excuse!) a high voltage probe so I could check the -2kV test point as well.

You're very welcome. Always glad to help!