Tek 475 trapezoidal display distortion

[Redcat]

Hello everyone and a happy new year  ,

I have still a little problem with my wonderful old Tektronix 475 oscilloscope.
(Found out tody it's from 1979 - 1979 written on some ics,1971 and 1978 on some pcb's - but I think it's 1979  ).

There is a terrible trapezoidal distortion of the waveforms, on the bottom of the screen they are wider than on the top.
I have tried to paint a little sketch, so you can see it better than in a picture:



It is a little more on the right side.
My guess is, that the beam is not centered (little up and left ?), but I dont't know how I could test it.
Voltage rails are spot on (50V spot on,-8V,15V... only some mV away, but 50V unregulated is 65V?).

Is there anyone who knows how this distortion could be calibrated? Is there any pot for this form of distortion or one for centering? (No, it's not the geometry pot, I tried it.)
Perhaps it's some other problem?

This would be very kind.

Nice greetings, Tom

[Redcat]

Are really all Tektronix experts still in winter holidays?  just fun 

Is there anybody out there who knows a little more about these old scopes?

nice greetings  , Tom

[JackOfVA]

First, I assume the waveform being displayed has been verified as correct when looked at with a known good oscilloscope.

I looked at the 475 manual and do not see an explicit adjustment or calibration process for "trapezoid distortion" unlike magnetic deflection CRT computer monitors. However, oscilloscopes of this era have many potentiometers and trimmer caps and variable inductors that can drift out of adjustment over time.  There's about 30 pages of detailed calibration and adjustment instruction in the 475 manual that might be helpful. Sorry I can't be more specific.

[Redcat]

Thank you very much  JackOfVA for your reply. Yes I remember my old crt computer Monitor 15 years ago had such an adjustment pot on the front  .
Yes the waveform is good, I have 3 scopes and on two it looks absolutely fine.
As a very simple test I did switch to ground and shifted the horizontal line (so that the end of the line ended with the most right/left gradicule) up and down and it moved some mm sideways (so with this I could eleminate the waveform as the reason for errors). Tested on both sides. And on a real sqarewave you can see it like in my sketch. So yes, sadly it seems to be really a trapezoidal distortion.

I have the manual of the 475 as a PDF file and I think have to dig a little deeper  (it's not so easy to find, as my english is not the best  ).
Have nothing found so far, but of course there could be some hidden information as you say. Hope I can find something.

Have even removed the metal grid filter from the front of the crt, because I thought it could be some magnetic issue, but wasn't.

I thought there would be someone, who has reworked or calibrated some old Tek scopes (and had the same problem) in the past and knows much more  . So i won't give up asking.

Nice greetings, Tom

[Rerouter]

found an adjustment that may work for you, its called geometery  (vertical curvature of the vertical timebase) its adjusted with r1390 front right side of the outside of the HV can

[Redcat]

Hello Rerouter, no sadly it's not the geometry pot 1390. I mentioned this in the first post, but rechecked the manual. I also tried pot 1385 Y alignment (near 1390). You can tweak it a little bit with both pots together but the problem is still there. It's not so much a curvature but trapezoidal.
But thank you very much for your effort in helping me  .

[philpem]

Has the scope been dropped?

You sometimes see this fault in scopes and electrostatically deflected CRTs which have been dropped -- the deflection plates either break away from their mountings or warp. Rapping your knuckles on the top of the scope chassis might help narrow this down -- if the trace shimmers or appears to move in relation to the graticule, it's possible that the CRT might have been damaged.

[Redcat]

Hi philpem,
this is one possible reason I thought about too. But the scope is 35 years old and I don't know if it was dropped before I got it (I didn't).
I tried to knock on the scope and watched the trace as you suggested, but nothing happened. So I don't think that anything is loose or broken.
But thank you for your guess  .
I also still think, that the crt can not be produced so that it (and the plates) is 100% accurate and the trace and possible distortions have to be able to align (in factory or later).
Just did a photo with my telephone when I tried the knocking test, so that you can see, how the wave is distorted (it's like in my sketch). Perhaps it's better to see as on my sketch.

The distortion looks the same on both channels, just tried it again.

Nice greetings, Tom

[grumpydoc]

It's not quite as bad as your sketch

If you look carefully at the faint up/down strokes then you can see, at least in the lower half of the screen, that the trace actually lines up vertically with the graticule quite well. I can't quite tell how it lines up at the top of the screen from your photo.

Sometimes you can get an effect with square waves which can make them look a bit odd - basically if you have the brightness turned up so that the trace "blooms" a bit you get a bit of overshoot on the horizontal portions of the trace and it can make the trace look a bit off - greatly exaggerated it winds up looking like this:



Where the green line is the path the trace takes, the light green shows the blooming on the CRT screen and the black line is what your eye perceives is the  shape of the wave.

I would start with a copy of the manual and do at least the following from the calibration section

Power supply calibration part 1 - DC levels (ripple will need a 2nd 'scope)
Power supply calibration part 2 - HV power supply but only if you have a HV probe and know how to use it

Display and Z axis part 1 - CRT grid bias
Display and Z axis part 2 - ASTIG control
Display and Z axis part 6 - Trace alignment
Display and Z axis part 7 - Y axis alignment
Display and Z axis part 8 - Geometry

and possibly
Vertical system 1 - Vertical Output bias
Vertical system 2 - Vertical shield voltage

None of these are terribly hard - you'll need a DMM, a sine wave source for the astigmatism adjustment, and if you don't happen to have a "time mark generator" for the Y axis adjustment a 10kHz square wave will do (even more so if you can adjust the duty cycle down to get a narrow pulse).

[w2aew]

My first thoughts would be that this is due to stray magnetism or a damaged CRT (bent deflection plates), rather than some vertical amp influence on the sweep linearity.  My inclination would be the former.

Is there something with strong magnetic field associated with it, somewhere in the vicinity of the scope?

Are the mu-metal shields still in place around the CRT?

More importantly, have you run a degaussing coil around the scope to remove any magnetism from the chassis?

[Redcat]

Hi  grumpydoc,
yes I did a little much on the scribble  , but it is 3-4 mm from bottom to top, so you can measure nothing.
My other scopes are so absolutely straight.
Actually I have done some of the procedures you suggested (tried at least  ).
The regulated voltages are fine (50V spot on and the other voltages,which depend on that are really good - as I said, only the unregulated 50V is 65V - but I don't know if this is important?), checked them with a scope and a good DMM. I set the trace of the image only a little brighter to show the bending better, the brightness doesn't change anything.
I definitely will not put my hand on the high voltage  , as I wouldn't on a big cap. But with the 10kHz sqarewave is really a nice tip, I will try it tomorrow (have only a little AD9850 signal gen, but it will do). Trace alignment and astig are good and the geometry pot bends it on the top+bottom, so it does not solve it (really tried several times, together with the Y alignment, but will try again).
Thank you very much , as I said I will try some things of this again tomorrow (have still winter hollydays this week  ).


Hi Alan,
Yes magnetic influence could have to do with it (was also my guess, so I removed the front metal grid filter - but didn't help).
The CRT tube is in a metal "pipe" and on top over the hole thing is a big metal plate and then the whole metal case.
I really hope that nothing is bend inside the CRT  - then I would be really sad. I can see nothing because of this metal around the CRT.
There should not be a strong magnetic field nearby (the other scopes are right near the 475 and are ok). My TV is from my grandma (CRT) and it is 3m away, but it is off (this would be the only source).
I have no degaussing coil (and I have to say that I don't know how it looks or works - know the function only from my old computer monitors).

Is there a way I could build such a tool? This is really something I want to try. Could you help me out a second time and explain it? 
(And  for your latest videos.  )

Very nice greetings to both of you, Tom

[GreyWoolfe]

You can actually use a trigger type soldering gun if you're very careful.  An uncle of mine, many years ago, showed me how to degauss old TV sets with it.  Keep it about 75 mm from the screen, pull and hold the trigger in and keep it moving so as not to damage anything by keeping it in 1 place to long.  A few loops around the CRT should do the trick if it is going to work.

[Redcat]

Oh thank you GreyWoolfe  .
I have no trigger solder gun, but a trigger desolder gun  . Hmmm.. I will try.

Sadly I have just found a video on youtube (yes I can not sleep  ) which shows a problem like mine on another Tek scope (but my trace does not jump when I knock on the scope (will try without the metal case tomorrow).
I will post the links to this and the discussion about it here. It turned out that one deflection plate was broken off.
I really really hope that is not in my case .

http://www.edaboard.com/thread265808.html



It is worse than on my scope so I have at least a little bit of hope left.

[Andy Watson]

Have you tried the yahoo tek group ?

http://tech.groups.yahoo.com/group/TekScopes/

[Redcat]

Hi Andy, no, i didn't know about. Thank you for posting the link, I will definitely take a look at this  . 

[w2aew]

If you try the degaussing, do the entire chassis and length of the CRT, not just the face of it. The metal around the CRT neck is Mu metal, designed to shield magnetic fields. But, it can be degraded also.

[Redcat]

Hi Alan,
took me half the day to test all the things you all suggested. Found out, my desoldering iron has a pistol grip, but the switch is only for the vacuum pump  .
But I tried my hairdryer, it's the only thing I could think of working for this.
Also tried again to tweak the pots I already tweaked  , measured the voltages again  , read the manual and looked at the scematics, but nothing changed.
Even looked at the back of the crt tube with a lamp, but couldn't see anything (there is not much glass to see).

Then I assembled the case just some minutes ago. I put on the front cover and slided the case on from the top (from the back of the vertical standing scope) - first time I did this (the other times I slided it on and off, when the scope was lying flat).
And when I then turned it on again... what should I say...look at it  :



Something in the crt tube is really broken and fall off to the front of the screen so that I could see it. 
I whished it wasn't like this.

But now we know all the answer.sad.

Have now stand it on it's feet and tried that the pieces fall back, so there is now again not much to see. I think I will use it as it is. Sad but ok, it's old.
And something interesting I found when I took some photos of the inside - I found the real production date : 4779 stamped with something like "production date"  , so I guess 47th week in 79, only 5 months younger than me  .

Many many thanks to you all, who tried to help me with this    .
Nice greetings, Tom

[Andy Watson]

If I remember correctly, the trace rotation is performed by magnetic coils. The "picture" is very sensitive to the location of these coils. Before giving-up on the tube it might be worth checking the integrity of the trace rotation coils.

[SeanB]

Sure there must be somebody with a CRT that will fit that scope.  At least with Tek making all their own CRT units you should be able to find a donor that has a good tube to use in it.

[Rigby]

Surely you can find a matching tube on eBay.  They seem to be common for popular scopes like the 475.

The instruction and/or service manuals for that model will list the part number.  Search for that on eBay and it is likely you'll find a tube soon if not immediately.

[SeanB]

I have a nice microchannel tube that is doing nothing in a partial mainframe. Perfect for anybody with a storage scope who needs one. Must dig it out and see what one it is.

[Redcat]

Hi Andy,
how can I test if the trace rotation coils are good (or in the "right" place)?
The tube is completely inside some metal tube, so I can't see anything - but perhaps I could measure?
If I rotate the trace with the pot on the front, I can rotate it a little bit more clockwise than counterclockwise (the trace goes up/down 0.8 to 0.5 cm at the edge).

Hi SeanB and Rigby,
yes, I think I have read, that some models even use the same tube. Won't be too complicated to find the part number (I think I have seen a website from US some weeks agoe which sells crt tubes - no option for me - but I remember they had listed them with the partnumber and Model in which they fit in). I will look for a new, but it depends on the cost if I will change it.
I don't consider to buy a used one, because I don't know if it's even worse  .

[JOERGG]

Ask Jan first may be a chance.

http://www.die-wuestens.de/dindex.htm

Here is his list of CRTs

http://www.die-wuestens.de/rd.htm

[Redcat]

Hi Joergg, I can't see a crt which fits by number, but made me a link to check from time to time (the scope clock looks nice ), will also check on ebay (wanted to buy a new one but it's waaaay too expensive, even a used is... ).
Have also read that they are realy hard to get for the 475. I wonder if one from a 465/B/M would fit.
And I have found the website with the part numbers I mentioned earlier, it's from Canada, not US  .
http://www.sphere.bc.ca/test/tek-crts.html

Have still a little bit hope left, since Andy mentioned the trace rotation coil. We will see  .

Nice greetings