Tektronix 475, is this likely to be a straightforward repair job?

[Chris Wilson]

TEK 475 with only a dot as the visible display, capable of vertical movement by knob manipulation, but no horizontal. likely to be a straightforward fix please? Thanks.

[amspire]

It may be reasonably straightforward finding the fault, if you know a bit about fixing electronics.

If the horizontal position does nothing, then it sounds like the problem is in the horizonal Amp stage that is all discrete transistors - no custom IC. Probably a fair chance of a repair.

The manual is at http://bama.edebris.com/download/tek/475/475.pdf

Some other documents at http://bama.edebris.com/manuals/tek/475/

Richard

[jgbena]

I have the same scope, and i just want to make sure what you are seeing.

1. you can move the scope trace up and down on each channel using the knobs next to the input right?
2. When you move the horizontal adjust knob below the screen, the dot still stays in the middle?

can you answer a few more questions?

1. was this scope working before and just stopped?  Oh and is this scope new to you or do you have experience with it? (just asking)
2. when you change the time base does anything happen?

Jon

[Chris Wilson]

It's a scope I have left a low bid on, not sure if I will get it or not. The info I gave is all I have to go on I am afraid. If its beyond hope I can break it for parts, it won't have cost me much. Thanks for the reply.

[Chris Wilson]

It may be reasonably straightforward finding the fault, if you know a bit about fixing electronics.

If the horizontal position does nothing, then it sounds like the problem is in the horizonal Amp stage that is all discrete transistors - no custom IC. Probably a fair chance of a repair.

The manual is at http://bama.edebris.com/download/tek/475/475.pdf

Some other documents at http://bama.edebris.com/manuals/tek/475/

Richard

I have downloaded these, thanks very much for the links Richard, appreciated. And thanks for the tip as to where to start looking.

[jgbena]

It's a scope I have left a low bid on, not sure if I will get it or not. The info I gave is all I have to go on I am afraid. If its beyond hope I can break it for parts, it won't have cost me much. Thanks for the reply.

Oh OK!!

Well a lot of those people selling scopes on eBay arent exactly "users"... it might be something silly.  But then again it could be something as the fella mentioned above.  I bought mine as a fixer-upper and the only thing wrong with it was that someone played with the knobs and put both scope traces out in the woods!

If you win.. keep us posted!

[Rerouter]

make sure that triggering is set to auto, and if you have a probe try feeding in the cal signal, also make sure its set to channel A or B, depending on what your using, not X,Y

and try every horizontal setting with a fer vertcal ones, the attenuators on the 475 can mask a signal if they have become a little loose in there sockets,

finally, if your game, take off the case, and measure the voltages at the test points, expecially at 50V, if its out by more than 0.2V, then either try another meter, or that may on its own be the culprit, being how all other rails are regulated off the 50V rail

edit: oh, missed the point you where still buying it, still if you get it, that would be the general approach to get a trace,

[grumpydoc]

I bid on the same 'scope, didn't get it as I guess I'm just too stingy but the photo looked as though it was set to x-y mode.

[Chris Wilson]

Well, I am picking it up tomorrow, so I guess I'll be back, as they say. I had a little spending spree today and now I am officially an oscilloscope collector I have bought a Fluke / Philips PM3380B digital / analogue scope that works just fine. I  really nice physical condition with a user manual, but no probes and no service manual, so looking for a service manual and a pair of the special probes that are like some Tek ones that "tell" the scope what attenuation they are from an extra connection ring on the back of the BNC. I have some normally terminated cheapo probes for now though. It has more functionality than I really know what to do with, but it will be a workshop scope for some auto diagnostics along with my USB scope. I also saw a brand new just out of its ex government storage packing 7633 mainframe with no plug ins, but I could see it working perfectly with some plug ins borrowed from the vendors own 7 series. Being unused, split mint, and cheapish it was irresistible. I am now looking for useful plug ins! I will need guidance as to what's useful and what's probably not, I never realized there were so many It even smells nice

And then there''s the Ebay 475, for which the "you have won" e-mail was awaiting me when we got back. I took the wife today, she had a great time sat outside the vendor's place reading the paper, after a long journey, and the promised pub lunch again degenerated into an eat while you drive KFC. She was quite pragmatic, she's getting used to me after 25 years. The 475  is just something to challenge me to get working. It's not far away and I just put a bid in and waited to see what occurred. It's a project that is well outside my comfort zone, but a New Year's resolution to quit smoking and give the local landlady at my favourite pub less profit margin, has left me with some extra spending money and some more free time. Whether the challenge of the 475 is sensible to take on I don't know, probably NOT, my knowledge level is pitiful, but I can hopefully check voltages and some capacitors and stuff, and maybe get lucky, or perhaps involve a friend who is a proper techie to help me out, yet again.

I am starting another thread on testing capacitors, and one on what to look for in a cheap but effective soldering / de-soldering station, and what solder to buy. I have read of a way to check capacitors with a basic signal generator, (which I already have), and a working scope. I may build myself a basic ESR meter, too. To try the scope method I need a BNC "T" fitting, which I blagged a pair of free, from the scope vendor, and a 50 ohm terminator, which I don't have, so Ebay has taken a tiny bit more dosh for 5 of these, this evening It's quite nice to rekindle an old interest at my age, makes a change from building engines and race cars, which is my day job. Thanks for the great advice, I'll post up about the 475 when it's back here and I have a chance to play with it.

[tekfan]

I also saw a brand new just out of its ex government storage packing 7633 mainframe with no plug ins, but I could see it working perfectly with some plug ins borrowed from the vendors own 7 series. Being unused, split mint, and cheapish it was irresistible. I am now looking for useful plug ins! I will need guidance as to what's useful and what's probably not, I never realized there were so many It even smells nice

Congratulations for your purchases. I think you'll find the 7000 series Teks particularly interesting.

These are some of the more useful plugins that will suit the 7633 nicely:
7B53A (100MHz delayed time base)
7A12 (100MHz dual channel 1Mohm input amplifier)
7A13 (100MHz differential amplifier)
7A22 (1MHz differential amplifier - very useful for low level measurments)

I would stay away from the 7B50 and 7B51 time bases. They are an all discrete transistor construction which makes them easy to repair but also have two tunnel diodes which often fail and are almost impossible to get.

For other plugins you may want to check this site:
http://amplifier.cd/Test_Equipment/Tektronix/Tektronix_Test_Equipment.htm

[alm]

Instead of the 7A12, other dual channel 1 Mohm plugins like the 7A26 would also work.

7A13 (100MHz differential amplifier)

Watch out for cracked gears in the version with mechanical readout. The LED version doesn't have this problem.

7A22 (1MHz differential amplifier - very useful for low level measurments)

Note that even though the bandwidth is limited, this one goes down to 10 µV/div with an excellent CMRR. It also has a selectable low-pass and high-pass filter to filter out any noise.

For other plugins you may want to check this site:
http://amplifier.cd/Test_Equipment/Tektronix/Tektronix_Test_Equipment.htm

Yep, that's going to help his desire to collect more scopes .

[Chris Wilson]

If I found a plug in with more bandwidth capability than 100 megs would it still work, but just be a bit wasted? Or do they have to  match the bandwidth of the mainframe? Thanks for the links. Do they do anything useful for testing capacitors? I read about using a basic scope and a signal generator, and testing them that way, but I have also a video about using a curve tracer? The curve tracer plug in appears to be just for testing semiconductors.

[tekfan]

Every plugin works in every mainframe but the bandwidth is limited by the minimum bandwidth of the system. It can either be the mainframe or the amplifier. So yes, as you said you would waste a bit of bandwidth.

The only plugins that are not compatible with the standard mainframes are the 7A16P and 7B90P. The P stands for programmable and are used exclusively by the Tek 7912 digitizers (although other amplifier and time base plugins can be used).

Sadly there are no plugins for testing capacitors. You could do it with the 7CT1N curve tracer plugin but the range would be very limited and percision wouldn't be so great. You'd be far better off with a proper LCR or ESR meter. The 7CT1N is actually very useful for troubleshooting circuits (Huntron tracker style) and matching transistors.

[bullet308]

The number one place on the web for sorting out Tek scope problems:


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

[Chris Wilson]

Thanks for the info re the plugins, I can see me scouring Ebay in my Tektronix anorak soon I couldn't understand how anyone could need or want several scopes unless they were a fully fledged electronics engineer. Now I am afraid to say they have a rather magnetic fascination, and still have me Ebay scope searches live, which is a BAD sign!!

Here's the latest on my 475 machine, which I have also posted to the TekScopes yahoo forum, now I have sorted out why I hadn't received group acceptance (my e-mail address was defaulting to a Yahoo one I never use, so i didn't get any reply..)

Right, I picked the 475 up this evening and have had a word with the seller and a quick look inside it. It was stored under far from ideal conditions for some years in an old building undergoing restoration. It has slight signs of this damp on one board, some of the metal canned transistors have the very first signs of green verdigris / corrosion on their cases, very very minor, mind you. One of the vertical side boards has very minor corrosion on some of the ground foil near a vent on the case, but the tracks still appear fine, it's very localised and would clean off I am sure. When the scope was purchased five years ago it worked fine save for the second channel losing the display if turned up to what he described as 2 megs?? Below this ?? it worked fine, and channel one worked fine in all respects. After coming out of storage and being put in a warm dry environment the problems were apparent with which it was sold.

It will focus quite well, in the centre of the pot. The dot almost appears to be two very short lines one above the other though, hard to say. Intensity needs to be in the last tenth of clockwise rotation to get any display to show. Horizontal control needs to be fully clockwise to bring the dot to the centre of the screen, a 1/4 or so turn anti clockwise and the dot is off the left of the screen. Beam finder creates a very bright bigger dot, with a lot of illumination on the phosphor around this dot.

The vertical control fully clockwise has the dot off the top of the screen, fully anti clock has it a maximum of a 1/4 of the way down. Fiddling with the control will make the display occasionally flash below the vertical half way point instantaneously, as if noise in the pot is doing it.

Randomly probing a few of the voltage test points which have their expected voltages shown on the main board give sensible readings, what the ripple is like I have no idea. There's a point marked "5 V unreg" on the main board, this is showing 62.9 volts. Others are within 0.01 volts of the stamped values on a reasonable hand held DMM, referenced to ground tags on the main board.

Fan works, three or so of the panel switch plastics are broken, Bezel of main front panel has some corrosion, fascia is so so, but perfectly legible and usable. Would you recommend I attempt more diagnostics, or is the above enough to condemn it as a source of spares only? Thanks. I can post photos. I would like to fix it as it would be my first attempt at fixing anything electrical of any complexity, and very different from working on new automotive engine and chassis control systems.

I don't NEED this to work, it won't change my life it it's a white elephant, but I know I'd get a buzz if someone were willing to treat me like a normal Joe passenger who ends up in the cockpit of a plane when the pilot has a seizure and needs talking into landing the thing, however inelegantly All I have are a 150 meg dual channel USB scope, a PM3380B 100 meg dual channel Fluke analogue / digital scope, a bench and a hand DVM, a signal gen that goes only to 1 meg. I have some soldering gear anyone with a true interest in PCB work would probably laugh at, but I don't mind buying a cheap fairly basic soldering un-soldering station, I have fancied something better for years. If I blow it up (further...) I will be annoyed, but would put it down to experience.

[vk6zgo]

You are pretty well set up to work on the 475 & should be able to do any checks needed with the  PM3380B & your DVM.
You quoted the dc volts on some test points as being OK,"but you don't know about the ripple".
Ripple  can be measured with PM3308,or even with the DVM on AC.

"5V unreg",should most definitely not be at 62.96 volts.--something to watch out for,is that some DVMs will read any AC component as a DC voltage when in the DC mode.--Another job for the Philips 'scope!
I would definitely concentrate on that area.
You really need a Workshop manual to make really meaningful checks.--It  can be done without,but entails a fair bit of circuit tracing.
Good Luck,
VK6ZGO

[amspire]

It will focus quite well, in the centre of the pot. The dot almost appears to be two very short lines one above the other though, hard to say. Intensity needs to be in the last tenth of clockwise rotation to get any display to show. Horizontal control needs to be fully clockwise to bring the dot to the centre of the screen, a 1/4 or so turn anti clockwise and the dot is off the left of the screen. Beam finder creates a very bright bigger dot, with a lot of illumination on the phosphor around this dot.

The vertical control fully clockwise has the dot off the top of the screen, fully anti clock has it a maximum of a 1/4 of the way down. Fiddling with the control will make the display occasionally flash below the vertical half way point instantaneously, as if noise in the pot is doing it.

Both horizontal and vertical have similar position problems. Sounds like a sick supply?

Anyway, since the horizontal position responds to the position control, it is more likely the problem is before the horizontal output amp.

Randomly probing a few of the voltage test points which have their expected voltages shown on the main board give sensible readings, what the ripple is like I have no idea. There's a point marked "5 V unreg" on the main board, this is showing 62.9 volts. Others are within 0.01 volts of the stamped values on a reasonable hand held DMM, referenced to ground tags on the main board.

I am pretty sure the "5 V unreg" is actually the "50V Unreg" line. Perhaps the "0' has rubbed off. It comes straight from the rectified output of one of the transformer secondaries, so the only way it can be high is if the mains selector settings are not right for you. You might have the mains set to L (110V nominal) when you should have it set to H (120V nominal). Or as VK6ZGO, your meter may be reading high due to the AC ripple. If the meter is OK and the transformer settings are OK, then I would have to say that the 62V has to also be OK for the 50V Unreg rail.

Richard

[Rerouter]

it is 50V unreg, as for the illumination, follow the calibration guide in the manual for grid bias / vert sheild volts, that fixed mine after i fixed a rail short,

there are alot of supplies, and you really need to make sure all of them apart from the unreg are good, expecially 50V

 -8V, 15V, 50V and 110 are the ones effecting illumination,

as for 2mV div setting, it has a seperate gain pot as well as the 5mV for the remainder on the vert board, used to bring 5mV and 2mV as close as possible, with these you can bring your lines into a more centered position, and dont heavily effect the cal,

[Chris Wilson]

Whoa, my bad, a typo in my previous post, sorry. 5V unreg should have been 50 volts unreg reads 62.9. I played with it until 2 am last night and got a result. Some of the switches are playing up. the bandwidth limiter push / pull one has iffy contacts. I was messing with it and in only full out did I get any display, then half out it started to show signs of intermittent display. I had a good look at how the fingers worked with the push out pegs and cut a clean piece of fine texture paper into a small rectangle. I put the switch fully in, so all contacts were closed and put the paper between the U shaped fingers and the board contact area, applied a little finger pressure and pulled the paper out. the paper came out dirty!. I repeated until it was coming out perfectly clean. the display was more reliable. I did the other contacts the same way, and the display, (dot, such as it was), was pretty much 100% reliable. I then looked to see what other switches were of the same design. the Auto Trigger and below ones seemed similar, but access needed the top board removing, so I decided to give them a two minute work out of switch position changes. I was rewarded with a horizontal sweep Timebase now started working, as well, and the vertical and horizontal position controls work properly now, with the full range of control, and mid  screen = roughly mid position of the knobs The intensity control is now fine too, I can't quite believe my luck, a trace is visible almost from fully anti clock, with a brilliantly bright one on full clock. I believe a lot of the issues are damp related storage mechanical / electrical switching ones. I fell asleep reading the shop manual, as to what methods to use fro switch checking and servicing / cleaning. It says WD40 and Isopropyl alcohol. I used to have the alcohol, for cleaning other things, but it has evaporated in storage, but it was years old. I need to source a supply, I was given the last lot by someone i have lost touch with. I have WD40, of curse! Before I take the top board off for better access to the other switches what do you recommend for switch cleaning, follow the manual or is there something better these days? The traces look a little wide, does the 475 usually have a fine trace? I will use my other scope to measure the power supply ripple when I get chance to have another play, and before taking the top board off, thanks for the tips there.

Hopefully there's a chance the old girl can be revived, if I get it working electrically. Cosmetically a few new knobs and maybe, if removing the cast alloy front frame orbezel, call it what you will, isn't a total nightmare, I might bead blast it and either re anodize it, or powder coat it. The blue texture finish case isn't too bad, but has the odd flaky bits, I would be tempted to bead blast this too, and respray it, is anyone "anoraky" enough to know a paint code? Being in the race engine game has it's benefits here, as I have a big bead blaster, powder coating gun, and paint spray gear.

What I do need are some nice new fine probes for my meters, and some of those little spring loaded, push button grasping things to grip onto test terminal eyelets. I have seen some on Ebay, but they look like they take banana plugs of much less than my stock 4 mm ones. Is there a smaller standard? Can I adapt my leads down to these little grippers?

Thanks everyone.

[tekfan]

I would stay away from WD-40. It's good for repairs that don't need to last. It then just gums everything up to the point where it was worse than before.

If you can, probably the best way is to remove the cam switch assembly and then clean the contacts on the board and the U shaped finges with isopropyl alcohol. Or if you can get underneath the contacts with a paper soaked in alcohol.

This link probably might be of interest to you: http://amplifier.cd/Test_Equipment/Tektronix/Tektronix_other/475.htm

Something like this:

[ModemHead]

@Chris: Wow, your experiences sound almost identical to mine while working on a 465.

https://www.eevblog.com/forum/index.php?topic=6511.0

For those littel gold leaf switch contacts, I used paper soaked in IPA (91% from the drugstore), and then finished off with paper soaked in Caig De-oxit. No WD-40, I was afraid of leaving a film.  I too had lots of trouble with the trigger-view switch, requiring more than one attempt to clean it.  There are also contacts on the back side of the PC which are tough to access.  I resorted to blasting the area with De-oxit in some cases.

Assuming your 475 is reasonably similar to my 465, I would expect the traces to be pretty sharp. Dirty vertical attenuator switches can make the traces fuzzy. You have to take off the shields at the front of the vertical board to gain access.  (See attached photo.)  There are little plastic attenuator modules (marked 100X and 10X) in the way, but they are removable by pulling them straight out of their sockets (with much care.)  If you do this, don't mix up the channel 1 and channel 2 modules.

I ended up doing this operation with the vertical board completely removed from the scope. The attenuator section is passive, so I fed the BNC input with a square wave from a function generator and kept cleaning until the attenuator outputs looked nice and square. They started off looking rounded.

[ModemHead]

<snip>
What I do need are some nice new fine probes for my meters, and some of those little spring loaded, push button grasping things to grip onto test terminal eyelets. I have seen some on Ebay, but they look like they take banana plugs of much less than my stock 4 mm ones. Is there a smaller standard? Can I adapt my leads down to these little grippers?

You may find a pair of Pomona 3782 leads handy.  I certainly do. Similar but cheaper ones from Hong Kong are pretty common, too.

[Rerouter]

hmm, interesting, the 465 only has 2 attenuator modules, while the 475 has 4,

and i thank you for the tip on cleaning contacts, may help fix the one lurking issue on my repaired one

[Chris Wilson]

OK, spent another half hour with it today. I have found more issues. I can't get a stable trace from my sig generator. I can get it to scroll forwards or back, slowly, with painfully careful messing, but it;s nit right. There's something wrong with the triggering I believe. I have also found neither of the voltage rotary controls will go into the highest two voltage positions, they physically stop before the last two divisions, the 20 volt and 60 volt ones. I thought it had the wrong knobs on at first, but looking at ModemHead's lovely restoration thread it seems the 465 has these divisions, am I missing something?

Other issues are with the horizontal display push buttons I can get no sign of a trace with either A (lock knobs) depressed, or with Normal or Single Sweep depressed. Neither the Ready nor Triggered lights can be persuaded to come on.

On anything higher than a few kilo hertz I cannot get a  usable waveform whatever I do.

I haven't used my good scope to look at the voltage ripples, (got a question re safety there), but using my multi meter set on AC all bar the -15V pin give a low 0.004 volts AC on the test pins. The - 15 volt test pin has the decimal point going wild and voltages changing too fast to read. On DC on the -15 volt pin I very occasionally see a very fast fluctuation, too fast again to read.

Is there anything to be aware of using my good bench or USB scope to measure these voltages for ripple, earth loop wise? I would rather ask and be safe than blindly probe and do some damage.

I have asked in ModemHead's restoration thread about a microswitch behind the Time / Division know, but will repeat the concern here:

Driven off the spindle of the Time / Division knob, inside the casing, is a microswitch, with a single round alloy ring with a peg sticking it of it, to activate it. My microswitch moves little, if at all, whatever is done turning or pulling the Time / Division knob. I will get the meter on it later, and see if it actual changes state, I suspect not unless it's incredibly sensitively set. Can you briefly describe how the microswitch should be triggered, and what it does please? I have lifted and reseated all accesible socketed components and plugged leads.

I feel a bit out of my depth, but will persevere, it's a good learning curve on something that hasn't cost too much and is complex enough to be a real challenge to me. Thanks to everyone for the replies and support, appreciated.

[ModemHead]

Yes, you have trigger problems.  Switch the trigger source to "line" and try a 50/60Hz signal (touching the input wire should suffice.)  Is that stable or does it still free-run?

The last two three V/div steps are for 10X probes. There are two lamps behind the knob skirt, one at the 11 o'clock position (1X) and one at the 1 o'clock position (10X). Tek 10X probes have an "actuator pin" on the BNC shell that contacts a ring around the input BNC connector and switches lamps, making it always indicate the correct V/div.

My advice, crawl before walking.  Stick with a 1KHz test signal. Use "A-lock knobs" mode and "Auto" triggering.  When that works, move on to other functions/settings.

When hooking up other scopes, only attach their ground to the 475's ground, or one of the labeled "Ground" test points.