Tek 475A with DMM44 service manuals / schematic

[calin]

I can get an 475A - nice 250 Mhz oldie - for about 50$ . Unfortunately the scope is broken (dead CRT) so I want to check if I found the right manuals. Just to do a bit of homework before I "blow" 50$ . I want to fix this oldie and give it a new life ... just thinking I will have an Jim Williams favorite on the bench .. well sounds nice

All I managed to dig out off the interwebs are the Bama manuals and a manual exactly for 475A with DMM44 option - here

http://bama.edebris.com/manuals/tek/475
http://bama.edebris.com/manuals/tek/475adm44/index-9595.html
https://dallasmakerspace.org/w/images/e/e5/TEK_475A_Service.pdf (no schematics scanned unfortunately)


I am pretty sure all 475 models are very similar .... so the schematic for 475 should be more or less OK to fix 475A. Or is my thinking completely wrong ?!! 

Does anyone know of an service manual for 475A with complete schematic ?

[Bored@Work]

I am pretty sure all 475 models are very similar .... so the schematic for 475 should be more or less OK to fix 475A. Or is my thinking completely wrong ?!! 

I have never compared a 475 with a 475A, so I can't give you a direct answer. What I know from other Tek oscilloscopes is that Tek engineers worked for their money. I.e. when Tek released an A or B version of an oscilloscope there were big changes inside. Not just a few components changed and a new sticker slapped on the case.

[alm]

http://artekmanuals.com/manuals/tektronix-manuals/
http://www.qservice.tv/vpasp/shopexd.asp?id=120&bc=no

Both will provide you with complete, high quality scans for a modest price.

[PaulAm]

That's the right price for a dead 465/475.  Odds are high that it's a power supply problem that's a relatively easy fix.

I always liked the looks of the DM piggyback.  It's not a great multimeter, but it adds a digital display for the delayed sweep, which beats reading a 10 turn pot.

I have a 475/dm44 in my "to be repaired" pile.

[calin]

Hmm that one @ qservice.tv looks good alm - thanks.

I am kind of leaning to think that they are different because all other 475's are 100MHz while the 475A is a 250MHz machine. And also why would Tek bother put out a separate service manual if it was not different ... chicken head deduction :-D

I am going to get the thing today and I will see .. i suspect dead power supplies and blown tantanlum caps all over. On the other side I am really getting it for the thrill of fixing.

[grumpydoc]

because all other 475's are 100MHz while the 475A is a 250MHz

475 is 200MHz, 475A and 475M are 250MHz

The 100MHz 'scope in that series is the 465.

[calin]

Right, gotta revive the thread

I got the 475 and all fine in dandy .. obviously the power supply electrolytic caps are .. boiled. Funny , one of the caps did the death "hisssss" while I was testing the supply rails .. literally it blew the blue smoke (or I should say white in this case) in my face. Just for an added hint that yes the caps are busted.

So now I am out to shop for cap but man .. this sucker has some exotic values that are almost impossible to find stocked.  For example 350uF/75v or 5500uF/35V .. good luck finding these stocked at  normal prices... What I found @ mouser digikey et all with these values were like 75$ for a cap. That kills it because remember .. i paid 50$ for the whole scope.

Given tolerances most likely I can replace them with more common close values and match capacity also by hooking in parallel:

For example:

350uF/75V = 22uF + 330uF / 100V
5500uF/35V = 2200uF + 3300uF / 35V


So here is my thing, since I am not a master. Will paralleling affect in any way negatively the behavior of these caps? From what i've read many say that is actually better to make up a big fat capacitor out of smaller ones  I know that higher rated voltages are OK ... but what about paralleling ? I have  plenty of space to wire these however I need inside the scope .. these old caps are huge compared with the new ones.

[Bored@Work]

Forget about paralleling. Use the nearest standard value, at the right or higher voltage, rated for the same or higher temperature and be done with it.

[calin]

Hmm I was thinking @ that .. but I wonder.

Normal close values I found stocked @ non insane prices are: 330uF instead of 350uF and 5600uF instead of 5500uF .. nice brands like Chemicon or Panasonic or Nichicon all rated to 150 degrees and needed voltages.

[SeanB]

350uF 75V = 470uF 100V 105c electrolytic.

5500uF 35V = 6800uF 35V.

Both should be the same size or smaller than the originals and should fit. Buy Rubycon units or Panasonic, not cheapie brand ones.

[calin]

I went the close route .. no paralleling etc.

350uF is got 470uF, 5000uF and 5500uF i g. got 5600uF/35V. Order is on its way from Digikey (about 16$ worth of caps) - once I get them up and see the results I-ll post more detailed what I had each of the old caps replaced with so the thread will help others that may have to fix a 745A Tek.

I hope are just the caps ... remains to be seen.

[Lunasix]

I don't know for the 475, but I have had a 465 which had a failure and tried to repair it with 465B schematics : totally different ! I then found (from Tektronix directly !) the right schematics and solved my problem.

[calin]

You are right. 475A is different from the "standard"  475 .. i got the schematics for both 475A and the DM44 so I am set. Of course there are similarities but they are different - for example the dreaded power caps are different values in 475A.  At least that is different .. from what I've seen so far. After I get the power supply going again will see what comes next. I am hoping stops @ the power suppy  

[PaulAm]

It's very rare that a power supply fault will take out other components in the 4xx series.  Doesn't mean everything works though, but you might get lucky.  On the whole, they are very reliable scopes.

When you remove the caps, the cans have 3 terminals along with the center + terminal.  It's a lot easier to get to them if you take out the trigger board, but it can be done.  You need to bridge the 3 ground holes on the PCB because Tek used the cap as a jumper on one or more of them.  Also exercise care when removing the old cans - it's very easy to lift the traces on the board.  This is a job for good desoldering equipment, although it can be done using a solder sucker, solder wick and care.

Check the bridge rectifiers while you're in there.

When I do this repair, I use 1 inch round PCBs that convert the standard snap cap into the 4 terminal mounting and provide the required bridging.  Makes for a neater repair.

[calin]

Will see .. i hope the rest is OK because everything looks neat and clean no traces of bad abuse anywhere. Switches and all seem clean - for example the trigger source switches i have them out are clean and shiny - and these are probably the most exposed things in the whole scope.

To get the caps out I used my side cutters and cut 3 legs off and then desoldered the 4'th because you can't get with the cutter to all pins. Then I just heated the cut offs a tad and they came of easily. The trick to not lift the traces and pads is to heat the pin not the pad. After that wick off excess solder and done nice clean work without drama. I did took the trigger board out.

I did noticed the can is used o bridge GND traces and i like the round PCB idea but i believe these  need to be custom made .. so simple black jumper wires will probably do the job while not look very fancy. But meh .. they are inside

[calin]

Long time no updates .. but hey I am on the good track to restore the machine.

Replaced the big fat caps with new ones  ... and I managed to get back to life the power supplies. i had to replace opamps and tantalum caps, a bridge .. but hey all power supplies are spot on smooth as a baby's bottom. Even more all front controls for teh scope look like they work perfectly fine, i can see trigger active etc. which is a pretty good sign.

I used the black wire GND connection - actually you have to bridge only in two places . So far the repair looks neat.

Now i have one question - if you look at the LV power supply schematic - CR1415 - is blown. I disconnected it after it took out the smoothing cap on the 15V rail supply out , after I fixed them (so i cooked again an electrolytic ).  After diconnect I got the supply's going again ; of course with the required fixes etc. The service manual says

152-0107-00  SEMICOND DEVICE:SILlCON,400V,400MA     G727<--(this is the manufacturer part number)

that is a Si rectifier diode with 400V reverse voltage and 400mA forward current found no eqiv for it. G727 is some old diode made by Texas Instruments from what I managed to dig out, and i did not find a data sheet.

Like I said, i am quite a newbie, but I read up and learn as i blow things up ... now can someone illuminate what is the role of that diode between the 50V unregulated and 15V unregulated rails ? Some kind of surge suppressor - discharge thing ? Everything works fine without it but I wonder what can be the long term effects of it missing. I guess i can replace it with something similar also.

Back to the restoration/fix status ; power rails are ok, all looks good on the front panel but one huge missing thing - CRT is literally not turning on. Confirmed that is the HV generator; that uses a HF generator ~50Khz and and fly-back transformer to generate the 2KV to drive the CRT. Sometimes when I turn the scope on I can hear some little crackling. Just more fun ahead .. i hope to get away without needing a HV probe

[tautech]

For the diode a 1N4004 to 1N4007 will be fine.

The HV flyback circuit can be a temperamental thing to get going. Let alone the measurement and adjustment of Grid voltage which is at usually minus 2 KV or more.
The Flyback also feeds a Quadrupler or Quintupler for PDA at 10 KV or more. Don't think you must measure this in fact do not!
Often it is just the power BJT but not always.
You have the schematic, you will win!

[calin]

Yeah, I will try to not stick my fingers in 2KV or more. What I meant to say it was that the HV oscillator is not working. It is supposed to oscillate @ around 50KHz  but is not. I am hoping to be able to fix the low voltage and the HV part to come alive by itself.  If not I will go an buy a HV probe for my 87V and yes .. the i will MEASURE that 10KV.

The HV supply docs and schematic are gorgeous .. there are plenty of test places and voltage values listed and in top of that nice waveforms in various points to help with troubleshooting.  Nice "operation theory" to describe how it works. I literally love what Tek guys did back in 76 - yes that's when my machine seems to have been built.

I wish we wold get on the form an old Tek engineer .. I saw that we got someone from Fluke . That thread was a really nice read yesterday. We should put the kids to read that instead of watching utter crap on Nickelodeon

[SeanB]

Petition Dave to get W2AEW as a guest again.

Listen to ep117 to hear him.

[tautech]

Yeah, I will try to not stick my fingers in 2KV or more. What I meant to say it was that the HV oscillator is not working. It is supposed to oscillate @ around 50KHz  but is not. I am hoping to be able to fix the low voltage and the HV part to come alive by itself.  If not I will go an buy a HV probe for my 87V and yes .. the i will MEASURE that 10KV.

The HV supply docs and schematic are gorgeous .. there are plenty of test places and voltage values listed and in top of that nice waveforms in various points to help with troubleshooting.  Nice "operation theory" to describe how it works. I literally love what Tek guys did back in 76 - yes that's when my machine seems to have been built.

I wish we wold get on the form an old Tek engineer .. I saw that we got someone from Fluke . That thread was a really nice read yesterday. We should put the kids to read that instead of watching utter crap on Nickelodeon

If you can measure to 2.5 KV that is all that is needed.
PDA can bite even when off ! You must discharge it before you do anything with the CRT.
Much of the circuitry to do with the HV oscillator and its secondary is under stress and is not uncommon to give problems. HV caps, BJT's, high value resistors and diodes all should be checked along with the ordinary components. HV diodes have a high Vf often 5 V or more.
Check transformer windings too.
Most discretes are available and one can substitute if needed.

[PaulAm]

Check the ceramic resistor network for cracks.  It's used in the regulator feedback loop and may kill the HV if it's open.

If you have to replace any of the HV components, you'll need a 2KV probe for your dmm to set the HV correctly.  Some of the functionality requires that to be set within +/- 50V (I haven't looked at the manual, it might be closer than that).

[calin]

Darn .. i am back to square 1.. power rails. So after i got them going I said to get the machine running for 20-30 mins as it is to see if they blow up. I did that the other day and .. well the 5V supply blew.

So I checked for shorts .. and yes 0.42 something ohm between GND and 5V test point .. dead short.  Normally the manual says it should be around 46 ohm. So I set out to fid where it is .. usually such dead short are tantalum caps and blown diodes .. more likely blown diodes cause caps still have some resistance left in them.  So I started with these ... but .. i took literally out all .. and I mean all parts that make the 5 V regulator and still did not managed to find the damn thing. SAme dead short.

So I .. must be some other board that shorts the 5V supply right ?  I even used the 4 wire short tracing method to no avail .. well I got lost on the board and to do that right you have to take the whole board out .. which I might do but I just need more beer and patience .

Then I said hell .. brute force approach .. bench supply into the 5V rail with controlled current. So I do that .. nice 5V .. and start ramping current slowly up to see the magic smoke escaping. I had to ramp almost to 3A  to se the supply starting to drive the short but no smoke .. not even heat ... WTF moment at this point !!  I don't have a FLIR or thermal camera so the simple just feel the heat or see the smoke method I tried is no big use. But still .. to drive a short 5V @ almost 3A is a lot of power .. that has to go somewhere .. make things glow .. but is like current is going in a black hole !!!

Any idea what the heck can suck current in such an efficient  way ?

JOKE .... Maybe I can make an extremely good and effecient bench load out of this scope .. black hole current sink

[tautech]

Biff Tant caps anyway, that old stuff will only give you more grief. Replace with new at higher voltage spec or good quality low ESR electrolytic. Tants can test ok but still break down.
Try cutting tracks to find the area where the 5 V rail measures low ohms.

Power up through an incandescent bulb so all rails are receiving some power together.
Bulb will start bright then dim if all is OK. Have a few different bulb Wattages.

[PaulAm]

Having an ohmmeter than can read to the milliohm range is useful to track down shorted caps.  Just hook one lead to ground and follow around the +rail until you hit a minimum; that will be at the bad part.  Some ESR meters will do this as well.  You really do need 0.001 ohm resolution for this.

Another tool is a current probe.  Hook up a pulse generator and use the current probe to follow the current path around and through the board.  On a multilayer board this can follow internal traces as well.  It will eventually lead you to the bad part.  HP used to make a really cool probe, the 547a, that you could use for this.

It usually takes me less than 10 minutes to find a shorted cap using one or the other of these.

[calin]

I am 100% is not something on the 5 v power supply because like I said I took out everything that makes the 5V regulator.
; literally I left only two resistors that are the 5V reference from 50V rail. I have 50.001V on the 50V rail, -7.9987 on -8V, 111.02 on 110V rail (that is normal), 14.997 on 15V ..... and 0.1 mV something on 5V

So I started to diagnose by taking out load from the 5V rail one module at a time; so far the HV supply and Z axis are not shorted so must be somewhere on another module. But this take its own sweet time because the 5 V rail goes to almost all modules ... I think I will take the main board out and do it that way. But at the same time I go module by module, when you take out the shorted one the 5V will spring to life, so I get a hit on which module is the short to close in. What puzzles me why the heck when I pushd 3 amps in the 5V rail nothing got smoked out ... does not makes sense. Hell the leads/wires from the bench supply to the scope almost meted and nothing on the scope ... Is like I have the chassis connected between + and - and I am trying to melt that

Yeah it wold be nice to have a current tracer ... but even the old ones are like 200$ on ebay... not worth the hassle. I have a Fluke 8842A that does 0.1m ohm, with 4 wire probes, I will go with that one after I narrow things down to a module.  I am really restoring this thing for the sake of history and beauty .. I have better scopes ... but I want one of these old Teks just for the cool factor