Tektronix TDS 460A repair(power supply

[zupta82]

Hello everyone, happy holidays. I’m using the time I have off from work to try and fix a bunch of stuff that’s been piling up. I have a Tektronix 460A that a friend initially bought just for the power supply to repair his 460A unit. I have had it for a couple of years. The first time when I powered it up it worked for several minutes then shut off. I opened it up and found some corrosion on the solder side and component side. I desoldered a whole bunch of heat-sinked diodes. Here are the current problems I am having:
 
1)  the unit has four sb540 schottkey diodes, two sb560 schottkey diodes, two UF5402 ultrafast diodes. I screwed up royally by losing the last two diodes since the last time I took it apart several years ago. Stupid of me I did not write down numbers or remember the polarity orientation. I’m not really good with power supplies I don’t know what I could use for a replacement?? I replaced all the diodes with new ones while the unit was taken apart. It seems there is no schematic anywhere on the web, some people say it’s close to a LeCroy scope. It’s a really nice scope and I don’t want to fry anything. Here is a pic

[scopeman]

UF5402's are available from Mouser:
https://www.mouser.com/ProductDetail/Vishay-General-Semiconductor/UF5402-E3-54/?qs=%2Fha2pyFaduitxDRewUACpHJGjb9SRBXYJ2%252BxgnZZ2FA%3D
SB560 (Can replace the SB540's as well)
https://www.mouser.com/ProductDetail/ON-Semiconductor-Fairchild/SB560/?qs=mVVXn4M53U%2FtgCVD382p0w%3D%3D

On the Diode Polarity trace the pad connections and find where they connect to the electrolytic capacitors. If the pad connects to the positive terminal then that is the end where the Cathode (banded end) of the diode goes. If you find that the pad connects to the negative side of the capacitor then the anode (the unbanded end) connects to the capacitor. You will find that the other end of the diode connects to the transformer. Remember that on electrolytics the Negative of the capacitor is labeled down the side.

Also for the PCB marking are there any other diodes that are standing up? If so what is the relationship to the marking on the silk screen on the board? I would suspect that all similar diodes are mounted in the same manner but I would make an ohmmeter check for the ones that you have in question.

Additionally you should be able to trace one end of each of those diode connections to the output connector that goes to the scope and determine the polarity of the output. If the supply feed from the connector is a positive DC output the banded end (Cathode) of the diode would trace to this output. If the supply is a negative one the unbanded end (Anode) would trace to this point.

On the similarity between this supply and the LeCroy one is because the same company made it. Also the long green ceramic coated 19 pin control module is the same part for both supplies. There are details and a schematic of the green module in the files section of the LeCroy group on groups.io. Failure of this unobtainable part is rare but a colne for this part has been developed and tested as a DIY project. Details on this also in the files section of the group.

There are also complete schematics for the power supplies used on the LeCroy 93XX series scopes which are similar in many respects to the Tek one. The LeCroy one has IIRC -5.2V, _5V, +15V and -15V outputs. I have attached the schematics for the two common LeCroy supplies so you can trace the circuits there.

Sam W3OHM
Owner and Moderator for the LeCroy Owners Group on Groups.io

[james_s]

When you bench test it, be sure to put a load on all of the outputs. I wasted a ton of time trying to work out why mine wouldn't regulate properly before I realized that it needs a load 

[zupta82]

Ok thank you, I think I can trace the polarity. I’m still kind of confused about which diodes to use. Looking at the data sheet for the UF5402 and SB560 they both have a similar IV curve. It seems the SB560 is an older semiconductor technology with a less of a knee in the curve. I did notice the UF5402 can be pushed a little harder before it flattens out

[zupta82]

The UF5402 is alot  less linear I don’t know why they used just two different diodes in thiis portion of the circuit maybe efficiency problems??

[scopeman]

With switch mode power supplies it's best to use the original parts if at all possible. I did not look in detail at the data sheets but there could be a different reverse recovery time. The designers must have had a reason for selecting different diodes.

Sam
W3OHM

[CaptDon]

The 'original' LeCroy power supply schematics are full of
errors!!! Make sure you have a corrected copy!!!!

[scopeman]

Hello CaptDon,

If you have more updated LeCroy Power Supply Schematics I sure would like to have them for my group. I have fixed quite a few 1715 and 1724 power supplies using those schematics.

I do know that the A1 100004 green module schematic has been updated and I created an entire folder on the site for that project including parts lists, schematics and PCB layouts. The design has been built and tested so we know that it works.

So in closing thanks for the input and feel free to join the group.

Sam
W3OHM
Moderator and Owner of
LeCroy Owners Group on groups.io
https://groups.io/g/LeCroyOwnersGroup/topics

[zupta82]

I’m kind of a beginner in terms of poking around and testing, I used to do more testing in my previous job it’s been about 5 years ago. I’m wondering if I could put a smaller ac voltage than the wall outlet voltage on the input, I don’t have a Variac. As far as testing the 19 pin output voltage connector there are some voltages printed on the large transformer to get started. Also, how would I safely place a load without shorting anything. Here is pic of the transformer . I’m guessing I would take all these measurements referenced to ground with my dmm

[zupta82]

Oh I forgot to mention there is a wall outlet rf noise filter on the scopes chassis. I was thinking do I need to be connected thru this point to get accurate measurements. Here is another pic

[james_s]

Oh I forgot to mention there is a wall outlet rf noise filter on the scopes chassis. I was thinking do I need to be connected thru this point to get accurate measurements. Here is another pic

No, that's just to prevent electrical nose produced by the power supply from leaking out, for testing purposes you can ignore that filter and power the supply directly.

In regards to your other question NO, absolutely do not try to feed these with a lower voltage. I tried powering one of mine up on the bench with a series bulb to limit the current and it blew up, I then had to spend more time repairing that new fault before I could get on with the rest of the project. Do not under any circumstances try powering it with a voltage outside of the 120 or 240V it is rated for.

For a load I used a 21W automotive light bulb, the sort of thing that is/was commonly used in brake lights. To connect to the power supply I used alligator clip leads clipped to the correct pins, using the LeCroy schematic for reference. You have to use the DC ground pins on the output connector, don't try to use earth ground.

Honestly if you are a beginner this is not really a good project. Switchmode power supplies are dangerous and notoriously tricky to work on and this one is a particularly difficult one. I have many years of experience repairing gear and I struggled with these quite a bit before I figured out the issues and fixed both of mine.

Oh and the LeCroy schematics, I don't know if they're full of errors but they are definitely for a different version of the power supply. The PCB is identical but some of the components are different in order to make some outputs the opposite polarity of what Tek used. This sort of thing is/was fairly common with a power supply like this, the OEM had a series of different base models to offer and then the customer could select customized variations of these, choosing the output voltage(s) and polarity and other optional features. LeCroy and Tek both used the same OEM but specified the configuration differently.

[CaptDon]

I gave the corrected schematics to a guy in Raleigh N.C. when I gave him
a TDS420 scope with a blown power module. I seem to recall a resistor that
the schematic called 470K was actually 470 ohms and if I recall there are
two capacitors in the schematic called C4, but one of them is actually C10.
I found some other minor stuff on my own but no longer have them to look
at in my lab. I got the following TDS420 (non-A), TDS644A and TDS644B
from the dumpster at work. I gave the 420 away although its new owner
paid the shipping (He did an extensive article about it here on EEV). The 644A
needed a new Acquisition Board. I still have the old board but it has a bad
MUX chip that is unavailable. I bought the replacement board from Ebay and
cleaned and recapped it, now working perfectly. The 644B had a bad video
display crystal and although the VGA worked perfectly the internal display
was blank. I bought a new crystal and the 644B had the upgraded capacitors
from the factory although there were a few that still need changed. It also
works good as new. The idiots in our CAL LAB threw out a TDS540 because
they couldn't get it to pass Aq self test. All it needed was recapped and we
were using it daily until they aid it won't cal and destroyed it. Dumb asses!!!
It's a union shop and they get really upset when contractors know more than
the rank and file guys.

 

[zupta82]

I installed the power supply and tried powering the unit up, nothing is happening.No smells or sound(no relays clicking or smoke or blown cap smells. Looking at the decal on the transformer the voltages indicate a current measurement as well I don’t have a high current probe for some of these larger currents.  I’m thinking of doing some poking around with my meter at the large white connector. I’m guessing some of these voltages are measured in standby mode. I kind of don’t know where to go next

[james_s]

I gave the corrected schematics to a guy in Raleigh N.C. when I gave him
a TDS420 scope with a blown power module. I seem to recall a resistor that
the schematic called 470K was actually 470 ohms and if I recall there are
two capacitors in the schematic called C4, but one of them is actually C10.
I found some other minor stuff on my own but no longer have them to look
at in my lab. I got the following TDS420 (non-A), TDS644A and TDS644B
from the dumpster at work. I gave the 420 away although its new owner
paid the shipping (He did an extensive article about it here on EEV). The 644A
needed a new Acquisition Board. I still have the old board but it has a bad
MUX chip that is unavailable. I bought the replacement board from Ebay and
cleaned and recapped it, now working perfectly. The 644B had a bad video
display crystal and although the VGA worked perfectly the internal display
was blank. I bought a new crystal and the 644B had the upgraded capacitors
from the factory although there were a few that still need changed. It also
works good as new. The idiots in our CAL LAB threw out a TDS540 because
they couldn't get it to pass Aq self test. All it needed was recapped and we
were using it daily until they aid it won't cal and destroyed it. Dumb asses!!!
It's a union shop and they get really upset when contractors know more than
the rank and file guys.

Yeah I remember you posting about that. I'm really glad I wasn't working there, I'd have ended up fired and in jail for finding the guy who did that and bashing him over the head with something.

[CaptDon]

You really shouldn't go anywhere next!! This is an extremely complicated
supply and easily damaged. It either starts up and runs, sits there dead,
or blows up in your face!!! Why would you want to measure the currents
individually? The thing is dead. The switching part of the 'switch mode' is
killed, there will be no currents and no voltages at the white connector.
It is an all or nothing deal. Because some of the voltages are inter-dependent
everything has to be exactly right for it to power up correctly. In most
SMPS there is an opto-coupler driven by an output voltage to set the
input side duty cycle. This supply is different. The opto-coupler is a shut
down device. If you don't know how to troubleshoot SMPS supplies then
this one is the worst example to simply play with. You won't accidentally
get it to run, but you will accidentally smoke more parts making repair
totally hopeless. You will probably find the switch transistor fried and the
ceramic 50 milliohm resistor is open. You will have to find and replace
EVERY bad component before powering the supply up or it will simply
blow up in a big way again and the more times it blows up while connected
to the rest of the scope the more you will possibly damage the scope.
It needs to be tested into a dummy load first to prove it is safe to connect
to the scope. Also the 'automatic voltage selector with the orange, blue
and yellow wire is often screwed up. If you never plan to run the scope
on 240vac it is best to look up the work around for removing the selector
module and replacing it with jumpers. It makes the input circuit a voltage
doubler on 120vac or non-doubling at 240vac and the B+ on the input
side runs about 280-320vdc and it is 'line power' meaning if you try to
troubleshoot the input side while powered up without an isolation transformer
you will likely blow up some of your line powered test equipment.

[zupta82]

Ok thank you I’ll go back and do some serious research before I go any further. I didn’t realize how extremely serious these can be. All I did was was replace some diodes and clean up some corrosion when I started this project. In my previous work experience I have repaired and calibrated JDSU TBerd bit error test sets and some Fluke lineman butt sets but it’s mostly been just board level and with some passive component replacement. We did have a guy there that messed with -47V DC telecom 3 phase supplies. They looked pretty complicated. I remember him wearing a face shield and and not having his ESD wrist strap on most of the time.

[zupta82]

I guess the general consensus here is SMPS usually are far more than just visual faults like solder joints and board corrosion. The bulk of the corrosion was on the diodes that I replaced. I just figured it wouldn’t be a bad idea to replace all of the similar components while I was deep in the board??

[james_s]

I guess the general consensus here is SMPS usually are far more than just visual faults like solder joints and board corrosion. The bulk of the corrosion was on the diodes that I replaced. I just figured it wouldn’t be a bad idea to replace all of the similar components while I was deep in the board??

Replacement of components for no particular reason is almost always a terrible idea. Every part replacement brings with it the risk of damage and mistakes. Maybe you misread the value and put the wrong part in there, or don't understand why a specific part was chosen and replace it with something "similar" or with a part that turns out to be counterfeit and not what it actually says it is, or maybe you damage the PCB or create a solder bridge without noticing that causes it to blow up when you apply power. These are things that can and do bite even experienced techs.

When repairing something like this it's important to first understand how it works, then you have an idea of where to look and with experience you start to know what sort of faults are likely to occur and how to track them down. With these power supplies the large electrolytic capacitors on the output side are known for failing and leaking electrolyte onto the board which makes a mess so I replace those. The next thing to look for is cracked solder joints where the daughter board mounts perpendicular near the output end. On the input side the circuit that automatically selects for 120/240 input is known for being problematic. They are complex and of an unusual design that is poorly documented and difficult to troubleshoot. I would not even touch one that someone has blindly replaced a bunch of parts already, they're tricky enough to fix even when you know nobody has messed with it.

[Hexley]

If you have more updated LeCroy Power Supply Schematics I sure would like to have them for my group. I have fixed quite a few 1715 and 1724 power supplies using those schematics.
Sam
W3OHM
Moderator and Owner of
LeCroy Owners Group on groups.io
https://groups.io/g/LeCroyOwnersGroup/topics

If you are collecting errata, here is something I found when repairing a TDS460 supply some years ago:

The Zener diodes + signal diodes series networks in all the voltage regulators do not go to VEE. The connections from CR110, CR210, CR310, and CR410 that are labelled "VEE" actually go to node E506.

Explanation: The zener networks create a signal that will be driven positive or negative if one of the respective outputs exceeds the limit, forcing the associated zener to conduct.  That signal is routed to node E506. The out-of-regulation condition will then be detected by the window comparator, U701 (outputs 13 and 14).   They probably meant to call that diode summing node Verror or something like that, and somehow it got noted as VEE instead.

Oh, and for the record the fault in the supply I repaired was C501, the 2.2uF/50V filter cap in the positive auxiliary supply. It had gone open circuit. No sign of leaking or bulging, BTW.

[zupta82]

I remember my boss at work who is 35 year experience EE saying it could be a capacitor. I was thinking this could be true because of inrush transient response that has to occur during start up. I guess this is why I could not hear any relays clicking. Somebody previously mentioned a device that is like an optoisolator. Would this possibly be the black box like device next to the really large electrolytic caps that says power systems inc?? Here is pic

[CaptDon]

C'mon man, that black box device with the orange, blue and yellow wire
sticking out is exactly what I told you it was in a previous post. The other
littler device is a bridge rectifier. This is hopeless.

[zupta82]

Oh sorry I had brain fart. I will pay attention next time

[james_s]

Oh sorry I had brain fart. I will pay attention next time

When working on a line powered switchmode power supply like this there won't be a next time if you don't pay attention. There is no second chance if you get electrocuted or shrapnel from an exploding semiconductor hits you in the eye. This is not a beginner project, it's dangerous if you don't know what you're doing and expensive if you screw it up, I highly recommend you set it aside and get a few more years of practice or send it out to somebody who knows what they're doing.