Tektronix TDS420/TDS460 SMPS Green Hybrid

[dzalf]

Dear Smoky and guys

I am currently working on exactly the same Power Supply for a TDS420 that I am trying to resurrect.

Please excuse me for posting here but you already have fresh experience with this PSU. Regardless, I will create a new post on my repair procedure but I wanted to ask you if you could help me out on something in particular regarding the LF1 module from this power supply (it's the potted black Line Filter).

Here's the main problem I am facing (in short): I received this power supply with an exploded varistor (the taped one) while the diode was in good condition. I replaced them with a P130L10 varistor and a 1N5406T-G diode. Then, I fully recapped the supply, fixed some traces and replaced all the ICs.

Once ready, I plugged the supply to a fused 240 VAC line (UK mains) and the varistor went kaboom! Big big explosion!  . (see attached images)

Based on my measurements around the primary side from the PS I suspect that the problem might be an internal short in LF1. It has 3 leads that go to the points marked as E12 (orange cable) E4 (yellow cable) and E3 (blue cable).

I measured a dead short between E12 and E4. E12 goes to the negative pin from C1 and E4 goes to pin 2 from the bridge rectifier.

I wanted to kindly ask you if you could please measure the resistance between these two points to confirm that this short is my main issue (which I believe it is).

Once again, my apologies to all for flooding the thread with a new issue. I just want to confirm my suspicion based on someone who has a working supply.

Best wishes to you all

Cheers

dzalf

[Smoky]

Dzalf,

I have two PS boards in front of me. One is a populated/restored/working SMPS and the other is stripped of its components. I see clearly the locations of E12 and E4.

On the working SMPS, and when I set my DMM to 2K Ohms (diode test), I get a quick beep as I alternate the probes between E12 and E4. Similar to what you'd get crossing a capacitor.

[dzalf]

Dear Smoky

I want to thank you for your help on this. I appreciate your time for measuring the resistance

Based on your measurement I can confirm that the potted component marked as LF1 is definitely shorted. I measured a dead short (0.35 Ohms) between the E4 and E12. This means that one of the AC nodes from the rectifier is directly connected via the short to the positive node from C2, which to my understanding is not right at all.

After some research, I found that this power supply is well known for being also used on some Lecroy Oscilloscopes. According to the diagram (attached) LF1 is only a line filter and removing it should allow me to further test the power supply (granted that nothing else shorted after the big kaboom).

This is where it gets interesting:

Based on the schematic, LF1 only bypasses the AC line. I removed the glue from the black potted part marked as LF1 for inspecting the area and I noticed that underneath there are two jumpers for the L and N from the AC line meaning that this component atop "LF1" is NOT a line filter rather a bank of capacitors (maybe?) which makes me wonder: what is the real purpose of this component then? Additionally,  I still do not understand why that part has 3 wires coming out of the potted body, this would prove my theory that it is a group of caps instead. 

Thank you once again for your time

EDIT: While browsing the forum I found this post: https://www.eevblog.com/forum/repair/tds-460-power-supply-problems/ and it turns out that the potted block is a custom rectifier! Chances are that I might have a hard time trying to find a replacement, so I am thinking of looking around eBay for a second hand PSU 

Stay safe

dzalf

[Smoky]

I pulled the Tektronix part number sticker off of LF1 and found another Power Systems Inc. part number under it.

I think bypassing it is a common practice from what I've read.

If you need any more measurements, just let me know but I think components would need to be lifted to get accurate results.

All of the help I received from Jame_S, Shakalnokturn, CaptDon, and the rest has been spot-on 

[james_s]

Isn't that potted block a circuit for automatically selecting a voltage doubler or a full wave rectifier configuration for 120/240V auto switching? Mine didn't have any issues in that part of the circuit so I didn't look too closely at it.

[Smoky]

Reading the service manual, there should be between 90-250 VAC between pins 1 and 3 on connector J1. That's the 3-pin plug that brings power to the SMPS from the Auxiliary Power Board A07. So something has to bring all of those voltages into line. I can see where I can be asking a ton of questions.

The potted block (LF1) appears to be just glued to the top of the PC board. It has no pins that go through to the foil side. I can tell because having two boards to compare, one of the LF1's is glued down crooked and angled. So it has just the three wires (yellow, orange, and blue).

I have repaired six of these now but they are from the U.S. market running at the lower region of the voltage scale (~120 VAC). No problems.

Maybe in the European market, where the 200+ VAC inputs are more common, the SMPS's were more prone to failure 

My guess is too that the eventual capacitor failures on the Acquisition board, the Attenuator board, etc combined with rotted traces and shorts, caused currents to rise across all LV outputs on the SMPS making it run hotter and working its little self to death

[shakalnokturn]

The 3 wired potted black box is the automatic line voltage selector.

It works like the voltage selector switch on ATX PSU's, allowing or disabling the voltage doubling during rectification only the switch here is a triac.
It's a good idea to remove it and hardwire to the correct line voltage anyway.

If you want to know what's in there out of curiosity or to rebuild one, there's a hand drawn schematic around internet. You can also download one of the low end LeCroy DSO service manual (9304 or 9314) where the voltage selector is not potted and appears on the PSU schematic.

[dzalf]

Isn't that potted block a circuit for automatically selecting a voltage doubler or a full wave rectifier configuration for 120/240V auto switching? Mine didn't have any issues in that part of the circuit so I didn't look too closely at it.

Dear James

You are right. Apparently it is the autovoltage selection circuit. Apparently, I could give it a try without it and it should work just fine on the fixed 240 VAC from the UK.

I will keep trying until I get happy results 

Cheers

dzalf

[dzalf]

Reading the service manual, there should be between 90-250 VAC between pins 1 and 3 on connector J1. That's the 3-pin plug that brings power to the SMPS from the Auxiliary Power Board A07. So something has to bring all of those voltages into line. I can see where I can be asking a ton of questions.

The potted block (LF1) appears to be just glued to the top of the PC board. It has no pins that go through to the foil side. I can tell because having two boards to compare, one of the LF1's is glued down crooked and angled. So it has just the three wires (yellow, orange, and blue).

I have repaired six of these now but they are from the U.S. market running at the lower region of the voltage scale (~120 VAC). No problems.

Maybe in the European market, where the 200+ VAC inputs are more common, the SMPS's were more prone to failure 

My guess is too that the eventual capacitor failures on the Acquisition board, the Attenuator board, etc combined with rotted traces and shorts, caused currents to rise across all LV outputs on the SMPS making it run hotter and working its little self to death

Dear Smoky,

Thanks for the info. You are right, the potted block is only glued to the board and the three wires go to the points we've discussed before. I agree that perhaps these power supplies were more prone to failure under the higher voltages here in the UK.

In my case, the varistor has exploded immediately after plugging the power supply to mains voltage and therefore I cannot consider over-heating as the main issue.

I am seriously thinking of removing the potted auto-voltage selector and give it a try. This test will have to wait until I get new fuses because I've blown all my stock already 

I will get back to you after running more tests before feeding electrons to the little beast. Meanwhile I will run more tests on the components from the primary side since I am a bit worries that something may have got damaged after the last explosion (hoping that the varistor gave its life for the sake of them  )

Cheers

dzalf

[dzalf]

The 3 wired potted black box is the automatic line voltage selector.

It works like the voltage selector switch on ATX PSU's, allowing or disabling the voltage doubling during rectification only the switch here is a triac.
It's a good idea to remove it and hardwire to the correct line voltage anyway.

If you want to know what's in there out of curiosity or to rebuild one, there's a hand drawn schematic around internet. You can also download one of the low end LeCroy DSO service manual (9304 or 9314) where the voltage selector is not potted and appears on the PSU schematic.

Dear shakalnokturn,

You just confirmed what I read in other forums regarding these power supplies: the potted block is an autovoltage selector or what they refer to as "custom rectifier".

In my opinion, manufacturers should NOT be allowed to use potted and/or proprietary components or subsystems which could make devices irreparable. Same case for the main transistor driver block (green thingy  ).

I actually found the reversed-engineered diagram from the auto-voltage selector however, to my knowledge, no one has built their own or at least they have not made it available to the market (and by now, it might be too late as these PSUs are more than "obsolete" in a way).

Finally, thanks for the info, I will check out that service manual right away.

As I told the other guys, I will try to get back to you all (hopefully) with better results. I will definitely try just removing it and testing again. I desperately want to make it work because I've spent many hours (and ££££) recapping, cleaning and re-soldering a bunch of stuff on it.

Stay safe y'all!

Cheers 

dzalf

[Smoky]

I thought it would be interesting to measure the current that is drawn while the PS-1690 SMPS is running inside of the oscilloscope.

I have a Tektronix A622 current probe and it easily clamps around the group of wires from each of the four voltage rails.








Let's see how each rail compares to the max current ratings of the SMPS.

The TDS400 series scopes I own are totally re-capped, including the backplane boards. They also pass their self-tests and SPC routines. I have one opened-up on the bench 

[Smoky]

Here is the -15 Volt rail, pins 1 and 2. The circuit is rated for .75 Amps.

I set the current probe for 10mV/A.





It appears it draws just under 100mA.

[Smoky]

Here is the -6.4 Volt rail, pins 6 and 7. The circuit is rated for 5.5 Amps.

I switched the current probe to 100mV/A.





This appears to translate to ~3.9 Amps.

[Smoky]

Here's the +15 Volt rail, pins 8 and 9. The circuit is rated for 3.5 Amps.

The current probe is set for 100mV/A.





I see around 2 Amps being devoured.

[Smoky]

And the last one, the +5.1 Volt rail, pins 16-20. The circuit is rated for 15 Amps.

The current probe is set to 100mV/A.





Looks to me to be a little over 10 Amps being drawn.

Hopefully, I figured these right. The current probe was a gift from a retired professor from NC State along with his broken TDS420

[Smoky]

When doing the current measurements, I made sure that the "Current" arrow (I) on the probe pointed away from the SMPS.





Looking back at the +15 Volt rail, it appears to have a bit of noise riding on it. I'll compare it to another scope later on

About the Varistor, I read that it is a Littlefuse part:

https://www.digikey.com/product-detail/en/littelfuse-inc/V150LA20BP/F3033-ND/1009262

[shakalnokturn]

Thanks.
Real consumptions are always good to have for those who need a replacement PSU solution.
What state was the TDS in during current measurements?
I'd expect some variation mainly on the -6.4V rail between stopped and sampling at the highest single shot speed.

[Smoky]

Hi ShaK,

The scope was sitting idle with Channel 1 displayed for all of the measurements.

Since you mentioned it, I went back to the -6.4v rail and re-started the scope.

During the time of the power-on self-tests, the -6.4v rail drew ~4.4 Amps, once completed, the current dropped back to ~3.9 Amps:





I then activated all four channels and tried both HiRes and Sample at its highest setting. The current draw was the same for both, ~4 Amps even:





On a side note, if anyone happens to have the Field Adjustment Software for calibrating these scopes, please let me know. I was curious so I contacted Tektronix to see how much it would cost to calibrate a TDS420 or 460. They quoted me $650 per unit plus shipping.