Author Topic: Restoring the Lambda LQ-532V power supply - end of project (Read 36031 times)

SnakeBite · « **on:** April 20, 2011, 09:59:47 am »

Hello,

i bought a few month ago this PS form Ebay in 50$ shipped (REAL BARGIN).
to anyone that doesn't know this PS , Lambda PS in the industry standard and it exist many years. this LQ series it there best series yet it's Linear and it's very reliable.
the main reason i bought this one because it has the best regualtion that i found - 0.005% and amazing ripple 150uV MAX (yes it's 150 micro volt max ripple at full load). this PS is 40v @ 5A .

the first time when i got this PS it seems rusty from the outside so i decided to open it before connecting it to the mains and i saw that the transformer and the transistor have rust on them and some of the resistors has corrosion on them and the electrolytic caps have some white leakage on there leads BUT i desided to connect it anyway and see if it works and surprisely it did. so i decided to restore it form the inside out. replace all the parts that need replacement and paint the chassis.

my questions are :

1. do i have to replace ALL the capacitors or just the electrolytic one (even that i check the electrolytic with a ESR meter and they all were fine)?
2. i know now that rust on the transformer it not damaging it performance (correct me if i'm wrong) . what is the best way to remove the rust and the corrosion form the pcb and components? i was thinking about alcohol (90%) or squeezed lemon juice. what's better?

here are some pictures:

Uploaded with ImageShack.us

front panel:

Uploaded with ImageShack.us

Excavatoree · « **Reply #1 on:** April 20, 2011, 11:50:20 am »

I don't see any need to replace any of them - they look dirty but that doesn't affect their serviceability. You've checked the ESR and found that to be OK, unless someone else knows a good reason to replace them, I'd leave them in there. The non electrolytics are generally non perishable; you would only need to replace them if damaged or non functional.

The surface rust on the TO-3 cases isn't a problem, either, in my opinion. The boards look good. Unless I'm missing something, I'd restore the rust as you see fit and leave the electronics alone, unless a repair is needed.

saturation · « **Reply #2 on:** April 20, 2011, 01:20:30 pm »

I concur with Excavatoree, don't change anything that isn't broke. You're lucky!

This unit has a very interesting look, it looks like military salvage, but spent a lot of time either in a hot ambient weather, and or humid or it was in a boat.

I wouldn't paint the unit either, but you can pretty up the housing with rust remover and sealant [Home Depot stuff] , and just air blow interiors. Although some connectors and joints look awful externally, its very likely the electrical connections are still pristine underneath.

To be sure nothing needs to be repaired, you need to performance test it; to run it at 40V at 5A and test its specs. It will run hot, and if something is off, it may blow. If nothing blows, you are good to go.

SnakeBite · « **Reply #3 on:** April 20, 2011, 01:35:08 pm »

i'm a Perfectionist, so i started to clean all the pcb not from all the rust and capacitor acid.
will post photos when i'm manage to clean the most of the rust and dirt.

Ido

p.s. i sent the PS to the LAMBDA representive here in Israel (the test was free so...) and there conclusive was:

1. the device doesn't manage to work full load.
2.rust and acid in the PS and the PCB.
3.dry capacitors (but i test them in circuit with ESR meter and they was fine)

this device is old and we don't have spare parts to fix it.

but i'm almost sure they just open it saw all the rust and corrosion and just look on the acid on the capacitors and just wrote that to cover there as* .because i looked everywhere but couldn't find any mod that they solder anything and even the fuses are the same.

about the caps,i don't want change them if they work either but what if they fail in a few years and they take with them a transistor. this transistors are very RARE and hard to find. should i take it under consideration??

Excavatoree · « **Reply #4 on:** April 20, 2011, 01:50:49 pm »

Just some additional thoughts:

I can't see the unit in person, but if there are leaking capacitors, of course they should be changed. The in-circuit ESR readings may look good because a good capacitor may be in parallel with a bad one.

I can't comment about transistor availability, intuition leads me to think that you should be able to at least find a suitable replacement part if the exact part isn't available.

scrat · « **Reply #5 on:** April 20, 2011, 01:52:21 pm »

Quote

about the caps,i don't want change them if they work either but what if they fail in a few years and they take with them a transistor. this transistors are very RARE and hard to find. should i take it under consideration??

In the while you could try finding some of those transistors, maybe you can have them for few bucks, since you're not in a hurry...

saturation · « **Reply #6 on:** April 20, 2011, 03:05:12 pm »

Hi Snakebite,

It seems you already know the status of its parts and performance from the best source, the Lamdba rep!

Given your location and if you can get a hold of them, replace the electrolytics as they do have a known limited lifespan and in your case case its better to do so, than risk damaging the power transistors.

Quote from: SnakeBite on April 20, 2011, 01:35:08 pm

i'm a Perfectionist, so i started to clean all the pcb not from all the rust and capacitor acid.
will post photos when i'm manage to clean the most of the rust and dirt.

Ido

p.s. i sent the PS to the LAMBDA representive here in Israel (the test was free so...) and there conclusive was:

1. the device doesn't manage to work full load.
2.rust and acid in the PS and the PCB.
3.dry capacitors (but i test them in circuit with ESR meter and they was fine)

this device is old and we don't have spare parts to fix it.

but i'm almost sure they just open it saw all the rust and corrosion and just look on the acid on the capacitors and just wrote that to cover there as* .because i looked everywhere but couldn't find any mod that they solder anything and even the fuses are the same.

about the caps,i don't want change them if they work either but what if they fail in a few years and they take with them a transistor. this transistors are very RARE and hard to find. should i take it under consideration??

FreeThinker · « **Reply #7 on:** April 20, 2011, 05:04:21 pm »

I would remove the transformer and get rid of the rust on the core. These soft iron cores are built up from layers of thinner metal separated by a varnish insulating paint. This gives you the 'laminated iron core' of the transformer and governs lots of things but mainly the eddy currents and associated iron losses. Rust will affect this by shorting out the laminations so a good clean will improve efficiency of the transformer at least and could even make it run cooler. Also check that the rust has not made the laminations separate as this can cause an annoying buzz and again reduce efficiency and create heat. If you find the laminations are loose a coat of varnish on the outside can help.

SnakeBite · « **Reply #8 on:** April 20, 2011, 10:21:47 pm »

thanks guys for all the comments.

i manage today to clean all the pcb that sit on the side of the PS (near the big capacitor)
also i scrape away almost all the rust from the top and sides of the transformer and i have to tell you it doesn't look that bad . the laminated are broken only at the sides of the transformer. tommorow i will continue cleaning.

here are some photos:

Uploaded with ImageShack.us

the rust you seeing on the transistor is AFTER i cleaned it .

Uploaded with ImageShack.us

Uploaded with ImageShack.us

you can see on both sides the iron core pops out for the rust and i'm not done cleaning it so i don't think i need to restore the transformer after all...

Uploaded with ImageShack.us

Uploaded with ImageShack.us

see ya
Ido

saturation · « **Reply #9 on:** April 21, 2011, 12:24:45 pm »

After reading all your good work, great cleaning job, and photos it appears like many capacitors are also custom made by Lambda? It would seem not only are the transistors difficult to replace, I'd expect these parts too. Thus, its probably best not to replace anything as I'd surmise these capacitors are far better than commercial grade. I've retracted my previous post.

The transformer is also wrapped in an uncommon way, very 'military' look. Again, seems best to just clean it and enjoy what you have, not replacing parts unless it does fail.

tekfan · « **Reply #10 on:** April 21, 2011, 04:44:03 pm »

Quote from: saturation on April 21, 2011, 12:24:45 pm

After reading all your good work, great cleaning job, and photos it appears like many capacitors are also custom made by Lambda? It would seem not only are the transistors difficult to replace, I'd expect these parts too. Thus, its probably best not to replace anything as I'd surmise these capacitors are far better than commercial grade. I've retracted my previous post.

The transformer is also wrapped in an uncommon way, very 'military' look. Again, seems best to just clean it and enjoy what you have, not replacing parts unless it does fail.

I had four Lambda LP-522-FM supplies. The construction is very similar. One of them had a leaky capacitor and wouldn't output the rated current. after changing it everything worked well.
The capacitors in this one still look good. Before you change any of them I suggest that you try and tweek the trimmer potentiometers inside and see if it makes a difference.

I can see three trimmer pots on the main regulator board. One is probably for the maximum voltage output and the other one for the max current output. The long blue multiturn one may be for the offset correction when the output is set to 0 volts. The red and blue vertically mounted trimmers on the meter board are also for the current and voltage indication.

I used a wire brush for cleaning transformers in the past with very good results.
The transistors can be easily substituted with new ones (2N3773, 2N3055, etc.)

saturation · « **Reply #11 on:** April 21, 2011, 06:27:48 pm »

This is great to know, tekfan. A few questions:

You say you had 4, where are they now? Have they died or have you just moved on and got newer units?

How has the transformer faired in its lifetime? Still working?

Can you replace the Lamdba labeled capacitors with any Digikey accessible capacitor and still get the legendary ripple performance of these units?

Substituting the board transistors, are these very similar electronically or just 'good enough' replacements and did replacing it affect performance?

I ask because as I googled these units, a number of old working units appeared for reasonable prices, most are over 20 years old, making the caps and transformers inline to fail 'any day' now.

Quote from: tekfan on April 21, 2011, 04:44:03 pm

Quote from: saturation on April 21, 2011, 12:24:45 pm
After reading all your good work, great cleaning job, and photos it appears like many capacitors are also custom made by Lambda? It would seem not only are the transistors difficult to replace, I'd expect these parts too. Thus, its probably best not to replace anything as I'd surmise these capacitors are far better than commercial grade. I've retracted my previous post.

The transformer is also wrapped in an uncommon way, very 'military' look. Again, seems best to just clean it and enjoy what you have, not replacing parts unless it does fail.

I had four Lambda LP-522-FM supplies. The construction is very similar. One of them had a leaky capacitor and wouldn't output the rated current. after changing it everything worked well.
The capacitors in this one still look good. Before you change any of them I suggest that you try and tweek the trimmer potentiometers inside and see if it makes a difference.

I can see three trimmer pots on the main regulator board. One is probably for the maximum voltage output and the other one for the max current output. The long blue multiturn one may be for the offset correction when the output is set to 0 volts. The red and blue vertically mounted trimmers on the meter board are also for the current and voltage indication.

I used a wire brush for cleaning transformers in the past with very good results.
The transistors can be easily substituted with new ones (2N3773, 2N3055, etc.)

tekfan · « **Reply #12 on:** April 21, 2011, 07:39:46 pm »

I had a rack of 4 supplies that I gave to a friend. They are still working like new. I'm running out of space so they had to go somewere. Besides, I've built enough supplies myself.

I once shorted one power supply out and the current limit was set to full. The supply began to heat up without me knowing it. Then I smelled something. It was a capacitor which I put on the supply so I wouldn't loose it. By this time it was melted beyond recognition. I couldn't touch the PS since it was so hot. Inside there is a thermostat that trips at 120°c. I know this is hot enough to boil water but up to this it still seems to be ''normal'' operating temperature. I disconnected the supply and let it cool down for a couple of hours. It worked fine afrerwards. They really don't make'em like they used to.

All the capacitors can be replaced with standard ones. The orange ones are mylar and have a tolerance of less than 10% but these almost have unlimited life. Electrolytics are also standard.

The TO-3 packaged transistors can be replaced by 2N3773 (it's self a very good transistor). Equal or better substitutes can easily be obtained. All the PS cares is weather it's NPN or PNP and the beta (hfe) of the transistor. No need for HFtransistors since everything is DC. There are some things that are on the main regulator board that are somewhat difficult to replace like the dual transistor pair (small round can with multiple leads on main board). It can be substituted with two small signal transistors that are matched for their beta (of example (2N3904/2N3906). The FM-52x-FM series have the MC1466 / MC1566 (mil spec) regulator IC. This is very hard to obtain but I've built several substitutes for this IC out of discrete transistors.
http://i1138.photobucket.com/albums/n537/tekscopes/boardMC1466.jpg?t=1300195211
http://i1138.photobucket.com/albums/n537/tekscopes/MC1466replacement.jpg?t=1300195263

The PS in this thread seems to have a LM124 (standard quad opamp) for the regulator which is easily obtainable. I wouldn't worry about the transistors because there is quite a bit of protection circuitry around to prevent voltage spikes, reverse voltge, overvoltage, etc.

Even the front panel potentiometers can be opened up and easily cleaned if dirty.

Richard W. · « **Reply #13 on:** April 21, 2011, 07:55:08 pm »

You should replace the thermal grease under the TO-3 transistors and clean the heat sinks to improve the cooling. (The mica-isolators can be replaced with the silicone pendant)

I would also replace the mounting-screws, nuts, washers, ... of the TO-3 transistors because they are the transistors collector. The contact resistance should be as low as possible.

To clean the boards you can use a mixture of 75% isopropanol and (distilled) water. (Tektronix recommend this for its oscilloscope boards). Please do not use methylated spirits, it leaves a white residue.

After all this treatment connect a high power load (car headlight bulb) to the PSU and let it heat up, to get rid of all the moisture.

Isolation resistance and protective earth resistance should be measured if possible. Every electrician can do that and check if everything is safe.

SnakeBite · « **Reply #14 on:** April 21, 2011, 11:56:08 pm »

thanks you for you comments tekfan and Richard i will take you comments under consideration.

ok guys i finish cleaning half the PS today.
tomorrow i will continue .

i took apart all the front part of the PS including the transformer and clean it completly (the yellow thing you see on the transformer is soe kind of glue).
tommorow i will clean the transistors (i started to disassemble the heatsink that hold them)

here are some photos:

Uploaded with ImageShack.us

Uploaded with ImageShack.us

p.s. tekfan , how to do know the part number of the power transistors? i even have the schamatics and don't know them.
you can find the schamatics here : http://128.238.9.201/~kurt/manuals/manuals/Other/LAMBDA%20LQ%20530%20Series%20Instruction.pdf
they has all the part number in there BUT lambda part number :-(

tekfan · « **Reply #15 on:** April 23, 2011, 01:31:45 pm »

Quote

p.s. tekfan , how to do know the part number of the power transistors? i even have the schamatics and don't know them.
you can find the schamatics here : http://128.238.9.201/~kurt/manuals/manuals/Other/LAMBDA%20LQ%20530%20Series%20Instruction.pdf
they has all the part number in there BUT lambda part number :-(

I don't know the exact part number but I would try and substitute Q2-Q6 (transistors on big heatsinks) with 2N3773, Q1 may be a different one.
But i wouldn't go replacing the transistors just yet. They are very rugged and don't fail easily. If you really want to check if any of the transistors is shorted,try to unscrew the blocks that the transistors are screwed on. The small blocks are glued on the main heatsink and don't come off easily. Then unsolder atleast two wires off the transistor and then use your multimeter and check with a multimeter (diode test function).

The transistors have a standard TO-3 package pin configuration. Put the red lead of the multimeter on the base and the black to the emitter and then the collector. you should get a voltage drop of about 0,6-0,8 volts. Put the multimeter on the ohms range and then probe the transistor from collector to emitter and then reverse the multimeter leads. You should get a resistance close to infinity in both directions.

This method can be used to check any transistor.

I wouldn't expect any of the transistors to be bad because you wouldn't be able to adjust the voltage. You wouldn't get any or just full output voltage. Try to adjust the trimmer pots to get full current capacity and output voltage.

If I had a curve tracer I would remove the transistors and compare them with different ones, but sadly curve tracers are very expensive.

saturation · « **Reply #16 on:** April 23, 2011, 01:47:17 pm »

A great thread for maintaining this PSU.

How about the transformer, tekfan, its been without issue?

I think it better to performance test the whole PSU after minor adjustment rather than break any electrical connections unless the unit failed testing.

tekfan · « **Reply #17 on:** April 23, 2011, 08:09:23 pm »

Quote from: saturation on April 23, 2011, 01:47:17 pm

A great thread for maintaining this PSU.

How about the transformer, tekfan, its been without issue?

I think it better to performance test the whole PSU after minor adjustment rather than break any electrical connections unless the unit failed testing.

Yes the transformer survived the capacitor melting. The enamel coated wire inside the transformer can take 120°c no problem. You can even have a hard time breaking trough the enamel with a soldering iron if you want to solder the wire

I've never had a mains transformer fail in any of my test equipment no matter how rusty or bad it looks.
Exceptions are the lowest grade cheap chinese transformers, many of which have failed. Either from internal short (followed by catastrophic meltdown) or internal open winding (followed by garbage bin).

I would also check the performance after adjustment. Remember, there's no need to replace components if it matches or exceeds the specs.

SnakeBite · « **Reply #18 on:** April 23, 2011, 09:16:40 pm »

thanks tekfan for you help.
i was able to get to the full voltage spec when i tested it before i started to clean it but it tested under NO load so the current was zero so i think the transistors are fine.

i have a few questions:

1. is it imortent to change the isolation from Mica to silicone or other?
2. i can buy mica ,silicone or ceramic isolation ? which one is beter?
3. what is the purpose of the isolation ? just to isolate or also to conduct the heat?

thanks
Ido

tekfan · « **Reply #19 on:** April 23, 2011, 10:45:31 pm »

Quote from: SnakeBite on April 23, 2011, 09:16:40 pm

thanks tekfan for you help.
i was able to get to the full voltage spec when i tested it before i started to clean it but it tested under NO load so the current was zero so i think the transistors are fine.

i have a few questions:

1. is it imortent to change the isolation from Mica to silicone or other?
2. i can buy mica ,silicone or ceramic isolation ? which one is beter?
3. what is the purpose of the isolation ? just to isolate or also to conduct the heat?

thanks
Ido

1. No, it isn't.

2. Personally I use only mica spacers. I had problems with the silicone ones because if you tighten the transistor down too much it punctures the silicone and touches the heatsink. Ceramic is also good it has slightly better heat conductivity and insulation than mica. On the other hand, mica sheets are thinner and provide a sort of capacitor between the transistor and heatsink (very good in this application since low noise and ripple is wanted). The mica sheets have approximately 750pF capacitance between transistor and heatsink, the ceramic ones only 20pF. I would think the ceramic spacers can come useful in oscillator circuits where low capacitance to ground is needed.

Personally I'd use mica sheets because they are more convenient (with thermal grease of course)

3. The insulators do just this. They insulate the transistors from each other. Without the sheets all the collectors of the transistors would be tied together and to ground, which would defeat the pourpose of a ''floating'' power supply.

And for thermal conductivity aswell. Mica is a very good heat conductor. If you put a mica sheet on your hand and press on it with your finger you will almost instantly feel the heat from the finger. If you take a piece of plastic of similar thickness and try this, it would take several seconds before you would feel anything.

SnakeBite · « **Reply #20 on:** April 24, 2011, 01:40:09 pm »

i started to work on the transistor side of the PS and i taked out one transistor for testing (i wasn't intended to test any transistor but i intended to replace the themal compound so i needed to get it out anyway) .

i checked from Base to Emitter on "Diode Test" and i have got a dropout of 0.536v is this OK?
from the collector to emitter and reverse in ohms range, i got infinite resistance in the two positions.

so i'm guessing it's OK ...

on the first image i took apart the three transistor from the heatsink and the black thing that stayed there that the thermostat.

and now..the photos:

Uploaded with ImageShack.us

Uploaded with ImageShack.us

more updates tomorrow.
Ido

p.s. since i am also Computer technician i have this thermal compound (the best on the market , also doesn't conduct electricity)

tekfan · « **Reply #21 on:** April 24, 2011, 01:53:00 pm »

Great! Soon you'll have a very nice power supply.

The 0.536V drop is within spec. Silicon transistors can have a voltage drop anywhere between 0.4 to 0.8 volts.

Cyclenorthman · « **Reply #22 on:** April 24, 2011, 05:19:37 pm »

You could also use a conformal coating on circuit boards to seal them after being cleaned.May also help seal other areas.We use it on avionics and other electronics.I have never tried varnish because of FAA regs but see what other members think about that.
Just my $0.02 worth.

CNM

tekfan · « **Reply #23 on:** April 24, 2011, 06:11:09 pm »

I think there was an option with these supplies where the factory could give the insides of it an anti fungus coating. This has to be for some really harsh conditions.

alm · « **Reply #24 on:** April 24, 2011, 08:16:54 pm »

Quote from: Cyclenorthman on April 24, 2011, 05:19:37 pm

You could also use a conformal coating on circuit boards to seal them after being cleaned.May also help seal other areas.

I would make sure to bake the boards before coating them to get rid of any residual moisture (this is directed at SnakeBite, I'm sure you're much more familiar with conformal coating than I).


EEVblog® Main Site	EEVblog® on Youtube	EEVblog® on Twitter	EEVblog® on Facebook	EEVblog® on Odysee

Author Topic: Restoring the Lambda LQ-532V power supply - **end of project** (Read 36031 times)

alm

Share me

Author Topic: Restoring the Lambda LQ-532V power supply - end of project (Read 36031 times)