Lambda LQD-421 repair

[BigJohn89]

Hi Everyone!  I just came across an old Lambda LQD-421 that I'm trying to get up and running.  One side functions OK, but appears to run away on voltage at times - I am already looking at re-capping that side.  The other side is stuck at minimum voltage, and will not raise voltage at all.  Checking at all three test points Va, b, and c give voltages that aren't too far out of whack, and so far I haven't come across anything burnt or failed - including the main control pots. 

I noticed some other messages related to repair of this PSU's brethren, but there wasn't much of advice for solutions or things to look out for (other than recapping).  Has anyone successfully repaired one of these, and have any kinds of good advice they would like to pass along?  I have a schematic for it that I have been tracing, however the unit itself doesn't seem to completely follow the design as printed...

Thanks in advance!

John

[coromonadalix]

Are your short jumpers on the black terminals screw block on the rear are okay ?

Could you post your schematic(s) ?

Schematic

http://bama.edebris.com/manuals/lambda/lqd420

Its somewat classical topology, not sure if ic101  is an lm723 ??

You have a floating supply section on top of the main supply, you have to be sure where you put your meter ground  for measurements

have you seen this thread ?
https://www.eevblog.com/forum/repair/datasheet-for-lambda-voltage-regulator/

The fbt-00-128 ic  seems to be an LAS3700 ic ?  an unobtanium part, i hope its not busted.

The lm723 to-100 case have a strange ressemblance minus the pin numbering ??
http://www.ti.com/lit/ds/symlink/lm723.pdf

I would not recap the psu immediately,    Repair before changing many parts.  The design fits many models with some parts change between them, you see the changes in the manual.


youre supposed to have :
ic 401       12v or 15vdc between  pin 10 and pin 1
ic 403       5v  on test point a16 and a20 for the display supply

And as you wrote
va = 41.5 volts
vb = 11 volts
vc  = 28.5 volts

[cnqhdszq]

http://bbs.38hot.net/forum.php?mod=viewthread&tid=100352

I had repaired a LQD-422-V  liner powersupply .First  you should check the main regulator, transistor  "FBN 36605N"  ,if  the transistor is OK ,then check the voltage of the volt adjust pot ,when turning the pot  the voltage will change ,commonly    the output voltage equals the pot voltage . Good luck!!
If you can speak chinese ,you may read my repair  post (above link).
 

the  PIC is my schematic by hand.

The resistor in the red circle is open ,both side .
One "FBN 36605N" is shorted ,and one 40K pot  is open .
and the switch of V - I on the front panel is not OK  ,and the num ，after I drop  a drop of lubrication oil ,it can do it's work. 

[cnqhdszq]

And a 40 UF  cap is dried up

[coromonadalix]

@cnqhdszq

You may add confusion here, the LQD-421 is not built like your shown schematic ....

[duak]

John, some of the older Lambda supplies (like the LP-4xx) will damage their voltage control pots if they are connected to a battery and the output voltage is turned down.  I don't know if the LQD series  do the same thing. Do you see the output voltage suddenly jump up when set to particular voltages?  If so, the voltage control pot may need cleaning or may be damaged.  I would shut the unit off and then measure the voltage control's resistance - it can be done from the back panel - RP (term 13) to +V (term 14)

I think the LAS-3700 chip is a custom Lambda part that doesn't seem to cross to anything else.  The chip in the LP-4xx series is a re-marked Motorola MC1466 floating regulator that is in a 14 pin DIP.  If the LAS-3700 is bad, it might be possible to substitute an MC1466, but these are obsolete and hard to come by.  Lambda is still in business and they sent me a manual when I asked about 10 years ago; maybe they have spares.

It might be possible to make up a replacement for the LAS-3700 using a dual op-amp, a current reference and a few other parts.

These are not bad supplies, they'll take a lot of abuse and keep on truckin' - I wouldn't give up yet.

[cnqhdszq]

Are you sure ?
If  you spend more 5 minutes on my schematic and your's,you will find them are the same .

When I repair my LQD-422-V   ,I had no service manual or schematic in my hand ,So I reverse the board .At the beginning I even thought the "FBT-00-128" is LM723 .One day a  member send me the same files you  post .

(sorry for  my poor english,but I really want to be helpful )
 

[coromonadalix]

@cnqhdszq

You miss some parts in the power stage section vs the lqd-421,   sure  it has some parts count differences since its not the same (model LQD-422)

But the service manual is found and posted

The OP will need to do some check ups  as mentioned in previous threads and report back to us

[BigJohn89]

Thank you everyone for your help and suggestions!  I've been plugging away at it for a while today, and wanted to report back a few things:

I have checked the resistance across the control pots on both sides, between RP and +V, and the pots seemed to work properly, though I noticed the resistance creep up even when I wasn't turning the pots.  I'll dig into that further, but since I get the same response out of both sides, I am not sure if it is having much of an effect on the side that is not working. 

I've checked the test point voltages, as well as others listed in the schematic:
va = 40.2v
vb = 10.4v
vc = 27.22v

For IC101 (this is the side that doesn't work, IC401 is the side that is working):
Between pins 10 & 1 = 15.27v (IC401 = 15.25v)
Between pins 9 & 10, which is also across the two mylar caps on the schematic (listed as 8v) = 9.5v (IC401 = 8.7v)
Across C102, which on the schematic should be 15v (and is pins 10 & 1, I just realized) = 15.27v

I am confused, because on the schematic there is a 12v designation on the connection along the top going into R101 - what is that 12v in respect to?  Is that between that point on R101 and ground?  I didn't check that because I didn't want to unintentionally cause more issues with this thing, so I figured it would be best to ask first. 

I'm going to look further into the datasheet on the IC101, and see if I can find a way to further test it.  But if you have any other ideas, I will be very grateful if you shared them with me. 

Thank you again for all of your help on this so far!

John

Edit: Oh yeah, forgot to mention testing between A16 and A20 came up with just over 5v between the two points.  Thanks!

[duak]

Here's some info and links to more data on the LAS3700:  https://www.eevblog.com/forum/buysellwanted/free-for-local-l-a-pickup-lambda-lqd-422-dual-output-power-supply/

The voltages aren't insanely wrong.  I've got to think about how the LAS3700 works some more.  BTW, is it warm or hot to touch?  Apparently it has a heater inside to stabilize the reference voltage.

I'll bet the resistance rises slightly because capacitors in the circuit are charging.

I think the 12.0 V is referenced to the +output terminal.  In floating regulators +V output is the pivot that everything is referenced to.  The typical non-floating supply uses the -V output.

[cnqhdszq]

 duak, Thank you very much ! 

Finally  ,I know what the FBT00128 is ,thank you .

[duak]

Are all the jumpers on the terminal strip on the back the same between the sub-units?  Are any missing or different?  Is there a connection from one to the other?

Is U101 warm or hot to the touch? The voltage between U101 pins 9 & 10 is high.  With the power off, what is the resistance between them?

Lambda supplies usually cut the output voltage off when the current limit control is set to minimum.  If the control is not set to minimum, there could be something wrong with the current limit circuit. 

With no load and current limit to maximum, what is the voltage across R126? With the negative lead of the meter on +V, what is the voltage on IC101 pins 1, 6, 7 & 8?  What is the voltage on IC101 pin 2 as the voltage control is varied?

[BigJohn89]

Hi Duak,

For the jumpers on the back, they are all in place in the standard configuration, with +V jumpered to +S, and -V to -S, as well as a 3rd jumper connecting two points labeled as ICS on the board. 

On both sides, IC101 (I think that is what you were referring to - are U101 and IC101 the same?) is definitely warm to the touch, I didn't have a temp probe so i couldn't get exacting figures though. 

The resistance between pins 9 and 10 was 7.41k , the working side checked out similar as well. 

With no load and max current, the voltage across R126 was .6 mV.  I need to recheck the working side, because I didn't have a figure there...

Here are the other checked voltages, working side is in the right-hand column. 

P1 = -2.483v  = -2.681v
P6 = -.3mv     = -.4mv
P7 = .4v         = 357.8mv
P8 = -.4mv (yes, negative)   = 7mv

P2 on the non-working side was .516v straight through the scale, with current limit set to none and maximum.  The working side was .512v through the scale, unless the limiter was up a small bit, then it dropped to 46.8mv.

I'm with you on the current limit circuit as well, and with pin 8 being a negative voltage, I'm going to start working my way back from there to see if I can test out components along the line.  It appears I may be looking at the Q1 - Q2 - Q5 transistors, would you happen to have any advice - or am I barking up the wrong tree?

I totally appreciate the assistance you are giving me, and can only hope to pay it forward some day!

Thank you,
John

[duak]

Hello John,

No hay problema.

Hmm, the manual I have calls it IC101.  I think I got it from here: http://bama.edebris.com/manuals/lambda/lqd420

If you're postive that the voltage across R126 is negative, it could be a clue that current is going through it backwards.  According to Table 1, R126 should have a value of 0.3 ohms - do you measure that?

With the negative lead of the meter on the +V output terminal, what is the voltage on IC101 pin 5? Q101 base, Q5 base and Q1 or Q2 base?  IC101 pin 5 should have about 3 V and as you work through the chain the voltage should drop by about 0.6 V every time it passes through a part.

Cheers,

[BigJohn89]

Hi Duak,

Thank you very much for your advice and taking the time out to assist me in this!

I think I made some progress, and added a little confusion for myself as well!  I have verified twice that I have an LQD-421, and the schematic reflects that - so far.  Now, going into the base of Q101 is a diode, CR124 (FBL-00-105) which is also connected to one side of R134 as well (at least on the schematic).  On the schematic however, Note 13 states that CR124 should be jumpered (and not present) on the LQD-421, and that R134 should be a 1.8K resistor.  In real life, CR124 is present and R134 is missing completely.  This is also the same on the other (working) side as well.  I have tested this diode in circuit (not the most reliable way, I'm sure) and got completely different measurements from the same diode on both sides.  I also tested CR138 on both the working and non-working sides and got very similar readings as the diode on the working side.  In those cases, I received a forward resistance, however reversing the polarity gave me an open circuit.  CR124 on the non-working side gave me resistance with both polarities, with numbers that were jumping all over the board (usually 3-ish M one way, and 700K the other).  I haven't pulled it yet, as I wanted to research further to make sure I wasn't missing anything else first - that and I don't have a suitable (or know of one yet) replacement anyway.

That being said, I checked R126, and the value on both working and non-working sides is 1.1 .

I went ahead and checked the voltages, here they are for both sides (non-working, then working):
P5 = -2.52v then slowly moved to -2.477v and stayed steady                                              |  2.448v
Q101 base = Testing here caused my meter to go "flat", it kept blinking but no voltage.         |  2.065v
Q5 base = -135mv                                                                                                           |  .926v
Q2 (or Q1) base = -134.5mv                                                                                             |  279mv

I also tested the voltage at the junction of C119 and CR124, which tested at -2.474v.  The same junction on the working side tested at 2.43v.

Thank you again for any assistance you (or anyone else reading this) can provide!
John

[duak]

John, I've got some good news and some bad news...

I think IC101 has expired.  Pin 5 should have about 3 V relative to +V output terminal to cause the pass transistors Q101, Q5, Q1 & Q2 to conduct and pass some current.  Looking at the Functional block diagram or the test circuit shown in the above files I see that both the V and I amplifiers can cause Pin 5 to have a voltage on it.  The V amplifier's + input is higher (0.512 V) than its - input (-0.3mV) so its output should be high.  Likewise, the I amplifier's + input is higher (357 mV) than its - input (-0.4mV) so its output should be high too.  Logically, pin 5 should be high too, but it's not.

According to the Catalog Pages info, U101 is supposed to heat itself to 115 C.  Is it hot?  Careful if using a digital thermometer (finger).

Clutching at straws now, the voltage between U101 pins 9 & 10 is high.  It may be that the resistance between the pins is too high.   Try comparing the value between the two channels to see if they're different.  With any luck, either R101 or R102 is bad and this is why U101 acting up.

If all else fails, it should be possible to replace U101 with a current reference and a couple of op-amps.  It would even be possible to use a few transistors with some loss in performance, but it would be useable.

Cheers,

[coromonadalix]

If ic101 is dead after all the tests,  the remaining circuit could be used with an lm723, it seems compatible with some drawings and pcb reworking

[duak]

Using a 723 is a good idea.  I would use two; one for the V reference & amplifier and another for the I reference & amplifier.  Their "Freq comp" pins would be tied together with a diode to allow one to override the other and reduce drive to the pass transistors.  The decision to be made is which of the amplifiers, V or I overrides the other.

[coromonadalix]

Each time i see this psu,  i'm seeing the lm723 in to-100 case, it is still sold today  and strangely compatible in some ways  ....  you have tons of schematics floating around and im sure this psu could be converted to a more recent design, while keeping a vast majority of parts.

[BigJohn89]

Thank you both very much!  I did run some more tests on the device, R101 was a little low at slightly over 5\$\Omega\$, while the working channel was closer to 7, but using R102 to compensate made no change in the voltage between pins 9-10 or coming off pin 5 whatsoever.  I did pull and test CR124, and while I had suspected it originally, it tested out fine.  So, at the end of the day, it does look like the FBT128 is a goner

Sadly, my engineering skills are much, much worse than my repair skills, so I have a lot of reading ahead of me to figure out how to make this work with the lm723!  If you have any advice, I am always open for suggestions.  It would be good to not only bring this device back from the dead, but I'm sure there are plenty of others like this one out there, so this may help keep those out of landfills as well.

Again, thank you very much, I really, really appreciate all of your help on this!
John

[coromonadalix]

I do understand you want to keep it genuine, it can be done       

I have personally designed and built some lm723 based psu's ... hacked some pyramid 12v power supplies ...


I dont know if someone has made a FBT-00-128 clone ?? or a small pcb ??

The problem is the fbt-00-128 aka LAS3700 contain the cc and cv crossover circuitry ...

[Kosmic]

You can find some sellers on ebay carrying what look like NOS FBT128. 10$ per chip. It's unclear if they are really new or recycled parts or just fake.

Coincidently, one channel of my LQD-421 just died this week. I traced the problem back to the FBT128. Ordered some from ebay. I should be in measure to tell you if they are really new parts in a month or 2 

[coromonadalix]

Apparently theses chips / ics heat a lot while working, that must be the failure reason over time

[duak]

I've attached a conceptual replacement for the LAS3700 chip using two 723 regulators.  The left side shows the Voltage Control section where the 723's reference is basically multiplied up by the ratio of the front panel pot's resistance to RVscale.  If RVscale is replaced by a current source, the regulation would improve because the closed loop gain would not change as much as the voltage setting was varied.

The right side show the Current limit section.  The error amplifier in the 723 compares the voltage across the current sense resistor R126 to a variable threshold voltage generated from the 723's reference voltage and the existing circuitry.  When the sensed current exceeds the threshold, the amplifier output goes low and through Dclamp pulls the Voltage control output low, reducing output current until an equilibrium is reached.

The advantages of using 723s is that they have a current source for a load for the error amplifier making wired-OR circuits easy to implement.  They also have voltage references in the right range.  Their error amplifiers probably have similar voltage gains and gain-bandwidth products to the LAS3700.

OTOH, for best performance, it'd be better to use current references and perhaps op-amps instead of the 723s although there will be more details to plow through.  In particular, the combining/crossover circuit and feedback stability.

I see from the LAS3700 catalog data sheet that it operates at 115 C (!!!) due to its internal heater.  I'm sure it'll be stable but maybe not so reliable.  I've got at least one Lambda supply that uses this chip so sooner or later I might have to replace it.

Cheers,

[coromonadalix]

@Duak   that seems doable   

There is a project like this idea
http://www.ve2ums.ca/chasse/Serge/Atelier/Projets/Membres/VE2EMM/alimentation_ang.htm

The same version pumped up to 10 amps in attached picture, could be adjusted for less amps.


But its not designed as @duak   suggested

To keep a constant current control and or the constant voltage   leds / operation  there are op amp based psus with the "or function"

see this schematic
http://www.seekic.com/circuit_diagram/Power_Supply_Circuit/Constant_voltage_and_constant_current_regulated_power_supply_circuit_diagram.html

you see the two diodes at the op amps output, it the "or function" switching between cc and cv

Or this one with relay switched xformer secondaries, some circuitry could be removed and you could still have the cc cv leds ...
http://www.seekic.com/circuit_diagram/Power_Supply_Circuit/High_precision_Constant_Voltage_And_Constant_Current_DC_Power_Supply_Circuit.html

The major problem with lm723 designs, they wont go down / less than 1.25 volt for the voltage output,  this has to be negated with an -1.2 volt supply in the voltage adjustment potentiometer ...