LKB (Bromma) 2197 Power-Supply

[tri-comp]

Hi,

Today I received this 20+ year old Swedish-made Elektrofocusing power-supply from eBay.
It's remarkable: 0~2500V/0~250mA ! Can you imagine this on you bench when building valve-amplifiers ?!
The price was fair, even considering shipping from the US.
The front 10-turn pot for setting the output voltage is worn-out/intermittent but a new Bourns 10-turn standard-pot will fix that.
I want to know if anyone has complete service-manual for this supply?
From looking at the innards it's apparent that it's a very interesting and alternative build.
I'd like to see exactly HOW it's done, stabilizing this wide output-range.
The thing is built like a tank, long before parts came from China and all electrolytic cap's are measuring like new.
Also, fortunately it's built w/o a microprocessor with discrete IC's.
Probably not very difficult to service if need be.

I'll attach what user/service-manual part I have and hope for more...

rgds,

/tri-comp

[Anthony]

Dear Tri-comp

What a blast from the past. I used one of these over 30 years ago for clinical research electrofocusing. It was very reliable.
Of course, the voltages are potentially lethal.
LKB was, I think, eventually take over by General Electric Life Sciences (GE Life Sciences).
I wonder whether a friendly engineer there might help.
Good luck

Best wishes

Anthony

[Alistair D]

I have been working on and off converting a similar supply for use as an HT power supply. As it stands YOU MUST NOT CONNECT THIS SUPPLY TO ANY EXPOSED METALWORK. Have a look at the block diagram(fig 4)The output is generated by 2 rails in series. The centre point is earthed(see the symbol just above the current sensing resistor) so that each output terminal will be at half the output voltage above and below ground. The only protection is that if more that 0.4mA passes to ground the earth leakage detector will switch the supply off.

Another possible problem(your supply may be different to mine in this respect) is that the time constants on the regulator circuit is very long. As a result when the output is switched on or the supply comes out of current limit the voltage goes very high for a fraction of a second.

Mods. I have sent tri-comp a PM, hopefully he will read this before connecting the supply up.

Edit: Just fol clarity. The internal ground does not float, it is linked to mains earth via the diodes and resistors that make up the earth leakage detector.

Al

[tri-comp]

Al,

I'm still alive
The power-supply is taken apart while waiting for 3 new Bourns pot's.
I'll be using it as a bench power-supply when breadboarding with valve-amplifier design.
It will never be used as a permanent supply.
I actually read about the symmetrical output in another user-manual for a similar supply before I settled on buying this one.
I suppose I'll just have to keep the power-supply chassis away from other instrument cases and I'll be fine.
Also don't reach for the scope with one hand while adjusting the power-supply with the other.
If it really fills the promises I don't see why these supplies aren't gobbled-up on eBay long ago.
I'll report back with my findings once it's re-assembled.

Any service-manual...?

rgds & thank you, Al

/tri-comp

[Alistair D]

Sorry I do not have a manualfor your supply. I don't even have a manual for my LKB 500/400 that is part of the reason it has taken me so long to modify it.

One thing puzzles me about your planned use of the supply. The earth leakage detector will trip it every time you connect an oscilloscope or sig gen to the circuit you are developing. How do you plan to get round this?

Al

[tri-comp]

So far I had it wired too a Variac and all was fine.
As a permanent fix I'll use a 1:1 transformer of relevant capacity.
The kind of power I'll be needing is sub 100W.

Shouldn't that plug the trip-circuit problem?

/tri-comp

[tri-comp]

...oh, and I dropped GE a letter and included your model in the plea

/tri-comp

[Radio Wrangler]

Um, if you use a mains isolating transformer to allow you to ground either end of the 2500v output, doesn't that mean that the isolation transformer is stressed at 1250 volts?

Also capacitive currents may trip your protection system.

David

[Alistair D]

An isolating transformer will will not help here. The PSU outputs are already isolated from the mains by the switch mode transformer and the transformer that generates the low voltage supplies for the regulation circuits. The centre point of the two halves is the key here. If the centre point tries to move more than a few volts away from earth then the leakage trip should switch of the output. Relating this to a valve audio circuit that needs repair and you want to substitute the HTof 400V with the power supply. Connecting the -ve of the PSU to the earthed chassis of the amp will take the centre point of the PSU to +200V above earth. The PSU should shut down. This begs the question, is the PSU earth connected. IF not then the case of the PSU will be floating at +200V. I just want to be sure that this setup is safe.

Tri-comp. Many thanks for your offer to help me find a technical manual for my supply.

Al

[tri-comp]

New pot's are installed but I haven't had the time to assemble the power-supply yet.
I did, before taking it apart, run it off the Variac with a BIG 2K cement-resistor as a load.
At the same time I checked the output with my Kikusui scope and didn't see anything odd except the jerks caused by the defective Voltage-pot.
The supply didn't shut-down and the resistor got HOT !

My LKB 2197 is built around the SG3524 PWM chip. Tracing the circuit from looking at the mother-board shouldn't be too difficult.

Also, today I e-mailed a co-tech who works at the regional hospital.
He has access to all sorts of nice documentation. Perhaps ....
On the other hand ... knowing him it's going to cost some beer

rgds,

/tri-comp

[tri-comp]

Today I received a reply from GE Healthcare and they regret that no service-documents are available any longer.
One down, one to go.
Yesterday I was on the phone with the tech at the hospital.
He activated his contacts and asked them to keep an eye open.

Keep your fingers x'ed.

rgds,

/tri-comp

[tri-comp]

Some development.
I gave up waiting on people it turned-out didn't have access to squat
I took my LKB-2197 apart and prepaired to reverse-engineer the circuits.
First thing is of course to locate the actual connection between the chassis and the electronics.
That wasn't hard; the ONLY connection, save the primary line voltage noise-filter de-coupling, is made through the left-most black wire in the 6-pin multi-connector at the left-rear of the mother-board.
ALL front-controls are insulated plastic-knobs so there's really no place to get in direct touch with electronics if you stay off the rear remote-plug.
Removing the black wire from the multi-plug should do the trick of deleting the semi-hot chassis construction permanently.
I can't believe it's that simple.
Your reflections, please....

Attached is the photo from the manual of the mother-board with a note inserted near the 6-pin multi-connector.

rgds,

/tri-comp

[tri-comp]

A few final notes.

The PSU seems to work normally even with the chassis reference connection gone.
Actually I don't know why it shouldn't; it's just a safety precaution when working with Electrophoresis.
When you're working with tube-amp's you're on your own; no automated safety what so ever here
Obviously sticking with regular safety rules will be vital as with all live electronics.

I managed to replace all 3 multi-turn pot's. RS Components # 692-8538, Bourns 3549S-1AA-502A fit exactly.
The wobble in output when setting the voltage disappeared immediately.
Now, when assembling the PSU I was going to throw away the old pot's and I noticed that they are actually held together using 2 opposite metal-clamps.
Prying them off and opening the pot was easy. I believe i good spray of DeoxIT would probably have fixed the original pot's. Waah !

Beware when buying these power-supplies off eBay.
Since this is an old-fashion heavy-duty power-transformer version, shipping will get you. Otoh you'll be in control as no fancy microprocessor with firmware you'll never be able to alter.
Watch out for broken-off knobs around the timer-section. However, it's not a vital feature, and if the price is right probably something you can live with.

Happy modding

rgds,

/tri-comp

[Alistair D]

Hi Tri-comp. Please check one thing before putting your power supply into service. My electrophoresis power supply has a very long time constant in the regulation feedback loop. As a result a step change in current of only a few mA causes the loop to lose regulation for a few milliseconds and the output voltage jumps to full. Could cause some damage to the load.

Al

[tri-comp]

Al,

You're right about this problem which can't be dismissed.
I hooked my 2197 up with a 2K/LOT's of WATT resistor and set it for 200Vout (=20Watt Load)
Paralleling a 1K/BIG-Watt resistor causes these spikes that you also noticed.
Connecting a 150uF/385V cap across the output made the regulation go south. The voltage rose slowly until just below 400V when some safety-circuit kicked-in and killed the output.
Actually, I thought: 'Killed the PSU'
Will, it didn't because after removing the charge on the big internal 4700uF/63V electrolytic cap's, the circuit reset itself and all's fine.
Still, this is serious as I'm expecting to use this PSU when working with tube-amp circuits and they will all have some B+ decoupling. 150uF seems not a lot here.

Now I need to locate that FB-Loop ! Will this never end..?

rgds,

/tri-comp

[tri-comp]

Somewhat of a miracle happened today when I received the full 2197 repair-manual.
Apparently LKB is still in business in Austria and they replied quickly and without charge to a cry for help !
Respect !!

I'm attaching the schematics only as the complete manual is too big for upload here.

About the FB-Loop issue i see some nasty 1,5uF cap's around N6/N7 op-amp's.
Would that be a place to start ?

Rgds,

/tri-comp

[Alistair D]

Hi tri - comp, can you PM me contact details. Hopefully I can get a manual for my power supply.

Al

[tri-comp]

Al,

Of course you may have a copy.
It seems easier if YOU PM ME with YOUR contact details, i.e. e-mail address, as I'm the one who's going to send the file and need some place to send it

Best regards,

/tri-comp

[Alistair D]

My supply is a different and less complicated model, EPS 500/400. Thanks for your help. After a fair bit of searching I found the site and have emailed them. My mistake was looking for LKB Bromma which kept directing me to the Swedish site rather than LKB in Austria. I would not have found it without your help.

I have not studied the circuit of your supply in detail yet but I have also had voltage overshoots when the output is switched on and also when the supply comes out of current limit.

Al

[tri-comp]

OK Al, I completely misunderstood

I can give you the name of the person I reached there and who helped me if you're not successful in your first go.

If/when you get the manual I'm all for swopping copies. Mine has detailled circuit explanations.

rgds,

/tri-comp