Bertan/Spellman 205A-05R High Voltage Power Supply Teardown

I got a non-working Bertan 205A-05R 5kV high voltage power supply and did an attempted repair video a while ago (see video towards the end). But I forgot to post my teardown pictures somehow. So here we go, better late than never.

Bertan 205A-05R is a 0-5 kV 5mA high voltage power supply. It is entirely analog, the voltage can be set via the front panel range switch (kV) and a ten-turn potentiometer (0~999V). Or it could be set remotely via a reference voltage input at the back panel. Besides the basic specifications, I could not find any service manual online for this unit.

205a05rfront

The inside looks surprisingly barren and the build quality looks very crude. Besides the main circuit board, everything else seemed to be using point-to-point construction.

inside_overall

Glued onto the side wall are two Curtis 520-LNA elapsed time indicators (ETIs). This is quite cool as I have not seen this kind of indicators used anywhere else. According to the datasheet, here is how they work:

The integral is achieved by passing current through a precision-bore glass capillary tube. The glass tube is filled with two columns of mercury, which are separated by an electrolyte gap. Application of current causes the mercury at the anode to be electrochemically transferred across the gap to the cathode at a rate which is proportional to the current. Thus the gap moves along the capillary tube and provides a visual indication of the current/time integral.

As you can see, this unit has been in active service for at least 10,000 hours.

servicelifeindicator

The pictures below shows the top and bottom side of the main board. The main active components here are some Op Amps (LM358, uA1458, TL082, TL083) and transistors. The metal can is a CA3028, which is a high bandwdith differential/cascode amplifier. The onboard oscillator is buffered and then drives the two 2N6578 Darlington transistors that switch the primary of the high voltage transformer inside the potted assembly. Judging from the datecodes, the unit was probably made sometime after 1989.

mb_top mb_bottom

To be honest, I have not seen such shoddy construction for a long time. Take a look at the interior of the front panel below, the transistor that drives the LED is glued directly on the case!

frontpanelinner

It’s getting even better! Take a look at the auxiliary power rectifier and filter circuit (for the main board), they are glued onto the transformer itself!

ps

Here is a picture of the output polarity switch and the MHV high voltage BNC output connector. You can also see the load balancing resistors soldered directly onto the pins of the transistors. A few standoff boards can also be seen in the picture.

polarityswitch

Finally, here is a picture of the potted HV transformer/rectifier assembly.

pottedhv

If you watch the video below, you will see that the reason this power supply is dead is that one of the HV capacitors inside was shorted. Unfortunately, this cannot be easily repaired.

Be Sociable, Share!

2 Comments

  1. STEVE BEST says:

    I enjoyed following along the write-up & video. Thanks for posting it.
    I just opened a Bertran 205A-10R to start troubleshooting…very similar construction except for polarity reversing switch and HV module of course.

    I don’t suppose you have ever happened by any schematics for these, have you?

    In the video you showed your partial hard-copy schematics of the remote/osc. drvr PCB.
    Any chance you still have that as a Jpeg or PDF that share with me or better yet, add as a link to web page. That might save me and others a bit of time??

    Also, just as an addition tip for other viewers for safety in future…
    When you were demonstrating the bad capacitor near the end, and you were attaching the alligator leads to parallel in the bad cap, IF the cap had not been bad, then the thin insulation on the high side alligator clip lead could have possibly failed (arc thru) to adjacent metal/leads/grounds. Given you knew the cap was bad and would breakdown low near 400V, this was not a concern,…but under other circumstances, it might have been.

    Thanks,
    …steve

Leave a Reply to kwong