Rohde&Schwarz NGAS 32/10

[najrao]

Got this bargain today, seemed to be in good condition physically.
Cleaned it out by blowing, no visible faults or missing parts.
Big caps are not shorted, transformer tests OK.
Bringing up voltage on variac and series lamp (both!) says dead short.
Pass transistor/s seem to be toast.
Problem is : no manual or schematic, yet. Assembly is awkward to say the least,  no connectors used, wires emanate directly from pcb's and terminate on other components. To get at vulnerable parts, needs lots desoldering with risk of wiring mixup.
Searching eev and elsewhere has offered no schematic or manual. Keen on restoring it and making it my main bench supply. Please help! Thank you.


Edit:  One of two pass transistors was shorted out. Effectively they are just in parallel, may or may not have  separate drives, difficult to trace. More importantly, they are linear mosfets of yesteryears, in TO-3 metal case, Motorola MTM40N20.  I have had to replace them by IRFP250's, a close match by spec. TO-247, so some bodge work inevitable.
The dead short at mains input was due to crossed wiring to the two-pole mains switch; strongly suspect that the idiot who who did this was ---  ME!
The voltmeter has had it too. Useless anyway, with a 36mm scale and +- 2.5%. I am scouting for cheap 3 1/2 digit substitutes (also the ammeter, for symmetry). Not easy, because the shunt feeding it is also the source resistor of the pass transistors, and so sits at the positive bus. Cheap digital meters use their neg supply as ref for Vin. Let me see.

[hscade]

thats where my heart jumps up and down 

[najrao]

Thank you, but what do you mean if anything?

[orbanp]

Hi Najrao,

This site is selling a unit like yours and there is a CD at the top of the PS.
It is possible that the CD has the the manual on it in electronic form. You could ask them:
http://www.messtechnik-colberg.de/common/hersteller-r/rohde-_-schwarz/geraete-r_s/ngas.html

Other than that, this PS is probably a simple circuit, a series regulator with voltage and current control loops, and with two voltage settings. You can trace out the schematics easily.
Take lots of pictures. Should not be that hard.
R&S usually makes quality equipment, it would be well worth it!

Regards, Peter

[najrao]

I copy a letter to  messtechnik-collberg.de I have just sent.
Can you or anybody else explain this riddle?

Dear Herr Bernd Collberg,

I am not even sure I am asking the right person, but your previous helpful reply prompts me to ask anyhow.

I have got the PS going after some repair, and want to use it regularly. Essentially, the two pass mosfets and the meters.Now, IRFP250 and a digital combo.

I don't see the heat sinks going beyond 75W or so, if that, even with the blower fan. The unregulated dc is 77.5 or 56.0V at no load, depending on the tap changing relay, and I expect these to drop to no less than 56 and 40V on 5A load. For regulated output below 41 and 25V respectively, the pass transistors will exceed the 75W limit continuously. How do R&S justify the claimed ratings?

To calibrate my new digital ammeter, I set the current limit at up to 8A; the load was essentially a short circuit. It took only seconds to overheat to 80C on the heatsink.

No, I did not set the voltage selector at 110V and operate at 230.

Is it common to rate linear PS's  in some peculiar (American) way to maximize the apparent horse power?

I thank you in advance for a cogent explanation.

J.

Edit:
There was no reply from Messtechnik.

After lots of experimentation, I had to plump for simply augmenting the cooling of the pass transistors. Any preregulator scheme to charge a 15mF capacitor to 78V from a stiff source, and sustain loads up to 16A from this dc bus, will involve peak currents to 150A, and loss to 75W. It just is not worth it. The best of the lot was trailing edge phase control using a 100V 47A 290W mosfet, as this involved the least complication in control; I gave up when it called for two such in parallel.


So, I simply transferred just under half the pass transistor heat to a heat sink outside the case, on which are fitted two more identical pass mosfets. Split source resistors, one on each, provide a semblance of current sharing equality. R&S have provided a thermostatically controlled two speed 220V blower, and my new heatsink also stands to get lots of air. See photo.


Test showed that at rated load (10A, 17.5 to 0V) into a short circuit in the CC mode, the blower stays at low speed for at least ten minutes. 215W of heat! I declare success!

There is a lot more detail to say, if there is going to be any interest.





Sent from my Galaxy Tab® S2

[najrao]

Please read my last edit. I posted it just to  bring the topic to a conclusion. More if there is interest.