Testing H7864 (MicroVAX II) PSU With No Load

[tony duell]

> 
> If it has a transistor at the primary side, and a high frequency transformer, it's a switching supply.
>  Those want a load, usually.  Especially the older ones.

If there are large-ish electrolytics (100's of uF) with a voltage rating around 200V and no reason to 
believe that there is a 200V output rail then it is very likely to be a switcher with those being the 
rectified mains smoothing capacitors. 

Some SMPSUs (Switchers -- it stands for Switch Mode Power Supply Unit)  need a load, other's 
are happy without one. Occasionally DEC put the specification of the supply in the printset -- they
did for the 11/730, and IIRC that supply is happy with no load on all outputs apart from something
odd like the +12V rail. I've yet to find a DEC PSU that fails destructively if not loaded, it is more likely
to just trip the overvoltaeg protection crowbar, shut down, then try to restart and repeat the cycle, 
tweeting as it goes 


> If it has a laminated iron transformer, big electrolytics for filtering, and transistors at the
> secondary side, it's probably a linear regulator supply.  Those don't need a load.  You
> may want to do full testing under load to observe correct regulation, but for initial testing 
> you can test them without.

Be careful!. There is a third type of supply, and it's one that DEC used a lot, going back at 
least to 1970 (that is, it turns up in most older PDP11s, etc). This has a big laminated core
mains transformer, rectifiers and smoothing and then _switching_ regualtors (not linear)
to produce the right output voltages. As with all switchers, that type might need a load
(but my experience is that the DEC ones don't).

> Linear supplies are older; I think by the time of VAX you're likely to see switchers (with the
>  possible exception of first generation stuff like the 780, I don't remember if those big
> supplies are linear or switchers).

Certainly the 11/730 is a mains switcher (it rectifies the mains to get about 350V DC, then
chops it). In fact it's 4 SMPSUs in that little box. 

The 11/780 draws a lot of power. I don't think they would have used linear regulators (which
by definition are inefficient) in that machine.

I would bet quite a lot that all MicroVAX PSUs are mains switchers

Also watch out that DEC somethimes had a little linear PSU to power the SMPSU control 
circuitry, at least before the SMPSU had rattled into life. The control circuitry (oscillator, 
regualtor, etc) is on the _output side_ of the isolation barrier in such supplies, the drive
signal is transformer-coupled to the base of the chopper transistor. Those turn up in
all sorts of things, the VT100 being the most common example. 

I hate to say this, but DEC seemed to like complicated and unusual power supply designs, 
and I would not want to try to repair one without a good schematic. 


-tony