[Amps] Re Direct rectification of AC mains to derive the amp, VDD, supply

[Manfred Mornhinweg]

Peter,

> Much as I respect Manfred, I think he is "pushing things" a bit far -
> at least for "Joe Q. Ham".

I can't help but keep defending my point of view!

Consider the following situations:

1.- A ham amplifier that uses the directly rectified mains, or perhaps 
the 350V output from a power factor correction circuit (non-insulated) 
to power a floating RF power module, insulated from the rest of the 
world by its RF transformers. The whole thing housed in a metal 
enclosure, connected to safety ground.

2.- A ham amplifier that uses a conventional transformer-input power 
supply to power an RF module that's connected to the metal enclosure. 
This enclosure is connected to safety ground.

3.- A ham amplifier that uses a switching power supply with power 
transistor circuitry directly connected to the line, and a 50kHz 
transformer providing insulation, to power an RF module that's connected 
to the metal enclosure. This enclosure is connected to safety ground.

4.- A washing machine that powers its motors, heater, valves, level 
sensors, etc, directly off the mains, with insulation provided inside 
each device. Housed in a metal enclosure, connected to safety ground.

Peter, can you tell me where's the difference, in terms of safety, 
between these four devices?  Each of them has a single level of 
insulation between the mains and ground, each of them involves windings 
in a transformer or motor, and lots of wiring, and each of them uses the 
grounded metal enclosure to drain off any leakage current, and trip a 
ground fault interrupter or even the main circuit breaker, if necessary. 
There is no fundamental difference between any of them.

But there is some difference in the details: The washing machine has 
many more possible points of insulation failure, than any of the 
amplifiers. Also the washing machine operates with water, will be 
touched by people with wet hands, possibly even standing on a wet floor, 
and will usually be operated by a housewife, who cannot be expected to 
know anything about electrical safety. Even Joe Q. Ham knows more about 
this! So the washing machine is BY FAR the most dangerous of these four 
devices!

And which one is the safest? The ham amplifier with the switching power 
supply has a small transformer packed full of magnet wire. I know from 
experience that it takes some good design and construction to achieve 
4kV of insulation in such a transformer. If badly designed or built, the 
insulation can fail by creepage at much lower levels.

The amplifier with the conventional power supply, just as the washing 
machine, use iron cores inside the magnet wire windings. So, in addition 
to direct wire-to-wire shorts in a transformer, these devices can suffer 
from wire-to-core shorts.

The amp with the 50kHz transformer is only slightly better, because the 
ferrite used for power conversion is slightly conductive.

But the amp using the RF transformers for insulation is the safest of 
all, because its transformer cores are non-conductive, and the windings 
that need to be insulated have only very few turns, and have enough room 
around them to accomodate a thick insulation.

Surprise!  ;-)

I think that the reluctance of some people to accept line-connected 
electronics stems only from gut feeling, and not from a cool-headed 
analysis of the advantages and disadvantages. Unfortunately all 
electronicians are taught to think in a grounded frame: There is ground 
in every circuit, everything is referenced to ground, bypassed to 
ground, dumped into ground, etc. Ground is our good friend who swallows 
all bad signals, can be trusted to be always at zero volts, and so on. 
It can be difficult for ground-dependent electronicians to accept the 
concept that a ground point isn't necessary at all for electronics, and 
that actually that single ground point that joins one side of ALL power 
and signal lines is a really big problem, simply because it's 
technically impossible to keep all grounded spots at exactly the same 
potential! It would be much healthier if electronic designers would 
consider BOTH poles of each signal and supply line, and handle both of 
them intelligently, instead of joining one pole of each signal and 
power, and calling that point "ground". This means differential signals 
throughout a circuit, and floating supplies, and the lack of a common 
point that could be called ground. This whole electronic circuit would 
be treated exactly like the wiring in a house: There would be one, and 
only one, point where the whole circuit would be connected to physical 
ground, to avoid problems of static build-up and the like. But there 
wouldn't be any single chunk of metal to which hundreds of components 
connect in common.

Many electronicians can understand this, and even like it, but in their 
daily life and work they fall back into the old concept of using 
single-ended signals and referencing everything to ground. And that's a 
real pity!

Manfred

========================
Visit my hobby homepage!
http://ludens.cl
========================