Re: Ampliphone HV FAQ is done!!!

<< You might want to re-think adding an insulator to VR1 (the 7824). The case
 of the 7824 is connected to GND. The heat sink on the Amplifone is also
 connected to GND, so there cannot be a short, or any current flow of any
kind,
 between the case of the 7824 and the heat sink -- they're both already
grounded!
 On the other hand adding an insulator *will* cut down on heat transfer,
which is
 not a good thing. >>

Hello Zonn,

     Thanks A Ton for the helpfull feedback!!
On the 7824 - I thought that since the heatsink was painted, that it wasn't
being a real good ground conductor, and that the 7824 and 7924 are using pin
2 as ground. I didn't like the idea of having an uninsulated 7824 because
I've seen a quite a few boards with the 1N4001's mounted to the solder side
of the regs, and I don't think you can allways count on the insulation of the
1N4001 leads - ( plastic tube insulation seems to get brittle and deteriorate
over time).I figured if one 1N4001 leads crossed a trace on the solder side
of the board, it would increase the likelyhood of blowing an uninsulated
7824. I felt insulating the 7824 was a good way to avoid this potential
problem. (I didn't really discuss this reasoning in the faq, I was starting
to get out of hand with my ramblings anyways).
     Also, I think anybody running a Amplifone HV board without the use of a
fan would be crazy. True, an insulated 7824 is not going to transfer heat as
well as one mounted directly to the heatsink, but I felt that the fan would
do more than compensate.
     I would love to get more feedback on the 7824 insulation issue, if
enough Vectorlist guys feel that it's better to NOT insulate the 7824, I'll
put it in the faq. So far I've got Zonn in at a "NOT INSULATE" vote.

>>#2) What does MC1 do?

It seems fairly apparent that the HV section of the Amplifone is regulated
much
like that of the old (and maybe some new) constant voltage AC supplies that
use
ferro-resonant transformers.

Basically the idea is that a transformer can only be driven to a maximum
voltage, after which it saturates and an increase in input voltage does not
correspond (nearly as much) to an increase in output voltage. This maximum
voltage, before saturation, is highly dependent upon the resonance frequency
of
the transformer.

The constant voltage AC regulators use a capacitor to tune a winding of an
isolation transformer to 60hz, causing the transforming to go into saturation.
After which voltages of around 70v to 140v input has little effect on the
output, and changes in current loading also have a much smaller effect.
Pretty
cool.<<

This is great!! Now the damn "red mystery can" is starting to make some
sense! With your permission, I'd like to put your info into the faq. You
really helped me here, many Thanks!

>>I have no way of knowing exactly what's inside MC1, but the only way for
this
circuit to regulate it's output voltage, given changing output currents, would
be for some type of ferro-resonant regulation to be taking place. MC1 must
be a
type of tuning coil, possibly a coil and capacitor, that along with the
inductance of T2 is used to set the resonant frequency of a ferro-resonant
regulator. (Which is probably why T2 is so expensive and hard for other
companies re-engineer. These companies are probably winding the HV to give
the
proper increase in voltage, but not taking into account it's inductance and Q
ratings, which in most HVTs are not nearly as critical. As a result, when it
comes time to test the transformer in the circuit, they fail at regulating the
output voltage. Just a guess mind you, I wonder if anyone could verify
this?)<<

My friend Ed and I took a bad one (on that was burned on the side from the 5w
50 ohm resistor) and we were going to cut it open to see what was inside. I
wonder if he did this yet?? Ed, are you there?? I also want to run some
comparison tests between a good and bad one to see what kind of readings I
get for ohms, diode test, and how it rings on a LOPT tester.

I'd like to put this e-mail to the Vectorlist for more input. Thanks Zonn!!
Received on Mon Jun 7 09:27:29 1999