GZ34
The PS transformer is wound serially so that the two sides of the winding are not the same resistance.
I did the work on this all of 20 years ago. The results got published in Glass Audio Volume 10, Number 3, 1998 & Electronics World (June 1998). And posted on RAT, several times over several years. Im surprised you are unaware of it. The resulting problem applies only to SET amplifiers, although PS hum sidebands do show up on the output off PP amps. Measured & posted that too.
The simple, low cost fix reduces the power frequency (60 Hz)hum component as much as 20 db. A simpler fix reverses the HV leads to the rectifier. That way for a filamentary rectifier such as the common 5U4G one of the HV wdg will be in series aiding while the other is series opposing to the 5V across the cathode surface. That is enough to yield a 15 db reduction in the power frequency component. And costs us nothing. Half of the amps built probably have those leads hooked up wrong way around.
Better measure some of your power transformer HV windings carefully. That is easy these daze with a digital meter. If you find any where the halves are the secondary are the same resistance, post the p/n & manufacturer here. Any Ive ever looked at were in all cases different resistance.
Most of the schematics show resistance for the halves of the secondary to be the same. But they are not. Quite a few of the old radio schematics out of the 30s show the actual HV resistances. For example, the Philco 37-93 is given as 230R & 250R.
I'm not sure how having a tube rectifier there, vs solid state, would in
any way improve the hum factor.
But there is a difference. The forward resistance of a SS diode is much less than a Toob. So the charging currents are larger, if all other things are left the same. The there is more hum resulting. Much of the work on that was done in the 30s by Terman & others, all in his textbooks. And copied into the Radiotron Designers Handbook (RDH4), commonly used by many of us old guys. All easy reading.
As for the PS transformer, they were obviously being cheap if that happens.
Not cheap at all, simply makes good business sense. Over many years in sales with HP & R&S I had quite a bit of contact with Hammond. One comment I recall had to do with the relative cost of a tap vs. starting another winding.. By the time that extra step filters thru to the end user it could be significant. Sales volume depends on reasonable cost as well as perfection.
one appeal of tube vs solid state rectifiers is the 'punch' effect you
get with a guitar amp. At high load you get an 'attack/decay' amplitude
effect because the capacitors discharge for a short time, giving you
slightly higher power levels than the power supply can deliver with a
sustained signal. This is due to internal resistance of the rectifier
tube. But if the behavior is non-linear, having a series resistor on a
diode won't give you the same effect...
Anyway, they should be using LC filters after the diode to minimize hum,
and a properly balanced power transformer secondary winding. That's how
the old Hi Fi systems were done. Just boost secondary voltage a bit to
compensate...
/me would consider adding a series voltage regulator using a transistor
and zeners and solid state rectifiers and a current limiting circuit, to
give you a nice clean primary power supply that doesn't 'surge' the
capacitors when it starts up.
SS circuitry makes life easy in many ways, so SS regulation is a possibility. Not so much when Williamson & the others got going.
A 25-50V drop would be sufficient, not
unlike what you get with a tube rectifier anyway.
but yeah, that's not very 'purist' of me. I just think the pre-amp and
output tubes are what matters for hifi tube audio, or guitar amp audio
for that matter. The power supply [other than the 'punch' effect] is
less important, but could be simulated if needed, using a derivative of
a current limiter circuit.
The guitar pickers & other electronic musicians have much greater problems to solve than PS hum. Ground loops can be a killer. It is somewhat of a surprise that we dont see a lot more balanced systems as used in broadcast where 600 ohms transformers are used. There certainly have the audio bandwidth required. But here again cost is a killer.
This note is not meant to flame you or anyone else. Just some information where Ive had a careful look at a potential problem & found a reasonable way out!
|