John Byrns wrote:
In article ,
"Phil Allison" wrote:
"robert casey"
Looks like the circuit is expecting the diode to start conducting when the
voltage gets higher than the 3KV,
** If the R3000 diode (string) ever conducts in the reverse direction -
that is the end of the diode.
It is NOT a damn zener - you clot - the voltage limiting effect works by
transformer action.
BTW:
There are in fact several ways in that Fender amp the R3000s CAN cop more
than 3 kV in reverse, both at the same time.
Then it is kaput for one of them.
See if you can figure even one of them out.
Leakage inductance between the two primary halves is one of them, I am
curious what the others are?
Speakers are inductive loads at HF, especially in guitar amps where
there is never any attempt made to place an impedance equalization Zobel
across the speaker.
A pair of Zobel networks are never placed network where they are
usually most effective, across each 1/2 of the primary on the OPT.
So a stray bit of noise with F content up to 20kHz can produce
high voltages at the anodes of a guitar amp in pentode mode because the
load
at the sec becomes high, as well as being decoupled from the primary
because of the
leakage L.
However, the high Ra of the pentoads and the shunt C of the OPT DOES
tend to shunt the
extreme HF gain of the pentoad.
I like to see a pair of Zobels beginning to load the output tubes at
20kHz at least.
I have often included the anode to 0V diodes to limit transformer
primary voltage excursions
with no load and have had no trouble with 3 seriesed diodes with PIV =
1,000V each.
The highest voltages geberated in an unloaded OPT seem to be with
pentoad amps,
and about the highest voltages I have seen are up to about 3 times the
maximum
Va-a with rated load at clipping.
So with an amp making 100 watts into 5k, this is 707Vrms, and with no
load
I have seen maybe 2,200Vrms a-a.
UL will also produce high no load voltages, and triodes not so much.
The amount of loading needed to prevent the strange non-commonsense
phenomena observed
when an amp continues to produce a really high output voltage with no
load
isn't anything like the rated load; it appears the load to prevent
more than 10% above high back-emf no load voltages is an RL of many more
ohm
than the rated normal load, but just what the general rule for such a
load is I don't know.
I've never made the trials to find out.
Some older amp schematics show what look to be a silly amount of R
across the output,
say 100 ohms, for where the load might normally be 16 ohms.
There is never a reason given for the 100 ohms, but perhaps its is to
guard against the
back emf voltage production with no load caused by the cut off of
current in one pentode
slowly turning off
while the other one which conducts until the anode Va goes down to 0V
and then it too cuts off sharply
because its anode becomes negative.
A loading of 100 ohms would not cause a huge power loss. 100W into 16
ohms = 40Vrms,
and this voltage across 100 ohms = 16 watts.
Perhaps 220 ohms would be OK, and the extra reliablity would be much
worth
the few watts wasted in the permamnent high ohm load across the output.
A better place to put such a load of say 33k would be across each 1/2
primary of the OPT.
People are welcome to investigate further, **cautiously!** and perhaps
not have
use any diodes.
But because the diodes don't waste any power, they are in favour.
Patrick Turner.
Regards,
John Byrns
--
Surf my web pages at, http://fmamradios.com/