Andre Jute wrote:
Patrick Turner wrote:
I have just uploaded the page on the 845 amps I completed last month.
I invite you all to try the link to see the amps at
http://www.turneraudio.com.au/monobloc845se55.html
Please let me know if there are any browsing problems.
Patrick Turner.
Stunning design and craftsmanship, Patrick. I like the liberal (but
not thoughtlessly stupid) use of chokes; they're an underrated
component these days. You will not be surprised to discover I have
already downloaded your protection schematic...
Chokes are good, no doubt about it. the alternative
woud have been more R&C sections, which I sometimes employ where
there isn't room to put a choke.
So if you have large C values, like the seriesed 470uF caps for 235uF,
it has 6.8 ohms reactance at 100 Hz. If R is 68 ohms, you have 100Hz
attenuation facto of 1/10.
3 such sections give AF = 1/1,000. so if Idc was 200mA, then Vdc drop
over 204 ohms = 40V,
and PdR = 8W, which isn't too bad, and Vripple at C3 = 1.9mV.
A choke of 4H has X = 2.5k, so one LC section after C1 has AF = 1/2,700,
so Vr = 0.7mV.
But there is an LC resonance between 4H and 235uF at 5.2Hz, and the amp
load isn't low enough
to damp it with its shunt loading across the C. So the other way to damp
it is to have
some added series R of approximately 1.4 x XL or XC at Fo, ie, 180 ohms.
The dcr of the choke is about 40 ohms, so an added 100 ohms gives a
dampiing effect to stop
LF noise from mains noise causing somewhat high undulations of the B+
centred around 5.2Hz.
Ad to assist further, I've added another R&C section of 100 ohms plus
235uF,
so the PS B+ ands B- rails are fairly quiet, and well damped against LF
transients
generated either from the damn mains or the amp itself.
Very slow frequencies in PS rails below 5Hz
are highly attenuated elsewhere in the amp circuit.
There are no browsing problems but there is a formatting solecism in
that your software has added a hard break at the end of lines. Would
you like me to download that page, fix it and email it to you to post
back up?
What is a hard break? formatting solecism?
The Mozzilla wyswyg composer does not promt you about such things, and I
just blithely
carry on ingnorantly sometimes.......
The trouble with computing is that the human cannot fully know or
remember what he's doing.
Yes, you may return a corrected version of the page, pehaps without the
images included
because of the email size.
Maybe its easier to somehow apply a simple command like "remove all hard
breaks" at this end.
BTW, KR Audio make a range of amplifiers using their own tubes.
One is a 30W + 30W integrated amp with a KRT100 tube, which is very like
the 845,
but Ra = 1.2k instead of 2.2k at the same Ea and Ia as the 845.
Biasing is about the same as the 845, and the KRT100 would work in any
circuit needing an 845
Except for one DAMN thing, the cathode heating needed is 2.5V x 2A,
which means the KRT100 is NOT
a plug in replacement for the 845.
There is one of these integrated KR 3030 amps in alocal hi-fi shop here
and the price is usd 14,400.
It will take them years to sell that amp here.
It uses a solid state input and drive amp circuit.
So it just ain't the full real Mc'coy tube amp.
However, the KRT100 will run happily at Ea = 800V, and with Ia at a safe
100mA for Pda = 80W,
and because RL could be 5.6k, then DF will be about 4 without loop FB,
and
nearly as good as an 845 with Ea = 1,050V, and RL = 12k, with DF = about
5.
The lower RL at the lower Ea means you could have a 50H choke fed anode,
and use a 50uF polypropylene
cap to couple a Hammond 1650P OPT which is rated for 6k6 : 4,8, and 16
ohms.
Ppl would not have to construct or buy a complex and expensive air
gapped OPT.
I thought of doing choke feed in my 845 amps, but chickened out in
favour
of doing things traditionally, with the only "compromise" of the split
rail
to avoid HV dc across the OPT insulations.
Prices for KR845 AND KRT100 are about usd $534 retail now from what I
see.
But availablity is whole other issue, and apparently there are no stocks
in Prague
where the KR company operates, and production of 845 is not due to
commence until October 08,
and lord only knows if any quanity will emerge.
Meanwhile, the most basic cheap version of Shuguang 845 works just fine
at less than 1/3 the KR price.
This Chinese triode seems like a faithful copy of the original US mades,
and comes without frills such a a copper base which seems to only add to
the price.
A pair of 845 used in SE parallel and Ea = about 1,000V can have a 50H
choke feed,
and have a cap coupled Hammond 1650P, OK for 60W into 6.6k, or maybe
better 1650R, rated for 100W into 5k 4,8,16 if you wish.
The 50H choke for the choke feed isn't very difficult to make.
In my OPT, only about 1/3 of the winding space in the 76mm x 25 mm
window is used for the primary wire.
That is because of the space needed for P-S insulation and the
secondaries.
The core has a 51mm tongue x 72mm stack, and if 3/4 of the winding
window was filled only with
fine wire as for a choke, there'd be about 6,000 turns, and the L value
would become huge.
Its probably better is to have two slightly smaller chokes in series to
make the Cshunt of the choke feed lower,
and the halve the ac voltages across the coils.
The chokes don't need to have GOSS, and any old lams you have laying
around or from
cooked old PT will do, and the winding need not be layer wound and can
be random wound but
using a fairly slow traversing speed to keep wire crossing angles very
shallow.
Varnish can be sprayed on from a can of clear varnish from a hardware
store
as you wind, and you don't need to count the turns, just fill the
bobbin, using about 0.45 Cu dia wire.
Patrick Turner.
Andre Jute
Mucho impressed