Dynaco ST70 KT88 Mod Question
Vincent Chong
My ST-70 had been mod to replace the EL-34 tubes to KT-88 by the previous
owner. I like the sound of this amp much more than the other ST-70 I have
with original circuitry.
I'd spent some time comparing
1. Values of the 4 coupling cap changed to 0.22.
2. One 10K resistor near the bias supply changed to 18K ( to the ground).
Maybe I'd missed the thread about the same thing discussed few months ago.
Is anyone there has any information regarding this?
What should be the correct bias setting?
Thanks,
Vincent Chong
Ross Ellingwood
Ross
"Vincent Chong" <vin...@proactiv.com.sg> wrote in message
news:3cb6...@news.starhub.net.sg...
Jim Doyle
: Maybe I'd missed the thread about the same thing discussed few months ago.
: Is anyone there has any information regarding this?
: What should be the correct bias setting?
I was chastisted for doing this myself, however I see no fundamental
reasons not to do this as long as you respect the original design
limitation of the power supply for filament supply draw and b+ idle..
Go no more than 50mA idle current on each tube, otherwise you are likely
to shorten the life or cook your power transformer.
I myself am using KT-66's (Shuguang) wired up as triodes.. 410V on the plate,
grid fixed bias around -40V. At idle, it draws around 50mA. I was just
fartin' around and discovered that KT-66s have a nice loadline push-pull
into 4300 ohms and 400V. ;) They were quite the surprise as far as sound..
On Klipsch speakers, these KT-66s sound stunning.. gorgeous midrange and
smooth and detailed highs. Aint never going back to EL-34s in this amp. :)
Get me some measurements, whats your fixed bias supply? And what is
driving them. 400V is alittle low to run KT88's, but that doesnt mean
they wont play music.
-- Jim
Vincent Chong
65-70mA but sound did not get as good.
There are significant tonal difference, KT88 has more detailed. The bass had
also improved.
"Jim Doyle" <do...@shell01.TheWorld.com> wrote in message
news:GuGr4...@world.std.com...
LiloLee
If so could you list them for us?
Cheers
Lee
Robin
Vincent,
Every Dyna power transformer I've seen runs very hot but of maybe 20
amps, or so, that have been through my shop for restoration or
repair/mods, only 1 had a burned out transformer. The heater current
for EL-34 is 1.5 amp, 6550 and KT-88 is 1.6 amp, KT-66 is 1.3 amp, I
think, and 6L6 is only .9 amp. All OK for the Dyna 70. The 18K bias
resistor is intended to add an extra -8 to -10 volts to the bias
adjust range. This allows the other tubes to be biased up correctly.
I wouldn't exceed 50 ma., as you said, you won't hear much improvment
and those transformers are scary hot.
On the other hand, if you lean the idle current down very much, (to 25
ma., or so,) thinking that going more into class AB1 for more output
power will something to try, I wouldn't do that either. The power
supply, if stock, does not have enough decoupling to keep all the
misbehaviour out of the audio signal. Not a pretty sound.
I really like those amps, though. I have a beat up one to build
experimental circuits in and I have another that is very pretty and
restored to like new condition.
Hope this helps,
Robin
Vincent Chong
I notice that Mark III uses 100K matched resistors instead if the stock
270Ks.
Do you think that the value depends on the type of OP tubes used?
Vincent
Robin
Vincent,
The 6550 and KT-88 types need a lower impedance from the input grid
back to ground or -bias. They have a tendency to build up a charge on
the input grid, probably the much larger metal structure produces more
gas and impurities, and the grid will go more positive. This causes
the tube to run hotter, increasing the problem and it may run away.
You'll see the plate go red hot, tinkling glass noise, etc.
However, I don't think this is likely with the 70, not at only 50 ma.
and 410 to 420 volts across the tube. If I remember right, the Mark
III runs at well over 500 volts. I've run KT-88s at 470-480 volts
with 200K to -bias and no problems, for about 8 or 10 years of
building stuff. (Of course, having said that, today everyone will
show up on the doorstep with a smoked amp.) Hm. Oh, well. I don't
think you'll have a problem. A high heater voltage, over 6.5 volts
will aggravate the problem but I don't think even that would cause a
problem with the 70. I'm not sure how much you could reduce the value
of the grid resistors and stay linear with the 70 but probably quite a
bit. Your ear would tell you as much as a distortion analyzer.
Robin
Vincent Chong
>
> The 6550 and KT-88 types need a lower impedance from the input grid
> back to ground or -bias. They have a tendency to build up a charge on
> the input grid, probably the much larger metal structure produces more
> gas and impurities, and the grid will go more positive. This causes
> the tube to run hotter, increasing the problem and it may run away.
> You'll see the plate go red hot, tinkling glass noise, etc.
>
> However, I don't think this is likely with the 70, not at only 50 ma.
> and 410 to 420 volts across the tube. If I remember right, the Mark
> III runs at well over 500 volts. I've run KT-88s at 470-480 volts
> with 200K to -bias and no problems, for about 8 or 10 years of
> building stuff. (Of course, having said that, today everyone will
> show up on the doorstep with a smoked amp.) Hm. Oh, well. I don't
> think you'll have a problem. A high heater voltage, over 6.5 volts
> will aggravate the problem but I don't think even that would cause a
> problem with the 70. I'm not sure how much you could reduce the value
> of the grid resistors and stay linear with the 70 but probably quite a
> bit. Your ear would tell you as much as a distortion analyzer.
>
> Robin
I tried replacing Mullard's EL34 quad with 4 brand new Svetlana SV6550Cs
($25 X 4 here). Running 50 ma idle current on each tube. I've got very
stable and good result. More revealing detail. I think I would stay on to
6550C for quite some time.
For those who are interested to try this mod, I suggest the bias supply 10K
resistor (to the ground)be replaced by 15K. So that you could switch between
El34 or KT88/6550 easier.
Jim Doyle
: Jim did you change any of the components?
: If so could you list them for us?
Oh yes, the driver stage is totally different - it uses a 12AX7 LTP
for split and gain. Still, with a 7199 driver, a KT66 in the
output tube sockets wire as a triode (100 ohm resistor between pins
3 and 4) should work just fine... Have a lookg at the triode
characteristics for EL34 versus KT-66 then KT-88. The KT-66's are
closer to the usual operating point of an EL34...
-- Jim
LiloLee
I really must try out triode mode on the ST70. May wire it with a switch
depending on whether it's jazz or rock.
Lee