Hi all, I bought two dozen of these back in May 2010 and have had a lot of fun with them since. First I pulled the EI KT90's out of my Mc 60's and plugged in the KT 120's. Wow did my TV system come to life. ( Mc 60 's were made to power C-8's so the extra filament is no issue with them.) Then I built a stereo McCopy using some old Mac transformers and a UTC plate and filament ( see my gallery ) and now I'm building a stereo 500 wpc version using 16 of the KT120's. You do need to use a cathode follower to drive them.
Ron
Ron
I would agree with this, design however so that if either wiper becomes intermittent you get full bias voltage rather than none. You also need to be able to assure that current is shared equally between tubes and you must therefore measure the cathode current of each output tube. Seems like a dedicated meter in the amplifier chassis that can be switched to measure bias current of each tube and overall balance too would be best.
In terms of bias stability fixed bias has no advantage as there is no feedback mechanism to regulate the operating point, however it does have have the advantage of eliminating a less than ideal component in the signal path (the large electrolytic cap) and makes the "tuning" the operating point for best performance quite easy. Given it is a bass amp I would probably go with fixed bias, and run each output tube at around 80 - 100mA on 400V - 450V B+.. I believe 25W should be achievable in SE UL with a pair of these tubes.
the panel meter is 'borrowed' from other designs, and a mix of several
with 'fixed', I dont understand why there still is a very big bypass cap on catode resistor
but its shown like that
with switch in 'neutral' the meter is out of active curcuit, and only switched in for casual checking
I suppose a multimeter will still be used for eventual adjustments
the 100ohm I inserted just before each KT120
I was hoping they could be used to adjust and do the 'balance act'
or else I suppose I will have to insert yet another 'fixed bias curcuit', one for each tube
now that its for instrument use, I thought it might be practical to have a power out meter
any ideas about that ?
but theres no grid to ground resistor with 'fixed bias' ?
with 'fixed', I dont understand why there still is a very big bypass cap on catode resistor
but its shown like that
with switch in 'neutral' the meter is out of active curcuit, and only switched in for casual checking
I suppose a multimeter will still be used for eventual adjustments
the 100ohm I inserted just before each KT120
I was hoping they could be used to adjust and do the 'balance act'
or else I suppose I will have to insert yet another 'fixed bias curcuit', one for each tube
now that its for instrument use, I thought it might be practical to have a power out meter
any ideas about that ?
but theres no grid to ground resistor with 'fixed bias' ?
Attachments
The simplest thing would be for each tube to have its own bias pot, which also means each must have its own coupling cap to the driver stage.
There is no grid resistor to ground with fixed bias, that resistor goes to the bias pot wiper. Note that there should also be something like a 47K - 100K resistor from the wiper to the most negative side of the bias pot to insure that the tube does not self-destruct in the event that the wiper goes open. I'd use a locking pot with a value of 5K or 10K for each bias pot. The grid bias resistor should be no larger than the maximum specified in the data sheet. It strikes me that a D3A or 7788 would make an excellent driver for this application.
The resistor in the cathode should be 10 ohms, or if you can get a very sensitive meter (say 200mV full scale) then a 1 ohm resistor will be fine.
Get a decent enough meter and nothing else will be required once everything is working properly.
In answer to the question you posed in post #72 Garter bias can be used with a pair of output tubes in SE.. Connect the plates in parallel, and capacitively couple each output tube independently to the driver stage. The Garter bias circuit works in the DC domain and doesn't care at all about the phase of the AC signals on the grids or plates, but each output tube must still have its own coupling cap back to the driver stage.
Last edited:
Nope, you need a big honking cap across each CCS to provide a low impedance AC path to ground in the cathode circuit.
Garter bias IMHO is much simpler and should work well enough if you want cathode bias.
I still prefer fixed bias.
some osmosis may be required hey, Kevin
in post 72, did you notice that in the first 'schematic' I suggested fixed bias, but with CCS in the catode
I thought that maybe the CCS would 'balance' the tubes
but maybe it doesnt work like that, with fixed bias
ok, I think we got the fundamental basic's of the Blumlein Garter version sorted
and just for show, here it is
lets forget that one, for now
on to fixed bias
I have some C3g, so it might be practical for me
but on Kevin's suggestion, I will take a look at D3a
I know some amps with D3a to study
I think it has been hard to get
but seems like its become available again
and just for show, here it is
lets forget that one, for now
on to fixed bias
I have some C3g, so it might be practical for me
but on Kevin's suggestion, I will take a look at D3a
I know some amps with D3a to study
I think it has been hard to get
but seems like its become available again
Attachments
C3G might also be a good choice parametrically speaking, but one potential concern with this tube is that some samples apparently have fragile filaments. Not sure how universal this is, however I was warned by some long time users not to disturb them in any way until the filaments were quite cold. (Mine are still sitting in the tube stash cabinet as a result..)
for a guitar amp in general, C3g doesnt sound like the best choise
whatever, I'm not going to carry it around
C3g, I think I was 'fooled' into that one by the low noise claims
anyway, D3a is claimed to be more linear in triode mode
so, thanks for 'reminding' me of that one
also seems like its a bit stronger driver, higher current etc, yes ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.