Nicely done Gtoro. What values you use for passive PSU filtering as it is now and does it play without hum and buzz? Sounds good? If it is, don't throw away more B+ because it is already on the low side. AC heating is natural and good as long as it does not intrude. In FFT it shows like spread spikes of harmonic noise that if they are not high enough to hear 100Hz buzz they can thicken the sound a bit. With good cable dressing there can be very good results but one has to see what's going on in the analyzer. For DC heating I recommend a current source. Is it the gain or the buffer version?
Hi salas! Thanks a lot you are my reference....
The pre is cf version Dc heater very quiet but you know...if you like a device you would improuve its perforances...so i take time to listen and enjoy it, but i would try to double B+ after rectifyng because transformer is big and i have current and use gyrator to kill the residual hum audible without song at maximum level
The pre is cf version Dc heater very quiet but you know...if you like a device you would improuve its perforances...so i take time to listen and enjoy it, but i would try to double B+ after rectifyng because transformer is big and i have current and use gyrator to kill the residual hum audible without song at maximum level
A gyrator is constant voltage anode loading and will fix the anode voltage to a level, but we must also have info about each such circuit's PSRR. It can prove good or not to your ears, still its an addition to the anode circuit. If you will voltage double then there will be much dissipation across it.
When you probably spare quite some voltage in the passive CRCRC filter now and still having little ripple maybe you could also consider a Maida style reg or an LR8 Supertex HV chip reg with external pass HV BJT/Mosfet. Those would be compact solutions since you say you don't have enough space left in the box. Should be good for 60-65dB ripple rejection and will provide stable voltage, without throwing away much. LR8 is happy with 12V minimum drop across to give you an idea. Thus you could minimize or discard the passive filters. No more space in your box sadly rules out bigger warmer solutions like an SSHV2 shunt for instance.
When you probably spare quite some voltage in the passive CRCRC filter now and still having little ripple maybe you could also consider a Maida style reg or an LR8 Supertex HV chip reg with external pass HV BJT/Mosfet. Those would be compact solutions since you say you don't have enough space left in the box. Should be good for 60-65dB ripple rejection and will provide stable voltage, without throwing away much. LR8 is happy with 12V minimum drop across to give you an idea. Thus you could minimize or discard the passive filters. No more space in your box sadly rules out bigger warmer solutions like an SSHV2 shunt for instance.
actyally i raised B+ reducing R to 320 V. And i'trying to use AC for heater to recover space for cleaning B+
Salas, still wrestling with how to do phase splitting. What would be your choice between the following options?
-6v6 gain stage feeding Jfet buffer that uses line output to split signal.
-6v6 gain stage feeding 6v6 cathodyne phase splitter
-6v6 gain stage feeding 6v6 cathode follower, feeding line output transformer.
I am trying to maintain the sonic signature of the gain stage. I know any addition affects this, but of the above, which do you think would be least impactful? Thanks and sorry for the trouble and thread pollution.
-6v6 gain stage feeding Jfet buffer that uses line output to split signal.
-6v6 gain stage feeding 6v6 cathodyne phase splitter
-6v6 gain stage feeding 6v6 cathode follower, feeding line output transformer.
I am trying to maintain the sonic signature of the gain stage. I know any addition affects this, but of the above, which do you think would be least impactful? Thanks and sorry for the trouble and thread pollution.
As I understand it, higher GM tubes lead ultimately to lower output impedance. While as low as possible is attractive, I figure the 6v6 would be around 600ohms. If this is incorrect, how do I go about figuring the Zo?
I used
Z=R*Rp/R+Rp(mu+2)
I lean towards staying with 6v6 simply because u it is the same tube, decreasing the chance of a different signature to the sound. It also allows use of well traveled operating points.
Is there another higher power tube, like 6v6 that should be considered? EL34 triode connected?
I used
Z=R*Rp/R+Rp(mu+2)
I lean towards staying with 6v6 simply because u it is the same tube, decreasing the chance of a different signature to the sound. It also allows use of well traveled operating points.
Is there another higher power tube, like 6v6 that should be considered? EL34 triode connected?
If i have scanned the thread properly, the plate of the gain version sits at around 230V, gate sits at 15V above the cathode. If using a phase splitter, the voltage divider would need to set the gate 15v above the voltage at the top of the cathode resistor. At 22mA bias with 2.5K for the cathode resistor, this means setting at about 70V. Is this correct?
Hi, this we i built gyrator and try it inside: the major thing was that now B+ is growth to 310 V ( I substitute the Gyro in place of a R); also I try AC heater and as you say (Salas) the sound was thicker and very pleasant.
The HUm remain similar with AC and DC Heater so I think I have to "clean" more the B+ rail (With LR8 as you propose and/or I suspect, work on mechanical vibration of chassis made by tansformer).
A work for next rainous we.
The HUm remain similar with AC and DC Heater so I think I have to "clean" more the B+ rail (With LR8 as you propose and/or I suspect, work on mechanical vibration of chassis made by tansformer).
A work for next rainous we.
Don't know from the top of my head right now. Does not the simulator show something roughly agreeable to those assumptions?
Yes. I am one to trust practical experience as much as sims, but they do offer a very good approximation I know for your application, you used a lower Rl to help reduce Zo. In looking at the this simulator, I can get better if loading the stage with 10k and running the rails at 400V. The bias current is still in the 22mA range. I figure if we are feeding a phase splitter, the higher load resistor is not a big deal. Anyway, I am at the point of prototyping the idea. I hope to share measurements soon. Thanks for the help.