Kalamazoo Amp Field Guide: Hum Reduction
The Valve Wizard
R1=220k
R2=~73k
c1=47uf
fill resistors = 100ohm
dcoffset = 66 volts
Sound good?
The Valve Wizard
R1=220k
R2=~73k
c1=47uf
fill resistors = 100ohm
dcoffset = 66 volts
Sound good?
Good simple circuit, but Sy, 100R in series with output???
This tweak illustrates the main weakness, simple Cathode
follower can and will distort (mostly 2nd) into low Z load...
My circuit is much the same, but your 10K is replaced by
another plate, actively driven as White Cathode Follower.
Into low Z load, it attenuates, but does not distort.
Into >10K load or where differential input is not needed,
and clean power is assured. I can see using your simpler
traditional circuit. But into <10K, especially where dirty
power, or common mode noise is a challenge, DWCF is
new king.
Complicated to explain everything that simultaneously
happens in the 47K bridge, but we ain't talking a whole
lot of parts for complicated construction. Good PSRR
could also mean reduced parts count in the supply.
Any supply 80V to 180V is fine, no circuit changes...
First non-so-obvious thing it does is regulate its own
voltage at the top plate. Load is attached at magic
nulpoint where power regulation activity is invisible.
This tweak illustrates the main weakness, simple Cathode
follower can and will distort (mostly 2nd) into low Z load...
My circuit is much the same, but your 10K is replaced by
another plate, actively driven as White Cathode Follower.
Into low Z load, it attenuates, but does not distort.
Into >10K load or where differential input is not needed,
and clean power is assured. I can see using your simpler
traditional circuit. But into <10K, especially where dirty
power, or common mode noise is a challenge, DWCF is
new king.
Complicated to explain everything that simultaneously
happens in the 47K bridge, but we ain't talking a whole
lot of parts for complicated construction. Good PSRR
could also mean reduced parts count in the supply.
Any supply 80V to 180V is fine, no circuit changes...
First non-so-obvious thing it does is regulate its own
voltage at the top plate. Load is attached at magic
nulpoint where power regulation activity is invisible.
Last edited:
Yeah, it increases source Z slightly, but can prevent oscillations. I've built a nearly identical circuit and got distortion well below 0.05% at 2V out with a 10k load. I think the CCS loading that I use in my more advanced projects is a better way to go, but the poster asked for something more suitable for a beginner.
Ya i kinda wanted something that i could understand and trouble shoot. Also running the numbers, i could pop in my 6n6p with no problem as long as i have a transformer to handle the heater current. On the other hand the ECC88 is still in production...
If some one could shoot Me a BOM of the Heretical Preamp or even just the watt requirements and silicon models i should be able to make the simplified version. Ive made the +-12vdc power supply before for other chipamps.
On another note, i have a left over power transformer and 5 henry choke. However its +- 365volts and after the clc filter the b+ is 300vdc right now. Can i just pop a 20k resistor in line to knock down the voltage or something? I dont think i can get the clc filter to go below 275.
If some one could shoot Me a BOM of the Heretical Preamp or even just the watt requirements and silicon models i should be able to make the simplified version. Ive made the +-12vdc power supply before for other chipamps.
On another note, i have a left over power transformer and 5 henry choke. However its +- 365volts and after the clc filter the b+ is 300vdc right now. Can i just pop a 20k resistor in line to knock down the voltage or something? I dont think i can get the clc filter to go below 275.
Sy: 1 Triode + 10 supporting parts
Ken: 1 TwinTriode + 12 supporting parts
If PSRR lets you lose 2 parts in the power supply somewhere, you break even.
As for understand and troubleshoot, I didn't understand my own creation for
months... Seriously... Accidentally on purpose did many good things beyond
what I had intended or thought possible for only two triodes? I've written to
Broskie more than once to ask if he would attempt a comprehensive writeup.
As all this magic balancing/cancelling out of bad stuff is sort of his territory.
I thought he could explain far more clearly than I could.
I've got two triodes doing at least 3 jobs at the same time...
Ken: 1 TwinTriode + 12 supporting parts
If PSRR lets you lose 2 parts in the power supply somewhere, you break even.
As for understand and troubleshoot, I didn't understand my own creation for
months... Seriously... Accidentally on purpose did many good things beyond
what I had intended or thought possible for only two triodes? I've written to
Broskie more than once to ask if he would attempt a comprehensive writeup.
As all this magic balancing/cancelling out of bad stuff is sort of his territory.
I thought he could explain far more clearly than I could.
I've got two triodes doing at least 3 jobs at the same time...
How do you deal with cathode-to-heater?
Actually, you can get 30-40dB of PS noise rejection from a simple CF. So if the power supply noise is 10mV (an easy figure to achieve), the noise at the output will be 80dB down from 2V nominal output. If you're willing to use that extra tube section, you can put it on top a la Tektronix and get another 30dB. But again, complications- the poster is looking for something simple.
Actually, you can get 30-40dB of PS noise rejection from a simple CF. So if the power supply noise is 10mV (an easy figure to achieve), the noise at the output will be 80dB down from 2V nominal output. If you're willing to use that extra tube section, you can put it on top a la Tektronix and get another 30dB. But again, complications- the poster is looking for something simple.
Buy a comparatively hefty 10 H. filter inductor and choke I/P filter the B+. The B+ rail voltage will be approx. 0.9X the RMS value minus losses. If that's too low a small "fudge factor" cap., to make the filter cLC, will bring the rail value up. Somewhere around 1 μF. is where behavior changes over from choke I/P to cap. I/P. An added benefit of choke I/P filtration, both pure and pseudo, is inherently good regulation.
Actually, not complicated, just unfamiliar.
Sides, overcomplicated *IS* the new simple...
Cathode to heater is within spec. Remember
top plate is self regulating, and middle cathode
hovers middle of regulated voltage, not middle
of full B+. Elevate heater if you worry, but not
necessary.
10VPP power ripple, an even easier figure...
Save you a big choke for nothing.
Sides, overcomplicated *IS* the new simple...
Cathode to heater is within spec. Remember
top plate is self regulating, and middle cathode
hovers middle of regulated voltage, not middle
of full B+. Elevate heater if you worry, but not
necessary.
10VPP power ripple, an even easier figure...
Save you a big choke for nothing.
6N1P has much lower transconductance than ECC88 (two to three times lower), so is a less optimum choice. The 180R cathode resistor may have to be adjusted to get 90-100V at the cathode, but otherwise, the simple circuit should work fine with that tube.
- Status
- Not open for further replies.