With 0.22uF input and 5.6uF output caps to your 20K DC input power amp there is -6dB at 1Hz. Try 47nF input cap for -12 dB. If with 2.2uF output cap also there will be -18dB at 1Hz for your setup. The higher near DC roll off should help pumping control.
Finally finished reading through this whole thread and heavily leaning toward making a cathode follower version to replace my Korg B1 > F6. Mostly going to feed it with a UFSP connected to a Hana SH (2.0 mV output).
I've been on the fence about making a lower gain version of the original build, but I think the CF should be sufficient (and adjustable by fiddling with the gain level on the UFSP) -- anyone disagree?
As Salas knows, I'm addicted to my Valve Itch Phono -- I just have to try another one of his tube designs.
I've been on the fence about making a lower gain version of the original build, but I think the CF should be sufficient (and adjustable by fiddling with the gain level on the UFSP) -- anyone disagree?
As Salas knows, I'm addicted to my Valve Itch Phono -- I just have to try another one of his tube designs.
To have a workable system gain with Valve Itch phono or UFSP phono and the cathode follower 6V6 line pre version, you would need no less than 26dB gain power amp and no less than 90dBSPL/2.83V sensitivity loudspeakers. If to overdrive at the source for compensation, better use the Valve Itch as it has tubes on HV supply giving it much more overload margin than the circa 35V PSU for the JFET UFSP phono.
Thanks Salas — this system would use the UFSP as my phono (VIP is in my main rig). Speakers are less than 90dB and the F6 is 14dB. I typically listen at low volume in this system, but I guess the gain version and the Korg B1 are pretty similar in gain (15 and 16db respectively)?
Gain of the triode 6V6 is 15 dB. Using this beast as a pentode will let you go well under this, or over it quite easily.
cheers.
Douglas
Cool -- just grabbed a quad of Winged C 6P6S just to stock up. Judging by current shipping times to the States from Ukraine/Russia, probably going to be a June-ish build.
I was originally planning to use my currently homeless Janus Shunt Regulator for this build, but I also have a SSHv2 on the way from the current GB.
I was originally planning to use my currently homeless Janus Shunt Regulator for this build, but I also have a SSHv2 on the way from the current GB.
Just for clarification of gain, and u; and triode wiring versus beam power wiring.
The 6V6 u = 9.8 (19.8 dB).
The gain of a triode wired 6V6 with a perfect current source plate load, and a cathode that is bypassed is 9.8 (19.8 dB).
Of course, the load also has to include the next stage (including the changing reactance versus frequency of the cable capacitance from preamp to amp).
The 6V6 gain of 15 dB (g = 5.6) is caused by u x (RL/(rp + RL)).
The gain of a Triode Wired tube is dependent on RL.
(just as the gain of a Beam Power wired tube is dependent on RL).
The gain of Beam Power is transconductance x RL (Gm x RL).
Again, the load of the next stage (including cable capacitance) causes the gain to be reduced.
What would you like the gain to be?
Calculate the required RL to get the gain you want (within limits).
Bandwidth is another matter . . .
RL may not be flat versus frequency.
Miller Effect in triode wired mode may be a factor.
Etc.
The 6V6 u = 9.8 (19.8 dB).
The gain of a triode wired 6V6 with a perfect current source plate load, and a cathode that is bypassed is 9.8 (19.8 dB).
Of course, the load also has to include the next stage (including the changing reactance versus frequency of the cable capacitance from preamp to amp).
The 6V6 gain of 15 dB (g = 5.6) is caused by u x (RL/(rp + RL)).
The gain of a Triode Wired tube is dependent on RL.
(just as the gain of a Beam Power wired tube is dependent on RL).
The gain of Beam Power is transconductance x RL (Gm x RL).
Again, the load of the next stage (including cable capacitance) causes the gain to be reduced.
What would you like the gain to be?
Calculate the required RL to get the gain you want (within limits).
Bandwidth is another matter . . .
RL may not be flat versus frequency.
Miller Effect in triode wired mode may be a factor.
Etc.
Last edited:
Thanks very much for this — it’s a really helpful and simple explanation. It’s been a couple years since I spent much time studying tube circuits, but I spent a good chunk of today drawing 6V6 load lines and I’ve compiled what looks like a pretty decent simulation of this circuit in LTspice. The simulated bias and anode voltages look fairly accurate and the gain is performing as expected in SPICE.
Some discussion not on the thread topic has been removed. Some posts were rules breaking and are being considered by the team.Further (civilised) continuation of this topic should be taken to its own thread.
I'm considering putting together a simple PCB design for this that anyone would be able to have made with the files freely available. Is anyone interested?
I want to make a couple personally so the PCBs would be useful for me.
I want to make a couple personally so the PCBs would be useful for me.
