Oh, on the bypass caps, the Janus I bought (tested only, didn't use, I wanted line regulation also it turned out) I bought it with the 1uf RTI caps supplied by tubecad, so I don't think that 4.7uf value is too critical. It did improve the sound of my ACF board a bit, but I ended up going for the other tubecad reg for line regulation.
1.5k is fixed and based on the characteristics of the tubes used. The Tubecad articles explain. I tried the janus reg on my ACF linestage, and it was an improvement over a passive CLC supply particularly in the highs. It did not solve my problem with line voltage fluctuations appearing on the output, so I built a PS-2 reg instead. That sounds good too!
Has anyone tried to use one paralleled 6n2p and 12au7 instead of 12DW7?
I have some ecc82 and 6n2p at home and was wondering how to use them paralleled in janus shunt reg but didnt figured out how the constructor get the 1,5K number for series resistor.
Can someone explain how this would alter the Rs...
Thanks...
I have some ecc82 and 6n2p at home and was wondering how to use them paralleled in janus shunt reg but didnt figured out how the constructor get the 1,5K number for series resistor.
Can someone explain how this would alter the Rs...
Thanks...
After shelving the Janus for years I decided to revive it and do it proper justice by running spice simulations on the circuit. Turns out the noise cancelling and voltage stabilization effectiveness is very sensitive to input voltage, transformer regulation (drop when load applied), the plate resistor of the 12AX7 section, and cathode resistor of the 12AU7 section, and the load.
I needed about 340V B+ with a 28-30mA load. With 390V input B+ I ended up using 120k for R4 instead of 200k, 1200R for R5 drop resistor, and 13k for R6. The ripple and noise rejection capability varies a lot with R4. tweak that value by 10% and you see noise on your scope go from 5mV (good) to about 100mV (very bad).
I now have renewed faith in this thing which actually seems to hold B+ pretty well even without a Vref.
Check your ballpark voltage of V-in once you get the thing going and run some numbers in LTSpice. This really helps.
I needed about 340V B+ with a 28-30mA load. With 390V input B+ I ended up using 120k for R4 instead of 200k, 1200R for R5 drop resistor, and 13k for R6. The ripple and noise rejection capability varies a lot with R4. tweak that value by 10% and you see noise on your scope go from 5mV (good) to about 100mV (very bad).
I now have renewed faith in this thing which actually seems to hold B+ pretty well even without a Vref.
Check your ballpark voltage of V-in once you get the thing going and run some numbers in LTSpice. This really helps.
I just completed a build for my aikido 6cg7 / 5687 running about 36 mA and B+ settled in at 275.
Was using an all SS ps-3 before same current but 300V. The pre was quiet before at least that's what I thought. Now it truly is dead silent. I always had a little buzz when cranked up past 50% which is too loud for normal listening levels so no biggie.
But I have never been able to go wide open and not hear a thing...
All I had to do was put in a 5V xformer.
Using 5y3 and 12dw7 and all standard values.
Was using an all SS ps-3 before same current but 300V. The pre was quiet before at least that's what I thought. Now it truly is dead silent. I always had a little buzz when cranked up past 50% which is too loud for normal listening levels so no biggie.
But I have never been able to go wide open and not hear a thing...
All I had to do was put in a 5V xformer.
Using 5y3 and 12dw7 and all standard values.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.