Wow, its been about six years since completing these mono amps with advice from many here. Always thought it sounded good, but just didn't wow me as much as the Pinkmouse Krell Clone and other SS amps I had...
Original Post:
Quicksilver 8417 Design - Stay with it or modify?
Anyway, I now have Quad ESL-988 and hearing that they love tubes, I decided to try the Quickies out...very clean and pretty much best I have heard.
Staying true to the QS 8417 circuit, but with SS rectifier, 47k /.47uf coupling caps, individual dual bias pots, and KT120 output tubes, it sounds great.
One thing I did differently this time is I pushed the bias on the KT120 as far as I could, all the way up to 100ma @400V on KT120 plates (the power tranny is good for 220 ma), and I measured around 210 ma out of power transformer (with the added current of 12AX7 and 12BH7 PS). I eventually backed it down to 88ma each tube. I always ran these tube at around 50ma not wanting to push anything...maybe running too cold was the reason it was just meh?
Anyway some questions -
(1) I have Edcor Output transformers that are 60 watts. No matter what I drive thru the KT120, I am only getting at best 60 watt output. Upping bias just puts more of those watts into Class A, correct?
(2) Since staying true to 8417 design, only half of the 12AX7 and 12BH7 are in use - is is beneficial to parallel the tubes? Add the diode to the output tube grid?
(3) Edcor Power transformer rated at 220ma - is it a problem running near that limit? The power transformer gets warm but not hot like my old ST-70 power tranny (like real hot). There is plenty of room on the KT120 at 400v as far as bias, and I am only at 35 watts on a 60w tube. Should last quite a bit longer, right?
I also now have a Doge LP8 as a preamp and may be one of the reasons I like this setup so much. I very impressed with the Doge.
Any advice appreciated!!
Thanks
Original Post:
Quicksilver 8417 Design - Stay with it or modify?
Anyway, I now have Quad ESL-988 and hearing that they love tubes, I decided to try the Quickies out...very clean and pretty much best I have heard.
Staying true to the QS 8417 circuit, but with SS rectifier, 47k /.47uf coupling caps, individual dual bias pots, and KT120 output tubes, it sounds great.
One thing I did differently this time is I pushed the bias on the KT120 as far as I could, all the way up to 100ma @400V on KT120 plates (the power tranny is good for 220 ma), and I measured around 210 ma out of power transformer (with the added current of 12AX7 and 12BH7 PS). I eventually backed it down to 88ma each tube. I always ran these tube at around 50ma not wanting to push anything...maybe running too cold was the reason it was just meh?
Anyway some questions -
(1) I have Edcor Output transformers that are 60 watts. No matter what I drive thru the KT120, I am only getting at best 60 watt output. Upping bias just puts more of those watts into Class A, correct?
(2) Since staying true to 8417 design, only half of the 12AX7 and 12BH7 are in use - is is beneficial to parallel the tubes? Add the diode to the output tube grid?
(3) Edcor Power transformer rated at 220ma - is it a problem running near that limit? The power transformer gets warm but not hot like my old ST-70 power tranny (like real hot). There is plenty of room on the KT120 at 400v as far as bias, and I am only at 35 watts on a 60w tube. Should last quite a bit longer, right?
I also now have a Doge LP8 as a preamp and may be one of the reasons I like this setup so much. I very impressed with the Doge.
Any advice appreciated!!
Thanks
Last edited:
1. Correct. If you want more power, you need more B+ and the correct Zaa.
2. Read this: Parallel operation of triodes.
3. If it doesn't get hot it should be ok.
FYI, this topology was widely used long before the QS. Bogen used this exact circuit in their M60, M120 and MO200 PA amps. They used SS rectifiers and high B+ for up to 60W/pair of 8417.
2. Read this: Parallel operation of triodes.
3. If it doesn't get hot it should be ok.
FYI, this topology was widely used long before the QS. Bogen used this exact circuit in their M60, M120 and MO200 PA amps. They used SS rectifiers and high B+ for up to 60W/pair of 8417.
Hmm,
Question begs an answer - if all positive and beneficial, why did QS not do it from the Get go? These were mono amps all along....
I see a benefit from paralleling the 12BH7 Phase Splitter (less that unity gain alone, so doubling gain may not matter, but gives more drive current, yes?), but wont paralleling the 12AX7 Amp stage trip up the overall gain design?
If gain increases 2X, then paralleling and putting in a 12AT7 would be a bit more gain as a single unparalleled 12AX7?
I was always kinda led to believe that a 12AX7 does not "belong" in Hi Fi - just Guitar amps and Phono stages... so going to a lesser gain 12AT7 or 6922 would be interesting...
But it sounds so Good!!
Question begs an answer - if all positive and beneficial, why did QS not do it from the Get go? These were mono amps all along....
I see a benefit from paralleling the 12BH7 Phase Splitter (less that unity gain alone, so doubling gain may not matter, but gives more drive current, yes?), but wont paralleling the 12AX7 Amp stage trip up the overall gain design?
If gain increases 2X, then paralleling and putting in a 12AT7 would be a bit more gain as a single unparalleled 12AX7?
I was always kinda led to believe that a 12AX7 does not "belong" in Hi Fi - just Guitar amps and Phono stages... so going to a lesser gain 12AT7 or 6922 would be interesting...
But it sounds so Good!!
Last edited:
Do what? Parallel? I don't know the details of the QS, but Bogen and others simply used a 7247 so there are no unused tube halves.
No, it doesn't work quite like that. You double gm, not mu. You can take advantage of that but you have to work through the stage design. But yes, more current drive if you can use it. The 8417 is a high gm tube so it's very easy to drive in AB1. No need to parallel drivers.
That notion has been debated and largely debunked. You can find several threads here on the subject. It can be very linear when properly applied. But so can the AT7.
Last edited:
Isn't mu = Gm x Rp
With same Rp (same tube), doubling Gm would double mu (be nice, I am only a mechanical engineer, and thenthitive)
With same Rp (same tube), doubling Gm would double mu (be nice, I am only a mechanical engineer, and thenthitive)
No. Not the same tube but 2 in parallel so Rp total is half. 2 equal resistors in parallel give half the resistance.
Think mechanical equivalent. 2 identical pipes in parallel give half the hydrodynamic resistance to flow.
Think mechanical equivalent. 2 identical pipes in parallel give half the hydrodynamic resistance to flow.
Last edited:
I went ahead a paralleled the 12BH7 Phase Splitter last night ... not sure I hear a difference, but done and still sounding great.
Thinking I will leave the 12AX7 alone...
Thinking I will leave the 12AX7 alone...
You can't just wire the triodes in parallel without adjusting the resistor values in the circuit.
I already had the two resistors at 22k...
From Triode.com...
4. (optional) This mod will allow more voltage drive and lower output impedence of 12AU7 or 12BH7 driver tube. (DNA to amps using 7247/12DW7 tube)
A. Install jumper between pins 1 and 6 of 12AU7 socket.
B. Install jumper between pins 2 and 7 of 12AU7 socket.
C. Install jumper between pins 3 and 8 of 12AU7 socket. The phase splitter resistors can be made any value from 22K to 47K (as low as 15K with 12BH7) ohms, so long as they're matched. Won't hurt to experiment with different values, so long as you're careful to get everything hooked back up right. Note that if this mod is performed, a tube with relatively well balanced sections is preferable.
From Triode.com...
4. (optional) This mod will allow more voltage drive and lower output impedence of 12AU7 or 12BH7 driver tube. (DNA to amps using 7247/12DW7 tube)
A. Install jumper between pins 1 and 6 of 12AU7 socket.
B. Install jumper between pins 2 and 7 of 12AU7 socket.
C. Install jumper between pins 3 and 8 of 12AU7 socket. The phase splitter resistors can be made any value from 22K to 47K (as low as 15K with 12BH7) ohms, so long as they're matched. Won't hurt to experiment with different values, so long as you're careful to get everything hooked back up right. Note that if this mod is performed, a tube with relatively well balanced sections is preferable.
Yes. Each of the 2 halves has 24k anode and cathode, so when you put them in parallel the resistors need halved.
Easiest way is to solder a second 24k in parallel on each existing resistor. That way if you want to go back to single section, you can just cut out the second resistor.
Easiest way is to solder a second 24k in parallel on each existing resistor. That way if you want to go back to single section, you can just cut out the second resistor.