You can definitely use just the one 6SL7 and you can keep the split load inverter if you want for the second stage. The advantage of the LTP is that is has gain. OTOH, I don't think you will need any more gain; it kind of depends on how much feedback you dial in. You will want to simplify the existing feedback circuit I think to make it easier to adjust.
You occasionally see a 12AX7 or 6SL7 as a cathode follower and I always wonder why. They would seem to be the exact opposite of what you would want for a CF: low gm and high gain. For a cathode follower I would use a high gm and low gain.
Have fun with your project.
You occasionally see a 12AX7 or 6SL7 as a cathode follower and I always wonder why. They would seem to be the exact opposite of what you would want for a CF: low gm and high gain. For a cathode follower I would use a high gm and low gain.
Have fun with your project.
Thanks for the response. Naturally, it raises more questions. Mind if I pick your, or anyone's, brain?
If I keep the split load inverter, can I just use 1/2 the 6sl7 (or any other high mu twin triode) for the driver and the other 1/2 for the inverter? this would be similar to how the current 6U8 is used and not require chassis holes, just circuit adaptation.
Could I use a lower mu twin tube like the 6CG7 (which I already have) in a cathode or even MOSFET follower as input into another 1/2 6CG7 splitter or a whole one as a LTP?
If I keep the split load inverter, can I just use 1/2 the 6sl7 (or any other high mu twin triode) for the driver and the other 1/2 for the inverter? this would be similar to how the current 6U8 is used and not require chassis holes, just circuit adaptation.
Could I use a lower mu twin tube like the 6CG7 (which I already have) in a cathode or even MOSFET follower as input into another 1/2 6CG7 splitter or a whole one as a LTP?
I wouldn't use the 6SL7 for a SLI for the same reason I wouldn't use it for a cathode follower: low gm and high mu. You need a low mu triode because you will get a better match top/bottom with one. Don't forget you are going to get a gain of <1 with the stage; it doesn't make sense to use a high mu tube. Further, you need a higher gm tube to effectively control the grid of the op tubes and overcome the Miller capacitance.
If you want to get off the beaten path and build something different try the 6EM7 as the front end tube. You have the voltage amp at a mu of 64 with modest gm and the SLI with a mu of 5.4 and very high gm. That would be my first choice; although with the 6EM7 or the 6SL7 as the voltage amp you will be restricted in the amount of feedback you can use depending on whether you use a preamp or not (feedback will reduce the sensitivity of the amp requiring more input voltage).
If you want to get off the beaten path and build something different try the 6EM7 as the front end tube. You have the voltage amp at a mu of 64 with modest gm and the SLI with a mu of 5.4 and very high gm. That would be my first choice; although with the 6EM7 or the 6SL7 as the voltage amp you will be restricted in the amount of feedback you can use depending on whether you use a preamp or not (feedback will reduce the sensitivity of the amp requiring more input voltage).
Ah, I am starting to understand.
So how much gain do I need for this kind of circuit?
I have a 2Vrms input, so I figure, without feedback and with ideal stage gains, I might need an input gain of about 7. Add feedback and less than 100% tube gains and a gain of 15 should be more than adequate.
Or am I off track.
So how much gain do I need for this kind of circuit?
I have a 2Vrms input, so I figure, without feedback and with ideal stage gains, I might need an input gain of about 7. Add feedback and less than 100% tube gains and a gain of 15 should be more than adequate.
Or am I off track.
That's correct: you have 2VRMS output from the CD player. Your bias on the EL34 is, from the schematic, 31VPeak or only 22 VRMS to drive the OP tubes to clipping. You don't want the driver stage to clip first, so say 25 or 30VRMS for the driver requirements. So without feedback you need a gain of <15.
I was only worried that you might run out of gain after feedback was applied. Don't forget that 6dB of feedback will reduce your gain by half (through the entire amp). You may have to breadboard this to get a good reading on your requirements. I usually end up building a project several different ways to see what works best.
I was confused above with the talk of the 6SL7 I thought there were octal sockets on the amp. Now I see the original tube was the 6U8 so you have 9 pin sockets. So the 9 pin equivalent of the 6EM7 is the 6DR7 which is not a standard audio tubes so prices are cheap: $5.10 at AES.
I was only worried that you might run out of gain after feedback was applied. Don't forget that 6dB of feedback will reduce your gain by half (through the entire amp). You may have to breadboard this to get a good reading on your requirements. I usually end up building a project several different ways to see what works best.
I was confused above with the talk of the 6SL7 I thought there were octal sockets on the amp. Now I see the original tube was the 6U8 so you have 9 pin sockets. So the 9 pin equivalent of the 6EM7 is the 6DR7 which is not a standard audio tubes so prices are cheap: $5.10 at AES.
Last edited:
Since I am ripping out the PCB and re-building P-P, I will have to replace all the sockets anyway. I have some chassis mount octal sockets that should do.
Basically, I can pretty much do what I want to this amp design within the confines of the parts I wish to keep, such as the Power and output Transformers, chokes and so on. I could even redesign the PSU, as long as I'm withing the limits of the transformer.
My next step, is to redraw the schematic as I want to rebuild it and model the PSU in PSUD2 to see what I might do there. I will post that when...when I finish the home remodel, help my wife deliver our first child and there is no more powder to ski...
Actually, I hope to do it this week while I am on call and stuck near home.
Thanks for the valuable help! I do think I will go with the 6EM7 and try triode connections for the output.
Think there is any point to puting a CCS on the input anode?
Basically, I can pretty much do what I want to this amp design within the confines of the parts I wish to keep, such as the Power and output Transformers, chokes and so on. I could even redesign the PSU, as long as I'm withing the limits of the transformer.
My next step, is to redraw the schematic as I want to rebuild it and model the PSU in PSUD2 to see what I might do there. I will post that when...when I finish the home remodel, help my wife deliver our first child and there is no more powder to ski...
Actually, I hope to do it this week while I am on call and stuck near home.
Thanks for the valuable help! I do think I will go with the 6EM7 and try triode connections for the output.
Think there is any point to puting a CCS on the input anode?
The CCS is cheap fun as far as I am concerned. I have one in my present amp which sounds great to me and I liked the amp fine without it too.
K & K Audio - Lundahl Transformers, audio DIY kits and more
I have built a few amps with that front end using unusual tubes with respect to typical audio. The last one was a 76 VA with a 6S4A as the SLI. The 6S4A is a horizontal output tube for TVs just like the 6EM7, 6BX7, 6BL7, 6DR7 and dozens of others. All these can be used in audio with the 6BX7 being the most popular. I just bought 10 6CS7 for a couple bucks each. They look great, just like a 7044 or 7119, I am psyched to try them.
Go wild and try different things. Make sure you at least have a cheap 'scope, a couple of DVMs, and a good solder station.
K & K Audio - Lundahl Transformers, audio DIY kits and more
I have built a few amps with that front end using unusual tubes with respect to typical audio. The last one was a 76 VA with a 6S4A as the SLI. The 6S4A is a horizontal output tube for TVs just like the 6EM7, 6BX7, 6BL7, 6DR7 and dozens of others. All these can be used in audio with the 6BX7 being the most popular. I just bought 10 6CS7 for a couple bucks each. They look great, just like a 7044 or 7119, I am psyched to try them.
Go wild and try different things. Make sure you at least have a cheap 'scope, a couple of DVMs, and a good solder station.
I am trying to model the current power supply in SPUD II and have a few questions.
I have measured the Transformer, which I believe is not center tapped and using a bridge.
Primary DCR 2.5R
Secondary DCR 16.8R
Primary (mains) 120VAC (right on the money here)
Secondary open load 311VAC
turns ratio 2.591
SI 33.59
Looking at the schematic a few pages back, there is a resistor to ground after the bridge of 330K R. This seems like a voltage limiter/divider for some reason. Probably since the basic simulation shows more than the 375VDC measured after the CLC.
Could i just take this out lower C1 and run the amp a little hotter with some circuit adjustments??
there is also a lot of ringing and some slow turn on, but I am guessing at the inductor values because I havent had time to set up a circuit to measure them.
Finally, How do i determine the "load" for this supply 9in terms of the resistor for the SPUD II circuit? Or does it realyl matter and I can use any reasonable value like 5K or 6K?
I have measured the Transformer, which I believe is not center tapped and using a bridge.
Primary DCR 2.5R
Secondary DCR 16.8R
Primary (mains) 120VAC (right on the money here)
Secondary open load 311VAC
turns ratio 2.591
SI 33.59
Looking at the schematic a few pages back, there is a resistor to ground after the bridge of 330K R. This seems like a voltage limiter/divider for some reason. Probably since the basic simulation shows more than the 375VDC measured after the CLC.
Could i just take this out lower C1 and run the amp a little hotter with some circuit adjustments??
there is also a lot of ringing and some slow turn on, but I am guessing at the inductor values because I havent had time to set up a circuit to measure them.
Finally, How do i determine the "load" for this supply 9in terms of the resistor for the SPUD II circuit? Or does it realyl matter and I can use any reasonable value like 5K or 6K?
The 330k resistor is a bleeder resistor to discharge the ps caps when you turn off the amp. It only draws 1mA so has no effect on the actual performance of the PS.
I am not sure I understand the question about the "load". Draw out the PS just as it is in the amp. Use current taps to represent the tubes. I would also use the stepped load function to give a more accurate representation of the startup curve. So for instance with your OP tube current tap, you could have it draw 5mA for the first 3 seconds and then 100mA as the tubes warm up and the rectifier conducts. PSUD is a useful tool, but in the end it is only a best guess.
I am not sure I understand the question about the "load". Draw out the PS just as it is in the amp. Use current taps to represent the tubes. I would also use the stepped load function to give a more accurate representation of the startup curve. So for instance with your OP tube current tap, you could have it draw 5mA for the first 3 seconds and then 100mA as the tubes warm up and the rectifier conducts. PSUD is a useful tool, but in the end it is only a best guess.
Thanks, that's helpful. I do really appreciate your willingness to answer my questions.
I figured out the 330k resistor shortly after I posted the question. I missed the obvious and now I know I can't take it out.
Are there any basic resource to learn PSUD available?
With a diode bridge, maybe the load current is not stepped? Then I may want a slower turn on.
I measured the inductors and it is only 2H, 52R.
The PSUD curve rises fast, rings a bit, then looks pretty flat, but I haven't played with he program enough to try and measure the calculated ripple.
I don't have a high voltage oscilloscope, so it is trouble measuong directly.
I guess I can play around a bit with the program, then try a few things.
I figured out the 330k resistor shortly after I posted the question. I missed the obvious and now I know I can't take it out.
Are there any basic resource to learn PSUD available?
With a diode bridge, maybe the load current is not stepped? Then I may want a slower turn on.
I measured the inductors and it is only 2H, 52R.
The PSUD curve rises fast, rings a bit, then looks pretty flat, but I haven't played with he program enough to try and measure the calculated ripple.
I don't have a high voltage oscilloscope, so it is trouble measuong directly.
I guess I can play around a bit with the program, then try a few things.
I don't think there is any kind of manual for the PSUD. You may have to google that. I think most learn the application by using it and trying different things.
With the diode bridge, the voltage will appear on the plates immediately, but the tubes will not conduct until the heaters are at the proper temp. That takes a few seconds. If you use the stepped load you can do away with the big current spike at turn on that is not really there in practice. 2H of inductance is fine, especially as the chokes are low resistance.
BTW: I found the plate curves for the 6DR7 and played around with the load lines. The first triode will work fine, but the second low mu, high gm, triode will not work without a lot higher B+, and even then it will require too much current to be practical. So my bad, these dissimilar triodes are much more practical for driving a big power triode SE. That's too bad as I still think the principle is good i.e., a single tube with dissimilar triodes that could be used for the two different functions of the topology you have, voltage amp and phase splitter/driver. The 12DW7 is the only one that comes to mind (12AX7 & 12AU7 in a single bottle), if you just want to get things running with the existing sockets I would think this tube would be the place to start. The 12AT7 or 6SL7 driving the 6SN7 or 6CG7 would probably be a better option if you want to punch a hole in the chassis.
With the diode bridge, the voltage will appear on the plates immediately, but the tubes will not conduct until the heaters are at the proper temp. That takes a few seconds. If you use the stepped load you can do away with the big current spike at turn on that is not really there in practice. 2H of inductance is fine, especially as the chokes are low resistance.
BTW: I found the plate curves for the 6DR7 and played around with the load lines. The first triode will work fine, but the second low mu, high gm, triode will not work without a lot higher B+, and even then it will require too much current to be practical. So my bad, these dissimilar triodes are much more practical for driving a big power triode SE. That's too bad as I still think the principle is good i.e., a single tube with dissimilar triodes that could be used for the two different functions of the topology you have, voltage amp and phase splitter/driver. The 12DW7 is the only one that comes to mind (12AX7 & 12AU7 in a single bottle), if you just want to get things running with the existing sockets I would think this tube would be the place to start. The 12AT7 or 6SL7 driving the 6SN7 or 6CG7 would probably be a better option if you want to punch a hole in the chassis.
- Status
- Not open for further replies.