final questions
ok... Should I:
A) Use my current 7500uf 200vdc caps (I have 4 of them... 2 per channel) ?
B) Should I include an inductor to make a CLC filter?
C) Should I scrap the inductor and make a Cap Multiplier?
D) Should I not use the 7500uf 200vdc caps for maybe the Panasonic TSHA line? Maybe smaller 400uf /200vdc caps at each tube?
E) Any other way???
Remember I will probably use 20db feedback, back to a EL84 input stage.. DC coupled to a choke loaded 6080 cap coupled to the 6AS7's with a grid choke..... or something like that.
What is the BEST way of doing it without going to Class A PP?
p.s. I actually have a large toroidal tranny with 2 115v windings on it... which I will use for input and output.. and there are a couple of other low voltage / high current windings that I could boost the secondary voltage if needed...
The toroid is about 6-7inches in diameter, about 3inches high and weighs a lot. should do for both channels.
Thanks!!!!!!!
ok... Should I:
A) Use my current 7500uf 200vdc caps (I have 4 of them... 2 per channel) ?
B) Should I include an inductor to make a CLC filter?
C) Should I scrap the inductor and make a Cap Multiplier?
D) Should I not use the 7500uf 200vdc caps for maybe the Panasonic TSHA line? Maybe smaller 400uf /200vdc caps at each tube?
E) Any other way???
Remember I will probably use 20db feedback, back to a EL84 input stage.. DC coupled to a choke loaded 6080 cap coupled to the 6AS7's with a grid choke..... or something like that.
What is the BEST way of doing it without going to Class A PP?
p.s. I actually have a large toroidal tranny with 2 115v windings on it... which I will use for input and output.. and there are a couple of other low voltage / high current windings that I could boost the secondary voltage if needed...
The toroid is about 6-7inches in diameter, about 3inches high and weighs a lot. should do for both channels.
Thanks!!!!!!!
Richard,
Some observations:
A capacitance multiplier loses more voltage than other smoothing methods (same as series pass reg). Inductors lose the least; use one if you have one of sufficient current capability.
There is danger in using the 2 115v windings. It is likely that the insulation between them is not up to the standards we expect between primary and secondary.
I still don't understand the topology. Are you saying this is SE OTL? If so is there an upper active device? Is it driven? How?
Am I being thick?
Cheers,
Some observations:
A capacitance multiplier loses more voltage than other smoothing methods (same as series pass reg). Inductors lose the least; use one if you have one of sufficient current capability.
There is danger in using the 2 115v windings. It is likely that the insulation between them is not up to the standards we expect between primary and secondary.
I still don't understand the topology. Are you saying this is SE OTL? If so is there an upper active device? Is it driven? How?
Am I being thick?
Cheers,
Hi Richard,
If it were me, I'd run the 6AS7's at a higher voltage, with a much lower current. That way most of your PSU issues simply go away. After all, what you're after is the 6AS7's low Rp and high transconductance - not necessarily amperes of plate current.
I would work with the 250V, 50mA, -125V bias operating point to start with. This is in a linear region.
Just a thought.
Joel
If it were me, I'd run the 6AS7's at a higher voltage, with a much lower current. That way most of your PSU issues simply go away. After all, what you're after is the 6AS7's low Rp and high transconductance - not necessarily amperes of plate current.
I would work with the 250V, 50mA, -125V bias operating point to start with. This is in a linear region.
Just a thought.
Joel
No free lunch
This is certainly no free lunch , with -125V bias , the problems are then moved to the driver stage which will have to provide 250V p-p (at least! , this does not include headroom
) . The driver would then require it's own power supply and the driver valve would have to be something quite special . At this 6AS7 operating point transconductance is way down etc , also these 6AS7/6080 things fall apart in terms of sectional matching as anode volts are increased . 
Personally I would keep away from multiple paralleled 6AS7 and go with fewer 6C33 instead 
Whatever , but I can see a project like this being a major pain in the neck !
316a
This is certainly no free lunch , with -125V bias , the problems are then moved to the driver stage which will have to provide 250V p-p (at least! , this does not include headroom
) . The driver would then require it's own power supply and the driver valve would have to be something quite special . At this 6AS7 operating point transconductance is way down etc , also these 6AS7/6080 things fall apart in terms of sectional matching as anode volts are increased . Personally I would keep away from multiple paralleled 6AS7 and go with fewer 6C33 instead Whatever , but I can see a project like this being a major pain in the neck !316a
316a,
You are far too pessimistic! I just built exactly such a driver and power stage, and you don't need a special driver or seperate PSU.
The driver is a run-of-the-mill 6SN7, with choke load (I use my old friend the hammond 156C again). Following the B+ for the power stage (which must be +375V since you will drop 125V for the cathode resistor) you will have +320V or more on the plate of the 6SN7 to work with. Plenty of swing, plenty of headroom - since when choke-loaded the tube is free to swing above and below the power supply.
For a power transformer, I used a hammond 300-0-300, with a 10uF->15Henry->40uF->5Henry->40uF filter. 5AR4 rectifier.
The amp is a parafeed output actually, but the basics are relevant to this discussion.
You are far too pessimistic! I just built exactly such a driver and power stage, and you don't need a special driver or seperate PSU.
The driver is a run-of-the-mill 6SN7, with choke load (I use my old friend the hammond 156C again). Following the B+ for the power stage (which must be +375V since you will drop 125V for the cathode resistor) you will have +320V or more on the plate of the 6SN7 to work with. Plenty of swing, plenty of headroom - since when choke-loaded the tube is free to swing above and below the power supply.
For a power transformer, I used a hammond 300-0-300, with a 10uF->15Henry->40uF->5Henry->40uF filter. 5AR4 rectifier.
The amp is a parafeed output actually, but the basics are relevant to this discussion.
6AS7/6080 matching
As far as the conception that 6080's are somehow wildly varying in performance between triodes, I just don't see it in real life.
Here are my measurements:
Plate Voltage
Target 375 V
Actual 364/364
Grid Bias Voltage
Target -125 V
Actual -130/-129.2
Plate Current
Target 50 mA
Actual ~52mA/~52mA
As you can see, they are all well within 5% of the target values.
*All measurements taken to ground.
As far as the conception that 6080's are somehow wildly varying in performance between triodes, I just don't see it in real life.
Here are my measurements:
Plate Voltage
Target 375 V
Actual 364/364
Grid Bias Voltage
Target -125 V
Actual -130/-129.2
Plate Current
Target 50 mA
Actual ~52mA/~52mA
As you can see, they are all well within 5% of the target values.
*All measurements taken to ground.
Driverstages
Now try measuring the 6080 with fixed bias ! Joel , you have a lot to learn about designing driver stages . Although the 6SN7 is a particular fave of mine , I would not expect 250V p-p swing from one in an undistorted fashion , I guess you used GTB's ? Remember headroom is also important , you may achieve the required (borderline) swing but at the expense of distortion . I would place a 6BL7 or 6BX7 in place of the 6SN7 in this particular application , or maybe even a 12B4a . We're both barking up the wrong tree regarding the thread . Richard was planning an OTL output stage .
316a
Now try measuring the 6080 with fixed bias ! Joel , you have a lot to learn about designing driver stages . Although the 6SN7 is a particular fave of mine , I would not expect 250V p-p swing from one in an undistorted fashion , I guess you used GTB's ? Remember headroom is also important , you may achieve the required (borderline) swing but at the expense of distortion . I would place a 6BL7 or 6BX7 in place of the 6SN7 in this particular application , or maybe even a 12B4a . We're both barking up the wrong tree regarding the thread . Richard was planning an OTL output stage .
316a
Re: Driverstages
Ahh, back to the grunty/meaty argument again, I see? Let's agree to disagree on that one.
Hmmm... why not? Seems quite easily done to me.
Actually, this is relevant if he chooses that operating point. The driver stage and bias options remain the same whether he goes OTL or not.
Joel
316a said:
Ahh, back to the grunty/meaty argument again, I see? Let's agree to disagree on that one.
Hmmm... why not? Seems quite easily done to me.
Actually, this is relevant if he chooses that operating point. The driver stage and bias options remain the same whether he goes OTL or not.
Joel
huh???????
Why would I want to have my 6AS7's at -120v bias in an OTL ???
Thats totally a waste of time.
Plus the fact remember that its a CF output stage... so the actual voltage gain is less than 1.
I need a bias around 20-30v.....
I am talking about a total of 2.5 amps in the output stage not milliamps.
How many 6AS7's would I need if I biased them at 50mA????????
anyway......
Why would I want to have my 6AS7's at -120v bias in an OTL ???
Thats totally a waste of time.
Plus the fact remember that its a CF output stage... so the actual voltage gain is less than 1.
I need a bias around 20-30v.....
I am talking about a total of 2.5 amps in the output stage not milliamps.
How many 6AS7's would I need if I biased them at 50mA????????
anyway......
keep it lean, not meaty...
As I said, we're just going to disagree on that issue.
As far as driving a 6080, this is not a subjective argument over "sonics"... a medium mu tube, such as a 6J5/6SN7 with at least 250V on its plate, and choke or transformer loaded is more than sufficient to swing the grid of a 6080 at that op point. Please see the RCA data sheets on the 6AS7G, where they explain all this.
Also, note that fixed bias is not recommended.
And of course one would need an additional stage of gain before the driver tube. Another 6J5, or a 6C5 would work good in that position.
regards,
Joel
As I said, we're just going to disagree on that issue.
As far as driving a 6080, this is not a subjective argument over "sonics"... a medium mu tube, such as a 6J5/6SN7 with at least 250V on its plate, and choke or transformer loaded is more than sufficient to swing the grid of a 6080 at that op point. Please see the RCA data sheets on the 6AS7G, where they explain all this.
Also, note that fixed bias is not recommended.
And of course one would need an additional stage of gain before the driver tube. Another 6J5, or a 6C5 would work good in that position.
regards,
Joel
JUST A Q....
Hi,
Operating points aside...is this SE OTL erequired to drive a real world speaker?
Reason I ask is that I don't see that happen with just a 6080 ...well, not for a full-range system anyway.
For a FR system you better target a Zo of below 1R if you value bass performance and that is going to take some large amount of NFB.
Just my 2 cents,
Hi,
Operating points aside...is this SE OTL erequired to drive a real world speaker?
Reason I ask is that I don't see that happen with just a 6080 ...well, not for a full-range system anyway.
For a FR system you better target a Zo of below 1R if you value bass performance and that is going to take some large amount of NFB.
Just my 2 cents,
Re: plate curves.
Richard,
Since this is a cathode follower circuit, and you are shunt feeding the load after a capacitor, the plate curves are not applicable. We want our gain as close to one as possible, and maximum power transfer, because a CF used as a power amp is going to be under 5% efficiency.
At any bias point, whether you use this one or not, you are going to need a large swing on the grids due to the degeneration losses.
And, max power tranfer is when the load=rp and Rk=sqrt. of 2rp
Actually, a 6080 can be run at 230V and 125mA quite safely. It's not as simple as multiplying Vp times Ip. Plate dissipation is equal to the DC plate dissipation minus the AC power output.
RichardJones said:
Richard,
Since this is a cathode follower circuit, and you are shunt feeding the load after a capacitor, the plate curves are not applicable. We want our gain as close to one as possible, and maximum power transfer, because a CF used as a power amp is going to be under 5% efficiency.
At any bias point, whether you use this one or not, you are going to need a large swing on the grids due to the degeneration losses.
And, max power tranfer is when the load=rp and Rk=sqrt. of 2rp
Actually, a 6080 can be run at 230V and 125mA quite safely. It's not as simple as multiplying Vp times Ip. Plate dissipation is equal to the DC plate dissipation minus the AC power output.
which is why
Which is why my input stage consists of a EL84 pentode 250v plate, 420v supply, 12K load resistor, 150v ? screen..
DC coupled to a EL34 running as a triode choke loaded, cap coupled to the grids of the 6080's with a grid choke.
Gain of input stage like 100, gain of triode stage around 11?, with the possibility of a 4v input to the EL84... I have plenty of gain to use up with feedback.
If this driver stage can deliver the voltage swing to the 6080's after the feedback... I don't see why I need to bias at -100v.
Granted the gain of the output stage is like 0.3 due to the low load of the 8ohm in parallel with the cathode resistor....
So what am I missing?
Which is why my input stage consists of a EL84 pentode 250v plate, 420v supply, 12K load resistor, 150v ? screen..
DC coupled to a EL34 running as a triode choke loaded, cap coupled to the grids of the 6080's with a grid choke.
Gain of input stage like 100, gain of triode stage around 11?, with the possibility of a 4v input to the EL84... I have plenty of gain to use up with feedback.
If this driver stage can deliver the voltage swing to the 6080's after the feedback... I don't see why I need to bias at -100v.
Granted the gain of the output stage is like 0.3 due to the low load of the 8ohm in parallel with the cathode resistor....
So what am I missing?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.