Take a look at this amp : http://www.geocities.com/mydesign_99/PP6L6.JPG
- Does it realy going to draw 200 mA over the power transformer ? Cause I would like to use a edcor transformer for 2 channels of this amp... and I think a 500 mA is a biiiit overpower.
- Does it realy going to draw 200 mA over the power transformer ? Cause I would like to use a edcor transformer for 2 channels of this amp... and I think a 500 mA is a biiiit overpower.
Hi quitvinny ,
Except for the input stage , the output schematic PP 6L6 Class A
is a classic , and it REALLY DRAWS about 170 - 200 mA at the full
power . and for two channels , you will really need about 400 -
450 mA of useful current .
The total output power is 18 Wats of pure Class A per channel .
But the sound is very amazing , very smooth and I'm sure that you will like it .
Regards
Carlos
Except for the input stage , the output schematic PP 6L6 Class A
is a classic , and it REALLY DRAWS about 170 - 200 mA at the full
power . and for two channels , you will really need about 400 -
450 mA of useful current .
The total output power is 18 Wats of pure Class A per channel .
But the sound is very amazing , very smooth and I'm sure that you will like it .
Regards
Carlos
For the output tranny, I check on http://tdsl.duncanamps.com/show.php?des=6L6
Zout is the impedance right ? do I look in the good column : Class : A PP Va : 270 .....Zout 5000 ??
Zout is the impedance right ? do I look in the good column : Class : A PP Va : 270 .....Zout 5000 ??
Hi quitvinny ,
As I said , this 6L6 PP Class A schematic is a classic and the correct Zout , that is the total load ( plate to plate ) is 5000 ohms , and the +B correct value is
270 – 300 Volts . Use a output tranny that can “hand” 25 Watts or more .
Preferably use a matched pair or a same brand / same lot pair of 6L6’s in each channel , because the quiescent current is high ( about 170 mA total or aprox. 85 mA per tube ) and if the current through one tube is very diferent from the current through another tube , this fact will cause a lot of distortion at the output.
( don’t forget : the output tubes are cathode coupled by the 125 ohms resistor )
Don’t try to change the cathode resistor , the correct value is on the schematic
( 125 ohms ) , and this value will give an aprox. grid bias value of ( - ) 20 Volts .
Don’t worry about the high value of quiescent current , because the tubes are
connected in the push-pull way , and the magnetic field created by one tube is
entirely canceled by the opposite magnetic field created by another tube .
The circuit is highly stable , and you won’t have any kind of trouble .
Good luck ,
Carlos
As I said , this 6L6 PP Class A schematic is a classic and the correct Zout , that is the total load ( plate to plate ) is 5000 ohms , and the +B correct value is
270 – 300 Volts . Use a output tranny that can “hand” 25 Watts or more .
Preferably use a matched pair or a same brand / same lot pair of 6L6’s in each channel , because the quiescent current is high ( about 170 mA total or aprox. 85 mA per tube ) and if the current through one tube is very diferent from the current through another tube , this fact will cause a lot of distortion at the output.
( don’t forget : the output tubes are cathode coupled by the 125 ohms resistor )
Don’t try to change the cathode resistor , the correct value is on the schematic
( 125 ohms ) , and this value will give an aprox. grid bias value of ( - ) 20 Volts .
Don’t worry about the high value of quiescent current , because the tubes are
connected in the push-pull way , and the magnetic field created by one tube is
entirely canceled by the opposite magnetic field created by another tube .
The circuit is highly stable , and you won’t have any kind of trouble .
Good luck ,
Carlos
Hi mod_evil
No you can’t use the 6SN7 in the place of 12AT7 in anyway , they are totally
different . The 6SN7 is a medium mu and the 12AT7 is a high mu dual triode .
The 6SN7 Class A grid bias is ( - ) 8 Volts and the 12AT7 Class A grid bias is
( - ) 2 Volts at 250 Volts plate voltage , the plate resistance and transconductan-
ce are different too .
If you follow the above gave schematics , will result a very amazing and a very
smooth sound .
Carlos
No you can’t use the 6SN7 in the place of 12AT7 in anyway , they are totally
different . The 6SN7 is a medium mu and the 12AT7 is a high mu dual triode .
The 6SN7 Class A grid bias is ( - ) 8 Volts and the 12AT7 Class A grid bias is
( - ) 2 Volts at 250 Volts plate voltage , the plate resistance and transconductan-
ce are different too .
If you follow the above gave schematics , will result a very amazing and a very
smooth sound .
Carlos
The reason the 12AT7 was chosen for the first stage, I suspect, was that it can draw enough current to drive the LM317 and also has reasonably high gain. A 12AX7 has more gain but would draw too little current. 12AU7 (or 6SN7) would be the opposite.
If I were you, I would not use the LM317 as a CCS - it's reputed to have a poor high frequency response. Choose a better CCS, maybe using a couple of transistors of appropriate type. If you do that, it would also allow you to use a higher gain but lower current tube for the first stage, such as a 12AX7 or, better still for sound quality, a 6SL7.
If you want to make it sound even nicer, use a choke in place of the 100 ohm resistor (R12) in the power supply, especially if you're planning a common supply for both channels.
If I were you, I would not use the LM317 as a CCS - it's reputed to have a poor high frequency response. Choose a better CCS, maybe using a couple of transistors of appropriate type. If you do that, it would also allow you to use a higher gain but lower current tube for the first stage, such as a 12AX7 or, better still for sound quality, a 6SL7.
If you want to make it sound even nicer, use a choke in place of the 100 ohm resistor (R12) in the power supply, especially if you're planning a common supply for both channels.
Hey, thank you all,
For CCS, where can I find good desing information ? or if you have shematic lying around send it to me.
And if I do that, can ou tell me ray_moth what I will have to change to use 6SL7 ?
Last question : I saw silver-steel bar on ebay (dia : 1 mm) does anyone have ever work with this material ? I would like to use it as a bus bar.
For CCS, where can I find good desing information ? or if you have shematic lying around send it to me.
And if I do that, can ou tell me ray_moth what I will have to change to use 6SL7 ?
Last question : I saw silver-steel bar on ebay (dia : 1 mm) does anyone have ever work with this material ? I would like to use it as a bus bar.
Hi quitvinny ,
No problems , you can use 6L6 ( metal tube ) , 6L6 GC , 6L6 GB , etc .
without modification .
Now I have 2 suggestions to do for you .
I strongly recommend that you try one OR other and listen !!!
Suggestion # 1
- Ignore the screen grids taps from the output tranny
- Connect to each screen grid terminal , a 220 ohms x 1 watt resistor
- The other side of the two resistors , will be connected direct to +B
- Now you have a normal pentode operation of the out tubes
- Listen !!!
Suggestion # 2
- You must to assemble the unit exactly like the schematic
- Use an output tranny with screen grids taps at 33 % in turns ratio , or 10% in impedance ( that means 250 ohms to each side from center tap , the +B tap )
- After the assembly completed , try to change the screen grids resistors , starting from 100 ohms , changing to 220 ohms , 470 ohms , 680 ohms until 1000 ohms
- Now you have a ultralinear operation with different screen grids loads .
- Listen !!!!
Both ways of operation are correct , and the decision belongs to you !!!
Regards
CARLOS
No problems , you can use 6L6 ( metal tube ) , 6L6 GC , 6L6 GB , etc .
without modification .
Now I have 2 suggestions to do for you .
I strongly recommend that you try one OR other and listen !!!
Suggestion # 1
- Ignore the screen grids taps from the output tranny
- Connect to each screen grid terminal , a 220 ohms x 1 watt resistor
- The other side of the two resistors , will be connected direct to +B
- Now you have a normal pentode operation of the out tubes
- Listen !!!
Suggestion # 2
- You must to assemble the unit exactly like the schematic
- Use an output tranny with screen grids taps at 33 % in turns ratio , or 10% in impedance ( that means 250 ohms to each side from center tap , the +B tap )
- After the assembly completed , try to change the screen grids resistors , starting from 100 ohms , changing to 220 ohms , 470 ohms , 680 ohms until 1000 ohms
- Now you have a ultralinear operation with different screen grids loads .
- Listen !!!!
Both ways of operation are correct , and the decision belongs to you !!!
Regards
CARLOS
Hi guitvinny,
What Carlos is talking about are screen stoppers, not screen loads. They are there to prevent parasitics, which means they must be connected to the screen pin with a very short lead. Those resistors also serve to provide some voltage drop, so that the screen operates at a lower voltage than the plate.
Sorry not to have answered your question about this before. I think a good transistor CCS can be found in Gingertube's "Baby Huey" schematic, which he has posted to this forum in another thread. The link is here. The CCS is on the second page.
If you decide to use a 6SL7 as the long tail pair (first stage), you should adjust the CCS current to, say, 1.5mA and use 120k plate load resistors for each half of the 6SL7. Perhaps Gingertube might like to comment?
What Carlos is talking about are screen stoppers, not screen loads. They are there to prevent parasitics, which means they must be connected to the screen pin with a very short lead. Those resistors also serve to provide some voltage drop, so that the screen operates at a lower voltage than the plate.
Sorry not to have answered your question about this before. I think a good transistor CCS can be found in Gingertube's "Baby Huey" schematic, which he has posted to this forum in another thread. The link is here. The CCS is on the second page.
If you decide to use a 6SL7 as the long tail pair (first stage), you should adjust the CCS current to, say, 1.5mA and use 120k plate load resistors for each half of the 6SL7. Perhaps Gingertube might like to comment?
Hi quitvinny ,
In the first place I want to say to ray_moth , that I’m sorry but I DISAGREE entirely from you .
The different resistors connected to the scrren grids , ARE NOT grid stoppers they are REALLY SCREEN GRIDS LOADS .
When David Hafler and Herbert Keroes , published their articles in the 50’s , about the Ultra-Linear’s concepts , what they really wanted was to create a “virtual tube” that behaves in the middle point between a triode and a pentode .
Inside this concept , the screen grids have a fundamental importance because they are the responsible element or electrode to create a kind of “reactive feedback” or “negative feedback” at all , THEN , they are connected in the output tranny , at convenient taps .
The output stage has an amount of current circulating across the screen grids resistors ( because screen grids are positive (+) electrodes ) , and that current will intefere in the tube’s behavior , changing the tube’s operation point , always at a point between triode and pentode , as said above .
So , this way , since we have current circulating across the screen grids resistors , we will have a voltage dropping . Change the resistor and you will change the voltage dropping across it , changing the operation point.
The “stopper” or “grid stopper resistor” , is used in the Control Grid Electrode to prevent parasitcs oscillations ( as ray_moth said ) , but the current don’t circulate across that resistor , because in the Class A , and in the Class AB1 , the grid’s control voltage , is always a negative (-) value.
The resistors R 9 and R 10 are the actual grid stoppers in your amplifier schematics .
Quitvinny , did you understand the explanation ??? I think that I have answered , your question . If not , feel free to ask me again !!
About the CCS , ray_moth gave to you a good tip , and as I said , the decision belongs to you .
Regards for all ,
Carlos
In the first place I want to say to ray_moth , that I’m sorry but I DISAGREE entirely from you .
The different resistors connected to the scrren grids , ARE NOT grid stoppers they are REALLY SCREEN GRIDS LOADS .
When David Hafler and Herbert Keroes , published their articles in the 50’s , about the Ultra-Linear’s concepts , what they really wanted was to create a “virtual tube” that behaves in the middle point between a triode and a pentode .
Inside this concept , the screen grids have a fundamental importance because they are the responsible element or electrode to create a kind of “reactive feedback” or “negative feedback” at all , THEN , they are connected in the output tranny , at convenient taps .
The output stage has an amount of current circulating across the screen grids resistors ( because screen grids are positive (+) electrodes ) , and that current will intefere in the tube’s behavior , changing the tube’s operation point , always at a point between triode and pentode , as said above .
So , this way , since we have current circulating across the screen grids resistors , we will have a voltage dropping . Change the resistor and you will change the voltage dropping across it , changing the operation point.
The “stopper” or “grid stopper resistor” , is used in the Control Grid Electrode to prevent parasitcs oscillations ( as ray_moth said ) , but the current don’t circulate across that resistor , because in the Class A , and in the Class AB1 , the grid’s control voltage , is always a negative (-) value.
The resistors R 9 and R 10 are the actual grid stoppers in your amplifier schematics .
Quitvinny , did you understand the explanation ??? I think that I have answered , your question . If not , feel free to ask me again !!
About the CCS , ray_moth gave to you a good tip , and as I said , the decision belongs to you .
Regards for all ,
Carlos
Hi there, Quitvinny,
Now you have a debate on your hands! Firstly, regarding the screen serie resistors, both Ray_Moth and Carlos are correct. Since there is Miller capacitance between G1 and G2, and leads have inductance, any resistor in either of these electrodes will tend to quench rf oscillation resulting from circuit parameters; they thus are (also) stoppers.
But as to the advantage of large enough screen resistors, the jury is still out - meaning this will depend on the type of tube. With regard to changing the operating point - respectfully Carlos, I doubt whether that will make a significant difference. E.g. for the EL34, Ig2 is approximately 10mA which through a 1K resistor will cause a voltage drop of only 10V - to my feeling a little academic. The primary effect in UL would be one of changing the feedback factor. (If the effect of the operating point is a factor I would rather suggest a change in G1 bias.)
Distortion curves for the EL34 shows a definite advantage in the use of a (rather large) 1K screen resistor, but with beam tubes like 6L6 with their almost 4:1 change in Ig2 over the operating region it is a different scenario. This will reflect as a large variation in r.g2 (the "internal resistance" reference to g2), and it can be argued whether it is not better to drive such a non-linear load directly from the relatively low impedance (about 180 ohm) of the output transformer UL tap. Intuitively I would feel so, unless someone has measured data for us. (I have never seen such for beam tubes, except that most circuits use a very low Rg2; also GEC, who manufactured the KT66 - they are unlikely to distribute non-optimal circuits for their product.)
Further to this, Quitvinny, it is your amp and your freedom to use pentodes instead of UL, but I cannot see the advantage in not using a type of operation that has pentode output and efficiency but near-triode performance regarding distortion and internal impedance - unless you want your amp to become a musical instrument by itself. You may like what it adds, but it will definitely change your signal.
Although this is becoming lengthy, kindly allow a few further comments;
1. I agree with Ray_Moth: The LM317 is a no-no for an audio CCS, it definitely does not have the required frequency response, which is easy to get with discrete circuits. This may not be serious in the very high impedance CCS for a LTP, but why take the risk?
2. Don't be over-impressed by stuff like a silver-steel bar for a bus-bar. Its conductivity is only 5% better than copper (but its temperature co-efficient is slightly worse, if you want to be really pedantic!) Spend your money where it will make a practical difference.
3. If you decide to use fixed bias from the already existing source (a worthwhile suggestion; the voltage should be about right) add an extra stage of R.C filtering.
Hope some of this was informative.
Regards.
Now you have a debate on your hands! Firstly, regarding the screen serie resistors, both Ray_Moth and Carlos are correct. Since there is Miller capacitance between G1 and G2, and leads have inductance, any resistor in either of these electrodes will tend to quench rf oscillation resulting from circuit parameters; they thus are (also) stoppers.
But as to the advantage of large enough screen resistors, the jury is still out - meaning this will depend on the type of tube. With regard to changing the operating point - respectfully Carlos, I doubt whether that will make a significant difference. E.g. for the EL34, Ig2 is approximately 10mA which through a 1K resistor will cause a voltage drop of only 10V - to my feeling a little academic. The primary effect in UL would be one of changing the feedback factor. (If the effect of the operating point is a factor I would rather suggest a change in G1 bias.)
Distortion curves for the EL34 shows a definite advantage in the use of a (rather large) 1K screen resistor, but with beam tubes like 6L6 with their almost 4:1 change in Ig2 over the operating region it is a different scenario. This will reflect as a large variation in r.g2 (the "internal resistance" reference to g2), and it can be argued whether it is not better to drive such a non-linear load directly from the relatively low impedance (about 180 ohm) of the output transformer UL tap. Intuitively I would feel so, unless someone has measured data for us. (I have never seen such for beam tubes, except that most circuits use a very low Rg2; also GEC, who manufactured the KT66 - they are unlikely to distribute non-optimal circuits for their product.)
Further to this, Quitvinny, it is your amp and your freedom to use pentodes instead of UL, but I cannot see the advantage in not using a type of operation that has pentode output and efficiency but near-triode performance regarding distortion and internal impedance - unless you want your amp to become a musical instrument by itself. You may like what it adds, but it will definitely change your signal.
Although this is becoming lengthy, kindly allow a few further comments;
1. I agree with Ray_Moth: The LM317 is a no-no for an audio CCS, it definitely does not have the required frequency response, which is easy to get with discrete circuits. This may not be serious in the very high impedance CCS for a LTP, but why take the risk?
2. Don't be over-impressed by stuff like a silver-steel bar for a bus-bar. Its conductivity is only 5% better than copper (but its temperature co-efficient is slightly worse, if you want to be really pedantic!) Spend your money where it will make a practical difference.
3. If you decide to use fixed bias from the already existing source (a worthwhile suggestion; the voltage should be about right) add an extra stage of R.C filtering.
Hope some of this was informative.
Regards.
No need to be sorry. However, what I said is what I have always understood to be the truth.
Also, there are some UL designs that don't use any resistor between the screens and the primary taps. Examples are: Dynaco variation of Mullard amp, Acrosound KT88, Allen 60w organ amp, Dynaco modified Williamson, Eico HF87.
Hi quitvinny ,
Hi Potqieter , good to “see” you on the debate , but the only goal to be reached , is to help quitvinny to make a very good amplifier .
All of your commentary is wellcome , but with your permission I’ll
disagree from your suggestion # 3 , I strongly recommend that
quitvinny don’t use fixed bias in this circuit , as I said previously
the best solution for this circuit is the use of bias by a 125 ohms
cathode resistor , like the schematics .
About my suggestion to quitvinny , to experiment the pentode
operation wich has like base , my own experience .
When I assembled this circuit ( two years ago ) , and even being the # 1 fanatic of U.L. operation , I have to recognize that pentode
operation ( IN THIS CASE ) , sounds better . Don’t ask me why !!
Regards for all ,
Carlos
Hi Potqieter , good to “see” you on the debate , but the only goal to be reached , is to help quitvinny to make a very good amplifier .
All of your commentary is wellcome , but with your permission I’ll
disagree from your suggestion # 3 , I strongly recommend that
quitvinny don’t use fixed bias in this circuit , as I said previously
the best solution for this circuit is the use of bias by a 125 ohms
cathode resistor , like the schematics .
About my suggestion to quitvinny , to experiment the pentode
operation wich has like base , my own experience .
When I assembled this circuit ( two years ago ) , and even being the # 1 fanatic of U.L. operation , I have to recognize that pentode
operation ( IN THIS CASE ) , sounds better . Don’t ask me why !!
Regards for all ,
Carlos
Yes, I did digress a bit. There are always periferals in amplifier designs.
Regarding fixed bias you would have noticed an "if" in my reply. I was merely trying to point out a common fault with fixed bias, i.e. improper filtering. Argument for is that fixed bias will give a higher max. output with resulting lower distortion than cathode bias at intermediate outputs; but advantage of cathode bias is safety (self-regulation). And if a large enough bypass capacitor is used there will be almost fixed bias conditions for short enough signal bursts - a method I mostly use.
Which brings me to another point raised earlier: The power transformer rating.
Quitvinny, one must remember that in music the peak to average power ratio can be 2 - 3. That means that you do not need a power transformer that can continuously supply the maximum current. Your power transformer will still run cool at only 50 - 60% of maximum demand, and if the filter capacitors are large the slightly poorer regulation will not be detected - this unless you disco the amplifier flat-out, mostly into overload!
Carlos, one possible explanation for the better bass perceived with a pentode output stage, is that because the internal resistance is higher the damping is less and with certain loudspeaker designs this will cause a rise in response round the bass resonance.
Regards.
Regarding fixed bias you would have noticed an "if" in my reply. I was merely trying to point out a common fault with fixed bias, i.e. improper filtering. Argument for is that fixed bias will give a higher max. output with resulting lower distortion than cathode bias at intermediate outputs; but advantage of cathode bias is safety (self-regulation). And if a large enough bypass capacitor is used there will be almost fixed bias conditions for short enough signal bursts - a method I mostly use.
Which brings me to another point raised earlier: The power transformer rating.
Quitvinny, one must remember that in music the peak to average power ratio can be 2 - 3. That means that you do not need a power transformer that can continuously supply the maximum current. Your power transformer will still run cool at only 50 - 60% of maximum demand, and if the filter capacitors are large the slightly poorer regulation will not be detected - this unless you disco the amplifier flat-out, mostly into overload!
Carlos, one possible explanation for the better bass perceived with a pentode output stage, is that because the internal resistance is higher the damping is less and with certain loudspeaker designs this will cause a rise in response round the bass resonance.
Regards.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.