tube schem
Ok, I have this much (the drawing) and these sets of values:
Set1
P1- 250k
R1-275k
R2- 1k
R3- 680
R4- 100k
R5- 100k
R6- 330-680 1W
R7- 10k
R8- 1k
C1- 50uF 25V
C2- 50uF, 250V
C3- 50uF, 250V
C4- 10-20uF, 250+V
add coupling cap between 6L6+6SL7, after R5
Set1
P1- ? (100k maybe?)
R1- 1M
R2- 10k
R3- about 1k
R4- 200k
R5- not used (actually harms amp by shorting 6SL7?)
R6- 3k-5k 10W
R7- ? don't know
R8- 100
C1- ?
C2- not used
C3- ?
C4- 220-470uF 250+V
Of these sets of values, which are most likely to work best? Should I mix them? How?
Power Supply
C1- 560uF, 250V
C2- 470uF, 250V
C3- 270uF, 250V
D1-D4- 1N4007
L1- 2H, 100mA, 175-ohm, 300V (Hammond 154M)
R1- not figured yet
T1- 40VA, 125-0-125@100mA, 6.3V@2A heater (Hammond 269AX)
B+ is 150VDC.
What do you think about these? Is the PS adequate? Which values should I use for the amp? Headphones are 300-ohm.
Thanks
Ok, I have this much (the drawing) and these sets of values:
Set1
P1- 250k
R1-275k
R2- 1k
R3- 680
R4- 100k
R5- 100k
R6- 330-680 1W
R7- 10k
R8- 1k
C1- 50uF 25V
C2- 50uF, 250V
C3- 50uF, 250V
C4- 10-20uF, 250+V
add coupling cap between 6L6+6SL7, after R5
Set1
P1- ? (100k maybe?)
R1- 1M
R2- 10k
R3- about 1k
R4- 200k
R5- not used (actually harms amp by shorting 6SL7?)
R6- 3k-5k 10W
R7- ? don't know
R8- 100
C1- ?
C2- not used
C3- ?
C4- 220-470uF 250+V
Of these sets of values, which are most likely to work best? Should I mix them? How?
Power Supply
An externally hosted image should be here but it was not working when we last tested it.
C1- 560uF, 250V
C2- 470uF, 250V
C3- 270uF, 250V
D1-D4- 1N4007
L1- 2H, 100mA, 175-ohm, 300V (Hammond 154M)
R1- not figured yet
T1- 40VA, 125-0-125@100mA, 6.3V@2A heater (Hammond 269AX)
B+ is 150VDC.
What do you think about these? Is the PS adequate? Which values should I use for the amp? Headphones are 300-ohm.
Thanks
Headphone Amp.
Hi,
If going for the big output tubes, why not use a 6AS7G or its relatives?
It will give you plenty of power, low Zo and it is a real triode, to boot.
Moreover there are plenty of those around at about 10 $ a piece thanks to the military.
Drive it with decent current like 1/2 of a ECC99 and you're in business ...
Cheers,
Hi,
If going for the big output tubes, why not use a 6AS7G or its relatives?
It will give you plenty of power, low Zo and it is a real triode, to boot.
Moreover there are plenty of those around at about 10 $ a piece thanks to the military.
Drive it with decent current like 1/2 of a ECC99 and you're in business ...
Cheers,
Oh, how quickly we forget
He has a whack of 6L6s to burn...
Here's a better schematic, plus my interpretation of it.
The values in the linked schematic look okay, except for that resistor to ground. It will indeed load the preamp tube, DC as well as AC; a coupling cap would help a lot.
Also, the lack of a coupling cap will put a lot of DC on the 6L6's grid, biasing it much higher than 0Vg would put it with a 680 ohm cathode resistor. My schematic allows a choice; take the first shown value for direct coupled, second for cap coupled.
Note that I haven't graphed the operating points, and the 5k 10W is a total SWAG.
Tim
He has a whack of 6L6s to burn...
Here's a better schematic, plus my interpretation of it.
The values in the linked schematic look okay, except for that resistor to ground. It will indeed load the preamp tube, DC as well as AC; a coupling cap would help a lot.
Also, the lack of a coupling cap will put a lot of DC on the 6L6's grid, biasing it much higher than 0Vg would put it with a 680 ohm cathode resistor. My schematic allows a choice; take the first shown value for direct coupled, second for cap coupled.
Note that I haven't graphed the operating points, and the 5k 10W is a total SWAG.
Tim
Attachments
Tresspasser,
Nice and simple, but you may want to make a voltage divider with about 100 volts and attach it to one side of the heater circuit, otherwise you may get some unwanted hum due to some rectification effect between the heater and the cathode.
fdegrove,
Curious. Though I have never tried it nor read anything about it... but why would it make a difference what tube you use for a cathode follower as long as the current capability is enough?
I would think that they would all sound the same.
Gabe
Nice and simple, but you may want to make a voltage divider with about 100 volts and attach it to one side of the heater circuit, otherwise you may get some unwanted hum due to some rectification effect between the heater and the cathode.
fdegrove,
Curious. Though I have never tried it nor read anything about it... but why would it make a difference what tube you use for a cathode follower as long as the current capability is enough?
I would think that they would all sound the same.
Gabe
Re: Headphone Amp.
P.S. Something I read at tubecad the other day - If you use a tetrode or pentode as a follower and *don't* want it to be a psuedo-triode, (provided the scrren is fed with a resistor!) run a bypass cap from the screen grid to cathode so the screen to cathode voltage remains constant even though the cathode is jumping up and down.
The anode voltage versus current curves of a triode and a beam tetrode are way different of course, but in this case we have the load connected to the cathode. With a pentode, beam tetrode and to a lesser extent a triode, the anode appears as a current sink and the cathode as a voltage source to whatever you are driving. Imagine the situation where you hold the grid at a constant voltage and gradually pull the cathode down further and further. Plot the anode current increase as you do this and there you have the difference between a "real" triode and a beam tetrode in this instance. That's what really matters here, not the difference in the plate curves. Not *entirely* sure but I imagine the main difference would be the steepness of the transconductance slope, i.e. how many milliamps increase per volt of cathode-grid voltage change. It could be that a 6L6 might be better than a triode as a cathode follower.fdegrove said:
P.S. Something I read at tubecad the other day - If you use a tetrode or pentode as a follower and *don't* want it to be a psuedo-triode, (provided the scrren is fed with a resistor!) run a bypass cap from the screen grid to cathode so the screen to cathode voltage remains constant even though the cathode is jumping up and down.
Trespasser,
I was just thinking about the same type of circuit a few weeks back! How funny.
But, I was planning on using a choke as the cathode load for the 6L6 cathode follower. The key would be to use one that has a DC voltage drop across it equal to your necessary bias point.
If you use resistors for the load with so much current flow, you get too much heat in the chassis for my tastes.
I was just thinking about the same type of circuit a few weeks back! How funny.
But, I was planning on using a choke as the cathode load for the 6L6 cathode follower. The key would be to use one that has a DC voltage drop across it equal to your necessary bias point.
If you use resistors for the load with so much current flow, you get too much heat in the chassis for my tastes.
SAME, SAME BUT DIFFERENT.
Hi,
Why wouldn't you hear differences between tubes whether they're in AF or CF mode?
After all it's not because the signal is taken off the anode or cathode that all of a sudden they're all going to sound the same.
Then again not all AFs are nor sound the same and surely this applies to CFs too...
Cheers,
Hi,
Why wouldn't you hear differences between tubes whether they're in AF or CF mode?
After all it's not because the signal is taken off the anode or cathode that all of a sudden they're all going to sound the same.
Then again not all AFs are nor sound the same and surely this applies to CFs too...
Cheers,
fdegrove,
Well... Just for the sake of argument, cathode followers generally will have a gain of something like 0.9, give or take a few millivolts. What have others called it? 100% NFB? Whereas at the anode, not connected as a unity gain follower (since I didn't mention that) will have some positive gain. Depending on the tube, some more gain than others. The characteristics of the tube then will be more or less influential, depending on the gain, on the sound.
But with 100% or more NFB, there should be no influence at all... I guess. Hence why I wonder why a triode will sound different, not necessarily better, than a tetrode or pentode.
Still curious.
Gabe
Well... Just for the sake of argument, cathode followers generally will have a gain of something like 0.9, give or take a few millivolts. What have others called it? 100% NFB? Whereas at the anode, not connected as a unity gain follower (since I didn't mention that) will have some positive gain. Depending on the tube, some more gain than others. The characteristics of the tube then will be more or less influential, depending on the gain, on the sound.
But with 100% or more NFB, there should be no influence at all... I guess. Hence why I wonder why a triode will sound different, not necessarily better, than a tetrode or pentode.
Still curious.
Gabe
CF.
Hi,
CF's are widely misunderstood, one good reason to opt for a triode is lower noise and better linearity.
A triode is more linear than a penthode for the triode has inherent feedback built in.
If you wish, penthodes can make for better CFs, in the case o the headphone amp I'd chose a triode such as the 6AS7, it has great current drive, low distortion, low Zo and is very linear as a CF.
Also, when talking CF there are half a dozen toplogies for those out there...
Cheers,
Hi,
CF's are widely misunderstood, one good reason to opt for a triode is lower noise and better linearity.
A triode is more linear than a penthode for the triode has inherent feedback built in.
If you wish, penthodes can make for better CFs, in the case o the headphone amp I'd chose a triode such as the 6AS7, it has great current drive, low distortion, low Zo and is very linear as a CF.
Also, when talking CF there are half a dozen toplogies for those out there...
Cheers,
Re: CF.
Hi Frank,
AFAIK, noise only matters 1. in high-sensitivity circuits, and also 2. at high frequencies. And noise is determined more by Gm.
Sigh. The internal NFB, though it does reduce distortion and Zo (manifested as Rp), is nonlinear, and as such a pentode has lower distortion than a similar triode when sufficient NFB (to equalize gain) is applied externally.
I agree that the 6AS7 is a great tube, but we don't know if he has any, nor do we know if he wants to buy any.
There are? When I hear CF, I instantly think "plate to +V, grid input, cathode output". not much simpler than that, and any other CF, White for example, has the qualifier to specify what it is ("White").
And remember. The first post in this thread says he has the topology, is going to use 6L6s for output, and is only asking for R/C values and if the power supply is sufficient.
Speaking of which -- I suppose someone should answer his questions!!
I'll have to graph the tube operating points, but otherwise:
The 269AX will do a good 330V 60mA as shown. (I'm listening to my 6L6GC SE right now (now playing: AC/DC - Highway To Hell) which uses it.) Technically, that's over the 125-0-125V 100mADC published figure (which translates to 250VCT 50mADC, meaning I'm 40% over the limit), but it still passes the 10 second finger test with aplomb.
Your filtering is very much more than adequate, though I'd like to see no more than 100uF for the first cap at the rectifier - reduce charging spikes, make the transformer run a little cooler.
The final RC is unnecessary IMHO, unless you need to drop a bit of voltage. (There's no reason the design can't use 330V.)
I'll get back to you on the values...
Tim
Hi Frank,
fdegrove said:
AFAIK, noise only matters 1. in high-sensitivity circuits, and also 2. at high frequencies. And noise is determined more by Gm.
Sigh. The internal NFB, though it does reduce distortion and Zo (manifested as Rp), is nonlinear, and as such a pentode has lower distortion than a similar triode when sufficient NFB (to equalize gain) is applied externally.
I agree that the 6AS7 is a great tube, but we don't know if he has any, nor do we know if he wants to buy any.
There are? When I hear CF, I instantly think "plate to +V, grid input, cathode output". not much simpler than that, and any other CF, White for example, has the qualifier to specify what it is ("White").
And remember. The first post in this thread says he has the topology, is going to use 6L6s for output, and is only asking for R/C values and if the power supply is sufficient.
Speaking of which -- I suppose someone should answer his questions!!
I'll have to graph the tube operating points, but otherwise:
The 269AX will do a good 330V 60mA as shown. (I'm listening to my 6L6GC SE right now (now playing: AC/DC - Highway To Hell) which uses it.) Technically, that's over the 125-0-125V 100mADC published figure (which translates to 250VCT 50mADC, meaning I'm 40% over the limit), but it still passes the 10 second finger test with aplomb.
Your filtering is very much more than adequate, though I'd like to see no more than 100uF for the first cap at the rectifier - reduce charging spikes, make the transformer run a little cooler.
The final RC is unnecessary IMHO, unless you need to drop a bit of voltage. (There's no reason the design can't use 330V.)
I'll get back to you on the values...
Tim
Tim, fdegrove, sch3matic, EC8010,
I thank you for all your input to my queries, and I am sure the original poster thanks us as well. I am sure he learned much from this.
Trespasser,
As for values...
For the power supply: I would think that with the bridge, the output voltage is 250 RMS, not 150, at 100 mA, not 50 as someone else said. With the 100µF cap right after the bridge, the output voltage will be more like 353, but realistically, with load, 320 volts. Choke... you don't need much more than 1.5-3 Henries. The cap after... use as large as you can get. 100-300µF. Otherwise you will hear the hum. The RC after I presume will power the preamp/driver tube.
Don't use a choke load for the cathode, as it will linearly boost bass and reduce highs. Chokes have a linear roll off. You want a linear flat response, not reduction in highs. So get a nice juicy 25 or higher watt resistor of about 330 ohms to idle at a reasonable current. This way you could opt to remove the grid resistor and cap and attach it directly to the plate of the first tube.. or not. You want about 15 volts at the cathode at idle with the capacitor and grid resistor.
If you opt for the direct couple approach (which may introduce hum), then the grid's being positive doesn't matter, because the cathode's voltage will be a certain voltage different than the grid to make up for the difference. As long as the grid voltage is negative with respect the cathode voltage it will be fine.
As for what to do with the screen grid... it doesn't seem to matter. The output is taken from the cathode. Screen grid functionality only affects the output at the plate (I know I will get scolded here). You may get a little more current gain (where the Gm, or transconductance comes in) but not much. It may only determine the resistor needed for the same voltage.
As for noise... again, it is only a factor where there is gain. We have none here, unless you call a gain of .8 or .9 "gain".
Stick with what you have there... except get rid of R5.
As for C4... for a headphones of about 20 to 30 ohms, you want at least a value of 270 µF! Why??? unless you don't care for any music below 240 hertz, you ain't gettin no bass from the recommended value. (F=1/{2PiXcC}) So go for the isolated output tube, putting a capacitor and grid resistor at the grid of the 6L6. Then calculate for that 15 volts idle at the cathode so you can apend $1.25 for a 220 or even a 330 µF cap at 35-50 volts.
Also, for power supply and other caps... 350 or higher volts. Remember, 125-0-125 translates to 250 center tap... which center tap you are not using. With a bridge you get 250 volts RMS, 353 volts peak with no load (250X1.414), more like 320 with load.
FWIW,
Gabe
I thank you for all your input to my queries, and I am sure the original poster thanks us as well. I am sure he learned much from this.
Trespasser,
As for values...
For the power supply: I would think that with the bridge, the output voltage is 250 RMS, not 150, at 100 mA, not 50 as someone else said. With the 100µF cap right after the bridge, the output voltage will be more like 353, but realistically, with load, 320 volts. Choke... you don't need much more than 1.5-3 Henries. The cap after... use as large as you can get. 100-300µF. Otherwise you will hear the hum. The RC after I presume will power the preamp/driver tube.
Don't use a choke load for the cathode, as it will linearly boost bass and reduce highs. Chokes have a linear roll off. You want a linear flat response, not reduction in highs. So get a nice juicy 25 or higher watt resistor of about 330 ohms to idle at a reasonable current. This way you could opt to remove the grid resistor and cap and attach it directly to the plate of the first tube.. or not. You want about 15 volts at the cathode at idle with the capacitor and grid resistor.
If you opt for the direct couple approach (which may introduce hum), then the grid's being positive doesn't matter, because the cathode's voltage will be a certain voltage different than the grid to make up for the difference. As long as the grid voltage is negative with respect the cathode voltage it will be fine.
As for what to do with the screen grid... it doesn't seem to matter. The output is taken from the cathode. Screen grid functionality only affects the output at the plate (I know I will get scolded here). You may get a little more current gain (where the Gm, or transconductance comes in) but not much. It may only determine the resistor needed for the same voltage.
As for noise... again, it is only a factor where there is gain. We have none here, unless you call a gain of .8 or .9 "gain".
Stick with what you have there... except get rid of R5.
As for C4... for a headphones of about 20 to 30 ohms, you want at least a value of 270 µF! Why??? unless you don't care for any music below 240 hertz, you ain't gettin no bass from the recommended value. (F=1/{2PiXcC}) So go for the isolated output tube, putting a capacitor and grid resistor at the grid of the 6L6. Then calculate for that 15 volts idle at the cathode so you can apend $1.25 for a 220 or even a 330 µF cap at 35-50 volts.
Also, for power supply and other caps... 350 or higher volts. Remember, 125-0-125 translates to 250 center tap... which center tap you are not using. With a bridge you get 250 volts RMS, 353 volts peak with no load (250X1.414), more like 320 with load.
FWIW,
Gabe
Gabevee said:
Gabe, you are incorrect on this point. Use the formula 2*Pi*F*L and tell me how you arrive at the conclusion that bass will be boosted and highs will be ruduced?
The gain of the CF can only approach 1, which it will very soon when using any choke of reasonable size inductance. This increased gain will only increase with frequency, not decrease. And it can only ever increase up to the number one. If the "rolloff" you are referring to is from capacitance, this occurrs at frequencies many many times greater than audio.
As far as biasing, the DC drop across the choke (or resistor) in the cathode is what sets the op point.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.