Ok, I am not really that good with tubes... I am out of school on summer break, and need a project. Summer assignments can wait for a nice audio gadget. 
I was thinking about doing a solid-state class A amp, but I just have to keep myself focused... I have always wanted tubes, so I must do that. I have lots of time now, so it should be do-able.
I have drawn up the attached schem for an OTL tube amp... I neevr find a design to my liking, so I draw up my own. Approximate parts values:
C1- 3.3uF
C2- 470uF, 25V
C3- 1000uF, 25V
C4- 470uF, 400V
C5- 1uF
P1- 100k audio/log
R1- 1M, 1/2W
R2- 1k, 1/4W
R3- 15k, 1W
R4- 1k, 1W
R5- 100, 1/2W
R6- 4.7k, 1W
R7- 470, 1W
R8- 47k, 1/2W
R9- 22k, 1/2W
V1- 6SN7
V2- 6SL7
What do you think? I would prefer an octal-base tube to a 9-pin one, if you think I should change drivers, etc. Will it efficiently drive Sennheiser HD580's? They are 300-ohm cans. Thanks for any help you can provide...
I was thinking about doing a solid-state class A amp, but I just have to keep myself focused... I have always wanted tubes, so I must do that. I have lots of time now, so it should be do-able. I have drawn up the attached schem for an OTL tube amp... I neevr find a design to my liking, so I draw up my own. Approximate parts values:
C1- 3.3uF
C2- 470uF, 25V
C3- 1000uF, 25V
C4- 470uF, 400V
C5- 1uF
P1- 100k audio/log
R1- 1M, 1/2W
R2- 1k, 1/4W
R3- 15k, 1W
R4- 1k, 1W
R5- 100, 1/2W
R6- 4.7k, 1W
R7- 470, 1W
R8- 47k, 1/2W
R9- 22k, 1/2W
V1- 6SN7
V2- 6SL7
What do you think? I would prefer an octal-base tube to a 9-pin one, if you think I should change drivers, etc. Will it efficiently drive Sennheiser HD580's? They are 300-ohm cans. Thanks for any help you can provide...
Attachments
Get rid of C1. Why do you need R1 and R2, get rid of them and adjust P1 for desired impedance, although I think just the pot would do fine. Also, I think you should get rid of R5. I am also guessing that the second half of the 6SL7 is choke loaded? R9 not necessary, you headphones provide low impedance as it is. That's all I could find as I went along, please correct me if I am wrong on something.
Agreed, you don't need C1. A 6SN7 probably doesn't need R2, the grid-stopper. Technically, R1 isn't needed because its provided by the wiper of the volume control to ground, but wiper contacts can be intermittent, causing noise, so I normally leave it in. You definitely need R9, otherwise you will blow up your headphones the first time you plug them in as C4 and C5 charge through the coils of your headphones.
300 Ohms is a low impedance load for a valve, and although you are right to use a cathode follower, it needs to pass substantial current. Staying with Octal, you could use half of a 6BL7 or even 6080 for the cathode follower. They should be biased to pass 20mA, or more. Headphones don't need many volts to drive them, so V2 isn't really needed, and V3 could be directly driven from the anode of V1. Driving V2 directly from the anode of V1 would not have worked because V2's grid would have had half the HT on it, so it's cathode resistor would have needed to be very large to achieve the correct current, and there would have been hardly any volts left for the valve to operate. V1 has rather a small value of anode load, which will increase its distortion - you might want to consider increasing this value. Don't be disheartened by these, we all have to make our mistakes...
300 Ohms is a low impedance load for a valve, and although you are right to use a cathode follower, it needs to pass substantial current. Staying with Octal, you could use half of a 6BL7 or even 6080 for the cathode follower. They should be biased to pass 20mA, or more. Headphones don't need many volts to drive them, so V2 isn't really needed, and V3 could be directly driven from the anode of V1. Driving V2 directly from the anode of V1 would not have worked because V2's grid would have had half the HT on it, so it's cathode resistor would have needed to be very large to achieve the correct current, and there would have been hardly any volts left for the valve to operate. V1 has rather a small value of anode load, which will increase its distortion - you might want to consider increasing this value. Don't be disheartened by these, we all have to make our mistakes...
Way more gain than you'll ever need. Parallel both halves of the 6SN7 as a cathode follower and enjoy.
And yes, loose all the resistors and the cap in the front end. The volume pot's total value sets the load impedance (I'd use 500k), and acts as the grid resistor. Functions very well that way.
(I've built this exact circuit... feel free to search on the 10,000 posts regarding it - "complete 6SN7 preamp", etc)
Joel
And yes, loose all the resistors and the cap in the front end. The volume pot's total value sets the load impedance (I'd use 500k), and acts as the grid resistor. Functions very well that way.
(I've built this exact circuit... feel free to search on the 10,000 posts regarding it - "complete 6SN7 preamp", etc)
Joel
GAIN?
Hi,
Good point obviously. In fact the circuit as shown has so much gain it's virtually useless...
If you rethink it for headphone use you best lose some gain and let some more current flow through the output.
You could easily use half a 6SN7 for the voltage amplification and a paralelled one CF for the output.
If it were for my own use I'd think along the lines of the dissimilar triodes where you can have the entire headphone amp from a single tube and it could easily double as a CDP preamp.
Cheers,
Hi,
Good point obviously. In fact the circuit as shown has so much gain it's virtually useless...
If you rethink it for headphone use you best lose some gain and let some more current flow through the output.
You could easily use half a 6SN7 for the voltage amplification and a paralelled one CF for the output.
If it were for my own use I'd think along the lines of the dissimilar triodes where you can have the entire headphone amp from a single tube and it could easily double as a CDP preamp.
Cheers,
Octals look cooler. hehe
(though they are actually QUITE warm...)
Anyway, if another that is non-octal would sound better, then by all means... so, it is possible to have just ONE tube amplify the whole thing? That would really cut down on parts cost... I am intrigued. Can anyone elaborate more? Thanks
(though they are actually QUITE warm...)
Anyway, if another that is non-octal would sound better, then by all means... so, it is possible to have just ONE tube amplify the whole thing? That would really cut down on parts cost... I am intrigued. Can anyone elaborate more? Thanks
You could use half of a 6SN7 driving half of a 6BX7/6BL7 as a cathode follower (or 6SL7 first stage for more gain). The other halves, of course are for the other channel.
Or use dissimilar dual triodes, one per channel. 6EM7/6EA7 or 6GL7 for high gain, 6DN7 if you don't need so much. The output half is a 7-10W triode for plenty of drive.
All are octal. 6EW7, 6FD7, 6DE7, 6DR7 are some 9 pin possibilities.
Or use dissimilar dual triodes, one per channel. 6EM7/6EA7 or 6GL7 for high gain, 6DN7 if you don't need so much. The output half is a 7-10W triode for plenty of drive.
All are octal. 6EW7, 6FD7, 6DE7, 6DR7 are some 9 pin possibilities.
I apologize for now sounding noob-ish... but I must ask in order to learn...
What exactly is a dissimilar triode? I understand the triode part, and how to use a triode (dual), but not the dissimilar part. What makes it different?
I have heard of 6EM7, but not any of the others. 7-10W? For headphones? 
How? Out of one triode? Golly...
How high a gain are we talking about with a 6EM7, etc.? And what about the 6DN7? My solid-state headphone amp has a gain of 10. I notice the amplification factor of a 6SL7 is 70- is this meaning gain=70? That is insanely large...
Another question I just thought of- people say the Sennheiser HD580's and 600's, along with some Grado headphones are 'dynamic.' What exactly is meant by this? There are some amp circuits that are also labeled 'dynamic' and are meant for these types of cans. What distinguishes this, and would a headphone amp like this be 'dynamic?'
Thanks all
What exactly is a dissimilar triode? I understand the triode part, and how to use a triode (dual), but not the dissimilar part. What makes it different?
I have heard of 6EM7, but not any of the others. 7-10W?
For headphones? How? Out of one triode? Golly...How high a gain are we talking about with a 6EM7, etc.? And what about the 6DN7? My solid-state headphone amp has a gain of 10. I notice the amplification factor of a 6SL7 is 70- is this meaning gain=70? That is insanely large...
Another question I just thought of- people say the Sennheiser HD580's and 600's, along with some Grado headphones are 'dynamic.' What exactly is meant by this? There are some amp circuits that are also labeled 'dynamic' and are meant for these types of cans. What distinguishes this, and would a headphone amp like this be 'dynamic?'
Thanks all
Many dual tubes have two identical sections, like 6SN7, 6SL7, 6BX7. The 6EM7 has a high gain triode (like half of a 6SL7) and a low gain, high power one (like half of a 6BX7). The latter will make a nice low impedance cathode follower. You CAN'T get 7-10W out of it - that's the input rating, and efficiency is quite low.
6DN7 is more like half a 6SN7, half a 6BX7. 6EW7 is similar in 9 pin.
Dynamic is short for electrodynamic, I suppose - magnetic as opposed to piezoelectric (crystal).
6DN7 is more like half a 6SN7, half a 6BX7. 6EW7 is similar in 9 pin.
Dynamic is short for electrodynamic, I suppose - magnetic as opposed to piezoelectric (crystal).
Yeah you can do away with all that gain, and a 'SL won't do enough current for real good operation...get that 6SN7 on the cans.
Lesse... let's say you want 10mW (far more than your ears could ever handle I'm sure), at 300 ohms that's 5.8mA RMS, or 8.2mA peak. So for class A, it needs at least that much bias... maybe 10mA at 200V. Below that you could use a 10k 2W resistor for the cathode load, and use 300V for +V.
Grid bias voltage will be -5V, which can be set with cathode bias or DC coupling to a preamp stage biased at 100V plate..
Tim
Lesse... let's say you want 10mW (far more than your ears could ever handle I'm sure), at 300 ohms that's 5.8mA RMS, or 8.2mA peak. So for class A, it needs at least that much bias... maybe 10mA at 200V. Below that you could use a 10k 2W resistor for the cathode load, and use 300V for +V.
Grid bias voltage will be -5V, which can be set with cathode bias or DC coupling to a preamp stage biased at 100V plate..
Tim
So, you think a 6DN7 would be good enough to drive 300-600-ohm cans to sufficient levels? With headphones, I am not a loud listener, I mainly care about sound quality with my cans, which is why I bought these particular ones to begin with. My current amp is decent, but I feel I am missing something... somethign I may find in tubes, or at very least, a Class A solid-state amp. I may throw together an SS amp out of spare junk and see what goes.... but I am torn. I NEED a sub amp, but tubes are driving me much harder than getting my sub up and running... and I don't know if I want a tube power amp or headphone amp. Odds are, if I used 6DN7 (NOS $8 a piece from The Tube Store) I could get a tube headphone amp up and running for under $100 EASILY. Right now, that is looking best...
I am finding minimal info on this 6DN7 also- TDSL has something about it, but I am a tad rusty on some of the abbreviations, and others I just don't know... It seems like everyone recommends 6SN7 over 6SL7, so I would have to go with 6DN7 over the others to accomodate for this. So, when creating the schem, I should just treat it as a 6SN7 and 6BX7? And this will provide adequate gain and SPL for my cans? Hmm.... I shall go to work on this... Thanks in advance for answering any of the questions in there (there are lots, I know... just trying to grasp this whole concept) Thanks
I am finding minimal info on this 6DN7 also- TDSL has something about it, but I am a tad rusty on some of the abbreviations, and others I just don't know... It seems like everyone recommends 6SN7 over 6SL7, so I would have to go with 6DN7 over the others to accomodate for this. So, when creating the schem, I should just treat it as a 6SN7 and 6BX7? And this will provide adequate gain and SPL for my cans? Hmm.... I shall go to work on this... Thanks in advance for answering any of the questions in there (there are lots, I know... just trying to grasp this whole concept) Thanks
When I look at the Mu's and Ra's of these tubes... it almost seems like the 6EM7 is better suited... Sure, the first stage has a mu of 64, but the second is only 5 and has an Ra of 750, as opposed to the 2000 of the 6DN7. I am guessing Ra is output impedence? Am I remotely correct? Most of this is new to me... If I am thinking corectly there, then 6EM7 is looking better...
You can download a full data sheet for 6DN7 from the usual site:
http://home.planet.nl/~frank.philipse/frank/frank.html
The medium-mu section isn't quite a 6SN7 type (it's mu is slightly higher and anode dissipation is only 1W), but it's quite close. You can use the medium-mu section as a gain stage, DC couple it to the low-mu section (passing 20mA, or more, if you can afford it), and capacitor couple to your cans. You could put some global feedback around the whole thing if you wanted to pull the output resistance further down, but I expect I'll be howled at for that suggestion!
6EW7 is even nicer, and it's rather pretty because it's the diameter of an Octal on a B9A base. I've measured them, and the Sylvania ones are very good.
http://home.planet.nl/~frank.philipse/frank/frank.html
The medium-mu section isn't quite a 6SN7 type (it's mu is slightly higher and anode dissipation is only 1W), but it's quite close. You can use the medium-mu section as a gain stage, DC couple it to the low-mu section (passing 20mA, or more, if you can afford it), and capacitor couple to your cans. You could put some global feedback around the whole thing if you wanted to pull the output resistance further down, but I expect I'll be howled at for that suggestion!
6EW7 is even nicer, and it's rather pretty because it's the diameter of an Octal on a B9A base. I've measured them, and the Sylvania ones are very good.
So, as a bare bones schem, we are looking at something like the attached, right? Just a standard cathode follower... I notice the output impedence of the 6DN7 is way too high, and that of the 6EM7 is higher than needed (750-ohm), but that could easily be fixed with NFB... as bad as that may seem to some... as long as it sounds "tubey" then it will be fine with me... I haven't really heard a good tube sound playing hi-fi stuff.
Quote: 6EW7 is even nicer, and it's rather pretty because it's the diameter of an Octal on a B9A base. I've measured them, and the Sylvania ones are very good.
Hmm... what exactly do they look like? I will search for this- it could be pretty... I also like aesthetically pleasing tube designs.
Thanks all for the help
Quote: 6EW7 is even nicer, and it's rather pretty because it's the diameter of an Octal on a B9A base. I've measured them, and the Sylvania ones are very good.
Hmm... what exactly do they look like? I will search for this- it could be pretty... I also like aesthetically pleasing tube designs.
Thanks all for the help
Attachments
Your topology looks good.
Yes, Zo is typically around the tube's Rp. Don't worry about Zo or NFB, since a cathode follower has inherent NFB in it, and ought to bring the Zo below that of the cans.
I'll try running some numbers, humm.
For the 2nd half of the 6DN7:
(Feel free to doubt this information and recalculate it yourself, as doubt and independent and repeatable experiments are what hard science is based on )
The figure of 3.9% THD (actually that's 2nd harmonic only) is at maximum signal, 140mW (probably more than enough to smoke your headphones, and ears!). At more sane levels such as 1mW, distortion will be slim to nil.
I'll leave the preamp section to someone else; post results please!
Tim
Yes, Zo is typically around the tube's Rp. Don't worry about Zo or NFB, since a cathode follower has inherent NFB in it, and ought to bring the Zo below that of the cans.
I'll try running some numbers, humm.
For the 2nd half of the 6DN7:
An externally hosted image should be here but it was not working when we last tested it.
(Feel free to doubt this information and recalculate it yourself, as doubt and independent and repeatable experiments are what hard science is based on
)The figure of 3.9% THD (actually that's 2nd harmonic only) is at maximum signal, 140mW (probably more than enough to smoke your headphones, and ears!). At more sane levels such as 1mW, distortion will be slim to nil.
I'll leave the preamp section to someone else; post results please!
Tim
TDSL says that the Ra of the second half of a 6DN7 is 2000. Isn't this really too high? Can just the topology reduce the Zo that much?
Another question... should I have a cathode bypass cap on the preamp half of the tube?
About what plate voltages would be best? 250V and 150? Something like that?
Thanks, Tim, for the graph and thanks to all for the help.
Another question... should I have a cathode bypass cap on the preamp half of the tube?
About what plate voltages would be best? 250V and 150? Something like that?
Thanks, Tim, for the graph and thanks to all for the help.
Well it'll reduce it by the same amount as it does THD, which is 5.1 times (whoops, that should show 2.2% THD then). That brings Zo to 390 ohms, pretty close (damping factor of .77). I say, try it and see if you like it. If you don't you can try a different tube (rewiring the circuit as necessary!) with lower Rp (typical for the big triode on tubes like these is 800 ohms).
Tim
Tim
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.