Can I use triodes in place of Q13/14?
http://www.headwize.com/projects/showfile.php?file=gilmore4_prj.htm (hybrid electrostatic headphones amp)
Since that's doing most of the voltage gain, seems to me a good place for triodes.
http://www.headwize.com/projects/showfile.php?file=gilmore4_prj.htm (hybrid electrostatic headphones amp)
Since that's doing most of the voltage gain, seems to me a good place for triodes.
Well that could be raised (I could pull the cathode lower; I'm still winding the transformer so getting other voltages is not a problem). Or low voltage tube used.
You can see below what I was thinking.
Can someone suggest low voltage and power tube I can use instead of the 6DJ8 to get higher gain, while it's still nicely linear? I don't need much current capability since it's driving a MOSFET follower anyway.
Can someone suggest low voltage and power tube I can use instead of the 6DJ8 to get higher gain, while it's still nicely linear? I don't need much current capability since it's driving a MOSFET follower anyway.
An externally hosted image should be here but it was not working when we last tested it.
That is one helluva schematic for a tube amp! One reason you're not getting replies could be sheer intimidation. Truthfully, I wouldn't know where to start commenting on your design.
In my experience, tube designers start off with tubes only, then maybe (but not usually) put in some SS devices where they can really do some good, e.g. for constant current sources/sinks, B+ delay circuits or rectifiers. Tube designs tend to be very simple. That's my 2 cents.
In my experience, tube designers start off with tubes only, then maybe (but not usually) put in some SS devices where they can really do some good, e.g. for constant current sources/sinks, B+ delay circuits or rectifiers. Tube designs tend to be very simple. That's my 2 cents.
Umm, not my design. There's a reason I posted a link in my first post. It's an electrostatic headphone amplifier. And the only part that really matters here is the third stage...
And I had a very specific question:
And I had a very specific question:
High gain, good linearity, and low plate voltage don't often happen at the same time 🙁
ECC86/6GM8 may work better here than 6DJ8 - it is specifically a low-voltage design - but don't write off the 6DJ8. With a CCS load they can do quite well at low plate voltages, much better than you would guess looking at the curves in the datasheet.
I have a circuit running on the bench now at ~30V plate voltage (48V B+), 2mA CCS plate load for two parallel sections, giving 0.08% THD at ~3V RMS out, ~10V RMS max out at 5% THD.
The same circuit gives more output swing with ECC86, but an Amperex 6DJ8 has lower distortion. A modern JJ 6DJ8 is not as good at this low plate voltage but still works OK.
Beware fairly large variation in characteristics tube-to-tube at this unusually low operating point. Adjustable bias and/or paralleling multiple sections is a good idea.
(I'm working on a simple hybrid amp design).
Pete
ECC86/6GM8 may work better here than 6DJ8 - it is specifically a low-voltage design - but don't write off the 6DJ8. With a CCS load they can do quite well at low plate voltages, much better than you would guess looking at the curves in the datasheet.
I have a circuit running on the bench now at ~30V plate voltage (48V B+), 2mA CCS plate load for two parallel sections, giving 0.08% THD at ~3V RMS out, ~10V RMS max out at 5% THD.
The same circuit gives more output swing with ECC86, but an Amperex 6DJ8 has lower distortion. A modern JJ 6DJ8 is not as good at this low plate voltage but still works OK.
Beware fairly large variation in characteristics tube-to-tube at this unusually low operating point. Adjustable bias and/or paralleling multiple sections is a good idea.
(I'm working on a simple hybrid amp design).
Pete
Er, that tube only uses a 30 V supply. I'm running the 6DJ8 at 130 V in the simulation and not getting enough gain. When I said low voltage, I meant on the order of a hundred volts.
A plate voltage of 130V is not low for a signal triode. In fact, 130V is the maximum permissable plate voltage for the 6DJ8.
So, what exactly are you loooking for? What gain? What plate voltage and current? What output impedance?
There are only a few thousand possibilities... you need to analyze the design and figure out what the requirements are.
Pete
So, what exactly are you loooking for? What gain? What plate voltage and current? What output impedance?
There are only a few thousand possibilities... you need to analyze the design and figure out what the requirements are.
Pete
It needs more gain and swing than a 6DJ8. Load is a constant current source, between 5 and 15 mA.
Regarding the original schematic, I want to use high quality resistors in the most critical parts of the amp. I'm planning to get Vishay bulk metal foils, but due to their price I want to use them only in the most important positions.
So which would these be? My guess is the feedback resistors (which I have to split due to 300 V limitation of the resistors, whereas that position sees 375 V), the 50K resistors in the second stage, and (maybe) the input stage source degeneration resistors.
For the rest I'll probably use Roederstein resistors.
What do you guys think? Did I miss any other important ones?
So which would these be? My guess is the feedback resistors (which I have to split due to 300 V limitation of the resistors, whereas that position sees 375 V), the 50K resistors in the second stage, and (maybe) the input stage source degeneration resistors.
For the rest I'll probably use Roederstein resistors.
What do you guys think? Did I miss any other important ones?
"Can someone suggest low voltage and power tube I can use instead of the 6DJ8 to get higher gain"
To operate at low voltage you will likely need to use a high gm, frame grid design tube, like the 6DJ8, but higher Mu. The EC8010 might fit the requirements. Mu is 55 to 60, gm is 28000. Expensive. The Telefunken data sheet has curves.
Don
To operate at low voltage you will likely need to use a high gm, frame grid design tube, like the 6DJ8, but higher Mu. The EC8010 might fit the requirements. Mu is 55 to 60, gm is 28000. Expensive. The Telefunken data sheet has curves.
Don
I gave up on using triodes in that part actually. At least in simulation I saw very little difference. Just trying to figure out my resistor question now. The bulk metal foils are $8 a piece so I want to minimize where I use them.
- Status
- Not open for further replies.