Here's hoping you gents can give me a few pointers. I'm currently looking for a line stage with quite a bit of gain. I am restricted to using the 6DJ8/ECC88 tube and have a 180-volt HT available to me. I currently run the 6DJ8/ECC88 in SRPP with the lower kathode resistor bypassed with a capacitor which yields about ~28x of gain. However the mu-follower looks to have better gain and better distortion figures on paper, but information on how to calculate it is scarce (or I'm looking in all the wrong places).
Any tips, suggestions or examples are most welcome!
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
Any tips, suggestions or examples are most welcome!
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
Hi Sander,
it took a while to ger that old post (where fdegrove posted such a schematic in reply to a question of mine) but here it is: http://www.diyaudio.com/forums/showthread.php?s=&threadid=31573&perpage=10&highlight=&pagenumber=5
Cheers
Andrea
it took a while to ger that old post (where fdegrove posted such a schematic in reply to a question of mine) but here it is: http://www.diyaudio.com/forums/showthread.php?s=&threadid=31573&perpage=10&highlight=&pagenumber=5
Cheers
Andrea
Mu follower calculations are treated in good detail in Morgan Jones's "Valve Amplifiers."
If you're already getting a gain of 28, it will be difficult to get it much higher- the mu is the ceiling, unless you're willing to cascode. With 180V, it's possible as long as you don't need gobs of swing.
If you're already getting a gain of 28, it will be difficult to get it much higher- the mu is the ceiling, unless you're willing to cascode. With 180V, it's possible as long as you don't need gobs of swing.
Thanks SY,
It is an output stage for a dac, input is ~0.06Vrms and output should be as close to 2Vrms as possible. Load impedance is 10K ~ 100K hence it is not required to drive a very low load, nor deliver high voltage swing. THD and gain however are important, THD should be as low as possible and gain should equal Mu.
Currently the schematic looks as follows, also showing the passive I/V conversion and 3rd order Bessel filter for the dac.
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
It is an output stage for a dac, input is ~0.06Vrms and output should be as close to 2Vrms as possible. Load impedance is 10K ~ 100K hence it is not required to drive a very low load, nor deliver high voltage swing. THD and gain however are important, THD should be as low as possible and gain should equal Mu.
Currently the schematic looks as follows, also showing the passive I/V conversion and 3rd order Bessel filter for the dac.
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
Also, I do own Morgan Jones' book, but he doesn't go into great detail about the Mu-follower, or at least not to an extent that allowed me to work it into a working Mu-follower with the 6DJ8/ECC88.
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
Do you have the Third Edition? It has about 5 or 6 pages detailing the mu follower, with a worked example using an E88CC (which is an exact equivalent of the ECC88).
Why are you restricted to ECC88? That gives you a severe gain limitation. Cascoding isn't an option unless you can add a buffer to the output.
Why are you restricted to ECC88? That gives you a severe gain limitation. Cascoding isn't an option unless you can add a buffer to the output.
Hi SY,
What pages exactly? I just skimmed the book again and the only examples I could find using the E88CC are on page 115 - 118, but I can't seem to rework those into anything that's usuable with my HT, they use a 285-volt HT, which is >100-volts more then I have available.
I'm not restricted to the ECC88 though, it is just the highest Mu tube I have available. And according to Morgan Jones' book it is a excellent choice for a low-distortion valve with good linearity.
I'm actually upgrading the output section of a commercial dac, so I have to work within the confines of the dac, it has two noval sockets on the PCB and, as mentioned, a 180-volt HT. Here's a closer look at the PCB with all tube components removed.
http://hardwareanalysis.com/images/articles/large/11817.jpg
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
What pages exactly? I just skimmed the book again and the only examples I could find using the E88CC are on page 115 - 118, but I can't seem to rework those into anything that's usuable with my HT, they use a 285-volt HT, which is >100-volts more then I have available.
I'm not restricted to the ECC88 though, it is just the highest Mu tube I have available. And according to Morgan Jones' book it is a excellent choice for a low-distortion valve with good linearity.
I'm actually upgrading the output section of a commercial dac, so I have to work within the confines of the dac, it has two noval sockets on the PCB and, as mentioned, a 180-volt HT. Here's a closer look at the PCB with all tube components removed.
http://hardwareanalysis.com/images/articles/large/11817.jpg
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
It might be worthwhile to follow those calculations, but use the voltage restrictions you have. For a 2V max signal, you could go down as low as 40V on the lower tube. Or for better linearity, the beta follower might be more suitable and waste less precious B+. If it were me, I'd lose the fancy topology and just run the tube as a conventional grounded cathode, LED bias, CCS load, and direct couple to a CCS-loaded cathode follower. That will get you a gain of 30 or so with excellent linearity.
For higher gain, consider an ECC81 run the way I described; that will be good for a gain of 45-50. With 120V on the plate, 2mA, the distortion at 2V will be miniscule.
For higher gain, consider an ECC81 run the way I described; that will be good for a gain of 45-50. With 120V on the plate, 2mA, the distortion at 2V will be miniscule.
Sander,
No need for the pi inductor input filter as audio tube preamplifiers are not so susceptible to rf. Most AUDIO tubes act as a low pass filter in itself.
If you don't believe me look as most famous preamplifiers, none of them have a inductor low pass input filter.
You need this filter with solid state preamplifier esp. when your power amplifier(close by) is a Class D type...
No need for the pi inductor input filter as audio tube preamplifiers are not so susceptible to rf. Most AUDIO tubes act as a low pass filter in itself.
If you don't believe me look as most famous preamplifiers, none of them have a inductor low pass input filter.
You need this filter with solid state preamplifier esp. when your power amplifier(close by) is a Class D type...
ttan98 said:
They're made to amplify RF so...
Maybe adding a grid stopper forming a lowpass with the Miller input capacitance, but it won't be a 3rd order Bessel no more.
SY: gnd k CCS loaded + gnd k CF CCS loaded... I like your way of designing things, it's very similar to mine: I thougt the same thing. Or... I have learnt from you
Don't forget using higher gain tubes (ECC83 or 5751) with some NFB to linearize things, as you can see I became pretty expert in NFB in linestages 
That is fancy topology!
Oh, whoa! The initial conditions were 180V or less. This will work OK at 150, and the cathode bias can be reduced a bit by going to a red LED for better linearity at low voltages, but why bother? There's no reason not to use a decent voltage to run a tube; isolation and industrial step-up/step-down transformers are so cheap and plentiful on the surplus market that getting a 150-180V B+ is not an economic issue. Heaters need to be biased up about 50-70V above ground; see Morgan Jones "Valve Amplifiers" for gory details on how to do that properly.
If I may be a bit heretical, if I wanted to run something at 30V supply, I'd use a transistor circuit.
edit: The 317 is not a particularly good current source for this kind of use. A cascode bipolar would be far better and just about as cheap. Again, see diyAudio's documentation and the Jones book for design details.
If I may be a bit heretical, if I wanted to run something at 30V supply, I'd use a transistor circuit.
edit: The 317 is not a particularly good current source for this kind of use. A cascode bipolar would be far better and just about as cheap. Again, see diyAudio's documentation and the Jones book for design details.
I immediately thought that this was a Dac60 when I read your first post and when I saw the output stage schematic and then the picture, it is indeed a Dac60. I have one of these too. I am going down the same road as you.
I like Stuart's concept best. I have sketched something similar (minus some minor details) and some things more complicated. But trying to make an easy rework with the space and sockets available, Stuart's is about as good as it gets, IMHO.
About the 3rd order Bessel Filter on the output of the 1704, is this ideal for this DAC? This is where I don't know much at all. I know you are a digital guru as I've seen your many posts on another forum. Would it be better to do the filtering between the gain and the buffer stages in Stuart's drawing?
I like Stuart's concept best. I have sketched something similar (minus some minor details) and some things more complicated. But trying to make an easy rework with the space and sockets available, Stuart's is about as good as it gets, IMHO.
About the 3rd order Bessel Filter on the output of the 1704, is this ideal for this DAC? This is where I don't know much at all. I know you are a digital guru as I've seen your many posts on another forum. Would it be better to do the filtering between the gain and the buffer stages in Stuart's drawing?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.