|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
173.189.217.200
In Reply to: RE: now this might have promise...... posted by Caucasian Blackplate on July 02, 2015 at 20:33:32
My lack of theory knowledge is showing again I'm afraid. I thought gain was a good thing, didn't know you could have too much. I will also have to read up on why a high output impedance is not desirable.
I have always thought the less parts between the music and the speaker the better and have always tried to stay with as a low parts count as possible. Apparently that is not always the case, huh?
Follow Ups:
Although I can read schems and I understand the very basic concepts of theory, I have long ago decided that my mind is not able to wrap itself around most audio design concepts, and certainly not the finer points. I am not stupid, I just don't have any aptitude for electronic theory.
I have to depend on others to say "I have tried this and it sounds great", and here is the schem. That works best for me. I now have a couple to try using both the 27 and the 26, so that ought to keep me out of trouble and off the streets for a while.
Edits: 07/03/15 07/04/15
With a 7k ohms output impedance it will not take very much cable capacitance (shunt to ground capacitance) to cause an audible loss of high frequencies.
Normally preamps are designed to have 1k ohms or less output impedance.
A preamp has to be able to deal with not only the cable capacitance but also the Miller capacitance of the first tube in the power amp. Those two capacitance's add. As an example, if there is 150pF of cable capacitance and 50pF of Miller capacitance the preamp has to be able to "drive" 200pF.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Gain is not necessarily a "problem", but the output impedance of this circuit could be. It would be roughly equal to the parallel sum of the plate resistance of the Type 27 (~9K ohms) and the 22K plate resistor. The 470K resistor is also in parallel but won't have that much effect because of its much larger value. So, you're looking at something around 7K output Z. Ideally, you'd like around a 10X higher input Z for the amplifier you would be driving, something > 70K. But here I also wonder about the 0.1uF coupling capacitor, seems too low for good low frequency extension, if indeed the input Z of the amplifier is on the order of 100K or any lower.
I haven't thought about these issues for a long time. I'm sure I will be corrected if this is not entirely accurate info.
The 27 will have a high output impedance if used with a plate resistor. For a linestage I would either use a cathode follower or, much better, a CCS for the 27 plate load with a mu-follower output.
The 27 isn't a great choice for a mu follower. If you look at the datasheets, there is no insulation rating for the cathode to filament interface, and in my experience this insulation is only good for 10-20 volts.
I was talking about a single 27 with a cascode CCS on the plate, but with the output coupling cap connected not to the plate directly but rather to the "mu output" of the CCS. In other words, connected to the source terminal of the "lower" transistor in the cascode.
Or would it be defeating the sonic advantages of the Type 27? I ask as someone who has not tried anything like what you suggest. I would sooner use a CCS and take the output off the plate, but that would still beg for a buffer stage.
Lew, I have never done a fair A/B comparison. I have used the mu connection when I needed a low output Z, and it sounded very good. I am confident it sounds better than adding a cathode follower. But in a situation where the output Z is fine even with a plate connection, I don't know whether the mu output would sound better or not as good.
I have had really good results using a CF with a CCS on the cathode side. Better yet with Allen Wright's "Super Linear Cathode Follower". In my amplifiers, I went through a progression where I listened to a plain CF as a driver stage for several years, then added the CCS for a few years, then upgraded to SLCF. The speakers and upstream gear did not change, essentially, so I feel very confident that the improvements were due solely to these modifications. Difference between no CCS and CCS was greater than between CCS and SLCF. I guess if I did a mu-follower, I would use a dual section triode rather than transistors.
Interesting. The only time I tried a CF as a driver was my first 300B SE amp where I started with a 6SL7 direct coupled to a 6BL7 CF. It sounded pretty decent even with resistor loading on the cathode. Unfortunately I never tried a CCS on the CF driver. And since it was the first input/driver I tried for the 300B, I felt obliged to go on and try a lot of others and never went back to the first version.
On preamps with a CF the CCS made a major improvement compared to resistor loading.
Amen to that.
Lew,
My co worker, and good audio friend, used to get rid of cathode follower stages, from Audio Research SP3 A-1s, etc, and thought, into a tube amp, with short cables, the deleted stage sounded markedly superior !!
There are commercial products he told me about, it MAY have a C-J preamp series, where the only difference in the later model was an addition of a cathode follower stage, and most everyone prefers the early "simple " edition of the preamp, he tells me. We are referencing to the used market, and posted comments, after several decades of both versions' existence.
I can get the model numbers from him, if you wanna know.
Jeff Medwin
I and Salectric are both agreeing with you, that a "plain Jane" CF, at least all the ones I've heard, does impart an unpleasant coloration that is most evident when you "fix" it, at least a little bit, with a CCS. You can fix it even more, IMO, by adding a constant voltage source on the anode side of the CF, while keeping the CCS. That's the "SLCF" described by Allen Wright. If the interface requires lowering the output Z, a CF with a CCS or an SLCF is pretty darn transparent. However, I would not argue that avoiding the need for a CF is not better than any CF or variant thereof, when possible. Often that is one advantage of an "all-in-one" preamplifier, with phono stage on the same chassis as the linestage. In that case, one can usually dispense with any CF or other output buffer on the phono output.
Well posted !!! Xln't. Thanks Lew.
Jeff Medwin
The thing about using the mu output.....
It puts the power supply back into the signal path.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: