tom-vdl said:
Perhaps. Sowter offers such little information about their transformers I'm a bit hesitant to make any recommendations. However I think might opt for something like the 4935 under Pro Audio.
It's a 1:7 microphone/line transformer.
se
fdegrove said:
Aye, that he did. I was just musing over what could be considered a single active device and got off on a tube tangent.
se
Fuling said:I guess a choke loades source follower driven by a tube VAS might be a more reasonable approach to this whole thing...
It has been done before, with success.
Yeah, but the thing is, it has been done before. 🙂
I still think there's something to be said for getting your voltage gain passively. Keep the active devices as followers where they perform much better and use a step-up transformer for your voltage gain.
Since you're willing to consider doing the whole thing with more than one active device (depending on how you're wanting to consider what constitutes one device 🙂 ), how 'bout something like this:
Output of CDP into JT-13K7-A (or perhaps the Sowter I mentioned above) driving a simple JFET follower. The output of the JFET follower driving something like a 1k or 5k attenuator, which would drive a MOSFET source follower, either directly or, my preference, through a 1:1 input transformer such as the JT-11P-1.
With the input transformer, you could locate the MOSFET source follower right next to your speakers with the input transformer giving you excellent noise rejection.
se
By the way, the one caveat here is that in that scheme, the JT-13K7-A will always see the full output of your CDP. If your CDP has Red Book output voltage of 2 volts full scale, that's ok. If it's much higher than that, you might need to attenuate its output before it goes into the 13K7-A.
se
se
Uhm, those transformers ain´t particulary cheap, you know...
For me this was intended to be a "just for fun" project.
For me this was intended to be a "just for fun" project.
Ok I'm showing my ignorance here, but I have loads of BJT power transistors, (I don't like the sound of mosfets that much), and a few Sowter 4057 1K/5KCT input trannies. Anything I can bodge together with these?
Fuling said:Uhm, those transformers ain´t particulary cheap, you know...
For me this was intended to be a "just for fun" project.
The Jensen's certainly aren't. But that Sowter 4935 I mentioned previously is only 28.56 GBP which at the current exchange rate is 377.39 SEK or 40.90 EUR. The 1:1 at the output follower's input isn't any sort of requirement. Just my particular preference.
se
Well, 377SEK ain´t too much for a piece of iron.
We even have a Sowter distributor here in Sweden now.
Could be an alternative.
We even have a Sowter distributor here in Sweden now.
Could be an alternative.
pinkmouse said:
Not much with just a single BJT. The output impedance would be way too high. If you don't mind using a pair of BJTs, you could use a Darlington or Sziklai pair and that'd make a nice follower. I'm playing with a Sziklai pair and the 13K7-A with good results.
se
Oh, I´ve seen that one before!
Nice and simple, but I can´t understand how they keep control of the bias and the offset voltage?!?
Nice and simple, but I can´t understand how they keep control of the bias and the offset voltage?!?
Protection diodes are not as necessary in this design, because there is no stepup transformer. However, there is no gain, either. I like this new design, but it uses 2 devices, rather than one.
This new design has one really big problem, however. The effective input impedance, after the pot is only about 250 ohms! This loads the input pot bigtime, and is impractical as it will also load the source too much. The bias network needs to be rescaled at least 10 times, perhaps 100 times.
Just a joke, guys.
But I do tend to use reverse-biased diodes across PS outputs, just in case.
Why is there no gain? I'm something of a solid-state ignoramus, but this sure looks like a common source stage.
But I do tend to use reverse-biased diodes across PS outputs, just in case.
Why is there no gain? I'm something of a solid-state ignoramus, but this sure looks like a common source stage.
I take it all back! This is NOT a push pull output stage, it is single ended. It actually has a relatively high input impedance. Sorry, I 'saw' a connection where there was only a crossover. I will look first, in future, before opening up about someone elses schematic.
Hi,
Naughty,naughty....
Jean-Paul, I think SY was just kidding.
Not really, only one is handling the signal, the other is tracking current.
If gain is needed we can still add a xformer at the front and a bunch of diodes...just to make everyone happy, right?
Sure, nothing new. It's by Thierry Gibouin.
It basically tracks itself, offset should be almost nil.
Note, I am by no means knowledgeable with transistors so don't shoot me if I'm off base.
Cheers,😉
Naughty,naughty....
Jean-Paul, I think SY was just kidding.
Not really, only one is handling the signal, the other is tracking current.
If gain is needed we can still add a xformer at the front and a bunch of diodes...just to make everyone happy, right?
Sure, nothing new. It's by Thierry Gibouin.
It basically tracks itself, offset should be almost nil.
Note, I am by no means knowledgeable with transistors so don't shoot me if I'm off base.
Cheers,😉
This new design has one really big problem, however. The effective input impedance, after the pot is only about 250 ohms! This loads the input pot bigtime, and is impractical as it will also load the source too much. The bias network needs to be rescaled at least 10 times, perhaps 100 times.
- Status
- Not open for further replies.