What's wrong with the kiss, boy?

Jump to Latest
#1,781
Baldrick.jpg
 
  • Like
Reactions: Ylab
#1,783
buffers are set to have 100K Rin, matter of resistor value shunting input gate to GND

as that, I wouldn't use more than 25K for volume

10K would be my choice, considering that today sources are mostly having decent low Rout

now, if you're going for higher value for vol pot, increase said input resistors to 220K or 270K

everything, for signal energy preservation
 
#1,786
peppennino said:
Thanks ZM. My phono preamp is a 834p with 12AX7 cathode follower at output. 25K would be enough?
Click to expand...

ZM's Thumb of rule - eyeball ( good enough) calculus for cathode follower Rout is 1/S, where you'll find S value in datasheets [ma/V] ....... right way should be deriving actual values from exact working point, but datasheet read is good enough in most cases

for 12AX7 (blasphemy using 12AX7 for that role but, hey, that' Tim de P. - he can do anything and I'm just going to green) S is 1.5mA/V so Rout is ~ 670R

in ZM's book - better to leave 50K volume and alter Rin of your Iron Pre to 270K

take care of not too long interconnect cables and that's it
 
#1,788
Hi ZM,

I'm still playing around with different buffer preamps, and I've bought some used Edcor XS4400 (which I got at a good price) to try in the spirit of this thread. I'm planning to try them as a switched gain boost, as in the Iron Pre, although with a different buffer, probably. (Maybe more than one.) The datasheet is here, if interested.

https://edcorusa.com/products/xs440...ching-transformer?_pos=1&_sid=2bf28928f&_ss=r

Another idea would be to wire it as an autoformer, and use it as an AVC, although one with very few steps. The idea would be to connect pins 3 and 4, pins 6 and 7, and pins 9 and 10, to create an autoformer with nine taps, connect the signal from the buffer to pin 1, ground to pin 12, and take the output signal from the taps, giving eight different volume levels, plus a mute at the last step. This is too few steps for a practical volume control, but it's enough to try the principle out at low expense.

Thinking about this I've had a crazier idea, a kind of hybrid AVC-pot volume control. If you connect a potentiometer across one of the windings of the autoformer (say pins 7 and 9) and take the output from the wiper (ground still from pin 12) then you would have a continuous volume variation between the tap at pin 9 and the tap at pin 7. If you use a two-pole switch to switch the pot between various positions, you would have a two-stage volume control, with discrete intervals selected by choosing the winding on the autoformer, and the pot providing finer adjustment.

I've attached a scan of a (badly drawn) schematic of this craziness. To keep things simple I've drawn it supposing we're only switching between four options of winding (although in principle I guess you could use more). My best guess is that (a) you'd want to use a pot of high resistance compared with the impedance of each winding, and probably (b) you'd want to send the output of this hybrid-AVC to a second buffer.

So what do you think? Too crazy to bother trying? Likely to work in theory, but not very well in practice?
 

Attachments

#1,789
everything what you said doable, with few remarks:

- if you connect XS4400 in AVC fashion (in place of output autoformer), its input is buffered and its output is having low Rout , so there is a catch with "a pot of high resistance compared with the impedance of each winding" use for "crazier idea"

-meaning - every Log pot is practically high impedance comparing to Rout you have on output of said AVC, simply because of fact that I can't remember when I even saw any Log pot available to buy, with value bellow 10K; 10K is practically high enough to not load output of AVC dramtically

- but - problem is - then you need to mount buffer on output of that additional VOl pot, so certainly better option - if you want to have "crazier idea" implemented - just keep regular Vol Pot of Iron Pre in regular position (its Rin high enough to not load sources significantly) while its output is buffered then signal is sent to your XS4400 connected as AVC

I didn't check your drawing, I suppose you did it as explained in your post
 
#1,790
Hi ZM,

Let me leave aside the issue of log or linear pots, and for now just suppose either will do.

If I can simplify things a little, you are suggesting to stick with

(A) Input pot -> buffer -> AVC

whereas my idea was

(B) Buffer -> hybrid AVC/pot -> buffer.

Trying (A) was on my mind anyhow, and I may do so, but let me persist a little with my crazier idea (B). Elsewhere you've written about "Signal energy preservation" perhaps explaining why (to your ears) autoformer volume controls sound better than resistive attenuation. The idea in my hybrid AVC/pot is that most of the attenuation is being done by the AVC so hopefully most of the "signal energy" would be preserved; the pot only provides adjustment of output from one of the windings. So (maybe) I'd get most of the benefits of an AVC without spending so much money just to try it out. The downside is having two-stage control of volume, however in practice this might work quite well, since I find I mostly change volume in a relatively short range on the volume control anyhow. Soldering up the extra two buffers is not such a big deal.

My first idea was to try

(C) Buffer -> AVC -> AVC,

the idea being to daisy-chain two 6- or 8-step AVCs to get 36 or 64 step-attentuation, needing only relatively cheap transformers and switches. Intact Audio has a two-switch model, which is what made me think about it, but I don't think theirs uses two autoformers. I couldn't see how to wire things, though, so I dropped it.
 
#1,791
feel free to try any possible variation, then choose what's best for you

it's always bunch of compromises and what of these is ending as best ...... we may have different impression about these

my reasoning ( arrangement "A") is simply thinking of terms of impedances and their relation, also taking care of not introducing another buffer in game ....... simpler (if possible) is always better - less signal conditioning yadayada ........ that's one of main virtues of all Papamps he gave us in all these years, so you have your proof there

anyhow, if you're going to have resistive attenuator, there is damage (my stance), and you can't get a wonder if you feed it with AVC

why you don't simply try said Edcor as AVC, without any resistive attenuator in chain ( short its pads on Iron Pre) -- you can be surprised how small number of steps you really need for proper Volume operation
 
#1,792
Hi ZM,

Thanks. I take your point about simplicity and avoiding the second buffer. And I'll certainly try the Edcor as AVC without pot, as you suggest. Fun to try, even if the number of steps ends up too small. We'll see.

Incidentally, I'm not trying these things with the 2SK170/2SJ74 Iron Pre at the moment (although I might later). I've made up a pair of the Cedarburg boards for the M2X amp, which are a unity-gain buffer using the AD797, and am listening to those with a normal volume pot in front. Similar idea, though.
 
#1,796
Hi ZM,

I've wired up the Edcors as an AVC, following option (A) you recommended above. Works well, and sounds good! I have just eight steps on the AVC, but have adjusted things so that the volume pots on the input to the buffer are at max volume, so effectively bypassing them, and attenuating on the AVC as far as possible. You were right about eight steps being more effective than I expected, but I doubt I'd be happy with just eight in a "final" preamp build. (If such a thing exists!) On the other hand, both the input pots I'm using and the Edcor as an AVC are linear, so I have finer control over volume changes when I'm nearer full volume, either on the pot or the AVC. Makes it more practical than might appear at first sight, as you predicted.

If you're interested, I've posted a couple of photos in another thread, where I had been talking about Cedarburg boards.

Thanks again!
 
You must log in or register to reply here.
Top Bottom