Another quickie... theory

[johnheath]

I read and try to grasp the overall discussion... interesting.

So... I've found this preamp schematic and wonder where you would place a "volume knob" if you were to rack it up? A dual gang pot I suppose?

Best regards

// John

 

[ruffrecords]

This is a fixed gain amplifier designed to be used in what I call a 'Classic' mixer topology which is fixed gain blocks interspersed with attenuators to set overall gain. As such it is not really designed to have it gain  altered. This is because it uses NFB to set the overall gain and, because of the number of capacitors in the loop, stability can only be guaranteed over a small range of gains.

So in my view, the best way to make this a variable gain pre is to have a ganged balanced switched input and output attenuator.

At the highest gain both attenuators are arranged to have 0dB loss. As you reduce the gain you initially increase the output so the gain drops and so does the noise until you reach a point where you decide headroom may be compromised by the increasing input level. From this point on you start to increase the attenuation at the input as well until you reach the minimum overall gain required.

Cheers

Ian

 

[johnheath]

Thank you Ian

Cool and interesting explanation.

A couple of question arises though and firstly... these attenuators are they supposed to shunt the signal totally or is it better to have a minimum resistance left? Tricky way to express it maybe?

Best regards

// John

 

[emrr]

I've found this preamp schematic and wonder where you would place a "volume knob" if you were to rack it up? A dual gang pot I suppose?

Funny, I have a 1016 here.  Tiny little Peerless input transformer and a UTC output. 

Question 1:  WHAT will you be doing with it?  That determines the type of attenuator on the front.

R13/15 can be brought to a switch, to achieve that 10dB range.  With C2/3 value as-is, there's probably a bit more acceptable attenuation range, additional gain range determined by the point response begins to suffer unacceptably. 

Headroom with either tube suggests an output attenuator with 10-15dB range.  The cheap-o way Gates attenuates the Stalevel is a 10dB range option. 

IF the goal is to use it as a MIC PREAMP, you will find in many cases you will have to use a 20dB U pad on the front end to avoid overload, overall gain will then be low, so you will set the feedback control resistors for maximum gain to make up some of the loss, and the output attenuator will be set for no attenuation. 

If the goal is to use it as a LINE AMP, there can be an advantage in having something like a pair of Daven 'T' attenuators (or Hairball/CAPI) on the front and back to trim it as desired.  Or go for the expensive 'H' type if you REALLY have a need.  You may want the 12BH7 for higher headroom, you may  want the 12AY7 for a more crunchy approach. 

With the stock input transformer on a 1016 you would not want to replace R1/2 with a dual pot primarily because any LINE use will easily overload the input transformer before you get a chance to attenuate.  MIC use may be ok, but may lose treble if turned down very much.  It won't be a universal approach.

Something like this that MIGHT be used MIC or LINE, I like to build a 3 position 'U' pad covering 0/-20/-30/-40, then it covers all bases, providing there's interstage gain or sufficient headroom for an output attenuator to be useful trim.  IF there's insufficient headroom, there's little point in an output attenuator, as you'll be primarily fighting the overload point OR not needing to worry about it.  That's the closest thing you'll find for a universal MIC/LINE single control with an amp like this. 

Look at the McCurdy AU300 for a feedback gain control in a push-pull preamp. 

 

[johnheath]

Thank you Doug

Interesting and very informative.

Well fact is that I built one of these a couple of years ago together with my son...

We were experimenting a bit with gain control and attenuation and ended up with a dual gang stepped pot in front of the first gain stage. As discussed here maybe not the dest way.

Anyhow the measurements were very fine and likewise the sound that was very transparent and clear... sounds great for acoustic guitar I can say.

Anyhow I was thinking about this overall concern with where and how to place a poy and this one came to mind as well.

I would like to use it as a MIC amp and I have two Sowter TX, 9610 as input and the 8650 as output... remaining is a good attenuation gizmo as well... I have paid some serious money for it so far and I do not want to waste it on crappy gain control design.

Fixing with the feedback resistors has also been in my head and perhaps changing and adding some values?

How ever it won't make it very quiet, but then there are other stuff in the chain anyway... so maybe not the greatest problem still?

Best regards

// John

 

[emrr]

We were experimenting a bit with gain control and attenuation and ended up with a dual gang stepped pot in front of the first gain stage. As discussed here maybe not the dest way.

Fixing with the feedback resistors has also been in my head and perhaps changing and adding some values?

How ever it won't make it very quiet

A stepped pot is definitely better than a continuous type. 

Feedback control won't HURT noise, so long as layout is clean and short.  You should be able to implement that with the same noise result at the same gain. 

 

[johnheath]

A stepped pot is definitely better than a continuous type. 

Feedback control won't HURT noise, so long as layout is clean and short.  You should be able to implement that with the same noise result at the same gain. 

Yes, and I meant that having a gain control including the feedback resistors won't have the possibility to turn the gain all the way down... that the preamp is Noise free is perhaps another thing if that's what you meant.

Sorry for my english... the fingers are faster than my mind sometimes

Best regards

// John

 

[mjrippe]

A stepped pot is definitely better than a continuous type. 

Careful there - a stepped SWITCH is better than a continuous pot.  A stepped pot (detented, so you feel clicks as it turns) is a piece of crap that is of limited use.  Because pots are low precision (10% or worse), the detents on one pot will not likely match those on another.  Cartec used them and I think Thermionic Culture did too.  Your left and right will never be the same level at the same number of clicks.  Sure, use them for your mono device that you want to recall settings on, but then when you build a second one it will not match.

The only exception is a single center detent on linear pots, of course.

 

[mjrippe]

Thank you Doug

Interesting and very informative.

Well fact is that I built one of these a couple of years ago together with my son...

We were experimenting a bit with gain control and attenuation and ended up with a dual gang stepped pot in front of the first gain stage. As discussed here maybe not the dest way.

Anyhow the measurements were very fine and likewise the sound that was very transparent and clear... sounds great for acoustic guitar I can say.

Doug is a great resource, isn't he   

I would say that if you built one with the attenuator after the transformer and it worked well and sounded good then keep it that way!

There is one advantage to this design - it is impossible to overload the input stage because you basically have a variable pad in front of it.

 

[johnheath]

Doug is a great resource, isn't he   

I would say that if you built one with the attenuator after the transformer and it worked well and sounded good then keep it that way!

There is one advantage to this design - it is impossible to overload the input stage because you basically have a variable pad in front of it.

Yes he is... like many others here it's simply a fantastic source of knowledge if you throw out a questions.

Sometimes it's hard to follow all expressions and abbreviations and what not but taking the punch being the stupid guy and ask a couple of more questions you finally get a load of knowledge yourself (me that is)

Anyhow... yes, I like your idea. It is not bad at all in fact. It just came to mind when sitting here (stuck in Africa) without any DIY stuff to do... and I began revising some of my stuff and cleared up some schematics... just to stay sane. I have a lot of time to think to say the least.

I'll mock around with it when I get back to Sweden.

Best regards

// John

 

[emrr]

Careful there - a stepped SWITCH is better than a continuous pot.  A stepped (detented, so you feel clicks as it turns) pot is a piece of crap that is of limited use.  Because pots are low precision (10% or worse), the detents on one pot will not likely match those on another.  Cartec used them and I think Thermionic Culture did too.  Your left and right will never be the same level at the same number of clicks.  Sure, use them for your mono device that you want to recall settings on, but then when you build a second one it will not match.

The only exception is a single center detent on linear pots, of course.

True....I don't consider those to be stepped pots anyway, but most people don't get that difference.

 

[johnheath]

About the stepped pot...

No, I meant a dual wafer stepped rotary switch... 2pol/11pos... very well matched resistors in 6dB steps

Best regards

// John

 

[ruffrecords]

Thank you Doug

Fixing with the feedback resistors has also been in my head and perhaps changing and adding some values?


Best regards

// John

As I mentioned earlier, this type of circuit does not lend itself to wide ranging gain changes by manipulating the negative feedback because of stability problems. You might be able to change it safely by 10dB or so but not much more.

Cheers

Ian

 

[johnheath]

As I mentioned earlier, this type of circuit does not lend itself to wide ranging gain changes by manipulating the negative feedback because of stability problems. You might be able to change it safely by 10dB or so but not much more.

Cheers

Ian

Thank you Ian

Yes, according to the original schematic that's what they "do"... ranging from 35 to 45dB

When I built mine I got rid of the two 180k ... after cao from cathode V1 and that increased the gain to roughly 60dB. ( calculating from a Sowter 4935, 1:7 and the Sowter 8650, 9:1 and having 57dB).

No problems with stability there so all in all I guess that you could elaborate from those circa 60dB down to the mentioned 35dB (question mark here as well)

All in all the circuit seems very stable and transparent and I liked it even though it's been a while since I used it.

Best regards

// John

 

[emrr]

As I mentioned earlier, this type of circuit does not lend itself to wide ranging gain changes by manipulating the negative feedback because of stability problems. You might be able to change it safely by 10dB or so but not much more.

I find in practice you can get more range than you give credit for, with  most of these type circuits.  You have to experiment to find the borders.  AC coupling cap dictating max NFB, acceptable response dictating least.  Some things you can disconnect completely, depends on transformer quality and layout.  I've comfortably gotten ranges of  +/-15 relative to stock out of a few real life vintage tube circuits.  Sometimes that requires a larger AC coupling cap to avoid a resonant boost on the bottom. 

 

[mjrippe]

True....I don't consider those to be stepped pots anyway, but most people don't get that difference.

Exactly, they are detented pots, not stepped pots.  I have had to replace them in the aforementioned gear and I hated every minute of it.