cathode follower study and how to reduce volume - hints needed.

[Renegadrian]

Hello everyone, don't write so often lately, but always an active builder...

One of my latest studies is on the cathode follower, starting from the typical bassman configuration to the boostraped version that you can find at valvewizard.
http://www.valvewizard.co.uk/dccf.html
Then all converted to FETs as we all know is possible.

Let me tell you the boostrap version is killer! but, it has a huge output volume. I hear you say, just tack a volume pot at the end! naaaaa it seems it loses some of its "fullness" and "crispness".

Soooo can you guys reccomend a way to mantain that full boost but with a lower output volume!?

If interested I can put here my schematic (but it's just that boostraped CF converted to FETs) and vero.

[Gus]

If I understand the post you used the last circuit in the link and converted it to JFETS?

If so the issue is the bootstrapping increases the gain of the first JFET because of how it works with the drain of the JFET

Have you tried it without bootstrapping the first stage.

[Renegadrian]

Yes, I used the last schematic, converted to FETs.
Actually, you just need to lift one leg of that extra cap to have a regular CF.
I also built a regular bassman style FET CF with a switchable bootstrap cap, that cap gives more presence. But my question is not related to the boostrap part, both circuits have a great sound on full throttle, but lose something when you turn down the volume pot.

[alfafalfa]

With real tubes the gain goes down if you lower the (plate)anode resistor or lower the B+ = supply voltage to the plate.  Have you tried that path ?
That it sounds different for you might have something to do with the Fletcher Munson curve

https://en.wikipedia.org/wiki/Fletcher%E2%80%93Munson_curves

"It's all in the ear of the beholder"

And me too would like to see a schem if that's possible.

Alf


Alf

[teemuk]

A schematic would help tremendously but I assume he fitted a potentiometer after the DC coupled common cathode + cathode follower circuit. But I suppose he ultimately utilised FETs instead of tubes. I don't think device is the issue here, though.

If the output is not "buffered", output Z will effectively - at low volume settings - be very high, as it largely consists of pot's resistance to its wiper. The varying output Z could form a RC filter (with input stage of whatever device it plugs into), which attenuates high frequencies more at higher output Z values.

Or something...

Without knowing what kind of potentiometer was fitted there, and how, it is quite hard to estimate whether the potentiometer introduces a considerable load to cathode (source?) follower stage, or not. That could alternatively have audible effects.

[Renegadrian]

ok here's schem (see some values needed some adjustments) the 47n can also be doubled to 100n for a little more bass. trimmer 25k or 50k, adjust then by ear.
I have a 1M pot for volume but can/will sub it easily.

[Gus]

Have you tried a nonbootstapped first stage?

Next what value volume control are you using after the follower 100k like in the picture?

What is the circuit with a volume control connected to?  An effect, amp?  What cable are you using? The capacitance is part of the interaction

You need to figure out the interaction between the circuits. When the volume control is turned down you add more series resistance "top" of the volume control to wiper and that can make things dull depending on the interactions.

[Keppy]

^ What he said.

I have a 1M pot for volume but can/will sub it easily.

A 1M pot adds a lot of series resistance to the output as you turn it down. That, combined with cable capacitance, combined with Fletcher-Munson, might be what you're losing by turning the volume down. Try smaller volume pots before you go reworking the circuit.

Note to maintain the same bass response, you MIGHT have to increase the 47n output cap by the same factor as you decrease the volume pot. So, for a 100k pot try a 470n, for a 10k pot try 4.7uF. I say might, because the cutoff frequency on the output is currently in the subsonic range, but this can be affected somewhat by the input of whatever you plug it into. You probably won't notice a difference. Unless you do.

I suppose you could also add a treble bleed cap to the volume pot to maintain the highs as you turn it down.

[Gus]

I missed the 1 meg, I was looking at the schematics

Try a 50K to 100k volume

EDIT
Somethings I posted before.  390pf cable capacitance into a 470k resistive input

100k volume

100k volume changed to a 500k volume note how it rolls off the highs at lower volume settings

[PRR]

1 Meg output pot is sure to affect sound when not full-up.

100K would be a better test.

Despite appearance, this is NOT a cathode follower. V2 works as a common-cathode stage. Signal is applied grid-cathode, not grid-ground. Output impedance is large.

[merlinb]

You can tap-off the output from a lower point, and use a lower resistance pot, e.g like this:

[Renegadrian]

Wow so honored to have a reply from you Merlin, let me say that your site is so full of useful information, I spent several nights studying your ckts.

for now I took out the 1M pot and put a 50K pot, it seems to work better (at least on a quick try) - have to try it at a good volume. Let me tell you again this ckts rocks even with fets, I love it!

[Renegadrian]

My latest experiment - I wanted to apply a baxandall tone control after my cathode follower (bassman style + the switchable bootstrap cap) but I get a huge level cut...too much loss...I know that a CF and a bax should work very well togheter, what am I missing here!?
posting both snippets
this is the fet bassman, with switches for the booststrap cap and for the cathode bypass cap (more gain with it) - it works great as is in this schematic


this is a snippet from a Fender amp (showman 6g14) it has a bax from the plate, not from the cathode. The output cap of the pre is connected to the 220k/200p instead of that 47n cap. resistor between pots is 100k (original didn't have it) - volume pot is 100k. it goes straight into amp, no other pedals in between.
big loss...doesn't work as supposed...

[Groovenut]

A passive Bax is goiing to have around 20dB of loss. Without some make up gain somewhere before or after the bax, you're really going to notice it.

[Renegadrian]

so I should put another gain stage after that...gonna try it...

[PRR]

> passive Bax is goiing to have around 20dB of loss.

James, not Bax.

20dB for the stock James, but Fender used odd values and the midband mid-turn loss is more like 28dB or about 25:1. As Fender's input tube only has a gain of 50, the output from the tonestack is not much stronger than the signal at the input jack.

[Groovenut]

> passive Bax is goiing to have around 20dB of loss.

James, not Bax.

20dB for the stock James, but Fender used odd values and the midband mid-turn loss is more like 28dB or about 25:1. As Fender's input tube only has a gain of 50, the output from the tonestack is not much stronger than the signal at the input jack.

Paul,

What's the difference between the James and the Baxandall? I see them cross referenced all the time and assumed (wrongly as such) they were the same.

If I had to guess the difference, it would be James is passive Baxandall is active?

Thanks!

[PRR]

> James is passive Baxandall is active?

James is passive and Baxandall is active!

[teemuk]

Being passive, the circuit must attenuate considerably in order to have any range for the controls. If the range to boost is, say 20 dB, then it means the insertion losses (how much the circuit attenuates) at level response must be about -20 dB.



Without gain you can't "boost" past 0 dB mark (no insertion losses) so maximum output is either unity gain or less. If that is "boost" mode then level response and cut must be attenuating the signal considerably more.

The passive "James" circuit actually has quite high insertion losses. If I remember right, they were even higher or pretty much in the same ballpark as those of a generic passive Fender/Marshall/Vox -style "tonestack". The circuit has that "low-loss" moniker because its initial version didn't combine bass and treble circuit branches (in parallel) into a single tone control circuit but simply had them in series, one branch following the other and each introducing its own insertion losses. James's combined version therefore has only one insertion loss instead of two so it's "low-loss" in that nature but it's not so "low-loss" when compared to other tone control circuit designs. But of course people started to use that "low-loss" moniker entirely out of its original context.

"Baxandall" term refers to active version where (quite) similar circuit is fitted into feedback loop of an amplifier / gain stage. In this arrangement the circuit can have actual gain, so it can achieve level response at 0 dB and either amplify the signal further, or attenuate it. So it's the active version that provides considerably lower insertion losses.

The "core" circuit schemes for both James and Baxandall are quite similar but passive circuit has logarithmic ratio of component values like this:
http://www.preservationsound.com/wp-content/uploads/2012/01/SimpleToneControlStage.jpg
Potentiometers also need logarithmic taper function.

The active version relies on establishing "equal impedances" from input to input of the amplifier, and from output to input of the amplifier.

In a standard inverting negative feedback amp arrangement gain is unity when these impedances are equal. If this ratio of impedances changes it either results to attenuation or gain.

When the circuit is arranged like this, the ratio of the component values must be linear instead of logarithmic and potentiometers must also have linear taper function.

--

Oh, ideally one wishes to drive all these tone controls (whether active or passive) from a low impedance source (e.g. cathode follower) instead of high impedance source (e.g. common cathode amp). Why? Because source impedance also affects the circuit's operation and response. The effect to response becomes more distinct the higher the source impedance is.

[Renegadrian]

Right, thx for your reply, as you can see output is from the CF, so it should par well with the james or bax. now, your suggestion, would it be best a james with a recovery stage after it or a bax with a fet fb loop!?