long tailed invertor intervention

[Andy Le Blanc]

I had a very pleasant experience sorting out the resistance values of the phase inverter in a rig I've been working on.

Its a cathode biased rig with no global feedback, and the inverter has one side
referenced to ground like a vox or a wem dominator

I started with bassman/marshal values, 100k for each plate and a 10k tail/feedback resistor.

it was CRAP... just plain nasty, hard and harsh, I was going for a touch more gain, very simple amp really,
Just a single pre stage before the invertor, last chance for gain...

I remembered that balance was improved with plate resitors and next tried 82k/100k

and... what a difference, it was still a bit hard sounding so I increased the 10k tail to 22k, I didn't want to go too far and lose gain possibly with the 47k vox example...

and oooh.... aaahhhh. It got really , really nice.


I think the lesson for me was that global feed back tends to mask invertor issues, use your ears before you assume that feedback will fix it.
and that in an amp with no global feedback, the other sources of degenerative feedback become much more critical.
Both the pre and power side were more than adequately bypassed at the cathode.

It was very nice to hear the critical resistor values in the invertor

[matt h]

(deleted)

[roberto]

Hi, Rk was 470? And what about the voltage of the supply?

In my amps I give the possibility to switch between 10 and 22k tail resistor.

[Tone Lover]

Thanks Andy good thing to add to my bag of tricks.
Bill

[pdf64]

Regarding balance, bear in mind that what matters is the system; the overall balance may have been worse with the 82k and 100k plate resistors, better with both at 100k.
Variables being the particular characteristics of the LTP tubes, output tubes, OT primary accuracy, resistors etc.
It's feasible that a smidge of 2nd harmonic from an imbalance may sweeten the tone.

[tubeswell]

... I remembered that balance was improved with plate resitors and next tried 82k/100k

and... what a difference, it was still a bit hard sounding so I increased the 10k tail to 22k, I didn't want to go too far and lose gain possibly with the 47k vox example...

and oooh.... aaahhhh. It got really , really nice.

And a 12AT7 will sound even suhweeter than a 12AX7 with this setup. YMMV

[Andy Le Blanc]

the cathode coupled long tailed pair came out of early oscilloscope circuits.

picture a seesaw like an ampeg inverter, in the cathode side of the inverter instead of the plate.

an evolution of the schmitt cathode coupled inverter

So... the simplest descriptions are that the inverted side is driven by the difference in the plate voltage swings of the two side.

The only way to get equal swings at both plates of the inverter is to have unequal plate load resistors.

[tubeswell]

... So... the simplest descriptions are that the inverted side is driven by the difference in the plate voltage swings of the two side.

This is why the LTP can also be used as a mixer; it only amplifies the difference between its inputs

... The only way to get equal swings at both plates of the inverter is to have unequal plate load resistors.

You can get closer to balanced outputs the more you increase the tail resistor (but you also reduce overall gain, all other things (i.e. the HT) being equal, if you do this)

[pdf64]

Or use a constant current source as the tail http://schems.com/manu/marshall/9100_2x50w.pdf