What are typical voltages going into LTP?

[Steve A.]

My question concerns the last series of Crate V amps which use an IC buffer immediately before the LTP. The TL072A chip is powered by a +/-14.7vdc supply so my understanding is that would limit the swing of the signal being output to the LTP to +/-14.7vdc.

So I was wondering what the range of signal levels going into a LTP would be, from a blackface Fender to a crunchy Marshall to a Soldano SLO- to determine if this chip is compressing the signal a bit or what...

Thanks!

Steve Ahola

[Merlinb]

A typical ECC83 LTP needs about 8Vp-p before it really starts clipping (assuming it's roughly centre biased). If there is also global NFB then this could rise to 16V to 20Vp-p, or maybe a little more.
Also, the opamp won't swing all the way from rail to rail. It'll be more like 25Vp-p at most. Not quite as much as the typical JCM45 circuit would give you at max, but not unreasonable.

[ETR]

Originally Posted by Steve A.

My question concerns the last series of Crate V amps which use an IC buffer immediately before the LTP. The TL072A chip is powered by a +/-14.7vdc supply so my understanding is that would limit the swing of the signal being output to the LTP to +/-14.7vdc.

So I was wondering what the range of signal levels going into a LTP would be, from a blackface Fender to a crunchy Marshall to a Soldano SLO- to determine if this chip is compressing the signal a bit or what...

Thanks!

Steve Ahola

It depends on the tail resistor of the LTP. A 10K tail resistor needs a lot less drive than a 22 or 47K tail.

[Steve A.]

Originally Posted by ETR

It depends on the tail resistor of the LTP. A 10K tail resistor needs a lot less drive than a 22 or 47K tail.

Thanks to both you and Merlinb! There is no NFB in this amp (the cathode biased Crate V18-112) - here are the schematics:

http://www.blueguitar.org/new/schem/...rawings_v2.pdf

I think that having the IC buffer there keeps the signal level down enough that they can get away with using 470k grid loads on the EL84's, which I have been replacing with 220k resistors. And the LTP comes with a 22k tail which I have been replacing with 47k to bring it around to Vox specs. Only problem is that those mods seem to reduce the overall volume of the amp significantly, although they do give it a lot of chime...

Thanks!

Steve Ahola

[Merlinb]

The loss of volume will mainly be the 220k resistors in place of the 470k. Why change them?

[ETR]

Originally Posted by Steve A.

Thanks to both you and Merlinb! There is no NFB in this amp (the cathode biased Crate V18-112) - here are the schematics:

http://www.blueguitar.org/new/schem/...rawings_v2.pdf

I think that having the IC buffer there keeps the signal level down enough that they can get away with using 470k grid loads on the EL84's, which I have been replacing with 220k resistors. And the LTP comes with a 22k tail which I have been replacing with 47k to bring it around to Vox specs. Only problem is that those mods seem to reduce the overall volume of the amp significantly, although they do give it a lot of chime...

Thanks!

Steve Ahola

I'd be tempted to bypass the buffer. I love 47K tails in a LTP.

[Merlinb]

Wow, that's a really unbalanced PI they have in that amp!

[Steve A.]

Originally Posted by Merlinb

The loss of volume will mainly be the 220k resistors in place of the 470k. Why change them?

The stock Crate V18 sounds like crap so I figured that if I morphed the circuit over to more traditional designs it would sound better. Which it does, but at a loss of volume that I was not taking the buffer into account.

I guess that the Marshall 18 watter uses 470k grid loads but I think that most amps out there use 220k.

I've been redoing the PI with a 1k2 resistor on the cathode and a 47k resistor on the tail which sounds really good to me. I was encouraged to try a 820k resistor on the cathode, but while it adds some grind to the signal I find that it keeps the amp from cleaning up nicely.

By unbalanced are you referring to the 120k and 82k plate resistors? With no NFB what values do you recommend? Some designs use a pair of 100k resistors. I'm not sure I hear a difference when I change them out- is there something that I should be listening for?

And yes, I will be bypassing the buffer on the next V18 that sees my test bench, er, dining room table... Remove R19 and R26 and run a jumper between the outside terminals.

Thanks!

Steve Ahola

P.S. Here is the thread over at The Gear Page spelling out most of the different mods I've been doing to these amps (right now I think that the Soldano SLO conversion is winning hands down):

http://www.thegearpage.net/board/sho...3&goto=newpost

They are also posted as separate threads (easier to follow) over at ppwatt.com but I believe that you need to register and log on to view them:

http://www.ppwatt.com/crate

.

[Merlinb]

Originally Posted by Steve A.

By unbalanced are you referring to the 120k and 82k plate resistors? With no NFB what values do you recommend? Some designs use a pair of 100k resistors...

With no NFB the traditional LTP is pretty much perfectly balanced with 91k/100k. However, with 100k/100k its fine for guitar purposes.
With the common values of 82k/100k it's more unbalanced, UNLESS you add NFB, in which case it comes right. In this circuit it's 82k/120k with no NFB, which is off the map!! Waaay unbalanced.

[Gregg]

A typical ECC83 LTP needs about 8Vp-p before it really starts clipping (assuming it's roughly centre biased). If there is also global NFB then this could rise to 16V to 20Vp-p, or maybe a little more.

My observations are the most common LTP (470R/10k/4k7) with FB needs 5-6V to start clipping (anode voltage 380-390V). Especially if you have a high gain preamp you don't need more than 10V.

[ETR]

Not really unbalanced for guitar use.