John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Hi Ed,

They'll know as soon as they start! No damage to the cat necessary.

-Chris

Confusing a sleeping panter with a dead one may not cause damage to the cat, you will not be so lucky. Even a house cat may draw blood, anything larger and it wil be a lot more serious.

Panthers (aka mountain lions) are still spotted around here on rare ocassions. Most of the larger cats are very rare but have been known to slip out of their cages.

Tigers even in captivity can mark you at 25+ feet.
 
Disabled Account
Joined 2012
Confusing a sleeping panter with a dead one may not cause damage to the cat, you will not be so lucky. Even a house cat may draw blood, anything larger and it wil be a lot more serious.

Panthers (aka mountain lions) are still spotted around here on rare ocassions. Most of the larger cats are very rare but have been known to slip out of their cages.

Tigers even in captivity can mark you at 25+ feet.


This mountain lion is a cub. They are very smart and curious animals, even at this young age. The mother had killed and eaten a woman jogging in the woods about 1/2 mile from my house. The mountain lion mother was tracked and killed. But then they realized it had babies so the tracker eventually found a pair of cubs and took them to a Zoo keepers home not far from me. The zoo keeper kept one of them. It was in the living room and walked up, hopped on the sofa where I sat and looked at this strange thing I held in my hand (camera). It stuck its face right into the lens as if to see inside. Click.

mountain lion cub.jpg


While I visited and I made a special whistle sound. I used it to call my wild deer friends. A year later, I went to the Folsom city zoo where it was taken in and given a home with a few other mountain lions....... I was standing 10 feet away from the fence and whistled.... that not so little lion came running right over and looked at me and we studied each other for awhile. It had remembered me! I never went back after that. Sad to be locked up.


-Richard

Back to amps....
 
Last edited:
Seriously, I used this very design for years, and even gave it to Richard Sequerra as a design for a cheap radio amp. He loved it!
It really works much better than the sum of its parts. I once loaned it to a friend who used triodes exclusively with big horns. He liked it too!
As I said, I did not come up with the original topology, but it is something that you all could learn from.
 
AX tech editor
Joined 2002
Paid Member
Here is an amp that I designed back in 1977. It works in class A+C. Class A part, single ended, is represented by a couple of transistors, Q1 and Q3, as a CFA amp. It outputs through a resistor R1. When voltage drop on that resistor is enough for 2 more transistors Q2 and Q4 to start acting, they gradually increase their current output, gradually entering class C in parallel with class A single ended operation that continues up to the clipping.
Nested feedback loops through the opamp provide what is now called VFA. Also, due to a capacitor in feedback of the opamp loop, it acts also as a servo. Ratios of feedback loops were selected for the cleanest resulting sound quality.

Anatolyi, this concept has been very popular for some years at Elektor, in various guises. IIRC in the 80-ies. One mod they can come up was to add a single diode between the bases of Q2 and Q4 which brings the class-C take over a bit closer. Cut down the xover distortion a lot.

Jan
 
Last edited:
Kittens cannot be Mountain Lions. And as far as I know there are no Mountain Lions in captivity because the epigenetic triggers only happen in the wild, and not for all Cougars.

The average cougar weighs 90lbs. If the genetics trigger it will grow past 150lbs, and get a beautiful golden coat. It will be longer and rippling with muscles. There is a reason they got the title Lion.

Here is a pic. They are not common.

Did not know Cougars got this big! A guy on my facebook shot this in Oregon. - Imgur
 
Anatolyi, this concept has been very popular for some years at Elektor, in various guises. IIRC in the 80-ies. One mod they can come up was to add a single diode between the bases of Q2 and Q4 which brings the class-C take over a bit closer. Cut down the xover distortion a lot.

Jan; it is a misunderstanding. There are no crossover distortions. No abrupt Off/On state of transistors. Additional emitter followers just start assisting class A, gradually! Additional diode would not bring down any distortions, but just allow class A driver to run cooler, demanding less current from it. If there were some crossover distortions, that means resistors' value from class A to output was too high!
It should be no more than 2 Ohm!

Class AB amp has worse crossover distortions actually, caused by gm-doubling, and this region it has on lower loudness! Here on low loudness works pure single ended class A!
 
Disabled Account
Joined 2012
Kittens cannot be Mountain Lions. And as far as I know there are no Mountain Lions in captivity because the epigenetic triggers only happen in the wild, and not for all Cougars.

[/url]

I know you must think you are an expert in such things, but the 'kitten' shown is in fact a cougar, as I told you. As was her mother who killed a woman and thus had to be killed herself. Large with brown coat in zoo now. Obviously, my true story is too much for you. I don't know what I am talking about is amazing and didn't take a picture of a young cougar/mountain lion. Where do you get this idea from? How can you deny what I said is not true? Who are you to say that?

Cougar Is Killed For Jogger's Death In California Park - NYTimes.com
Mountain lion that stalked Colfax hiker was healthy, had food in belly | Auburn Journal

-Richard
Cool, CA.
 
Last edited:
AX tech editor
Joined 2002
Paid Member
Jan; it is a misunderstanding. There are no crossover distortions. No abrupt Off/On state of transistors. Additional emitter followers just start assisting class A, gradually! Additional diode would not bring down any distortions, but just allow class A driver to run cooler, demanding less current from it. If there were some crossover distortions, that means resistors' value from class A to output was too high!
It should be no more than 2 Ohm!

Class AB amp has worse crossover distortions actually, caused by gm-doubling, and this region it has on lower loudness! Here on low loudness works pure single ended class A!

Well I beg to differ - at the instant the additional emitterfollowers start to conduct (Vbe across the R) your loop gain rises massively. That happens at every zero crossing. And even though this is damped by the feedback, it is clearly seen. Even at carefully adjusted class (A)B you see it, how much more here!

In fact, the abrupt loop gain changes with the activation of the 'dumpers' (if I may call them that), was the origin of Peter Walkers' Current Dumping approach. He realised that no amount of feedback could eliminate it, and so he devised his very clever scheme to change the feedback factor at the moment the 'dumpers' came on, thereby keeping the loop gain constant!

Jan
 
Well I beg to differ - at the instant the additional emitterfollowers start to conduct (Vbe across the R) your loop gain rises massively. That happens at every zero crossing. And even though this is damped by the feedback, it is clearly seen. Even at carefully adjusted class (A)B you see it, how much more here!

In fact, the abrupt loop gain changes with the activation of the 'dumpers' (if I may call them that), was the origin of Peter Walkers' Current Dumping approach. He realised that no amount of feedback could eliminate it, and so he devised his very clever scheme to change the feedback factor at the moment the 'dumpers' came on, thereby keeping the loop gain constant!

Jan, you still did not get it. It is not class B amp slightly assisted by class A driver. It is a full blown single ended class A amp slightly assisted by class C output transistors. Why would loop gain raise massively, if class A works on 8 Ohm load through 1 Ohm resistor? It is 1/8Th of voltage loss only! I.e. you call 12% increase of gain "massive"? And you have to remember, that when class C starts conducting exponentially, the class A stage still supplies current through 1 Ohm resistor!
My amp delivers first 5W from pure SE amp, and the rest 100W assisted by class C emitter followers. What crossover distortions?! It is class AB amp that delivers crossover distortions that depend on temperature that in turn depends on the sound envelope, i.e. envelope modulated crossover distortions! My amp has absolutely no crossover distortions!

I can't speak for Peter Walker, but you misunderstood him obviously. His problem was, no fast power transistors! And no fast enough opamps to correct errors! That's why he invented that "clever L-C bridge", so higher frequencies were supplied by class A amp, instead of class C transistors.

Also, you can see that my design differs from Walker's. I did not see his when came up with mine, and I already had fast enough output transistors and opamps. And I was not afraid even then of nested feedback loops. I loved them! :)
 
AX tech editor
Joined 2002
Paid Member
Look at the situation when going through zero crossing. The 'dumpers' go off, and have to get turned on again. I agree that with the low R, the 'problem area' is quite small, but it is there.

At the point where the voltage across the 2 ohms reaches 0,6V, the loop gain increases significantly, maybe with a factor of 50 or 100. That happens with an Iout of about 0.3A. If my calculation is correct, assuming 8 ohms, that is at about 0.3W RMS out.
So at a very low signal level these changes occur, right within the critical 'first watt'.

Jan
 
Disabled Account
Joined 2012
Now what would be a more perfect design? Well, a balanced bridge output with CFA feedback would be perhaps the best approach. This is more complex, but more elegant than anything I now have out there.
Richard and I have been discussing doing this very thing. Who knows? It might develop into a product someday.

It will be great and a fun challenge.

:) :cool:


-RM
 
Status
Not open for further replies.