Newbie question: Why current feedback is not a goal in Mosfet power amps?

[Superfuzz]

Interesting.
Why those values?

Mm, being just a "ssguitar apprentice" made me copy in some way, something that I saw on some Peavey designs and stuff..then I read here about a member modding his alesis Power amp and..you know how it goes when you don't know so much about electronics

So you're telling me of, reducing the sensing resistor to 0.1 ohm and then take the signal directly from the node beetwen speaker - and resitor, without any resistor inbeetwen?

[Roly]

Thank you for that @JM, most interesting.   :dbtu:



A = voltage NFB, B = current NFB

This circuit was intended to deliver fully variable damping (c1955).  The connection of the current sense resistors R24 and R25 in a bridge arrangement means that damping factor, or current feedback, can be continuously adjusted from a high degree of damping, through zero damping, to negative damping.

[J M Fahey]

^^^^^   Wow, that's interesting, should try it.

[teemuk]

^ As far as I know, that scheme was pioneered by both Raymond G. Anthes, Herbert W. Sullivan and Charles A. Wilkins. Two latter gentleman worked for David Bogen & Co. and the scheme is naturally also employed in some of Bogen's amps.

A little Google searching with those names should find a handful of articles, patents and whitepapers from 1950's.

The idea was using -positive- current feedback to decrease (instead of increasing) output impedance (kinda opposite what someone eventually figured out to do with solid-state amps to fake tube amps). Voltage feedback could have achieved the same goal but due to limitations of tube circuits it was difficult to apply it in ample-enough amounts.

By the way, if you weren't already familiar with it, you might want to check out the power amp design of Seymour Duncan 100W Convertible. ;-)

[Superfuzz]

Interesting.
Why those values?

So you're telling me of, reducing the sensing resistor to 0.1 ohm and then take the signal directly from the node beetwen speaker - and resitor, without any resistor inbeetwen?

>>>>>>>>>>>calling mr. J M Fahey  <<<<<<<<<<<<<<<<

[Roly]

High source resistance, current feedback response.



High R, voltage.



Low R, current.



Low R, voltage.

[J M Fahey]

As simple a can be.
I used it in way more than 10000 amplifiers, be it discrete bipolar, MosFet or chipamp.
Does what I originally wanted in 1972: you pull the speaker out of the circuit, signal voltage doubles, "just like the Twins/Bassmans I had been building".
Quite gentle, adds "just a little bit", the good way.
Our friend Phatt  prefers external equalizing .
Fine with me, you can choose exactly what you want, tweak at will, etc.
But this one means adding just 1 (one) 0.1 ohms resistor, good return on such a little investment 

[Superfuzz]

Ok, now it's clear and it is even simpler than I supposed!!

Gimme some more spare time, and I will let you hear my impressions!  <3)

P.S.
I got some difficoulties reading all that graphics by Roly, I actually can't understand the schematic.. the 1M/12K5 resistors are the voltage feedback ones?

[phatt]

Thanks Jaun and Roly,
                           Good to test my brain again and yes I get (well some of it lol)
I know I'm not the wizz kid but one thing that really struck me when I first learned how to simulate circuits was that SS and Valve gear has almost opposite tone shaping.

No nothing to do with tone controls,, just the roll offs as you go through each stage (these do add up and may alter the outcome more than folks realize.

A classic Valve rig tends to have little bass roll off at the front end and depending on interstage cap values by the time the signal gets to the output this tone shaping can be extreme. I noted dramatic roll offs below ~100hZ

Well that's assuming I'm doing it all correctly but it's enough to tell me the output frequency response is likely far from HiFi flat.

Now SS rigs seem to be a very different animal where often the power section is kinda flat as a cow pat under a Massey Ferguson and the front end often has big steep bass roll offs right at the front end.

Surely with or without CFB on the power stage there is a high probability that just adding CFB on it's own may not reap enough to pull of a convincing sound.

I say this because my (not so famous) PhAbbTone pedal circuit shows two very different sonic results just by inserting it before and after a lot of distortion units.

From that alone it tells me just where in the sequence you change response curves does alter the outcome. (many pedal board freaks often speak of swapping their pedal sequences to attain a better sound,,hum?)

My thoughts are that just hanging  CFB on the end of any old SS power stage may not be enough. (As Juan noted earlier CFB on a hifi bandwidth power amp shows the designer is not guitar amp savvy)

Also some really old Valve rigs (some worth more than my house) were often loved or hated because when you cranked up the volume to high levels the tone controls no longer worked as efficiently as when the volume was low. So they only had one sound at high volume and if it was not your thing then you would sell it for a better amp.
(I'll leave that loading quirk explanation for the professionals, wink)

A vast majority of SS rigs do not suffer from this tone loss problem and if that pleases the player then read no further BUT for most may lead to a very long/expensive/frustrating learning curve to find the real problem. I've found most sound very harsh,,even if you are lucky enough to find one with darker speakers.

Oh how I wish I had time to spare to go dig up all my junk and test this all again but I've committed a terrible sin I've purchased a new keyboard a few months back,, (shush,, don't tell the guitar players) 
So very little guitar playing right now. I've only reached page 20 on the korg manual so it will take a while. lol
Phil.

[Roly]

1M/12K5 resistors are the voltage feedback ones?

No, you can ignore those as an artifact of trying to get both signals equal for phase comparison, but it didn't work as hoped.


JM's circuit is the same idea, perhaps better drawn.

[greenm01]

High source resistance, current feedback response.

Will you please post the LTSpice file?

[Roly]

Here you go...

Code Select Expand

Version 4
SHEET 1 996 680
WIRE 352 -112 -144 -112
WIRE -144 -96 -144 -112
WIRE 352 -64 352 -112
WIRE 352 -64 320 -64
WIRE 384 -64 352 -64
WIRE 496 -64 464 -64
WIRE 608 -64 576 -64
WIRE 704 -64 608 -64
WIRE 800 -64 704 -64
WIRE 704 -48 704 -64
WIRE 800 -48 800 -64
WIRE 608 -32 608 -64
WIRE -144 0 -144 -16
WIRE 352 48 320 48
WIRE 608 48 608 32
WIRE 608 48 352 48
WIRE 704 48 704 32
WIRE 704 48 608 48
WIRE 800 48 800 32
WIRE 800 48 704 48
WIRE 352 96 352 48
WIRE -144 224 -144 80
WIRE 352 224 352 176
WIRE 352 224 -144 224
WIRE -144 256 -144 224
FLAG 320 -64 Voltage_Feedback_2.828V
IOPIN 320 -64 Out
FLAG 320 48 Current_Feedback_35mV
IOPIN 320 48 Out
FLAG -144 256 0
SYMBOL voltage -144 -16 M0
WINDOW 123 24 124 Left 2
WINDOW 39 0 0 Left 2
SYMATTR Value2 AC 2.828V
SYMATTR InstName Vamp
SYMATTR Value SINE(0 2.828V 440 0 0 0 4)
SYMBOL res 336 80 R0
SYMATTR InstName R1
SYMATTR Value 0.1r
SYMBOL res 368 -48 R270
WINDOW 0 32 56 VTop 2
WINDOW 3 0 61 VBottom 2
SYMATTR InstName Rspk
SYMATTR Value 6.2r
SYMBOL ind 480 -48 R270
WINDOW 0 34 53 VTop 2
WINDOW 3 3 56 VBottom 2
SYMATTR InstName Lspk
SYMATTR Value 10mH
SYMBOL res -128 -112 M0
SYMATTR InstName Rsource
SYMATTR Value 5r
SYMBOL cap 592 -32 R0
SYMATTR InstName C1
SYMATTR Value 700µF
SYMBOL ind 688 -64 R0
SYMATTR InstName L1
SYMATTR Value 50mH
SYMBOL res 784 -64 R0
SYMATTR InstName R2
SYMATTR Value 44r
TEXT -48 280 Left 2 ;Feedback - Voltage and Current
TEXT 8 304 Left 2 ;Roly Roper 131013
TEXT 360 -120 Left 2 ;8 Ohm nom. sealed cab speaker equiv. cct.
TEXT 128 -96 Left 2 ;1W in 8r is 2828mV
TEXT -104 248 Left 2 !.ac oct 50 20Hz 20kHz
TEXT 464 72 Left 2 ;(Thanks to Rod Elliott)
TEXT 480 160 Left 2 ;1 watt in 8 ohms\nvoltage:\nE = (PR)^0.5\n(1*8)^0.5 = 2.828V(rms)\ncurrent:\nI = E/R\n2.828/8.1 = 0.349Amp\n0.35 * 0.1r = 0.035V = 35mV
TEXT 144 328 Left 2 ;131015



And also attached; of old Hi-Fi cct shown above.
http://www.ssguitar.com/index.php?topic=3098.msg23584#msg23584

[greenm01]

Here you go...

Thanks!