Output coil on power amp...

[AKSA]

Toino,

You are a citizen of Portugal, so you probably love anything which burns hydrocarbons.

Think of non-switching Class AB as a gasoline engine doing diesel engine tricks.... if none of the output devices ever switch off, you have a Class AB amp doing Class A tricks. Two Englishman, Blomley and Visch, first thought of this back in the early seventies. It has always been a pot of gold - efficiency with no crossover distortion. NP has also addressed it, and a few years ago took out a patent on a unique way of doing this. Sometimes it's called sliding bias, though that is not really the same.

Cheers,

Hugh

 

[TOINO]

Thanks by the explanation AKSA.
As a matter of fact we try to be good ambientalist here.

http://www.diyaudio.com/forums/showthread.php?postid=981652#post981652

 

[quasi]

I thought the main reason was to prevent RF picked up by speaker leads finding its way back into the amp. Surely if an amp becomes unstable when the coil is removed then we have an amplifier that is just bad. Prevention is better than cure ?
My $0.02

 

[Hartono]

Hi Quasi,

The coil doesn't help that much to prevent the RF going into amp

the coil is mostly used to prevent capacitive loading seen by the amp,
which if bad enough can cause instability in many design.

I miss four'n twenty meatpies in 7 eleven

 

[GK]

When used in conjunction with an output network such as that in this amplifier……

http://users.picknowl.com.au/~glenk/12W.HTM

…..very good rejection of RF picked up by the speaker leads can be achieved. The output capacitor shunts the RF to ground.

I have no trouble running my amps in the vicinity of things such as this…..

http://users.picknowl.com.au/~glenk/CTM2K.HTM

…..without a problem. And no, the inductor doesn’t act as an “antenna” because I house my designs in metal boxes.

Cheers,
Glen

 

[jacco vermeulen]

The output capacitor shunts the RF to ground.

Mr K,

an ordinary foil cap for C22 ?

(no visitor counter on the LandRover page ?)

 

[kASD]

Glen your LR restoration is more than AWESOME

 

[Hartono]

nice

 

[GK]

Mr K,

an ordinary foil cap for C22 ?

Metalised polypropylene.

(no visitor counter on the LandRover page ?)

I'm working on it (and lots more).

Glen your LR restoration is more than AWESOME

Thanks.

Cheers,
Glen

 

[Hartono]

Metalised polypropylene.

outer foil on the ground side I guess

 

[forr]

Hi Hugh,

The Blomley amplifier is is a real class B where the commutation between the two branches of the push-pull output stage which are always conducting is done ahead of it, in a previous low power stage.

The biggest audio stupidity I've ever done is to have built a Blomley, to have tested it with square waves and capacitive 2 µF load, and slightly unsatisfied, to have dismantled it. A few hours later I discovered I've not listen to it !

I've never heard of Visch. Any schematics ?

Cheers

 

[Bob Cordell]

Toino,

You are a citizen of Portugal, so you probably love anything which burns hydrocarbons.

Think of non-switching Class AB as a gasoline engine doing diesel engine tricks.... if none of the output devices ever switch off, you have a Class AB amp doing Class A tricks. Two Englishman, Blomley and Visch, first thought of this back in the early seventies. It has always been a pot of gold - efficiency with no crossover distortion. NP has also addressed it, and a few years ago took out a patent on a unique way of doing this. Sometimes it's called sliding bias, though that is not really the same.

Cheers,

Hugh

Hi Hugh.

Minor caveat: although assuring non-switch-off in the Class AB output stage may reduce crossover distortion, I will not generally eliminate it, since the net transconductance of the output stage will still usually go through variations as the output current swing goes through crossover.

Cheers,
Bob

 

[Conrad Hoffman]

Had a thought on my morning walk, maybe not original, maybe not correct, but here goes. Taking the Kleinschmidt amp circuit above as an example- I don't remember what the output inductor value is, but let's just use 5uH for talking purposes. 5uH in parallel with 6 ohms can be converted to a series model of close to 5uH in series with 164 uohms (1 kHz), or 5uH in series with 65 mohms (20 kHz). It seems like if one picks up a bit of resistance in the inductor by choosing smaller wire or using manganin or constantan, the parallel resistors can be dispensed with. Damping factor might go down by some minuscule amount, but that's not necessarily a bad thing. The calculated Q is surprisingly high either way, but I doubt real output networks ever achieve that- at least I hope not. Because the impedances are so low, I wonder if real circuits are as close to the simulations as one would hope? Or vice versa?

 

[GK]

Had a thought on my morning walk, maybe not original, maybe not correct, but here goes. Taking the Kleinschmidt amp circuit above as an example- I don't remember what the output inductor value is, but let's just use 5uH for talking purposes. 5uH in parallel with 6 ohms can be converted to a series model of close to 5uH in series with 164 uohms (1 kHz), or 5uH in series with 65 mohms (20 kHz). It seems like if one picks up a bit of resistance in the inductor by choosing smaller wire or using manganin or constantan, the parallel resistors can be dispensed with. Damping factor might go down by some minuscule amount, but that's not necessarily a bad thing. The calculated Q is surprisingly high either way, but I doubt real output networks ever achieve that- at least I hope not. Because the impedances are so low, I wonder if real circuits are as close to the simulations as one would hope? Or vice versa?

G'day

The resistors are not there to dampen the coil, they are there to provide an R-C Zobel in conjunction with the output capacitor.

Cheers,
Glen

 

[phase_accurate]

This configuration would work even better if you could guarantee that the inactive device never completely switched off....

Hi Hugh, how does such an amp sound like ?


regards

Charles

 

[AndrewT]

quote:
This configuration would work even better if you could guarantee that the inactive device never completely switched off....


Hi Hugh, how does such an amp sound like ?

Since one side of the push pull is in-active, then I reckon it sounds like any other ClassAB output stage.

I think the ClassAB bit is more accurately defined as using one active side to drive the load, whereas ClassA uses both sides ALL the time to drive the load. Just like single ended, where the active device/s is/are allways driving the load.

 

[Bob Cordell]

Had a thought on my morning walk, maybe not original, maybe not correct, but here goes. Taking the Kleinschmidt amp circuit above as an example- I don't remember what the output inductor value is, but let's just use 5uH for talking purposes. 5uH in parallel with 6 ohms can be converted to a series model of close to 5uH in series with 164 uohms (1 kHz), or 5uH in series with 65 mohms (20 kHz). It seems like if one picks up a bit of resistance in the inductor by choosing smaller wire or using manganin or constantan, the parallel resistors can be dispensed with. Damping factor might go down by some minuscule amount, but that's not necessarily a bad thing. The calculated Q is surprisingly high either way, but I doubt real output networks ever achieve that- at least I hope not. Because the impedances are so low, I wonder if real circuits are as close to the simulations as one would hope? Or vice versa?

Hi Conrad,

I'm confused. I think you are talking about the resistor that is traditionally in parallel with the output coil. Not sure why you want to dispense with that resistor. To my knowledge, it is not a source of a problem, and it damps the coil and assures a resistive interface at high frequencies.

I guess if you want to increase the amount of RFI ingress rejection, then maybe you want to get rid of that resistor. The smaller that resistor, the less rejection of external RFI. Is that what you had in mind?

Thanks,
Bob

 

[TOINO]

In my case, the amplifier is hi power and the coil was heating a lot.
That is the main reason that took me to try to take it off

 

[Bob Cordell]

Basic coil question #1

Here's a very basic question on the audibility of output coils that I would like to see some discussion on:

If the use of the R-L output coil combination on an amplifier causes a sonic degradation, do we think it is due to the inductance itself, or do we think it is due to the way in which the inductor is implemented?

If the former, it would seem to fall in the linear frequency response/transient response category. Ringing perhaps, for example.

If the latter, it would seem to fall in the category of nonlinearities, such as by using an inductor with a core that saturates, or one that is in too close proximity to metal, or one that is picking up RFI, or one that is disturbing other circuits with its radiated magnetic field.

Thanks,
Bob

 

[Hartono]

Hi Bob,

First,I don't like the nonlinearities you mentioned, that would ask for unreasonably large metal enclosure.

In voltage drive ,"perfect" output inductor + parallel resistor will change the damping factor at high frequency, this is my only real nagging, combine the impedance at HF + passive network + loudspeaker(tweeter?) thermal modulated resistance of voice coil, this should sound different compared to no inductor.

basically different damping factor sounds different, can you hear it ? I dunno it is measurable though.

Actually I'm converting to current drive

seriously