Easy, GC Current Amp piece-

[stefanobilliani]

2 days ago I went to the local store an purchase some chip amps just to play in current amplification mode .

I builded a pair of GMs MLTL speakers with the JX92s (if you like take a look at the Quarter Wave site gallery )and going to test also a pair of MLTL with the Fostex FE206E.
Reading the First Watt article by Nelson Pass , I decided to try the applications described. Actually, the tests are waiting for the FE206E to come although the enclosures are ready , by the way the JX92s gives a taste of what can be achieved following this different approach.

So the circuit attached is obtained following my tastes with the JX92 drivers in the GMs MLTL enclosure .
The circuit inspiration comes from the Elliot Sound pages variable impedance project.

As pointed out by Nelson I believe that the JX doesn't gain the full advantages from this kind of amplification and I will report back the results with the Fostex (they are more efficient) , but it is still worth trying and decide for yourself if it is good or not .

I like the results a lot anyway , and musically I prefear the current amp to the voltage one . They are just 2 different things

The chip used is what I find in the local store and more chip are in count to try.They doesn't equal the class A for sure but they are cheap and very easy to build. More to the point, the speaker is included in the feedback loop - where in a class A current source it can be avoided .

So ... be prepared to enjoy

PS . The amp is dedicated to fullrange drivers .

[Igla]

Hi Stefano !

I must say,that I am very interested to try this current chip amp...I have some spare LM3875 and full range loudspeakers to drive them.
Well,I must say I am not really expert for electronic stuff.
I mean...I've build GC but here I could use some help.Rod Elliot article didn't helped me much...I just need plane and simple explanation;for example
...what happens if I change R5 or R3 ; how to calculate the gain of the amp.
Any help would be appreciated.

Regards,
Matjaz

P.S.
Should this thread be moved to chipamp forum?

[Nuuk]

Yes, thanks Stefano. I am another who doesn't really understand the theory so trying somethnig will be the way forward and any pioneering work is much appreciated!

[azira]

His circuit is basically Elliot's "Mixed Mode Feedback" circuit. Click the link and scroll down half way and you will see the calculations for gain and output impedance. By the looks of it, he's setting it up for a higher output impedance than elliots circuit.

[Igla]

I have read the article and I know this circuit is a mix mode amp
but Rod has put there 10k , 560ohm and 0.2ohm resistor and
Stefano has 22k , 220ohm and 1.4ohm if I recall it right.
So,if R1 is bigger , R2 should be smaller and R3 should be bigger
or is there any other rule to change resistors value if I want to
change output impedance and have gain of about 22.
I guess that R3 should be as small as possible to minimize the power lost.
Regards

[azira]

There isn't an easy answer to how much "gain" this amp has. It varies with the loading. It's designed to be nearly load independant, so in other words it is capable of delivering 'x' amount of "power".

In low impedance loads, it puts out less gain than in high impedance loads because you need less voltage/current in order to drive lower impedances to the same power level.

I guess that R3 should be as small as possible to minimize the power lost.

No, if that were the case we could just omit R3 all together, 0 resistance or no resistor is as small as possible. R3 is the current gain resistor portion of the amp. It's value affects the Zo of the amp. The goal of this amp is to establish a particular output impedance. The reasons for this are varied. In Rod Elliots amplifier, he wants a Zo that resembles the Zo of a tube amp. In NPs article, he is using a different circuit to establish a Zo of 50ohms or so. So that it looks like a current mode amplifier or something like that.

I haven't really analyzed what stefano wants to do. Perhaps he is converting Rods schematic to 50 ohms output impedance? You could find out with a sim and a few calculations.
--
Danny

[Nelson Pass]

The F1 has an intrinisc output impedance of about 1 K ohm,
and I parallel 94 ohms across the output to establish about
80 ohms output impedance.

[stefanobilliani]

Originally posted by Igla

So,if R1 is bigger , R2 should be smaller and R3 should be bigger
or is there any other rule to change resistors value if I want to
change output impedance and have gain of about 22.
Regards

What I see is :
R1 tends to be in parallel with the speaker if R2 tends to zero.
The speaker is not a pure resistor.

Originally posted by azira
There isn't an easy answer to how much "gain" this amp has. It varies with the loading. It's designed to be nearly load independant, so in other words it is capable of delivering 'x' amount of "power".

In low impedance loads, it puts out less gain than in high impedance loads because you need less voltage/current in order to drive lower impedances to the same power level.



No, if that were the case we could just omit R3 all together, 0 resistance or no resistor is as small as possible. R3 is the current gain resistor portion of the amp. It's value affects the Zo of the amp. The goal of this amp is to establish a particular output impedance. The reasons for this are varied. In Rod Elliots amplifier, he wants a Zo that resembles the Zo of a tube amp. In NPs article, he is using a different circuit to establish a Zo of 50ohms or so. So that it looks like a current mode amplifier or something like that.

I haven't really analyzed what stefano wants to do. Perhaps he is converting Rods schematic to 50 ohms output impedance? You could find out with a sim and a few calculations.
--
Danny

Playing fiew ours listenig and changin the resistors values - with Jordans that are less than 90 db efficient - something told me that the differences are not so big in the end. There can be a sweet spot probably .
I have not analyzed my work too neither simulated .
I searced a more readable octaves contest in the bass guitar region. I guess a possible key here is experimenting with every set up and speaker and enclosure .
I do no count the gain as a possible problem playing with different speaker . Probably I am wrong . Again, I will let you know what happens with the Fostex Fe206e.
I am using the parallel eq network anyhow and wokrs well.

[stefanobilliani]

Originally posted by stefanobilliani

What I see is :
R1 tends to be in parallel with the speaker if R2 tends to zero.
The speaker is not a pure resistor.


Playing fiew ours listenig and changin the resistors values - with Jordans that are less than 90 db efficient - something told me that the differences are not so big in the end. There can be a sweet spot probably .
I have not analyzed my work too neither simulated .
I searced a more readable octaves contest in the bass guitar region. I guess a possible key here is experimenting with every set up and speaker and enclosure .
I do no count the gain as a possible problem playing with different speaker . Probably I am wrong . Again, I will let you know what happens with the Fostex Fe206e.
I am using the parallel eq network anyhow and wokrs well.

Sorry it seem a "telegraph road" reply .

Thanks for yours

Too late for an edit.

[Igla]

Yesterday evening I did some simulations of the circuit with
my CircuitMaker.
The results are not what I've been expecting...whatewer I did
(changing resistors) the output impedance of the circuit
stayed between 7 to 8 ohm from 20 to 20000Hz.
Changing R3 changed Zout just a bit...but bigger R3-smaller
gain.And changing R2 had no significant impact on Zout just on gain...smaller R2-bigger gain.
I was aiming for Zout to be 3 to 4 ohm and a gain of 20 to 22.
Perhaps it was just wrong simulation because did it with small
power op amps like ua 741 and similar.