Amp Camp Amp - ACA

[flg]

Hi Mike, did you see my post #213. Is that correct? For instance; Higher gate Voltage (more -) on the R085 will cause a shallower slope ie; higher Resistance. Lower Vgate, (closer to 0V) will be lots of current change real fast, lower R.

[Michael Rothacher]

I'm late to the party, so forgive me if I'm missing the context, but it sounds like you've got the right idea. Incidentally, I ran some curves for the R085 last summer over a more expanded range. I'll see if I can find them.

[buzzforb]

Sweet. I am behind. Base don what you suggested, the Rds on my oroginal choice is about 10R. Does that sound about right.

[flg]

RdsON is typically, the minmum resistance with a FET, with the gate slammed to 10V and the Vds will go almost to 0.0V. It is a parameter used for switching power supply design where you might have an inductor the FET and say 12V. You turn on the FET really hard and fast and current builds up in the inductor. Then you shut it off and the inductor still pulls current etc. etc. It might be analogous to Vcesat in a Bipolar.
It is not generally a parametor we care about because we operate in the resistive region between off and on.
Then there is Drain Resistance, that could actually be associated with RdsON but??? Then there is your load resistor on the drain???

[buzzforb]

Sorry, incorrect thought made too big a hurry, as usual. 10r drain resistance.

[Loudthud]

Originally Posted by flg

I am confused by these statements
If it is "an active component with an AC negative resistance" how can it's contibution be higher impeadance? It's negative, simply speaking, thats lower.

The negative impedance of the mu follower is in parallel with the 8 ohm load. To see it's effect on gain, remember to gain is roughly RL/(1/Gm).

The parallel combination of 8 ohms in parallel with -16 ohms is 16 ohms. Since RL is higher, the gain is higher. Using the conductace formula for parallel resistors, 1/8 + 1/(-16) = 1/16.

I built one channel using IRF640s, a 2N3904 and MPF102 (Idss=11mA). The only deviation from the ACA1 schematic was a 0.1 ohm resistor in series with the source of Q1. To measure open loop gain I just look at the gate of Q4 with a scope. Gain was 20 (26dB) with the 0.1 in place and 23.1 (27dB) with the 0.1 bypassed.

Looking at the IRF data sheet, the Gm is very non-linear at 1 amp. Isn't there a better part? This amp biases Q1 pretty hot, but that makes it more linear.

[stefanobilliani]

A pic of the proto ACA I did build here .

They are in 2 sided board style , p2p , with IRF044 , works great .

s

[Loudthud]

I looked at an IRF640 on the curve tracer using offset to get smaller steps around the (about) 1 Amp operating point. The second plot is the same part with 0.47 Ohms inserted in the source lead. The gain goes down but the linearity increases. I have no control over the camera shutter speed so some of the traces are fading out.

I replaced the MPF102 I was using with a PN4393. It gave slightly lower distortion numbers. Then I installed a 0.47 Ohm resistor between the Q1 source and ground and ran distortion plots with and without the source resistor. The distortion is higher above 4 Watts because of clipping, but at lower levels the distortion is lower inspite of the fact that the open loop gain is lower. Open loop gain measured 29.3 dB with the 0.47 shorted and 23.2 dB with the 0.47 enabled. DC operating point was adjusted to 10V at idle for each test.

Square wave response of the amp is very well behaved.

[Variac]

Well here we are!

MAKE | My Day at Amp Camp
http://blog.makezine.com/2012/07/20/...arousel-224348
MAKE | MAKE magazine

Loudthud, that circuit is quite recognizable! Even with the monster resistors!

[Loudthud]

Originally Posted by Nelson Pass

1) I wouldn't assume too much, as it depends on everything. There
generally will be a sweet spot where the 2nd nulls and where the total THD
is lowest, and on either side of that the 2nd comes back.

That was in reference to a question about R1, R2, R3, R4 and their ratios. (I'll just refer to them as R1 and R3.) I made a resistor string with a 0.12 ohm on each end and six 0.05 ohm resistors in between. Total resistance 0.54 ohms. The first attachment shows distortion plots. The comments are the two resistor values in order R1, R3. The surprising thing was that the lowest distortion was when R1 was zero! Looking at the analyzer output though, there was a lot of 3rd and higher harmonics. At any other position the distortion is a nice sinewave at 2x the frequency up to about 4 watts. (I didn't try R3 = 0.0 ohms.)

To investigate this I re-arranged the resistors so that all the 0.05 ohmers were on the Q2 source end and plotted three more curves in the second attachment. The third attachment is the same as the second but with three more curves with the 0.47 ohm resistor between the Q1 source and ground.