OK, it´s not a PASS project but...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
...I guess SE mosfet amps belongs here anyway.

A while ago I bough some 112mH 2A chokes (with airgaps, before anyone comments that).
Seems like a perfect source/drain load for a smallish SE amp.

The easiest thing to build would be a source follower with a separate gain stage, but that feels a little...boring:D

What do you guys think about this one:
 

Attachments

  • se12w.jpg
    se12w.jpg
    14.8 KB · Views: 1,805
Excellent. How does it sound?

Thanks!

Er, I don´t know yet.
I came up with this circuit just hours ago, inspired by an amp designed by my my fellow countryman Circlomanen.
I basically chopped his circlotron circuit in half and replaced the missing half with the choke.
:)

I´ll start building a prototype as soon as I can, I expect my power fets to arrive withing a couple of days.
 
Yes, the low driver current concerns me a bit but it´s a compromise.
Higher current would lead to very low resistor values in the feedback network (and thus a seriously big capacitor)

At least it´s only a single mosfet that is to be driven, not 2 or 4 or...
 
Today I finally decided to build a prototype of this circuit.
Took me about an hour.
Results? Probably a great midrange amp, cuz´it has neither bass or treble:smash:

EDIT: Forget what I wrote about the frequency respone, I had made a wiring mistake in the feedback loop which messed things up quite a bit.
Now it works muuuch better!!
I´m listening to a mono channel right now, and it´s definitely not bad!
 
Perhaps there is something I'm not noticing.

The way you have your feedback applied is actually giving you positive feedback. It seems that it would be negative feedback because the output signal is out of phase with the source of the first stage. But that will actually increase the gain of the amp in this circumstance, which is positive feedback, not negative. I only noticed because I designed a tube amp that had the feedback the same way, only to find out later what was really going on.
 
I must disagree on both points.
If the feedback was positive the amp would probably suffer from extremly high gain, very high output impedance and most likely severe oscillations.
This amp has a voltage gain of ~8, "tight" bass response and no oscillations that can be detected on my ´scope.

Regarding the choke, yes it would be nice with 10x the inductance but it would be practically impossible.
Some guy here made a fullrange amp with a pair of 85mH chokes some years ago, so I can´t see why it wouldn´t work.
But as you say, higher inductance would of course be a good thing.
 
Your right regarding the inductor. I just realized impedance of most speakers is higher at lower frequencies due to resonance. So that inductor should work. The inductor will be hogging current at low frequencies but the speaker will require less current at low frequencies due, to the higher impedance.

Thanks for pointing out the output stage is a P-Channel. Your right, I'm so used to tubes, I forget solid state has devices that conduct the other way.:headbash: Sorry, I'm a dumbass.
 
The reactance of an 112mH choke is 14 ohms at 20Hz, a bit more would be nice of course but size, weight and DC resistance (in particular, since it affects the operation point at DC) are issues that cannot be neglected.

Tubes...don´t mention them!
I was up all night (literally) trying to make an ECC86 capable of driving a simple Mosfet buffer. No matter how I tried with CCS´s, buffers and µ-followers they never made it to 20kHz.
 
Less distortion is always (?) attractive.
I guess the only way to find out is to try, just as you say.
Will do that some day, right now I´m busy assembling a couple of TDA2005-based kits for an experimental biamp project.
For some reason these kits don´t feel as solid as my choke-amp prototype:D
 
I just ordered some part for a 4 channel version of this amp, and this time I´m going to use BF245C instead of BF245B.
This should allow (with some changes in resistor values) higher current through the input stage and thus better capability to deliver whatever the Mosfets needs at high frequencies.
Speaking of Mosfets, I ordered some IRF9530 instead of 9640.
The lower power and voltage rating means lower input capacitance, yes?

I´ve also been shopping some alu sheets at the hardware store for the chassis, it´s time to get past the prototype stage.
 
refining things a bit

Since I´m intending to build four channels of this amp for my biamped speakers (FE167E´s with stereo subs) I´d like to adjust the bass and HF amps for their different purposes: The subwoofers requiers more power, higher damping factor and more than twice the gain compared to the HF amps.

Max output power is limited to about 10-15W by the chokes, they can´t handle more than 2A DC. 10W is though enough, and I prefer 10W SE class A over 200W from some cheap commercial AB amp...
For the FE167E´s I only need a few watts, 4W would be perfectly adequate.

Regarding the damping factor I believe it can be altered by changing the open loop gain, right?
Higher open loop gain = more feedback = higher DF, as I understand it.
Replacing the Jfet´s drain resistor (500R) with a CCS should increase the O-L gain quite a bit I figure.
Same thing if I decrease the Mosfets source resistanse (Rs), since Rs inserts degenerative (current) feedback in the output stage.
Closed loop gain can be adjusted by trimming the 33R resistor, if we choose 11R the gain should be about x10 which would be just right for my needs.

For the HF amps I want quite the opposite: low DF and about x4-5 gain. A plain resistor as a load for the Jfet and a quite high source resistance for the mosfet. 100/33R in the feedback loop should set the gain to x4.

The one with the cascoded input stage is just an idea to improve the HF performance, though I still don´t know how/if cascoding affects the gain of the input stage.
I believe that with the same bottom Jfet and load resistor the gain should be the same as for a non-cascoded stage, but I´m not sure.
 

Attachments

  • smallse.jpg
    smallse.jpg
    20 KB · Views: 386
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.