New Dynamp

[ronybc]

Hey all...

See Dynamp 2.6
http://www.ronybc.8k.com/mosfet.htm

I didn't construct this one yet. LTSpice simulations says its OK. Please share your opinion about this amplifier.

[darkfenriz]

i like it- resistive only feedback may sound fine
honestly i also projected(and am currently constructing) amp with feedback like yours and an op-amp,but:
1.you have no dc-feedback
2.adjusting pots to no dc at output combined with desired biasing may take some time (and nerves)
3.with not exactly that rail voltage (which happens very often when driving amp to near max power) bias will change
give small (like 0.33) gate resitors for stable biasing point

cheers

P.Si like your clever local feedback on r13/14 too

[Eva]

I have not experiences this exact topology with local feedback, but I have experienced similar topologies with only global feedback and bipolar output transistors instead of mosfets

The capacitive coupling of the op-amp output to the driver transistors causes very poor PSRR and unstable biasing

Also, the circuit has to be operated in class B to avoid strong cross-conduction problems when playing music

As the author states in his webpage, this is just a cheap amplifier circuit focused on low part count and simplicity

[darkfenriz]

Originally posted by Eva

Also, the circuit has to be operated in class B to avoid strong cross-conduction problems when playing music

Please explain me Eva why you think so??

[jacco vermeulen]

This looks like a stripped version of some end 80's designs, for DIY the part count seems silly to me.

A number of companies in Europe still produce opamp/mosfet combo's, the ones in Germany were the most dominant.
AVM in germany still produces a version, even uses the 630/9630 for it i think.

As these are only 75/80 watt mosfets and can be had for less than $ 0.75 its a waste of expensive heatsinks, transformers and capacitors to use only 2.

Use 6 to 8 of them and swap the 547/557's for BD139/140 for a few cents more and you have a powerfull output for some $ 4.

Secondly, why bother with so few gain stages as the opamp on the entrance has so many ?

It has no dc adjustment, and personally those 47uF's give me the creeps.
Use the TL07* for active DC offsetting , get rid of those caps and add a few gain stages with current regulation.

In my view someone is better of copying designs like the Sphinx project twelve, AVM M's, i can name a couple more.

As the author says: I have no scope !

[Eva]

As the instantaneous biasing depends on the instantaneous rail voltage of both rails and the instantaneous voltage across the coupling capacitors, and these variables fluctuate wildly when playing music...

This bias also goes wild when playing music, particularly when several channels share the same power supply

A couple of cascode transistors instead of the coupling capacitors will solve the problem, but this implies circuit redesign and increased part count

[jacco vermeulen]

Originally posted by Eva
As the instantaneous biasing depends on the instantaneous rail voltage of both rails and the instantaneous voltage across the coupling capacitors, and these variables fluctuate wildly when playing music...

This bias also goes wild when playing music, particularly when several channels share the same power supply

A couple of cascode transistors instead of the coupling capacitors will solve the problem, but this implies circuit redesign and increased part count

hear hear !

[john curl]

See Marantz 16 power amp from 1970.

[john curl]

See Marantz 16 power amp from 1970.

[ronybc]

Drops the weapon...

I understand the problems. The circuit have no other scope than total redesign..! Thanks for your reply.