Zero Feedback Impedance Amplifiers

[Graham Maynard]

Hi Susan,

What a column !

Would reverse biased shunting diodes from each output source to the drain/psu rail prevent possible back EMF induced Mosfet failure ?

Has anyone tried per-half damping resistors yet ?

Cheers ......... Graham.

[Circlomanen]

Hi Susan!
Would it be possible to use 2 30 VA input transformers and connect the 230 volt windings in series to get a midpoint and a greater stepup ratio??

[Susan-Parker]

Hi Circlomanen,

Originally posted by Circlomanen
Hi Susan!
Would it be possible to use 2 30 VA input transformers and connect the 230 volt windings in series to get a midpoint and a greater step up ratio??

Yes, using a pair of same type (and preferably same build batch) you can put the (24 volt?) inputs in parallel and the 230 volt outputs in series.

What you can't do is drive them coupled push-pull as an interstage transformer as the primaries are on separate cores, unless you wire them up so that each arm goes through both transformers.

E.g. 2 off transformers T1a and T1b with windings 12Va + 12Vb on each.

Q1 = (12va) T1a + (12va) T1b
Q2 = (12vb) T1a + (12vb) T1b

This should work.

A quick sketch...



Larger version http://www.susan-parker.co.uk/zeus-2stage-toroid-v1.gif

At the end of the day using specific transformers will give better frequency response and matching and lower distortion. But this will cost a lot more and the toroids seem to be more than adequate to get one going.

N.B. The 15VA toroid that I have has a slightly higher top end frequency response, although it is only a few kilohertz.

I have been working on a maximum of a 1:10+10 step up for my input transformers.

You probably also want to use two matched Rterm's (one for each secondary) rather than just one across the mosfets.

Note: the Rterm values shown in the schematic are nominal, and could be much lower in practice.

My two stage amplifier using the 15VA 12+12:115:115 toroid as the interstage does work from a line level input with a 1:2+2 input transformer and although this small step-up doesn't swing enough to drive much power it does provide a 600 ohm input.

Whereas in the single stage the 15VA toroid is well under 100 ohms input impedance (59 ohms series, 15 ohms parallel) and needs headphone type drive levels.

Hope this helps.

Best wishes,
Susan.

[Susan-Parker]

Hi Graham,

Hope you are keeping well.

Originally posted by Graham Maynard
Hi Susan,

What a column !

Would reverse biased shunting diodes from each output source to the drain/psu rail prevent possible back EMF induced Mosfet failure ?

Has anyone tried per-half damping resistors yet ?

Cheers ......... Graham.

Thanks

Not that I am counting you understand but I did "just" happen to notice that we are now number 5 in solid state for both replies and views

Yes, that should work - although for the circuit with the pre-reg protection circuit they would go to the main power rail. The idea of the pre-reg is to be able to turn off the power very quickly - as the current is still rising on the first cycle - so we don't want storage caps at this point.

I probably haven't notice this particular problem as I am using a transformer and when I break the load the winding goes open circuit and thus can't kick current back into the primaries. Or I have just been lucky!

If you mean the 33Rs across the output transformer/inductor primary windings, no not yet although I haven't forgotten. Waiting to do that when I have my testing setup sorted out so I can check distortion effects at the same time.

Work stuff is keeping me distracted at the moment

Best wishes,
Susan.

[Netlist]

Originally posted by Susan-Parker
Not that I am counting you understand but I did "just" happen to notice that we are now number 5 in solid state for both replies and views

I suspect that your 'special design' keyboards have an auto-reload function build in.

/Hugo

[Workhorse]

Originally posted by bobo1on1



Ehm what are back to back zeners? I did put in a 12 volt zener between the gate and the source, that should make sure the Vgs does not exceed 12 volts, or am I wrong?

Hi BOBO
Back to Back Zeners refers to the configuration which involves connecting 2 zeners in series but with same polarity on connecting middle tag.
This configuration eliminates both negative and positive overdrive of gate voltage.
hope this helps

[peranders]

Originally posted by Netlist

I suspect that your 'special design' keyboards have an auto-reload function build in.

/Hugo

.... Hugo, naughty boy

I have the real answer: Susan has chosen two magical trigger words, "zero feedback"! This creates interest regardless what's all about. "Class A" is also good

[darkmoebius]

For those who newbies (like me) who already have RCA or "single-ended" preamps, cd players, etc and still want to try Susan's Zeus amp without a major investment to convert to balanced/differential inputs - digi01 has started an excellent thread based around this single chip design that only uses 2 resistors and capacitors. Total parts cost including the regulators and power supply should be less than $10. The best part is that it may be possible to simply place this tiny SE to Differential converter inside the Zeus amp at it's inputs.

The DRV134 (and DRV135 in SO-8 package) converts a single-ended, ground-referenced input to a floating differential output with +6dB gain (G = 2).

(seems DiyAudio is not displaying pictures right now, so click link)

[Circlomanen]

I have two 1000VA transformers to strip down and rewind as centertapped chokes.
That should mean 250 watts without global feedback and with a very very nice sound.... My neighbours will love me, or DIE!!!!!!!!!!!!

[mikelm]

I can't help thinking that we may be losing the plot a bit resorting to op amps to this drive amp.

Even with my limited experience I have found that finding a an op amp that sounds as good as a good discrete design is not easy.

From the reports so far about this amp I think this amps deserves a better solution.

I can't help thinking that if we have to introduce another perhaps inferia stage to drive this cct, Hugh Dean's idea to have a valve stage instead of the i/p transformers seems increasingly attractive.

Any ideas about a good valve cct for this ?
I don't think it has to be balanced, does it ? The input transformer will do the convertion here I think.

GOOD NEWS I have just discovered that I have have four good sized toroid transformers and on inspection they all seem to have bifilar secondaries...

What do we think about the fact that the primaries are between the secondaries ans the core ?

I also have 16 IRFP044N's - will these be better IRF540's or IRF 530n's ?

mike