Zero Feedback Impedance Amplifiers

[Susan-Parker]

Hi Dartagnan,

Sorry for the delay in replying, been away for a couple of days.,

Originally posted by Dartagnan
Thanks for the replay, Susan.

Unfortunately, I can find no more specs on the THD 60 in Tannoy's literature. Yes, the laminations are welded and would require substantial grinding to seperate.

Um, thought that might be the case.

These welded laminations aren't really suited for wide band low distortion audio use as the welding messes with the magnetic properties of the laminations. And the laminations themselves are probably not of the best quality.

Anyhow, try using the speaker windings two transformers, connected in the same manner, with the speaker connected across the ends.

I will try to set up this configuration but regret that it will probably not be until the end of the weekend at the earliest due to pressures of work and personal commitments.

What made me think these transformers might be usable in a Zeus was that the wattage taps seemed to indicate they would lend themselves to split primaries. For example, using the common tap as the negative and the 60W tap as the positive, grounding the 120W tap would yield two 83-ohm primaries. Lower wattage taps would yield higher impedances.

Unfortunately it is likely that the windings will not be matched and whilst you may get some results it isn't going to be optimum.

But you have experimented with these transformers already? When you say they generally don't have symmetrical windings, do you mean for example that the 120W tap doesn't have exactly half the turns of the 60W tap?

Yes.

The issues with using these alternative transformers such as yours and toroids is that these can provide a good method of trying out the basic principles of how the amplifier works - BUT will not necessarily give an optimum configuration or best performance.

This is important to remember because the ultimate performance of the amplifier should not be solely judged on how well a test setup performs.

Sorry about the basic questions, but I don't have means of testing (or much know-how, really. )
Thanks,
Dart

Don't apologize for a sensible question

Best wishes,
Susan.

[Susan-Parker]

Hi Amp_man,

Originally posted by amp_man_1
Hi Susan-Parker MIEE

The amp is now under serious testing condition as follows.

Input Signal 100Hz sinewave
Load 2 Ohms heater element
duration 6Hours Continous testing.

Input Stage Consists of:

Dual Differential input Line Driver Balanced with V swing at 15 Volts

regards
Workhorse Technologies

Thanks for the update

Look forward to your further comments.

Best wishes,
Susan.

[Dartagnan]

I will try to set up this configuration but regret that it will probably not be until the end of the weekend at the earliest due to pressures of work and personal commitments.

Susan, you're very kind to offer! But don't go to any trouble unless you're really curious.

Cheers,

Dart

[Susan-Parker]

Dear All,

When I was originally doing the design work for the amplifier I tried to collect as much information as possible on wide band audio transformers (not much as it turned out) and spoke to a number of people who were knowledgeable in this field.

One Guru told me that the most I could get was the 1/3, 1/3, 1/3 split in power with only a third to the output.

As one respects the handed down wisdom from those who are practiced in the art I accepted this statement and happily included it subsequent information - including a post at the beginning of this thread.

I did always have a bit of a niggle with this but just put this down to my lack of proper mathematical understanding in these matters (and the use of DVMs).

On reflecting about these power requirements I note that I have been happily hitting 20 volts RMS into 8 ohms in my recent testing sessions. BUT the fuse I have in the DC supply line is only 5 amps rating.

Um?

Using the 1/3 split regime with an additional 1.5 amps bias for the Toshiba Lateral FETS and 2.5 amps for the load I should have a total current drain of 9 amps.

Somewhat higher than the T1 20mm 5 amp fuse's rating!

Oops!

My world view crumbles at this point. So much for listening to perceived wisdom.

From my notes using my 2:1 step down transformer configuration the average measured current for 20 volts AC rms into 8 ohms is a mere 2.75 amps.

2.75 amps minus 1.5 amps bias = 1.25 amps.

The 2:1 ratio transformer trades voltage for current, so 1.25 amps x 2 = 2.5 amps which is correct for 8 ohms.

Ergo the amplifier is running at far greater than 1/3 power efficiency into the load.

With a 1 amp total bias (which is what I am figuring on using for the powermesh mosfets) 20 vac RMS into 8 ohms should be 2.25 amps average max (which is over unity as far as current goes compared to a straight solid state amp direct drive output). Peak currents are higher of course.

This is a much more satisfactory state of affairs than the supposed 8.5 amps I "should" be taking.

Um, sorry about that. Humble pie etc.

I will set out to make definitive measurements the next time I have my test rig out as I think I may be getting some regenerative power back from the back EMF of the flux transitions as well. And use a moving coil test meter.

Best wishes,
Susan.

[Susan-Parker]

Dear All,

I have added a schematic to the toroid page:

http://www.susan-parker.co.uk/zeus-toroid-amp-1.htm

BW,
Susan.

[darkmoebius]

Hi Susan,

I'm in an ordering mood these days, but the Sowter 8160 is out of my price range and the 6-8 week wait is beyond my attention span.

The LL1922 sells for $80 ea. US and K & K Audio has them in stock. But, the big question is "Are they as good as the Sowter's?" A 30% savings is fine and dandy, unless I would end up ordering the Sowters anyway.

Did you ever get any new info from Lundahl or K & K?

Also, Jensen Transformers in Los Angeles is pretty famous for their professional audio transformers. Could you please take a peek at their products and see if any meet the needs for Zeus's input and output transfo's?

They are located only 30 minutes from my house (w/o traffic).

[darkmoebius]

I was specifically referring to some of their microphone input transformers like the $73 JT-115K-E:

MICROPHONE INPUT TRANSFORMER
1:10 STEP-UP FOR HIGH IMPEDANCE AMPLIFIERS
* Ideal for FET or vacuum tube input amplifiers
* Wide bandwidth: -3 dB at 2.5 Hz and 90 kHz
* 20 dB of voltage gain with Noise Figure of only 1.5 dB
* Input impedance of 1.4 kOhm for loading loss of 0.9 dB
* High common-mode rejection: 110 dB at 60 Hz
This transformer is designed for highest practical step-up ratio. Its secondary source impedance makes it ideal for use with low noise FET or vacuum tube input amplifiers. The primary is fully balanced and its leads may be reversed to invert polarity, if required. A 30 dB magnetic shield package is standard.

or

$80 JT-110K-HPC:

MICROPHONE INPUT TRANSFORMER
1:8 STEP-UP FOR HIGH IMPEDANCE AMPLIFIERS
* Ideal for a variety of IC or discrete input amplifiers
* Wide bandwidth: -3 dB at 1.0 Hz and 100 kHz
* 18 dB of voltage gain with Noise Figure of 2.4 dB
* Input impedance of 1.4 kOhm for loading loss of 0.9 dB
* High common-mode rejection: 115 dB at 60 Hz
This transformer has a turns ratio which optimizes the noise performance of many IC amplifiers. Its secondary source impedance varies less than 10% from 20 Hz to 20 kHz. The primary is fully balanced and its terminals may be reversed to invert polarity, if required. A 30 dB magnetic shield package is standard.

The biggest problem is no center tap, but they will supposedly do simple mods to existing transformer designs for a modest fee.

[Terry Demol]

Originally posted by darkmoebius
I was specifically referring to some of their microphone input transformers like the $73 JT-115K-E:
or
$80 JT-110K-HPC:

The biggest problem is no center tap, but they will supposedly do simple mods to existing transformer designs for a modest fee.

Folks, you might also want to check out Lundahls LL7905.
This is a very high level mic IP tranny with 1+1+1+1 : 5.6+5.6
So with parallel prim to series sec you can get 1 : 11.2 with a
CT on secondary.
Data sheet specs max sec voltage at +37dBu (54.5V rms) so it
looks like it may work fairly well at the levels required for Susans
amp.

Here is data sheet:

http://www.lundahl.se/pdfs/datash/7905.pdf

Please note that in general, for mic IP transformers to work best
they are specified for a source impedance of 150ohms for
optimum transient response. This is particularly true with
Jensens as they have an almost perfect Bessell HF roll off when
used as such.
Further to this point, the output capacitance of MOSFETs that
this transformer drives may have to be compenasted for.

The best way to do this is to feed a 10kHz square wave and
adjust a snubber at transformer output for least amount of
ringing.

Cheers,

Terry

[Susan-Parker]

Hi Darkmoebius,

Been looking into these posts.

Originally posted by darkmoebius
Hi Susan,

I'm in an ordering mood these days, but the Sowter 8160 is out of my price range and the 6-8 week wait is beyond my attention span.

Ordering is good

The LL1922 sells for $80 ea. US and K & K Audio has them in stock. But, the big question is "Are they as good as the Sowter's?" A 30% savings is fine and dandy, unless I would end up ordering the Sowters anyway.

Um. I haven't used any Lundahl's but have heard good things about them.

However note the bandwidth of the LL1922 in 1:8 is 10 Hz to 30 kHz +/- 1.0 dB.

I don't know what the distortion levels are without comparing the two side by side on the bench but a 15 VA toroid has a bandwidth that is not a lot lower than this.

Did you ever get any new info from Lundahl or K & K?

Only what's up on their website.

Also, Jensen Transformers in Los Angeles is pretty famous for their professional audio transformers. Could you please take a peek at their products and see if any meet the needs for Zeus's input and output transfo's?

They are located only 30 minutes from my house (w/o traffic).

Would be very convenient.

I have had a good look and I can't see anything really suitable.

Some of the line stuff is half way there but nothing with a big enough core cross section for the step up. You could use a pair if you had some in a box but it wouldn't be cheaper than the Sowter solution which was specifically designed for this application.

The Lundahl's look better but their bandwidth is much lower.

I would note from the specs for the Sony DVP-NS955 that...

Frequency Response: Standard CD is 2 Hz to 20kHz(±0.5ｄB)

Frequency Response: Super Audio CD is 2Hz to 100kHz (50kHz: -3dB ±1dB)

... so the Toshiba Linear FETs don't hack it either.

Best wishes,
Susan.

[mikelm]

.
A friend and I were lucky enough just over a week ago to get to audition Susan's Zeus amplifier along with the rest of her system and I have to say we were pretty impressed with what we heard ...

Susan explained that the system did not have any esoteric, hi end, expensive components - just stuff that she happened to have at the time when she built the system.

Of course, it is always difficult to know what is contributing what in an unfamiliar system but the overall sound cannot be better that the weakest link and overall sound was VERY good.

Over the last 5 years or so I have listened to quite a few home systems of different friends. Many DIY, some bought. I have also auditioned some very expensive stuff at dealers and attended a couple of hi-fi shows.

Susan's system seemed to combine good balance of detail, clarity and dynamics along with a lovely fluidity that made it really easy to listen to. I regard this combination as very rare. The sound was also quite big and effortless.

I have heard beautiful sounding systems that sounded somewhat compressed. I have heard very detailed systems that sound somehow too analytical or dry or rough. Then there are systems that sound nice and soft but lack detail.

For me, Susan's system had a special combination of fluidity, detail and dynamics that I regard as the best I have heard to date. ( not bad considereing some of the systems mentioned above cost £50,000+ ).

I will be ordering my transformers soon and I am really looking forward to hearing a zeus amp in my system ( although realistically I think it will be mid January before this is realised ).

mike