Zero Feedback Impedance Amplifiers

Re: Not to threadjack again, but....

Hi darkmoebius,

Thank you for your post.

darkmoebius said:
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.

Had a look, couldn't see my name anywhere :(

Okay, back to the regular programming :)

Please remember that the drive to the Zeus amplifier is not "just" 600 ohms, or 150 ohms, but that the mosfet gate load is reflected back through the transformers.

As such not all line driving stages are suitable even though they might quite happily drive a standard 600 ohm resistive impedance.

I don't want to put people off, but it is something that one should be aware of as it isn't guaranteed that any old op-amp can walk this particular walk.

Hence the OP275s which do work well, and the TI part that drives that as well as much lower load impedances quite happily.

Also if there is anything specifically to do with the Zeus amplifier then please keep it here as otherwise it will get lost very rapidly (I don't imagine the DIYaudio list people, wonderful as they are, will give us our own section to keep things organized).

Some of us don't get to skim other threads much, myself included, as I am spending 20 to 40 hours a week just supporting this thread (someone asks a question - I go off and research it, maybe order some parts from RS - come back hopefully with an answer or a suggestion).

Re the DRV134 I haven't tried this particular part, so can't comment on it's actual use. However I would fit series resistors of around 22 ohms on each output. And use very good decoupling caps right up close to the pins.

In case anyone was wondering I did test out the SSM2142 and regretfully (as I have a full tube of the things) that gets the thumbs down as far as transformer input stage driving goes.

Look forward to receiving any feedback on how the DRV134 part works out.

Best wishes,
Susan.
 
Susan-Parker said:
Had a look, couldn't see my name anywhere :(

Aaah, but you're always in our thoughts :angel:

Please remember that the drive to the Zeus amplifier is not "just" 600 ohms, or 150 ohms, but that the mosfet gate load is reflected back through the transformers.

As such not all line driving stages are suitable even though they might quite happily drive a standard 600 ohm resistive impedance.

Funny that you should mention this (again) because I took the liberty of posting on the forum of Lundahl's USA distributor, K&K Audio, asking about the input imped. & FR of the LL1689 or if Lundahl has something more appropiate . I supplied your schematic, the Sowter #8160 specs and a link to your website.

Kevin Carter(owner) responded:

I could find no data for the Sowter 8160 on the Sowter webpage. While the ratios are similar, the inductance data for the 8160 are very different than those for the LL1689/PP. I'm not sure what to make of that, frankly. MOSFETs are very high impedance devices, so presumably the input impedance is set by the reflected terminating resistance. I am curious to know what kind of a circuit you intend to use to drive the input transformer. That's a pretty challenging input impedance.

Kevin Carter
K&K Audio
www.kandkaudio.com

I'll let him know R-Term is 120 KΩ and see what he thinks.

Re the DRV134 I haven't tried this particular part, so can't comment on it's actual use. However I would fit series resistors of around 22 ohms on each output. And use very good decoupling caps right up close to the pins.

Look forward to receiving any feedback on how the DRV134 part works out.

Surprise, surprise!!! The UPS man dropped off four DRV134 this morning. I need to assemble a nice, quiet, 15V supply for a mockup. I also have quads of AD815 & AD826's on the way for testing. We'll see how they stand up to the stress.

I need to tackle the Zeus input and output transformer issue this weekend. I've got the PS transformers and matched Hitatchi 23j49 MOSFETS covered from Bear, although I won't be using those until the circuit is up and running smoothly for a few weeks on much cheaper IRFP150N's.
 
To Susan

When I looked at the inductance of my toroidal transformers, I used an HP4194A analyzer in L/Z mode, swept from 100Hz to 20kHz. Excitation voltage was around 400mV. The inductance on either side of the secondary CT was about 760 uH (matched to better than 1%!). The inductance steadily fell with frequency until resonance at about 2kHz - not very pretty, but I suspec, typical. Tomorrow I'll look at a 28-0-28V transformer (I have scads of them) if I can remember to slip it in my briefcase tonight. Just from the higher output voltage, it should have higher inductance (I hope). I suspect that the resonance situation will be no better, though.
 
Hi folks -

I just realized that I hadn't gotten back to this thread with news about Warren Gilson, the patent holder referred to earlier in the thread.

I had posted that he was listed in the local phone book, with an address just blocks from my home. I had copied the schematic from the first post in this thread and I was going to stop in and show it to him.

Alas, the phone book that I had here in the office was several years old, and when I checked the current one, Mr. Gilson was no longer listed. His prior address had been the location of either a nursing home or other retirement facility. I fear that Mr. Gilson is perhaps no longer with us.

- Pat
 
Anybody interested in a Toroidal kit?

Torid Corp of Maryland offers them in 5 standard configurations(all 117V primary). They supply everything needed to wind the transformer.

Here's a picture of the winding jig:
An externally hosted image should be here but it was not working when we last tested it.


The two possibly relevent to this design are:
80VA, 60W rating, 0.13V per turn, $27 US
200VA, 140W, 0.23V per turn, $38

Each kit contains:

* A toroidal core of grain oriented steel
* Mylar taped for insulation of secondary windings put on by the customer
* Built-in thermal fuse rated at 130 ¹ C
* Manual with step-by-step instructions on how to design and wind the secondary windings.
* Metal washer and 2 rubber pads for mounting the transformer
* Guide to calculate proper wire size and length required to generate your calculated output data.

If someone could help figure out the winding count for our purposes, I'd be willing to give it a try.

For those in the USA, mabe this is an option:
Torid Corp of Maryland
 
Hi wrenchone ,

wrenchone said:
To Susan

When I looked at the inductance of my toroidal transformers, I used an HP4194A analyzer in L/Z mode, swept from 100Hz to 20kHz. Excitation voltage was around 400mV. The inductance on either side of the secondary CT was about 760 uH (matched to better than 1%!). The inductance steadily fell with frequency until resonance at about 2kHz - not very pretty, but I suspec, typical. Tomorrow I'll look at a 28-0-28V transformer (I have scads of them) if I can remember to slip it in my briefcase tonight. Just from the higher output voltage, it should have higher inductance (I hope). I suspect that the resonance situation will be no better, though.

Thanks for the further info.

Probably best to put on a termination resistor with a 10k twiddle pot and do a quick sine sweep and adjust to flaten out the kick up.

The mains toroids as being used are not going to be the best one can do but hopefully good enough to try things out and get a bit of hands on.

You should be able to run a pair of 28-0-28 transformers together to give you a balanced center tap output on the 115 + 115 windings.

Best wishes,
Susan.
 
Hi darkmoebius

darkmoebius said:


...

I'll let him know R-Term is 120 K& and see what he thinks.


Okay, will be interesting :)

Surprise, surprise!!! The UPS man dropped off four DRV134 this morning. I need to assemble a nice, quiet, 15V supply for a mockup. I also have quads of AD815 & AD826's on the way for testing. We'll see how they stand up to the stress.

Always a nice surprise - the TI chips arrived here in England before I go the email notification that my sample order had been accepted and a shipping number!

I need to tackle the Zeus input and output transformer issue this weekend. I've got the PS transformers and matched Hitatchi 23j49 MOSFETS covered from Bear, although I won't be using those until the circuit is up and running smoothly for a few weeks on much cheaper IRFP150N's.

Looking forwad to hearing of your progress (as I do for everyone who is having a go at this amplifier).

Prudent to use the IRFP150N to start with. These are the parts I have been using so I do know that one can get good results with them.

Don't forget the fuse after the smoothing caps :)

Best wishes,
Susan.
 
Re: Anybody interested in a Toroidal kit?

Hi darkmoebius,

Thank you for the post...

darkmoebius said:
Torid Corp of Maryland offers them in 5 standard configurations(all 117V primary). They supply everything needed to wind the transformer.

Here's a picture of the winding jig:
An externally hosted image should be here but it was not working when we last tested it.


... and this information.

The two possibly relevent to this design are:
80VA, 60W rating, 0.13V per turn, $27 US
200VA, 140W, 0.23V per turn, $38

Also the 400 VA, 260W $54.00

For power ratings (i.e. flux density) one needs to derate for both frequency and also to keep lower down the flux squiggle as mains transformers are usually rated at 80% to 85% of saturation and the toroids curls over faster than EIs with their less contained magnetic circuit.

So for output transformers the 200 VA (at 60 Hz) is for the sort of powers that I use i.e. 35 watts (here in the UK we have 50 Hz so the equivalent VA rating is about 20% higher).

The 400 VA core is good for powers up to 75 watts with a bit of a decrease in margins.

Each kit contains:

* A toroidal core of grain oriented steel
* Mylar taped for insulation of secondary windings put on by the customer
* Built-in thermal fuse rated at 130 ¹ C
* Manual with step-by-step instructions on how to design and wind the secondary windings.
* Metal washer and 2 rubber pads for mounting the transformer
* Guide to calculate proper wire size and length required to generate your calculated output data.

If someone could help figure out the winding count for our purposes, I'd be willing to give it a try.

For the output transformer one doesn't need the mains winding so maybe the company will supply the cores unwound, otherwise they should be removed.

Quad-filar wind 0.8 mm enameled magnet wire (200 VA core) or 1.0 mm wire (400 VA core) with 120 to 150 turns.


For those in the USA, maybe this is an option:
Torid Corp of Maryland

Looks good :)

Many thanks for looking into this.

Best wishes,
Susan.
 
A little more transformer info....

I checked with Sowter for the US price(including VAT & shipping) on the 8160 input transformer - 124 UKP or $232 US.

Not cheap! But, hey, people drop a hell of a lot more than that on high-end interconnects. Not to mention NOS output tubes.

I'm still trying to track down an E-I kit, but there doesn't seem to be a lot of companies in the US offering such. Lots of Google hits for the UK, though.

As for the toroid kit, is there any special core type to request? Something the equivilent of M6 Othosil?
 
Re: A little more transformer info....

Hi darkmoebius,

darkmoebius said:
I checked with Sowter for the US price(including VAT & shipping) on the 8160 input transformer - 124 UKP or $232 US.

Not cheap! But, hey, people drop a hell of a lot more than that on high-end interconnects. Not to mention NOS output tubes.


I assume that's the unit price for one or two off?

This is why I am happy for people to experiment with small mains toroids first, and get happy with the how the amplifier works. Then spend the extra.

I'm still trying to track down an E-I kit, but there doesn't seem to be a lot of companies in the US offering such. Lots of Google hits for the UK, though.

Don't forget you can buy the laminations as required from Thomas & Skinner, and in a pinch make up the bobbins from heavy duty brown paper and varnish.

The point of the kits is that you get the mains windings pre-wound and insulated to current safety legislation. Plus all the hardware but not the wire you will want to wind as that has to be bought separately. Give that we will strip off the mains windings the kits are okay if nothing else is available, but not ideal.

Zeus EI-120 Output Transformer (100 Watts).

After a good deal of thought I have specified the bobbin and windings for the output transformer to Sowter last Thursday and expect a price back soon. They have told me that they will sell the wound bobbin as an item in it's own right or in a fully made up transformer.

This should help people in the US to keep shipping costs and import duty down as you can buy the laminations as required from Thomas & Skinner (I will be putting this information up on my website later on this weekend).

The Sowter output transformer is to have a two chamber bobbin giving two sets of 1.0 mm quad filar wound windings to give eight identical windings in total for maximum flexibility.

The transformer can be wired as a center tapped inductor with the loudspeaker across the ends, or across intermediate taps for better matching with lower impedance loads.

Or the transformer can be wired as per my normal usage as a proper transformer with four identical isolated output windings which can be wired up in series, parallel-series, or parallel depending on the speaker load.

As for the toroid kit, is there any special core type to request? Something the equivilent of M6 Othosil?

You want to have the Class A cores which are wound from a single strip, not the Class B which are made up from segments (different manufactures may use different names for this).

This company sells cores

http://www.alphacore.com

and their prices look reasonable. See:

http://www.alphacore.com/stcores.htm

for standard toroid cores in "off the peg" sizes.

Something to look at, especially because we are hand winding, are their new toroidal transformer O-cores which have significant advantages in not having any sharp corners and much easier for us to keep a good winding tension.

http://www.alphacore.com/ocore.htm

The OA-400 rated at 480-960 VA is $15.56

ID = 62 mm, 2.44"
OD = 134 mm, 5.28"
Dia = 37.5mm, 1.48"
Weight = 2.53 kg, 5.57 lbs.

Which is the one I would go for if you are wanting to push 100 watts.

They also do the mounting hardware

http://www.alphacore.com/webmount.htm

and accessory items.

Hope this helps.

Best wishes,
Susan.
 
Member
Joined 2004
Paid Member
i just ordered several cores and mounting hardware from alphacore the other day actually, and i must say that they are VERY reasonable prices. the phone sales staff is also friendly and helpful. stuff ships out same day too (the "in stock" items, anyhow). just thought i would chime in, as i came damn close to unwinding some avel toroids i have on hand to rewind them for output use. then i thought of how much i paid for them, and compared to the prices of alphacore...it makes sense to keep my avel toroids for a power supply! after all...it's the winding you're paying for! :D
 
Hi enochRoot,

enochRoot said:
i just ordered several cores and mounting hardware from alphacore the other day actually, and i must say that they are VERY reasonable prices. the phone sales staff is also friendly and helpful. stuff ships out same day too (the "in stock" items, anyhow). ...

Many thanks for your post and info on your experience with Alphacore.

Best wishes,
Susan.
 
Re: Re: A little more transformer info....

First off, thank you Susan for all the time and effort you are putting into this project, I'm having a lot of fun learning along the way. Judging from the thread views, it seems a lot of other people are too.

Susan-Parker said:
I assume that's the unit price for one or two off? This is why I am happy for people to experiment with small mains toroids first, and get happy with the how the amplifier works. Then spend the extra.

Yep, that price was for a single pair of input transformers shipped "across the pond". A little scary price at first blush, but if the cheaper experimental transfo's work well for a "proof of concept" amp, it's really not too much for finished amp if you compare it with a nice set of Magnequest, Peerless, Tango, or Tamura tansfos.

Don't forget you can buy the laminations as required from Thomas & Skinner...Give that we will strip off the mains windings the kits are okay if nothing else is available, but not ideal.

All the better(and cheaper) then, no need for kits :D

BTW, which transfos have the better performance - toroid or E-I?

This company sells cores http://www.alphacore.com and their prices look reasonable.
Something to look at, especially because we are hand winding, are their new toroidal transformer O-cores which have significant advantages in not having any sharp corners and much easier for us to keep a good winding tension.

Last night, I was trying to picture how to wind wire tightly around the edges of a standard toroid without damaging the insulation.

The OA-400 rated at 480-960 VA is $15.56

My speakers are 97dB/8 ohm, so I don't need 100w - just 20-35W max. Would a 200-250VA toroid do fine for that level output?.

Anyway, at the bargain basement price of $11 each, I'm just going to go ahead and order a pair first thing this morning(it's 2 am here).
 
Re: Re: Re: A little more transformer info....

Hi darkmoebius,

Thank you for your post...

darkmoebius said:
First off, thank you Susan for all the time and effort you are putting into this project, I'm having a lot of fun learning along the way. Judging from the thread views, it seems a lot of other people are too.


... and for your kind words.

I am very pleased that you are finding this fun - our hobbies should be for enjoyment and pleasure after all :)

It has been an interesting experience for me. I work mostly from home and sort of part time which by it's nature normally leads to quite an insular life.

This last month has been a bit different!

Yep, that price was for a single pair of input transformers shipped "across the pond". A little scary price at first blush, but if the cheaper experimental transfo's work well for a "proof of concept" amp, it's really not too much for finished amp if you compare it with a nice set of Magnequest, Peerless, Tango, or Tamura tansfos.

Okay just to confirm that's 124 UKP or $232 US for two transformers and shipping?

I do realize that compared to the cost of a pair of 15 VA toroid mains transformers for the same task (which would cost me about 24 UKP from RS (not the cheapest place to get stuff)) they do seem a little expensive.

Which is why I have been happy to see the satisfactory results one can get from these toroids to get to grips with the amp concept first. One does have to be a little careful though as not all of these small toroids may be suitable and it would be a pity if a lack of performance here was then attributed to the basic design.

BTW, which transfos have the better performance - toroid or E-I?

They both have advantages.

Toroids have a more contained magnetic field and are thus more efficient, but have a sharper knee to their saturation characteristics.

EI's have a loser field but with their slower saturation don't "crash" in the same way.

EI's are also more flexible in the choice of lamination materials and build up strategies. Note: If given a choice of lamination thicknesses for a given stack size, go for the thinner ones. A little more fiddly to assemble but will have better theoretical performance.

Using unity and near unity ratio multi-filar windings the EI transformers can go up to hundreds of kHz bandwidth (which is normally very difficult to achieve with valve type ratios which require very complex sectioning and build up).

I personally prefer the flexibility of the EIs for output transformers which I can build up myself (after much practice I would admit) in quite short order. I have a small lathe which I now mount the bobbin on with a 4 jaw chuck and turn the spindle by hand which makes for a slightly easier life.

However from the results we have been able to get from pressing mains toroids into service, and as long as one keeps a good margin in hand flux wise, there is no reason why toroids shouldn't be fine.

The O-cores look particularly attractive, I may even order a couple for myself :)

Last night, I was trying to picture how to wind wire tightly around the edges of a standard toroid without damaging the insulation.

With care!

Make sure the core is either resin dipped or first wound with some mylar or other tape that the wire won't cut into.

One won't be able to get the same tension as a machine wound part, but that shouldn't be a big problem. Try (and I know this is obvious but isn't that easy) to keep the windings as even as possible (with the tri/quad filar windings together in their set of three or four wires to maintain the coupling) and to end up with full layers i.e. start and end in the same place.

Once you have the transformers wound and happy that they work okay you can varnish them to stop the singing.

My speakers are 97dB/8 ohm, so I don't need 100w - just 20-35W max. Would a 200-250VA toroid do fine for that level output?.

Yes, that is more than adequate. Even in a fair sized room you will only need a watt of two for normal listening, the extra just needed for those loud scary moments :)

Note that for these sensitivity ratings you should run the transformer as 2:1 not just across the arms to get lower distortion.

And if you wind the toroid in for example two sections - remembering to keep the turns matched (1,2,3,4,5.... 57, !!!distraction!!!, where was I?, oh bother!!!) then for your sensitivity speakers you could also try 4:1 which will really bring the distortion levels down.

One of the great advantage of using inductors is the flexibility in load matching and being able to make these trade offs.

Anyway, at the bargain basement price of $11 each, I'm just going to go ahead and order a pair first thing this morning(it's 2 am here).

Night Owl.

I am still sipping my morning coffee - should be awake by midday :)

Thanks.

Best wishes,
Susan.
 
Confirmation on Affirmation

Hi SUSAN - PARKER

We are Going to TRY ur ZFA with 2 X APT20M18BVR Mosfets for Thundering 750 WRMS With TOROIDAL OUTPUT TRAFOS.

New Title To be given to this Version

ONE HORSE POWER ZERO FEEDBACK AMP

STAY TUNED I WILL POST THE PICTURES ALSO.
SO WISH US A GOOD LUCK.

Regards
Workhorse Technologies.
 
Re: Confirmation on Affirmation

Hi Amp_man,

amp_man_1 said:
Hi SUSAN - PARKER

We are Going to TRY ur ZFA with 2 X APT20M18BVR Mosfets for Thundering 750 WRMS With TOROIDAL OUTPUT TRAFOS.

New Title To be given to this Version

ONE HORSE POWER ZERO FEEDBACK AMP

STAY TUNED I WILL POST THE PICTURES ALSO.
SO WISH US A GOOD LUCK.

Regards
Workhorse Technologies.

I do indeed wish you luck.

I guess you are running about 85 / 90 volts power supply?

With the APT20M18s you should be good for a minimum of 400 watts continuious (with a really big heatsink or forced air cooling - which does make a big diference).

With the APT20M18s the Crss is only 700 pF so you should be able to drive the gates okay from your 50 VA input transformer.

I hope you get to see the 746 watts, even if it's only peak. That's a lot of power.

Into 8 ohms impedance?

I have a pair of APT20M20LLL in TO-264 Packages, and they do work. Just don't have a zillion amp power supply, or a megawatt load to dump the power into :)

I really do wish you good health, a long life, and good luck.

Look forward to the updates and pictures.

Best wishes,
Susan.