Thank you, Susan! Just the kind of feedback I was hoping for.
Yes, 10 Hz to 50 kHz would be plenty of bandwidth.
I hadn't realized the ribbon output level was similar to that of an MC phono cartridge. So an MC phono preamp (minus the RIAA EQ, plus balanced output) could serve well as my ribbon head amp? Or would the execution be somewhat different?
Cheers,
Bill
Yes, 10 Hz to 50 kHz would be plenty of bandwidth.
I hadn't realized the ribbon output level was similar to that of an MC phono cartridge. So an MC phono preamp (minus the RIAA EQ, plus balanced output) could serve well as my ribbon head amp? Or would the execution be somewhat different?
Cheers,
Bill
AD815s as line drivers
For those who are wincing at the cost of the transformers for the Zeus VMOS Line Driver...
Further to my previous post the AD815s have arrived and I have lashed something together (not pretty, but it works).
I use the circuit as per the front page of the AD815 datasheet, but all three gain resistors are 1K0s. Plus 27 ohm series resistors on each output.
For simplicity to get the balanced input I have used a SSM2142 as a splitter/ driver.
Regret it is too late in the evening to start trying out with the amp, have to start tidying up.
More on this soon.
Best wishes,
Susan.
For those who are wincing at the cost of the transformers for the Zeus VMOS Line Driver...
Further to my previous post the AD815s have arrived and I have lashed something together (not pretty, but it works).
I use the circuit as per the front page of the AD815 datasheet, but all three gain resistors are 1K0s. Plus 27 ohm series resistors on each output.
For simplicity to get the balanced input I have used a SSM2142 as a splitter/ driver.
Regret it is too late in the evening to start trying out with the amp, have to start tidying up.
Best wishes,
Susan.
Re: AD815s as line drivers
Part II
Initial testing showed the AD815 had LF insability so I have moved the two feedback 1K0 resistors from the output (as shown on the datasheet schematic) to the far side of the 27 ohm series resistors. Now drives ruler flat down to 20 Hz.
The SSM2142 splitter/driver needs to be driven from a low impedance (lower than I was expecting) otherwise it is very noisy, so I am adding a buffer (variable gain) stage before it.
Initial testing here:
http://www.audiophonics.com/audiophonics-zeus-line-driver-ad815-1.html
... which shows line driver and Zeus75 combo.
N.B. The two graphs further down the page show the single ended "hot" EMu 1212M source (I normally use this output balanced). There is significant noise which is clearly visible on a scope with normal signal levels.
I have left the basic EMu loopback tests at the bottom for comparison.
Best wishes,
Susan.
Part II
Initial testing showed the AD815 had LF insability so I have moved the two feedback 1K0 resistors from the output (as shown on the datasheet schematic) to the far side of the 27 ohm series resistors. Now drives ruler flat down to 20 Hz.
The SSM2142 splitter/driver needs to be driven from a low impedance (lower than I was expecting) otherwise it is very noisy, so I am adding a buffer (variable gain) stage before it.
Initial testing here:
http://www.audiophonics.com/audiophonics-zeus-line-driver-ad815-1.html
... which shows line driver and Zeus75 combo.
N.B. The two graphs further down the page show the single ended "hot" EMu 1212M source (I normally use this output balanced). There is significant noise which is clearly visible on a scope with normal signal levels.
I have left the basic EMu loopback tests at the bottom for comparison.
Best wishes,
Susan.
Zeus transformers from Sowter
Dear All,
I am please to be able to say that all current Zeus transformers available from Sowter can now be ordered online...
http://www.sowter.co.uk/acatalog/SOWTER_CLASSIC_AMPS_15.html#a6
as well as the 9063c type...
http://www.sowter.co.uk/acatalog/SOWTER_INPUT_TRANSFORMERS_11.html#a59
Hope this will help.
Best wishes,
Susan.
P.S. I used their online ordering this afternoon for a pair of 9063s for the 6C33C triode version of the VMOS PP Line Driver. The system works.
Dear All,
I am please to be able to say that all current Zeus transformers available from Sowter can now be ordered online...
http://www.sowter.co.uk/acatalog/SOWTER_CLASSIC_AMPS_15.html#a6
as well as the 9063c type...
http://www.sowter.co.uk/acatalog/SOWTER_INPUT_TRANSFORMERS_11.html#a59
Hope this will help.
Best wishes,
Susan.
P.S. I used their online ordering this afternoon for a pair of 9063s for the 6C33C triode version of the VMOS PP Line Driver. The system works.
Hi Bill,
If I have done the sums right (and not misscounted the fingers and toes) then as far as I can see the basic amplification chain could be of a similar topology (although there are allowances for losses in the filtering which we don't need to worry about).
I am not sure about hacking up an existing MC phono-preamp without seeing the schematic as I am sure there are going to be a few wrinkles (of the type that can be easily overlooked in bold sweeping statements) in that.
Simplest test if you happen to have a MC preamp handy is to connect up the ribbon element and see how much signal you get (making allowences for the RIAA curve).
Best wishes,
Susan.
Bill F. said:
If I have done the sums right (and not misscounted the fingers and toes) then as far as I can see the basic amplification chain could be of a similar topology (although there are allowances for losses in the filtering which we don't need to worry about).
I am not sure about hacking up an existing MC phono-preamp without seeing the schematic as I am sure there are going to be a few wrinkles (of the type that can be easily overlooked in bold sweeping statements) in that.
Simplest test if you happen to have a MC preamp handy is to connect up the ribbon element and see how much signal you get (making allowences for the RIAA curve).
Best wishes,
Susan.
AD815s as line drivers
Part III
The SSM2142 splitter/driver is now driven from a LM4562 buffer (variable gain) stage before it. N.B. needs a unity gain stable opamp.
Picture, schematic and further testing added:
http://www.audiophonics.com/audiophonics-zeus-line-driver-ad815-1.htmll
... which now shows basic line driver performance as well as the line driver and Zeus75 together.
I have also added further open/short and EMu loopback tests for comparison.
Best wishes,
Susan.
Part III
The SSM2142 splitter/driver is now driven from a LM4562 buffer (variable gain) stage before it. N.B. needs a unity gain stable opamp.
Picture, schematic and further testing added:
http://www.audiophonics.com/audiophonics-zeus-line-driver-ad815-1.htmll
... which now shows basic line driver performance as well as the line driver and Zeus75 together.
I have also added further open/short and EMu loopback tests for comparison.
Best wishes,
Susan.
Hello Susan,
your Zeus is a nice and simple design... alas not quite cheap
if one buy the transformers.
Do you know his ancestor from Dynacord?
http://www.audiophonics.com/audioph...r-ad815-1.htmll
this link seems not to work!
Regards
Heinz!
your Zeus is a nice and simple design... alas not quite cheap
if one buy the transformers.
Do you know his ancestor from Dynacord?
http://www.audiophonics.com/audioph...r-ad815-1.htmll
this link seems not to work!
Regards
Heinz!
Attachments
Hi Heinz,
Thank you for your post.
Thanks. I agree the transformers do addup if one has to buy them all at once. However one can tryout the principles using mains power toroids and a headphone output which is very cost efective.
This gives a easy introduction and enables one to get a feel for the "Zeus sound".
No, never seen this one before. Thanks for the schematic.
Ops, seems to have an extra l on the end.
Should be:
http://www.audiophonics.com/audiophonics-zeus-line-driver-ad815-1.html
Thanks for pointing that out.
best wishes,
Susan.
Thank you for your post.
powerbecker said:
Thanks. I agree the transformers do addup if one has to buy them all at once. However one can tryout the principles using mains power toroids and a headphone output which is very cost efective.
This gives a easy introduction and enables one to get a feel for the "Zeus sound".
No, never seen this one before. Thanks for the schematic.
Ops, seems to have an extra l on the end.
Should be:
http://www.audiophonics.com/audiophonics-zeus-line-driver-ad815-1.html
Thanks for pointing that out.
best wishes,
Susan.
Wow, 109 pages of posts! 
Hello Susan, thank you for this addition to the DIY world. I've recently come upon it after thinking I wanted to build something similar to the First Watt F2. I have been reading your website (old and updated) intently over a few days, as I am intrigued by this wonderfully clean and simple design.
I'm wondering if your design can be adapted to serve as a high output impedance power source simply enough by choosing a different (step-up?) output transformer? I'm hoping for at least 40 ohms, but a bit higher would be fine too. I'm sorry if this was answered already ...but 109 pages of posts! whew.
p.s. for others: I modeled the circuit in LTSpice, but can't seem to get any current through the MOSFETs to the output xformer, even though I can get voltage if I raise the biase to 4.1v or over. All I get is super low level noise. Anyone use LTSpice? (I know you use B2 Spice v5 Susan. )
Hello Susan, thank you for this addition to the DIY world. I've recently come upon it after thinking I wanted to build something similar to the First Watt F2. I have been reading your website (old and updated) intently over a few days, as I am intrigued by this wonderfully clean and simple design.
I'm wondering if your design can be adapted to serve as a high output impedance power source simply enough by choosing a different (step-up?) output transformer? I'm hoping for at least 40 ohms, but a bit higher would be fine too. I'm sorry if this was answered already ...but 109 pages of posts! whew.
p.s. for others: I modeled the circuit in LTSpice, but can't seem to get any current through the MOSFETs to the output xformer, even though I can get voltage if I raise the biase to 4.1v or over. All I get is super low level noise. Anyone use LTSpice? (I know you use B2 Spice v5 Susan.
)
Thank you for your post and interest.
Not quite sure why you need the higher impedance... do you need the higher voltage swing?
If you use a 1:1 output then the impedance (depending on the mosfet type) will be about 3 to 4 ohms, however the output distortion levels will rise proportionality unless the load also is of a proportionally higher impedance.
To increase the impedance beyond this requires either a different ratio output transformer or adding series resistors.
Best wishes,
Susan.
m8o said:Hello Susan, thank you for this addition to the DIY world. I've recently come upon it after thinking I wanted to build something similar to the First Watt F2. I have been reading your website (old and updated) intently over a few days, as I am intrigued by this wonderfully clean and simple design.
I'm wondering if your design can be adapted to serve as a high output impedance power source simply enough by choosing a different (step-up?) output transformer? I'm hoping for at least 40 ohms, but a bit higher would be fine too. I'm sorry if this was answered already ...but 109 pages of posts! whew.
Not quite sure why you need the higher impedance... do you need the higher voltage swing?
If you use a 1:1 output then the impedance (depending on the mosfet type) will be about 3 to 4 ohms, however the output distortion levels will rise proportionality unless the load also is of a proportionally higher impedance.
To increase the impedance beyond this requires either a different ratio output transformer or adding series resistors.
Best wishes,
Susan.
Susan-Parker said:
Howdy, Susan!
The answer to that question can be found in this pdf article by Nelson Pass:
Current Source Amplifiers and Sensitive / Full-Range Drivers
se
Well, speak ob da debbil!
You're just in time to join the Insane Santa Posse.
Saddle up!
You're just in time to join the Insane Santa Posse.
Saddle up!
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Thanx for the reply. Ah, I see. That (hi-watt series resistors), I wanted to avoid. I was hoping to gain high impedance through a "side-effect" if you will of the transformer choice. I guess I should consider the heat dissipation in the transformer too then though... Perhaps a bad idea.
To answer why, if you'll indulge me with this off-topic explaination into my insanity... I have Klipsch La Scala speakers. I have some really nice power. Audio Research Classic 150 monoblocks, and a Butler 5-channel tube/ss hybrid amp. I wish to do this experiment however:
- Use my directly heated triode tube headamp (Litle Dot III+)
- an old 3-way fully configurable active x-over that I have
- followed by a transconductance amplifier, with either 0 gain or low gain (doesn't really matter). Could use a hi-gain amp, then reduce the gain through a resistor as you note.
- X3 ; one amp for the bass horn, one amp for the mid horn, one amp for the tweeter. Of course amp power to the tweeter would be a fraction of that to the bass horn with power to the mid in-between.
Reason being...
1) I understand Transconductance amps to mate well with not only hi-efficiency lightweight full-range drivers, but also horn loaded drivers.
2) I notice on impedance graphs that horn loaded drivers often have peaks up the frequency spectrum ; which just so happens to coincide with dips I note in their frequency response. As a constant current amp will provide more power into the transducer when impedance is high, I believe that will effectively flatten the response of the horn loaded driver ; and this 'smoothening out' is what I wish to experiment with.
So if I like what I hear, I planned on building a completely custom 2 channel 3-way integrated amp comprised of
- a stereo 3-way tube-based x-over
- 3 stereo tube line amps with perhaps some nominal tone control per low/mid/high (perhaps pre & tone functions folded right into the x-over stage)
- 6 high impedance solid state monoblock amps or power followers (would probably want a step-up x-former for the latter, which is what brought me to your design Susan).
I made Frankensteins out of my Martin Logan Sequel IIs, where I ripped out the passive x-overs & equalization, and bi-amped them driven by a digital eq & room correction (a TaCT unit), then added a 4 foot ribbon tweeter into the mix, all with great success.
....This (transconductance drive of my horns) is my new obsession.
Instead, I may just do one stereo amp and build new passive x-overs with x-over points calculated to the high impedance of the amp + horn drivers for the test. I could just do simple series 1st order if I keep the evaluation to low level (1st order ... good). But then I'd need to know exactly what the amp's output impedance is at the x-over frequency I believe ... not sure how to measure that but I have some idea.
Either way though, high impedance is necessary for this experiment.
To answer why, if you'll indulge me with this off-topic explaination into my insanity...
I have Klipsch La Scala speakers. I have some really nice power. Audio Research Classic 150 monoblocks, and a Butler 5-channel tube/ss hybrid amp. I wish to do this experiment however:- Use my directly heated triode tube headamp (Litle Dot III+)
- an old 3-way fully configurable active x-over that I have
- followed by a transconductance amplifier, with either 0 gain or low gain (doesn't really matter). Could use a hi-gain amp, then reduce the gain through a resistor as you note.
- X3 ; one amp for the bass horn, one amp for the mid horn, one amp for the tweeter. Of course amp power to the tweeter would be a fraction of that to the bass horn with power to the mid in-between.
Reason being...
1) I understand Transconductance amps to mate well with not only hi-efficiency lightweight full-range drivers, but also horn loaded drivers.
2) I notice on impedance graphs that horn loaded drivers often have peaks up the frequency spectrum ; which just so happens to coincide with dips I note in their frequency response. As a constant current amp will provide more power into the transducer when impedance is high, I believe that will effectively flatten the response of the horn loaded driver ; and this 'smoothening out' is what I wish to experiment with.
So if I like what I hear, I planned on building a completely custom 2 channel 3-way integrated amp comprised of
- a stereo 3-way tube-based x-over
- 3 stereo tube line amps with perhaps some nominal tone control per low/mid/high (perhaps pre & tone functions folded right into the x-over stage)
- 6 high impedance solid state monoblock amps or power followers (would probably want a step-up x-former for the latter, which is what brought me to your design Susan).
I made Frankensteins out of my Martin Logan Sequel IIs, where I ripped out the passive x-overs & equalization, and bi-amped them driven by a digital eq & room correction (a TaCT unit), then added a 4 foot ribbon tweeter into the mix, all with great success.
....This (transconductance drive of my horns) is my new obsession.
Instead, I may just do one stereo amp and build new passive x-overs with x-over points calculated to the high impedance of the amp + horn drivers for the test. I could just do simple series 1st order if I keep the evaluation to low level (1st order ... good). But then I'd need to know exactly what the amp's output impedance is at the x-over frequency I believe ... not sure how to measure that but I have some idea.
Either way though, high impedance is necessary for this experiment.
Sorry for the delays in my posts showing up. I'm new and still under moderation.Steve Eddy said:
Correct, that fine work was my main inspiration most recently. I had heard about "constant current" many many [many] years ago. But to tell you the truth, I couldn't "get it". I was taught ohm's law; that voltage is delivered and for a particular resistance [or inductance] current is "drawn" by that law. But I couldn't get my head around approaching it from the opposite direction and 'driving' the current. I knew in that case how that could oddly affect power delivery into complex frequency depedant impedances compared to a voltage amp. But from the fine work of Nelson and numerous others across many board throughout the world, I get how it operates now and how that could prove to have sonic benefits in certain circumstances.
Here's an interesting thread if you want to go tubes: http://www.diyaudio.com/forums/showthread.php?s=&threadid=42347&highlight=
Or a cheap way to try: http://www.diyaudio.com/forums/showthread.php?s=&threadid=20108&highlight=
Sheldon
Or a cheap way to try: http://www.diyaudio.com/forums/showthread.php?s=&threadid=20108&highlight=
Sheldon
hmmm.... the possibility or building a CC amp from one of those gainclone chips or another is -too- simple, attractive, and cheap to pass up! ... at least as far as my 'test' of the idea is concerned. That seems the way I'll go for now. I may return to Susan's design if my trial of driving the horns with a constant current amp proves fruitful.
Best Regards all,
Steve
Best Regards all,
Steve
1:1 output transformer question
Hi Susan!
I would like to ask you for some help!
I would like to try a SS amp with 1:1 output transformer. I have heard so many oppinions pro and contra about output transformers, that now I would like to test it myself.
I now the sound of the direct coupled and also the capacitor coupled amps, but not the transformer coupled. So, I would like to try it, for example with the good reference amp the JLH69 (if it is possible at all.... ) or with another ca. 15W capacitor coupled Mosfet follower amp instead of the output caps.
I know your Zeus PP design and I saw that you built also a SE version of the Zeus.
Could you help me how can I calculate the parameters of such an 1:1 output transformer???
I could try EI types and toroidal isolation transformers too.
What do you suggest?
Greets:
Tyimo
Hi Susan!
I would like to ask you for some help!
I would like to try a SS amp with 1:1 output transformer. I have heard so many oppinions pro and contra about output transformers, that now I would like to test it myself.
I now the sound of the direct coupled and also the capacitor coupled amps, but not the transformer coupled. So, I would like to try it, for example with the good reference amp the JLH69 (if it is possible at all.... ) or with another ca. 15W capacitor coupled Mosfet follower amp instead of the output caps.
I know your Zeus PP design and I saw that you built also a SE version of the Zeus.
Could you help me how can I calculate the parameters of such an 1:1 output transformer???
I could try EI types and toroidal isolation transformers too.
What do you suggest?
Greets:
Tyimo
Re: 1:1 output transformer question
Hi Tyimo,
Thank you for your post.
SE needs a gaped transformer, so standard toroid won't be suitable.
However you can use the secondaries of a toroid in PP for a 1:1, although the distortion levels will be higher.
This is a size comparison between gapped EI for single ended and standard EI interleaved for push-pull:
For SE you will need higher bias, so the windings will need to be of a heavier gauge (or several in parallel).
What power are you looking for and what sort of load (4ohms, 8ohm, ...)?
Best wishes,
Susan.
Hi Tyimo,
Thank you for your post.
Tyimo said:
SE needs a gaped transformer, so standard toroid won't be suitable.
However you can use the secondaries of a toroid in PP for a 1:1, although the distortion levels will be higher.
This is a size comparison between gapped EI for single ended and standard EI interleaved for push-pull:
An externally hosted image should be here but it was not working when we last tested it.
For SE you will need higher bias, so the windings will need to be of a heavier gauge (or several in parallel).
What power are you looking for and what sort of load (4ohms, 8ohm, ...)?
Best wishes,
Susan.