Here is the Siliconix audio amplifier (published in October 1979, when VMOS hit the mass market).
Here is the Siemens BUZ20 application-note audio amplifier (published in 1982 maybe, when SIPMOS hit the mass market).
See attached files.
I guess audio freaks will love them, as they are non symetric, non complementary. If global feedback doesn't ruin everything, one can expect a smooth harmonic content with H2, H3, H4 and H5 showing in a nice undisturbed parabolic profile.
But, what about dV/dt ?
But, what about Class-B crossover distorsion ?
But, what about thermal stability ?
The Siliconix audio amplifier may not work, as F. de Dieuleveult (French Journalist) was known for making mistakes (distraction mainly) upon publishing articles. A proof of this is that the bias setting procedure he is describing, is completely nuts, referring to non-existing components. No doubt he re-numbered the components on the schematic, after he finished writing the article !
If somebody can provide a link to the genuine Siliconix application note entitled "DA76-1 - The VMOS Power FET Audio Amplifier", it would be fantastic.
What do you think about the Siemens BUZ20 application-note audio amplifier ?
I'm quite sure this worked, as it got published as a DIY mini-amplifier by the magazine Le Haut-Parleur (France and Belgium), authored by Etienne Lemery, who was (is still !) one of the best audio journalists here in France, always in serach for excellence.
Any comments, any suggestions welcome.
Here is the Siemens BUZ20 application-note audio amplifier (published in 1982 maybe, when SIPMOS hit the mass market).
See attached files.
I guess audio freaks will love them, as they are non symetric, non complementary. If global feedback doesn't ruin everything, one can expect a smooth harmonic content with H2, H3, H4 and H5 showing in a nice undisturbed parabolic profile.
But, what about dV/dt ?
But, what about Class-B crossover distorsion ?
But, what about thermal stability ?
The Siliconix audio amplifier may not work, as F. de Dieuleveult (French Journalist) was known for making mistakes (distraction mainly) upon publishing articles. A proof of this is that the bias setting procedure he is describing, is completely nuts, referring to non-existing components. No doubt he re-numbered the components on the schematic, after he finished writing the article !
If somebody can provide a link to the genuine Siliconix application note entitled "DA76-1 - The VMOS Power FET Audio Amplifier", it would be fantastic.
What do you think about the Siemens BUZ20 application-note audio amplifier ?
I'm quite sure this worked, as it got published as a DIY mini-amplifier by the magazine Le Haut-Parleur (France and Belgium), authored by Etienne Lemery, who was (is still !) one of the best audio journalists here in France, always in serach for excellence.
Any comments, any suggestions welcome.
Attachments
Last edited:
If my memory is good, Dieuleveult s amp was published in the french
review Radio Plans.
A refined schematic, as often with him despite some mistakes
as his 2 X 150W amp that did blow the output devices automatically.
Hey, he did use a single pair of 2N5631/2N6031 in his first try.
A month later ,he upgraded to four pairs..
He published in the same review a preamp using discrete op amps
that was truly wonderfull.
The Spice simulation shows that the Siemens - Etienne Lemery amplifier delivers clean results with excellent bandwidth and slew-rate.
See attached .zip file. It is containing two Tina 7 Texas Instruments files ready to be simulated.
Tina 7 Texas Instruments is free and downloadable from SPICE-Based Analog Simulation Program - TINA-TI - TI Tool Folder
See attached .zip file. It is containing two Tina 7 Texas Instruments files ready to be simulated.
Tina 7 Texas Instruments is free and downloadable from SPICE-Based Analog Simulation Program - TINA-TI - TI Tool Folder
What do you think about the Siemens BUZ20 application-note audio amplifier ?
I'm quite sure this worked, as it got published as a DIY mini-amplifier by the magazine Le Haut-Parleur (France and Belgium), authored by Etienne Lemery, who was (is still !) one of the best audio journalists here in France, always in serach for excellence.
Any comments, any suggestions welcome.[/QUOTE]
I interested to see this DIY mini-amplifier, I made German Elektor BUZAMP long time ago, and yes he work.
I'm quite sure this worked, as it got published as a DIY mini-amplifier by the magazine Le Haut-Parleur (France and Belgium), authored by Etienne Lemery, who was (is still !) one of the best audio journalists here in France, always in serach for excellence.
Any comments, any suggestions welcome.[/QUOTE]
I interested to see this DIY mini-amplifier, I made German Elektor BUZAMP long time ago, and yes he work.
Last edited:
apexaudio,
it would have been appropriated to tell people that you started a thread about this particular Siemens BUZ20 NMOS amplifier one week after I started this one.
http://www.diyaudio.com/forums/solid-state/163159-150w-mosfet-amplifier-irfp250x2.html
Started on March 14th, 2010.
The simulation of this amplifier is here since long ! Why don't you asked me ?
Now that you encourage people to build this remarkable mis-estimated amp, why don't you explain in the exposition of your thread, that thermal runaway is going to happen unless you put a temperature sensor (thermistor symbol) in the definition of the 2nd differential pair current source ? Etienne Lemery showed a proper practical solution ages ago. Read the original schematic (provided above). There is no Vbe multiplier ! The quiescent current is defined by the voltage drops on the 1K resistors.
Once it is understood how the quiescent current gets defined, this amp doesn't deserve the bad reputation he got ages ago.
Be more careful next time when you "quote" me. Your previous message is confusing / misleading. You were in such a hurry ... so you see, that's quite transparent now ...
cheers and respectful,
Steph
(oops, with an attitude ?)
it would have been appropriated to tell people that you started a thread about this particular Siemens BUZ20 NMOS amplifier one week after I started this one.
http://www.diyaudio.com/forums/solid-state/163159-150w-mosfet-amplifier-irfp250x2.html
Started on March 14th, 2010.
The simulation of this amplifier is here since long ! Why don't you asked me ?
Now that you encourage people to build this remarkable mis-estimated amp, why don't you explain in the exposition of your thread, that thermal runaway is going to happen unless you put a temperature sensor (thermistor symbol) in the definition of the 2nd differential pair current source ? Etienne Lemery showed a proper practical solution ages ago. Read the original schematic (provided above). There is no Vbe multiplier ! The quiescent current is defined by the voltage drops on the 1K resistors.
Once it is understood how the quiescent current gets defined, this amp doesn't deserve the bad reputation he got ages ago.
Be more careful next time when you "quote" me. Your previous message is confusing / misleading. You were in such a hurry ... so you see, that's quite transparent now ...
cheers and respectful,
Steph
(oops, with an attitude ?)
Last edited:
Hello Mile, seems a good idea as the other thread has drifted, and is not concentrating anymore on the lovely simple Siemens BUZamp. I'm definitely sure there is a future for the Siemens BUZamp. Shall we design an add-on for getting a Class AB Soft Non-Switching Siemens BUZamp, following kenpeter research ?
What makes you saying this ? If you are basing on another Siemens BUZamp implementation, it would be nice to share the schematic.
lol
I agree that it'd help more with bias stability connected from the source to negative rail.No idea what it's doing there on the drain other than looking like it makes things look symetrical at the output, it's just an illusion
Not your amp Mile, we are refering to the Siemens BUZ20 amp schematic
Yours looks pretty interesting btw & i already sussed out it's modus operandi.
I interested to see this DIY mini-amplifier, I made German Elektor BUZAMP long time ago, and yes he work.[/QUOTE]
The V-FET Amp (based on Siliconix application - only N-Channel versions in the output) was published in Germans Elektor 7+8/1978 and is very close to an similare topology used by several Yamaha V-FET models, e. g.
http://www.amplimos.it/images/2sk77 amp YAMAHA B-1.gif
The Siemens BUZAMP was also published in German's Elektor, in this case in a special edition "Plus 7" from 1989. A used version with some photos are to find about ebay item No 270513251935
About No 1 and 2 (by post #1) you will find some other circuit sources with the same topology:
http://www.diyaudio.com/forums/soli...better-audio-non-complements-audio-power.html
Ask formerly's Creek developer and diyaudio member X-PRO about his statement - his own concepts are large improved "BUZAMP's" - i. e. also only N-Channels for better audio in the output and additional reduced THD values.
Ckeck out his threads:
http://www.diyaudio.com/forums/solid-state/110771-improving-linearity-n-ch-mosfet-output-stage.html
http://www.diyaudio.com/forums/solid-state/14320-n-channel-only-output-devices-power-follower.html
http://www.diyaudio.com/forums/solid-state/94030-quasi-not-quasi-question.html
and this thread:
http://www.diyaudio.com/forums/solid-state/154388-its-cheap-its-n-its-dirty-its-circlomos-4.html
Last edited:
My amp is based on BUZAMP, Steph just not draw complete schematics.Not your amp Mile, we are refering to the Siemens BUZ20 amp schematic
Yours looks pretty interesting btw & i already sussed out it's modus operandi.
Aha, so what you are possibly saying is that the drain resistor probably has a use for current limiting & we haven't got the full schematic
Ok, fair enough. But the resistor is useless where it is unless it does something & right now with the circuit as drawn it does nothing other than lower the efficiency
About the 0.27 ohm resistor in the lower NMOS drain,
seems some of you are forgetting that the bias current is defined by the voltage drop that's developing on the 1K resistor.
Look the top NMOS. Clearly, from what is stated above, the 0.27 ohm resistor doesn't affect the bias, the 1K resistor being returned to the source, and not encompassing the voltage drop on the 0.27 ohm resistor.
Look the bottom NMOS. Knowing that the bias, over there, is also defined by the voltage drop that's developing on the 1K resistor, one gets the choice where to place the 0.27 ohm resistor. I guess Etienne Lemery wanted to deliver something conventional, looking symetric. Therefore he put the 0.27 ohm resistor in the drain of the bottom NMOS.
I agree that the real symetry, once the topology is understood, was to place the bottom NMOS 0.27 ohm in the source, same layout as the upper NMOS.
But I repeat, this doesn't affect the bias, in such circuit.
One may simulate and test a different circuit, putting the bottom NMOS 0.27 ohm resistor in the source, and returning both 1K resistors not to the sources, but to the output for the upper NMOS, and to the negative rail for the bottom NMOS.
This different circuit will exhibit some degree of quiescent current negative feedback.
That's easy to understand : if the 0.27 ohm is developing a voltage (strong quiescent current), this voltage gets substracted from the actual NMOS drive voltage (voltage developing across the 1K resistor, less voltage developing across the 0.27 ohm resistor).
But wait a minute. When the current is about 5 amp, the voltage on the 0.27 ohm will be 1.35 V. The NMOS drive will get significantly decreased. This makes the output stage not working as voltage source anymore, but working as a controlled current source. Don't know if this is a good idea. The open-loop gain of the amp will need to battle against this kind of current drive mode.
Am I right ?
seems some of you are forgetting that the bias current is defined by the voltage drop that's developing on the 1K resistor.
Look the top NMOS. Clearly, from what is stated above, the 0.27 ohm resistor doesn't affect the bias, the 1K resistor being returned to the source, and not encompassing the voltage drop on the 0.27 ohm resistor.
Look the bottom NMOS. Knowing that the bias, over there, is also defined by the voltage drop that's developing on the 1K resistor, one gets the choice where to place the 0.27 ohm resistor. I guess Etienne Lemery wanted to deliver something conventional, looking symetric. Therefore he put the 0.27 ohm resistor in the drain of the bottom NMOS.
I agree that the real symetry, once the topology is understood, was to place the bottom NMOS 0.27 ohm in the source, same layout as the upper NMOS.
But I repeat, this doesn't affect the bias, in such circuit.
One may simulate and test a different circuit, putting the bottom NMOS 0.27 ohm resistor in the source, and returning both 1K resistors not to the sources, but to the output for the upper NMOS, and to the negative rail for the bottom NMOS.
This different circuit will exhibit some degree of quiescent current negative feedback.
That's easy to understand : if the 0.27 ohm is developing a voltage (strong quiescent current), this voltage gets substracted from the actual NMOS drive voltage (voltage developing across the 1K resistor, less voltage developing across the 0.27 ohm resistor).
But wait a minute. When the current is about 5 amp, the voltage on the 0.27 ohm will be 1.35 V. The NMOS drive will get significantly decreased. This makes the output stage not working as voltage source anymore, but working as a controlled current source. Don't know if this is a good idea. The open-loop gain of the amp will need to battle against this kind of current drive mode.
Am I right ?
Attachments
Last edited:
hi tiefbassuebertr,
thanks for your fantastic post. Nice contribution. There is a mass of information inside, I still need to process.
From your post, I grabbed one of the pictures on eBay. It would be nice to read the schematic of Elektor BUZamp ! From the picture, I see :
- 4 x BUZ23 NMOS
- one differential pair at the input
- 6 small signal transistors in the right side (two current sources ?)
- 2 medium power transistors (2nd differential pair ?)
- two temperature sensors on the heatsink (linked to the current sources ?)
- left side maybe protection circuit (4 small signal diodes, 3 small signal transistors, 1 medium power transistor)
Looks like a bloated Etienne Lemery - Siemens BUZ 20 amp.
I am nearly sure that in the links you provided, we'll find some people trying to reach a Class AB Soft Non-Switching operation.
This amplifier structure should not be confused with others NMOS amplifiers, that we may call semi-complementary. Here, we are concentrating on an amplifier's structure that's definitively turning away from the semi-complementary concept. It is a pure non-complementary concept. And I like it, and I think there is a future for it.
Cheers,
Steph
thanks for your fantastic post. Nice contribution. There is a mass of information inside, I still need to process.
From your post, I grabbed one of the pictures on eBay. It would be nice to read the schematic of Elektor BUZamp ! From the picture, I see :
- 4 x BUZ23 NMOS
- one differential pair at the input
- 6 small signal transistors in the right side (two current sources ?)
- 2 medium power transistors (2nd differential pair ?)
- two temperature sensors on the heatsink (linked to the current sources ?)
- left side maybe protection circuit (4 small signal diodes, 3 small signal transistors, 1 medium power transistor)
Looks like a bloated Etienne Lemery - Siemens BUZ 20 amp.
I am nearly sure that in the links you provided, we'll find some people trying to reach a Class AB Soft Non-Switching operation.
This amplifier structure should not be confused with others NMOS amplifiers, that we may call semi-complementary. Here, we are concentrating on an amplifier's structure that's definitively turning away from the semi-complementary concept. It is a pure non-complementary concept. And I like it, and I think there is a future for it.
Cheers,
Steph
Attachments
Are you joking, apexaudio ? That's a lazy answer, and detrimental for me. What game are you playing, my friend ?
I was about to reply that your schematics are barely complete, never showing the bias current thermal feedback arrangement, but I have not replied it because of risking a 3 day ban for irrespectful behaviour.
(oops, I just wrote it)
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.