The Cello's use a choke input filter, like selected Ayre models. True choke input filters are quite large and expensive, but they do wonderful things for the power supply behavior, eliminating the large charging pulses and providing another energy storage mechanism besides the output cap. They work best in class A designs or AB designs running very high bais current. The Ayre V1x uses a choke input power supply, but some of there other newer models, like the V5 and V6, don't. They are built to a more affordable cost target.
The custom chokes I had Plitron build are each larger than the power supply transformer they are to be used with. But then, the use of a choke input supply improves the power transformer utilixation by nearly a factor of two, so the transformer can be smaller for the same output power.
You can do a search of the posts at DIYAUDIO.COM and come up with schematics for a lot of older "classic" amps, including Mark Levinson, etc. This may give you some good ideas for project, depending on your current level of experience.
Beyond that, you have the matter of circuit topology. If you have access to the FM acoustic topology details and like it, then why not build it to start off with? Listen, and learn, and see if you can tweak it up some more.
Keep in mind that selection of components is probably nearly as important to the final sound as the circuit topology. Just this week I'm evaluating four different caps of the same nominal value for rail bypass on the main PCB; though published specs are similar, actual measurements differ- in this case, some Panasonic 680 uF FC caps I measure are out of spec on ESR (~40-45 milliohms, compared with a spec of 36 milliohms, for ), while Nichicon 820uF 100V HE types the same size are measuring ~ 20 milliohms, and staying below 30 milliohms out to 500 kHz. Measurements aren't everything; I'll listen to them, too, but obviously I'm leaning towards the Nichicon at this time.
If you use bipolar output devices, the extended beta types which Toshiba invented and On Semi copies are probably the best to use for reasonable output power. There are some more rugged On Semi devices, but they're usually a bit slower and don't have as flat an Hfe curve; i.e., they have higher distortion open loop.
I generally favor the no loop feedback approaches such as C. Hansen uses, Theta Digital, and Ayre use.
Last words, you'll probably be happiest with a DIY approach if you value the learning and development process as much as the end result. Access to components and availability of good tools (including measuring equipment) are usually big factors for success.
If you're more interested in having a high performance amp for the lowest price, it's probably better to buy a good used amplifier.
Regards,
Jon
The custom chokes I had Plitron build are each larger than the power supply transformer they are to be used with. But then, the use of a choke input supply improves the power transformer utilixation by nearly a factor of two, so the transformer can be smaller for the same output power.
You can do a search of the posts at DIYAUDIO.COM and come up with schematics for a lot of older "classic" amps, including Mark Levinson, etc. This may give you some good ideas for project, depending on your current level of experience.
Beyond that, you have the matter of circuit topology. If you have access to the FM acoustic topology details and like it, then why not build it to start off with? Listen, and learn, and see if you can tweak it up some more.
Keep in mind that selection of components is probably nearly as important to the final sound as the circuit topology. Just this week I'm evaluating four different caps of the same nominal value for rail bypass on the main PCB; though published specs are similar, actual measurements differ- in this case, some Panasonic 680 uF FC caps I measure are out of spec on ESR (~40-45 milliohms, compared with a spec of 36 milliohms, for ), while Nichicon 820uF 100V HE types the same size are measuring ~ 20 milliohms, and staying below 30 milliohms out to 500 kHz. Measurements aren't everything; I'll listen to them, too, but obviously I'm leaning towards the Nichicon at this time.
If you use bipolar output devices, the extended beta types which Toshiba invented and On Semi copies are probably the best to use for reasonable output power. There are some more rugged On Semi devices, but they're usually a bit slower and don't have as flat an Hfe curve; i.e., they have higher distortion open loop.
I generally favor the no loop feedback approaches such as C. Hansen uses, Theta Digital, and Ayre use.
Last words, you'll probably be happiest with a DIY approach if you value the learning and development process as much as the end result. Access to components and availability of good tools (including measuring equipment) are usually big factors for success.
If you're more interested in having a high performance amp for the lowest price, it's probably better to buy a good used amplifier.
Regards,
Jon
FM Acoustics
The FM amps look all too simple and its hard to believe how expensive they are. However they are an example of obsessive commitment to parts selection and evaluation. I spoke to Manuel Huber about this and he indicated that they sort the transistors for beta linearity to an extreme degree. He gets something like a 10% yield at best. You can attempt this if you have a curve tracer and a gross of transistors to sort through.
The same obsession is true for the other components. (They are made like a Patek Phillipe and cost like one.)
The circuits are not published and unavailable.
Other premium designs are available but you might get further if you apply the principles of quality to a design that you have good knowledge of. A well executed Gain Clone will sound much better than a botched Krell or Levinson. if you don't know the underlying rational for each decision you can find yourself in a deep hole if something doesn't work.
-Demian
The FM amps look all too simple and its hard to believe how expensive they are. However they are an example of obsessive commitment to parts selection and evaluation. I spoke to Manuel Huber about this and he indicated that they sort the transistors for beta linearity to an extreme degree. He gets something like a 10% yield at best. You can attempt this if you have a curve tracer and a gross of transistors to sort through.
The same obsession is true for the other components. (They are made like a Patek Phillipe and cost like one.)
The circuits are not published and unavailable.
Other premium designs are available but you might get further if you apply the principles of quality to a design that you have good knowledge of. A well executed Gain Clone will sound much better than a botched Krell or Levinson. if you don't know the underlying rational for each decision you can find yourself in a deep hole if something doesn't work.
-Demian
Of the designs you cite, the Levinson ML-2 has schematics
and is simple enough that it would be a good choice. The
early Krells, such as the KSA50 are also a good potential
choice. The post-Curl stuff from Madrigal I wouldn't touch
with a 10 foot pole, and I suggest that the Halcro is in the
same category for DIYers.
and is simple enough that it would be a good choice. The
early Krells, such as the KSA50 are also a good potential
choice. The post-Curl stuff from Madrigal I wouldn't touch
with a 10 foot pole, and I suggest that the Halcro is in the
same category for DIYers.
Transistor amp DIY FM or ..?
Similar to tubes amp, I do think that in SS amp we need to knw the sound of each transistor and how the sound changes when we tune the working condition of the amp as well as with diff beta of the active devises.
I am curious to see FM. Manuel did tell me that his basic design did not change since the early 90s, only with better components. Over the year, I also do not see much different with any particular topology as long as they are not overly complicated. Implimentation is most important I guess.
In modifying my Ss amp, usually, I just improve on the PS, short cut the safty protection parts such as the fuses etc. For example, this will worked evry well in Mac SS amp. For powr amp, I will try to get better match output devises by measuring the voltage drops accross the emitter resistors. An increase in the bias current usually will change the sound by a lot, but for most amp, beyond certain sweet spot the sound become a bit too dark and powerless, may be the devises are under too much stress...
I have repair my Audio Lab 8000A amp half way. Blew the emitter resistors when I tweaked the pre, by pass some output resistors in serial, and forgot to instal back the feedback loop.
As the circuit is overly complicate, I guess this will be a good candidate to cut out some stages to get better sound.
Perhaps I will just use the Macs circuitries whch is simple and good. But curious to see how FM circuit sounds like with the off the shelf components.
I always wanted to see how the SS amp sounded like if we use simple high voltage circuit to increase the head room.
Best rgds
William Lee
Singapore
Similar to tubes amp, I do think that in SS amp we need to knw the sound of each transistor and how the sound changes when we tune the working condition of the amp as well as with diff beta of the active devises.
I am curious to see FM. Manuel did tell me that his basic design did not change since the early 90s, only with better components. Over the year, I also do not see much different with any particular topology as long as they are not overly complicated. Implimentation is most important I guess.
In modifying my Ss amp, usually, I just improve on the PS, short cut the safty protection parts such as the fuses etc. For example, this will worked evry well in Mac SS amp. For powr amp, I will try to get better match output devises by measuring the voltage drops accross the emitter resistors. An increase in the bias current usually will change the sound by a lot, but for most amp, beyond certain sweet spot the sound become a bit too dark and powerless, may be the devises are under too much stress...
I have repair my Audio Lab 8000A amp half way. Blew the emitter resistors when I tweaked the pre, by pass some output resistors in serial, and forgot to instal back the feedback loop.
As the circuit is overly complicate, I guess this will be a good candidate to cut out some stages to get better sound.
Perhaps I will just use the Macs circuitries whch is simple and good. But curious to see how FM circuit sounds like with the off the shelf components.
I always wanted to see how the SS amp sounded like if we use simple high voltage circuit to increase the head room.
Best rgds
William Lee
Singapore
Hallo Mr. Pass,
"a 10 foot pole" - what is this?
regarded FM Acoustics I have start this thread:
http://www.diyaudio.com/forums/soli...amplifier-absolute-state-art.html#post1963472
About 3 meter long wooden stick for "remote evaluation"... isn't it Nelson?!
Until this days amps with outdoor power supplies and choke filters are numbered (i. e. very rarely), mostly even in the highest available price class - e. g.
http://www.diyaudio.com/forums/soli...ke-pure-classa-se-despite-low-idle-power.html
Thus an universal outdoor amplifier power supply is helpful to check out the real (or the actually) quality standart of the used topology.
Without such a thing mostly the sonic character of power supply (in association with the GND management) overlapped the sonic character of the actually amp topology.
BTW - according my own experiences a cheap integrated amp between 30-50W output power sounds better in most cases together with a right designed outdoor power supply than an expensive high power amp device in genuine condition (i. e. with the inddor supply).
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.