Not a group buy, that was the minimum orderBut I also have a company to put it though, and are planning to maybe sell some boards. And maybe some special parts.
In all circumstances I can spare some for you....
Soren
Hi Soren,
Great. Would you be so kind to send me a PM when the trannies are arrived?
Thanks in advance.
Cheers,
E.
open question
Hi guys -- there is nothing veiled or hidden in this question - just my curiosity - interesting thread and I have been trying to keep up.
I'm probably seen by Mr Cordell as a bit of a Luddite - I'm one of those ignominious BJT journeymen
My question really is to do with my perception of the final product. I admit times may have moved on - or there are more refined design principles beyond what Hitachi set out in 1970
In my concert rig past we used Power MOSFET - specifically 2SK176 2SJ56. That may give some of you who have been around long enough an indication of my age.....ok ok ok ....I'm an old f.a.r.t.
The amplifiers we played with wer anywhere between 150W to 3KW - my career moved on just as PWM (switch explode) amplifiers were just coming out.
The amplifiers makers I remember the most were; Rauch, Beta, C-Audio, Monotech, Amcron/Crown, H|H, Carlsbro, Hill Audio, JBL, Hafler, Yamaha, Citronic, Peavey - "Oh god! ! ! no, not Peavey, pardon me while I throw up", Carver, Quad....just to mention a few (mix there of bipolar and FET)
At the time we had grown accustomed to FET Sluggish but powerful bass and crisp but tizzy top end. Where as the BJT variants of the day wernt quite as bomb proof as the FETs but there was a distinct quality difference once couple to various speaker combinations such as JBL - ElectroVoice, Turbo-Sound etc. The whole community of engineers said the same at the time. We used to use the FETS as Tractors and the BJTs as Buses to convey the audio across to a mass audience. BJTs seemed to have tighter more controlled and realistic bass - against FETs booom and overhang, and the BJT treble was sweeter against FETs tizzzzz and gritty top. This was especially noticable on the ones driven "full range" - the problems seemed to disapear when used on three way (Bass-Mid-Top) active crossovers.
In the past when these issues relating to impedance - transconductance and all the other technical anomalies you mentioned, the designs seem to go bonkers complicated (see pic)
Which is why when I came out of commercial side of audio and became a hobbiest, I stuck with BJTs.
Why, if the FET is superior in all the technical respects did the designs ming quite so much? - I remember always having to strap more 220pFs across the 2nd stages of their VX range because when you run 30 or more meters of multicore, that antenna picked up RF + creating its own parasitics on that PCB made out of cornbread and wood chippings (joke-- but they were rubbish), which then disturbed the quality even more. Just one of those areas where it's awesome speed was not required. Like having a military jet in a traffic jam.
I have used many amplifiers including my own designs for the various recording studios I am connected with. Not one is FET based - studio owners choices tend to favour Krell KSAs, ubiquitous Quad 405/606/909s, Mark Levinson's
I have a Hafler DH250 right here, I just cant listen to it - it's awful
I realise sound is a subjective thing - but out of all the community of PA system integrators, the consensus was pretty much the same.
I'm curious to know if things have improved and one could actually design something that sounds high-end without all the concomitant crudities? Could be something I choose to discover as a next project after my Transciptor Skeleton Amplifier.
Hi guys -- there is nothing veiled or hidden in this question - just my curiosity - interesting thread and I have been trying to keep up.
I'm probably seen by Mr Cordell as a bit of a Luddite - I'm one of those ignominious BJT journeymen
My question really is to do with my perception of the final product. I admit times may have moved on - or there are more refined design principles beyond what Hitachi set out in 1970
In my concert rig past we used Power MOSFET - specifically 2SK176 2SJ56. That may give some of you who have been around long enough an indication of my age.....ok ok ok ....I'm an old f.a.r.t.
The amplifiers we played with wer anywhere between 150W to 3KW - my career moved on just as PWM (switch explode) amplifiers were just coming out.
The amplifiers makers I remember the most were; Rauch, Beta, C-Audio, Monotech, Amcron/Crown, H|H, Carlsbro, Hill Audio, JBL, Hafler, Yamaha, Citronic, Peavey - "Oh god! ! ! no, not Peavey, pardon me while I throw up", Carver, Quad....just to mention a few (mix there of bipolar and FET)
At the time we had grown accustomed to FET Sluggish but powerful bass and crisp but tizzy top end. Where as the BJT variants of the day wernt quite as bomb proof as the FETs but there was a distinct quality difference once couple to various speaker combinations such as JBL - ElectroVoice, Turbo-Sound etc. The whole community of engineers said the same at the time. We used to use the FETS as Tractors and the BJTs as Buses to convey the audio across to a mass audience. BJTs seemed to have tighter more controlled and realistic bass - against FETs booom and overhang, and the BJT treble was sweeter against FETs tizzzzz and gritty top. This was especially noticable on the ones driven "full range" - the problems seemed to disapear when used on three way (Bass-Mid-Top) active crossovers.
In the past when these issues relating to impedance - transconductance and all the other technical anomalies you mentioned, the designs seem to go bonkers complicated (see pic)
Which is why when I came out of commercial side of audio and became a hobbiest, I stuck with BJTs.
Why, if the FET is superior in all the technical respects did the designs ming quite so much? - I remember always having to strap more 220pFs across the 2nd stages of their VX range because when you run 30 or more meters of multicore, that antenna picked up RF + creating its own parasitics on that PCB made out of cornbread and wood chippings (joke-- but they were rubbish), which then disturbed the quality even more. Just one of those areas where it's awesome speed was not required. Like having a military jet in a traffic jam.
I have used many amplifiers including my own designs for the various recording studios I am connected with. Not one is FET based - studio owners choices tend to favour Krell KSAs, ubiquitous Quad 405/606/909s, Mark Levinson's
I have a Hafler DH250 right here, I just cant listen to it - it's awful
I realise sound is a subjective thing - but out of all the community of PA system integrators, the consensus was pretty much the same.
I'm curious to know if things have improved and one could actually design something that sounds high-end without all the concomitant crudities? Could be something I choose to discover as a next project after my Transciptor Skeleton Amplifier.
Attachments
Last edited:
Hi guys -- there is nothing veiled or hidden in this question - just my curiosity - interesting thread and I have been trying to keep up.
I'm probably seen by Mr Cordell as a bit of a Luddite - I'm one of those ignominious BJT journeymen
My question really is to do with my perception of the final product. I admit times may have moved on - or there are more refined design principles beyond what Hitachi set out in 1970
In my concert rig past we used Power MOSFET - specifically 2SK176 2SJ56. That may give some of you who have been around long enough an indication of my age.....ok ok ok ....I'm an old f.a.r.t.
The amplifiers we played with wer anywhere between 150W to 3KW - my career moved on just as PWM (switch explode) amplifiers were just coming out.
The amplifiers makers I remember the most were; Rauch, Beta, C-Audio, Monotech, Amcron/Crown, H|H, Carlsbro, Hill Audio, JBL, Hafler, Yamaha, Citronic, Peavey - "Oh god! ! ! no, not Peavey, pardon me while I throw up", Carver, Quad....just to mention a few (mix there of bipolar and FET)
At the time we had grown accustomed to FET Sluggish but powerful bass and crisp but tizzy top end. Where as the BJT variants of the day wernt quite as bomb proof as the FETs but there was a distinct quality difference once couple to various speaker combinations such as JBL - ElectroVoice, Turbo-Sound etc. The whole community of engineers said the same at the time. We used to use the FETS as Tractors and the BJTs as Buses to convey the audio across to a mass audience. BJTs seemed to have tighter more controlled and realistic bass - against FETs booom and overhang, and the BJT treble was sweeter against FETs tizzzzz and gritty top. This was especially noticable on the ones driven "full range" - the problems seemed to disapear when used on three way (Bass-Mid-Top) active crossovers.
In the past when these issues relating to impedance - transconductance and all the other technical anomalies you mentioned, the designs seem to go bonkers complicated (see pic)
Which is why when I came out of commercial side of audio and became a hobbiest, I stuck with BJTs.
Why, if the FET is superior in all the technical respects did the designs ming quite so much? - I remember always having to strap more 220pFs across the 2nd stages of their VX range because when you run 30 or more meters of multicore, that antenna picked up RF + creating its own parasitics on that PCB made out of cornbread and wood chippings (joke-- but they were rubbish), which then disturbed the quality even more. Just one of those areas where it's awesome speed was not required. Like having a military jet in a traffic jam.
I have used many amplifiers including my own designs for the various recording studios I am connected with. Not one is FET based - studio owners choices tend to favour Krell KSAs, ubiquitous Quad 405/606/909s, Mark Levinson's
I have a Hafler DH250 right here, I just cant listen to it - it's awful
I realise sound is a subjective thing - but out of all the community of PA system integrators, the consensus was pretty much the same.
I'm curious to know if things have improved and one could actually design something that sounds high-end without all the concomitant crudities? Could be something I choose to discover as a next project after my Transciptor Skeleton Amplifier.
That design you have right there is a fine amp, difficult to sort out, but when you get it going its very good indeed.
Hey, thanks homey
I thought no one would notice with all the blameless activity going on
Hey, thanks homey
I thought no one would notice with all the blameless activity going on
A blameless can sound very good indeed, depending on how you compensate it but that design by stochino is excellent sounding amp, use different mosfets like toshibas 2sj201 and compl. This amp is in my top 3 of good sounding class AB amps of this forum, the irf640 s battle in this application, they just cant dissapate enough heat, that amp is very fast, needs carefull pcb design or it will oscilate. Some of my family members, they are classical music musicians prefer that amp to my electrocompaniets which cost me an arm and a leg. There is a couple of threads and a website that has some details about it. I could dig out the origianal articles of that amp if youd like to read through it. Its called the stochino.
You're right 15 years is a long gap -- but as I said in my rather long post to Mr Cordell (If indeed he is still reading) my expectations were set in a different epoch; and in more "industrial" circumstances.
Perhaps this is the right time for me to learn a bit more about FET design and the compromises that lay therein. Jacco, I'll order a couple of those devices, thanks
But at the end of the day, at the pinnacle top end of audio and hi-fi--- it doesnt matter how much we bang on about transconductance, charge carriers, co-valent bonds, high gate impedances, lower gate voltages and drain induced barrier lowering....if the final amplifier sounds crap at the end.
Last edited:
Hi guys -- there is nothing veiled or hidden in this question - just my curiosity - interesting thread and I have been trying to keep up.
I'm probably seen by Mr Cordell as a bit of a Luddite - I'm one of those ignominious BJT journeymen
My question really is to do with my perception of the final product. I admit times may have moved on - or there are more refined design principles beyond what Hitachi set out in 1970
In my concert rig past we used Power MOSFET - specifically 2SK176 2SJ56. That may give some of you who have been around long enough an indication of my age.....ok ok ok ....I'm an old f.a.r.t.
The amplifiers we played with wer anywhere between 150W to 3KW - my career moved on just as PWM (switch explode) amplifiers were just coming out.
The amplifiers makers I remember the most were; Rauch, Beta, C-Audio, Monotech, Amcron/Crown, H|H, Carlsbro, Hill Audio, JBL, Hafler, Yamaha, Citronic, Peavey - "Oh god! ! ! no, not Peavey, pardon me while I throw up", Carver, Quad....just to mention a few (mix there of bipolar and FET)
At the time we had grown accustomed to FET Sluggish but powerful bass and crisp but tizzy top end. Where as the BJT variants of the day wernt quite as bomb proof as the FETs but there was a distinct quality difference once couple to various speaker combinations such as JBL - ElectroVoice, Turbo-Sound etc. The whole community of engineers said the same at the time. We used to use the FETS as Tractors and the BJTs as Buses to convey the audio across to a mass audience. BJTs seemed to have tighter more controlled and realistic bass - against FETs booom and overhang, and the BJT treble was sweeter against FETs tizzzzz and gritty top. This was especially noticable on the ones driven "full range" - the problems seemed to disapear when used on three way (Bass-Mid-Top) active crossovers.
In the past when these issues relating to impedance - transconductance and all the other technical anomalies you mentioned, the designs seem to go bonkers complicated (see pic)
Which is why when I came out of commercial side of audio and became a hobbiest, I stuck with BJTs.
Why, if the FET is superior in all the technical respects did the designs ming quite so much? - I remember always having to strap more 220pFs across the 2nd stages of their VX range because when you run 30 or more meters of multicore, that antenna picked up RF + creating its own parasitics on that PCB made out of cornbread and wood chippings (joke-- but they were rubbish), which then disturbed the quality even more. Just one of those areas where it's awesome speed was not required. Like having a military jet in a traffic jam.
I have used many amplifiers including my own designs for the various recording studios I am connected with. Not one is FET based - studio owners choices tend to favour Krell KSAs, ubiquitous Quad 405/606/909s, Mark Levinson's
I have a Hafler DH250 right here, I just cant listen to it - it's awful
I realise sound is a subjective thing - but out of all the community of PA system integrators, the consensus was pretty much the same.
I'm curious to know if things have improved and one could actually design something that sounds high-end without all the concomitant crudities? Could be something I choose to discover as a next project after my Transciptor Skeleton Amplifier.
You raise a lot of good questions. Things have indeed improved a great deal over the last twenty years, in both MOSFET and BJT devices and design techniques. It is also true that it is equally easy to make a bad-sounding amplifier from MOSFETs or BJTs. With good engineering and attention to detail, it is possible to make exceptionally good amplifiers from either technology. Each technology has its own advantages.
I would not be inclined to paint either technology with too broad a brush with generalities like "the bass of the MOSFET amplifiers is thin" and the like. Those kinds of characteristics generally are more a function of the designer and the specific implementaion than the transistor technology.
Cheers,
Bob
Hi Bob
Thank you for responding - I hope you are well and where you are is sunny
Yeah - I realise it was a generalisation - at the time it did become an industry concensus about practical application of the technology
Lets not forget these amps were pretty much variations of the same theme. We just couldnt get rid of that tizzziness in the treble - at power it would hurt compression drivers and the feedback threshold on stage was reduced. So we opted for the BJTs to always do tops. You'll still find the same activity at concerts today.
But back into high end land - this thread has inspired me to keep an open mind and have a go - see what loony mistakes I can make with them
Just as an amusing aside - Actually the bass always seemed to be hefty with FET technology - - With Citronic 4x PPX1200 Bridged run off of 3rd Generation active crossorver set at resonant "A" < 440 hz -12db/octave drop into JBL "W" bins (a familiy of 4 could live inside this chamber) (12x 15" Subs) as sub re-inforcement - what we could do with that heart beat at the beginning of Pink Floyd's Dark Side of the Moon, which was to make all the other systems people (lighting, effects, H&S officials) feel physically queeezy with the undamped natural subsonic - especially after breakfast.
Half chewed bacon sandwiches everywhere
H&S officals sending in written complaints to our managers .....ahhhh they were the days
Last edited:
Hi Bob
Thank you for responding - I hope you are well and where you are is sunny
Yeah - I realise it was a generalisation - at the time it did become an industry concensus about practical application of the technology
Lets not forget these amps were pretty much variations of the same theme. We just couldnt get rid of that tizzziness in the treble - at power it would hurt compression drivers and the feedback threshold on stage was reduced. So we opted for the BJTs to always do tops. You'll still find the same activity at concerts today.
But back into high end land - this thread has inspired me to keep an open mind and have a go - see what loony mistakes I can make with them
Just as an amusing aside - Actually the bass always seemed to be hefty with FET technology - - With Citronic 4x PPX1200 Bridged run off of 3rd Generation active crossorver set at resonant "A" < 440 hz -12db/octave drop into JBL "W" bins (a familiy of 4 could live inside this chamber) (12x 15" Subs) as sub re-inforcement - what we could do with that heart beat at the beginning of Pink Floyd's Dark Side of the Moon, which was to make all the other systems people (lighting, effects, H&S officials) feel physically queeezy with the undamped natural subsonic - especially after breakfast.
Half chewed bacon sandwiches everywhere
They were indeed the days. I have always enjoyed Dark Side of the Moon and its sonic acrobatics. Really deep, clean bass that is not over-blown makes everything else in the spectrum sound better.
The MOSFET designs might have been better in the bass as a result of higher peak output current capability and possible lack of protection circuits. There is also a chance that there was less influence of ripple and other garbage from the power rails, as for a given wattage rating MOSFET amplifiers often need higher rail voltages. This is all just speculation.
Cheers,
Bob
What do you recommend
I havent played with FETs from a design aspect properly for donkeys
I like to start in miniature and work my way up
Rather than leap up to doing some stonking great amp - I'm thinking of perhaps trying a FET based Pre-amp with RIAA - Line and other inputs + A FET H'phone out at low rail PD
This way I don't have my eyebrows removed again and end up looking like my avatar
My preference in BJTs at signal level are BC547,549,550,560,557,638, 2N3816,2N3838, - 2SC2705, 2SA1145, ZTX650,750....etc...
Now I have absolutely no idea what are the best small signal FETs would be in such an application...anyone any ideas?
I havent played with FETs from a design aspect properly for donkeys
I like to start in miniature and work my way up
Rather than leap up to doing some stonking great amp - I'm thinking of perhaps trying a FET based Pre-amp with RIAA - Line and other inputs + A FET H'phone out at low rail PD
This way I don't have my eyebrows removed again and end up looking like my avatar
My preference in BJTs at signal level are BC547,549,550,560,557,638, 2N3816,2N3838, - 2SC2705, 2SA1145, ZTX650,750....etc...
Now I have absolutely no idea what are the best small signal FETs would be in such an application...anyone any ideas?
20/10 for attitude, Loopy! 
...as opposed to clueless noobs wanting to build a 1000W amp as a first project.
Sorry, can't help with the actual question. Never did quite get the hang of them new-fangled fet things myself.
I used to make MY coils from 14Ga oxygen free copper. So there!
IMO -- or FWIW: 18 awg or 1600 circ.mils is a very good compromise number.... this gives a suitably high skin depth starting at 17KHz and suitable current for high peak current levels. But, more importantly to myself is the size of the field radiated from the PA output coil and cross-talk. So care in placement of this coil is needed. As would be the left-right speaker leads and power leads and pcb layout.
Outside the chassis, L-R speaker leads in close proxy to one another and power (ac) and signal leads needs to be examined for cross-talk under two channel load conditions or even better is measured in-place. Most all ac wiring to and signal in/out of preamps/power amps/sources are a rats nest all near one another.
THx-RNMarsh
Last edited:
I agree, #18 is a good compromise, including regarding skin depth and the size of the radiated field.
A 2-layer #18 coil in a radial orientation may be smaller and have a smaller magnetic footprint, but this approach is probably not necessary if the coil value is only 1uH.
Coil diameter of 1/2 inch is good.
Cheers,
Bob
The best low noise, cheap JFET available now is the BF862 from NXP. This was confirmed independently by Dimitri who measured a number of different devices types. The BF862 came out at c. 0.8 nV rt/Hz and had a low 1/f corner.
You can get them from Mouser or DK for really good prices
There is no P complement. However, single ended designs are still excellent performers (see Self's single ended MC preamp and Syn08 BF862 for some SOT examples).
You can get them from Mouser or DK for really good prices
There is no P complement. However, single ended designs are still excellent performers (see Self's single ended MC preamp and Syn08 BF862 for some SOT examples).
Are these devices available in TO92 package? Just looked at Mouser/ DK and they seem to be all SMD.
Although not cheap, the JFETs from Linear Systems are quite good in the noise department. The LSK389 dual monolithic is of particular interest. I used these in my VinylTrak preamp with very good results. The single version is the LSK170. They also now have the LSJ74 single complement.
For a lower-transconductance, lower capacitance dual, the LSK489 is also quite a good part with very low noise given its lower transconductance.
Cheers,
Bob
Fairly low noise FET alternatives.
The BF862 does look very good. Also, onsemi has an interestiing line of JFETS.
The CPH3910 has a narrow (20-40mA, and a high gm(40mS typical) -- fairly low noise. The 2sk932 has a narrow Idss ranges(7.3-12,10-17,14.5-24ma) and a very high gm (50mS typical.) The 2sk3557 is speced with low LF NF, good gm (35mS, and has versions with 10-20mA Idss and 16-32mA Idss) They have several different JFETS with several different optimized characteristics. the noise spec for the 2sk3557 looks especially good -- but is specified in terms of NF instead of en/in. Even so, the NF of 3dB at (10Hz, 1kOhm, Id=1mA) and 2dB at (rin=100 Ohms,1kHz.id=1mA) seems pretty good.
Havent tried them yet -- until a few weeks ago, they had a TO92 version, but seems to have disappeared. Price of the JFETs is at the jellybean level.
John Dyson
The BF862 does look very good. Also, onsemi has an interestiing line of JFETS.
The CPH3910 has a narrow (20-40mA, and a high gm(40mS typical) -- fairly low noise. The 2sk932 has a narrow Idss ranges(7.3-12,10-17,14.5-24ma) and a very high gm (50mS typical.) The 2sk3557 is speced with low LF NF, good gm (35mS, and has versions with 10-20mA Idss and 16-32mA Idss) They have several different JFETS with several different optimized characteristics. the noise spec for the 2sk3557 looks especially good -- but is specified in terms of NF instead of en/in. Even so, the NF of 3dB at (10Hz, 1kOhm, Id=1mA) and 2dB at (rin=100 Ohms,1kHz.id=1mA) seems pretty good.
Havent tried them yet -- until a few weeks ago, they had a TO92 version, but seems to have disappeared. Price of the JFETs is at the jellybean level.
John Dyson