My latest project is another SIT follower amplifier. You may have noticed that I favour SIT followers, since I had built a 2SK180 choke loaded follower and a THF-51S CCS Mu Follower common drain amplifier, and now I am building another SIT common drain amplifier. It's just that in my system, the SIT followers sound the best. After listening to the followers, I realized that the single ended common drain SIT amplifiers that I had built in the past had a bit too much harmonic distortion. With followers and their lower distortion, the music has a clarity that I did not hear from my common drain amplifiers.
The two latest diyAudio Sony VFET amplifiers designed by Mr. Pass are also followers and are much loved by their owners, so another VFET follower amplifier can't be a bad thing. This will be a bit different, being choke loaded instead of Constant Current Source (CCS) loaded.
I decided to go with choke load for a few reasons. Deciding factors include low power dissipation because the choke dissipates very little power, unlike a CCS. Therefore I can re-use my 22VDC power supplies since the choke drops very little voltage, again unlike a CCS. And I really enjoy the sound of my choke loaded 2SK180 follower amplifier, so another choke loaded follower is welcomed. The lower heat dissipation will be appreciated when I use it in the summertime. Also by choke loading, the build will be very simple. There are not many parts. The amplifier circuit board will be very simple, with a resitor or two, a Zener diode, and the VFET. It will be mainly used to provide connection points for power, bias voltage, input signal, and speaker out wiring. I will have a board for the bias supply, but the input and output coupling capacitors will be chassis mounted. My Luminaria preamp has enough voltage output and low enough output impedance, so a separate voltage gain stage and an input buffer are not required.
As mentioned the choke load drops very little voltage, allowing an operating point of 21V Vds and 1.5A Iq from a 22V power supply. This operating point is similar to that of the diyAudio Sony VFET amplifiers. The typical 18V power transformer and 22VDC power supply that is used in many First Watt clone amplifiers can be used in this amplifier. The diyAudio power supply would need a modification since V- is not required. The V- half of the board would need the capacitors repositioned so that V+ is outputted, allowing each half of the supply to power one channel. It would be nearly dual mono with only the power transformer primary shared by the two channels. So if you have a pair of Sony VFETs and an unused amplifier, this amplifier can be quite easily built.
I will be re-using the chassis and power supply of an existing amplifier that started life as a Hammond 193V choke loaded VFET L'Amp, and then repurposed to accommodate a F2J. This new amp is reminiscent of Michael Rothacher's L'Amp in choke load configuration, except this will be common drain instead of common source. So it is back to the beginning for the chassis.
I have ordered the parts that I need, and they should arrive tomorrow. So let the build begin!
The two latest diyAudio Sony VFET amplifiers designed by Mr. Pass are also followers and are much loved by their owners, so another VFET follower amplifier can't be a bad thing. This will be a bit different, being choke loaded instead of Constant Current Source (CCS) loaded.
I decided to go with choke load for a few reasons. Deciding factors include low power dissipation because the choke dissipates very little power, unlike a CCS. Therefore I can re-use my 22VDC power supplies since the choke drops very little voltage, again unlike a CCS. And I really enjoy the sound of my choke loaded 2SK180 follower amplifier, so another choke loaded follower is welcomed. The lower heat dissipation will be appreciated when I use it in the summertime. Also by choke loading, the build will be very simple. There are not many parts. The amplifier circuit board will be very simple, with a resitor or two, a Zener diode, and the VFET. It will be mainly used to provide connection points for power, bias voltage, input signal, and speaker out wiring. I will have a board for the bias supply, but the input and output coupling capacitors will be chassis mounted. My Luminaria preamp has enough voltage output and low enough output impedance, so a separate voltage gain stage and an input buffer are not required.
As mentioned the choke load drops very little voltage, allowing an operating point of 21V Vds and 1.5A Iq from a 22V power supply. This operating point is similar to that of the diyAudio Sony VFET amplifiers. The typical 18V power transformer and 22VDC power supply that is used in many First Watt clone amplifiers can be used in this amplifier. The diyAudio power supply would need a modification since V- is not required. The V- half of the board would need the capacitors repositioned so that V+ is outputted, allowing each half of the supply to power one channel. It would be nearly dual mono with only the power transformer primary shared by the two channels. So if you have a pair of Sony VFETs and an unused amplifier, this amplifier can be quite easily built.
I will be re-using the chassis and power supply of an existing amplifier that started life as a Hammond 193V choke loaded VFET L'Amp, and then repurposed to accommodate a F2J. This new amp is reminiscent of Michael Rothacher's L'Amp in choke load configuration, except this will be common drain instead of common source. So it is back to the beginning for the chassis.
I have ordered the parts that I need, and they should arrive tomorrow. So let the build begin!
Attachments
I was lucky enough to win the VFET an it opened up all new horizons for music listening for me on the horn speakers. All old records sound good, best amp I have ever listened to in my system. Cannot be replaced at this point.
As I am about to put a couple of tokin thf51s to use in a cpouple different ways, I look forward to your listening impressions since you have done all that already.
As I am about to put a couple of tokin thf51s to use in a cpouple different ways, I look forward to your listening impressions since you have done all that already.
I have been using very similar amp for 108dB horn tweeters for 3 years. The good thing about it is it does not require gain stage, so I can just feed the output of my DAC.
Mine is with an adjustable self bias, and currently operating at 20V 1.8A. Mine does not have D2 diode, but no problem so far.
Mine is with an adjustable self bias, and currently operating at 20V 1.8A. Mine does not have D2 diode, but no problem so far.
The Choke
The Hammond 193V that I had used in my choke loaded L'Amp are now in my 2SK180 choke loaded follower amps so for this project I will be using a different choke. One big issue with my 193V L'Amp was that it was very, very heavy. It was built in a stereo chassis and I could barely lift it. For this build I will be reusing the same stereo chassis but I want it to be lighter weight, so I want to use a smaller and lighter choke.
Looking at Hammond catalog two candidates appeared - the 193T and the 159ZC. Both have inductance of 60mH, maximum current of 2A, resistance of 0.7 Ohm, and weight of 2.5 pounds. The main difference is the 193T is enclosed and potted whereas the 159ZC is open with no covering. Comparing them to the 193V, the 193V has inductance of 150mH, maximum current of 3A, resistance of 1 Ohm, and weight of 9 pounds.
So the main issue with the two candidates is inductance. The lower weight and lower resistance are welcomed, and their maximum current is not a limiting factor. The lack of enclosure for the 159ZC should not be an issue either, as it is no different than having open frame chokes that I use in CLC filters in the power supplies of my single ended SIT amplifiers. The only question is whether the lower inductance is too detrimental to the low frequency response of the amplifier.
To investigate, I did some LTSpice simulations and the results were favourable. The 193V loaded circuit's low frequency response was flat down to about 20 Hz and the 159ZC loaded circuit was about 0.3 dB down at 20 Hz. That is good enough for me. Of course, the results of the simulations should not be taken as gospel, but they do give an indication that at least the circuit will work, and hopefully the actual performance will be somewhat similar to the simulation. So based on that I decided to go with the 159ZC. The 159ZC is also the lowest price of the three, US$30.82 at Mouser.
I also ran some 159ZC circuit simulations for harmonic distortion but again, actual performance may differ from simulation results.
The Hammond 193V that I had used in my choke loaded L'Amp are now in my 2SK180 choke loaded follower amps so for this project I will be using a different choke. One big issue with my 193V L'Amp was that it was very, very heavy. It was built in a stereo chassis and I could barely lift it. For this build I will be reusing the same stereo chassis but I want it to be lighter weight, so I want to use a smaller and lighter choke.
Looking at Hammond catalog two candidates appeared - the 193T and the 159ZC. Both have inductance of 60mH, maximum current of 2A, resistance of 0.7 Ohm, and weight of 2.5 pounds. The main difference is the 193T is enclosed and potted whereas the 159ZC is open with no covering. Comparing them to the 193V, the 193V has inductance of 150mH, maximum current of 3A, resistance of 1 Ohm, and weight of 9 pounds.
So the main issue with the two candidates is inductance. The lower weight and lower resistance are welcomed, and their maximum current is not a limiting factor. The lack of enclosure for the 159ZC should not be an issue either, as it is no different than having open frame chokes that I use in CLC filters in the power supplies of my single ended SIT amplifiers. The only question is whether the lower inductance is too detrimental to the low frequency response of the amplifier.
To investigate, I did some LTSpice simulations and the results were favourable. The 193V loaded circuit's low frequency response was flat down to about 20 Hz and the 159ZC loaded circuit was about 0.3 dB down at 20 Hz. That is good enough for me. Of course, the results of the simulations should not be taken as gospel, but they do give an indication that at least the circuit will work, and hopefully the actual performance will be somewhat similar to the simulation. So based on that I decided to go with the 159ZC. The 159ZC is also the lowest price of the three, US$30.82 at Mouser.
I also ran some 159ZC circuit simulations for harmonic distortion but again, actual performance may differ from simulation results.
Attachments
Thank you, everyone. 
analog_sa, I can't say that I prefer the sound of inductive loading to CCS loading. Although I am listening to my choke loaded 2SK180 follower right now and I really enjoy the sound, I also have my mu follower THF-51S that is not in use right now because it is still too warm to put it into the system. It dissipates about 200W per channel so it probably won't go in for another few weeks, when I need to heat my room. Then I will listen to it during the cold winter months. Since my room heating is provided by electric baseboards, supplemental heat from Class A amplifiers is great. I have to pay for electricity to heat my room anyways so if the amplifier provides heat and music, the music is free.
For this build I chose choke loading so that I can reuse an existing power transformer and for the lower heat output since I intend to use this amp in the summer, and also for its simplicity. And it's a different implementation from other Sony VFET amplifiers on diyAudio. It's like playing with single ended triodes, which I did until I discovered SITs/VFETs. So my preference is actually single ended SITs/VFETs. I like to try different configurations. I started with common source and now I am at common drain.
analog_sa, I can't say that I prefer the sound of inductive loading to CCS loading. Although I am listening to my choke loaded 2SK180 follower right now and I really enjoy the sound, I also have my mu follower THF-51S that is not in use right now because it is still too warm to put it into the system. It dissipates about 200W per channel so it probably won't go in for another few weeks, when I need to heat my room. Then I will listen to it during the cold winter months. Since my room heating is provided by electric baseboards, supplemental heat from Class A amplifiers is great. I have to pay for electricity to heat my room anyways so if the amplifier provides heat and music, the music is free.
For this build I chose choke loading so that I can reuse an existing power transformer and for the lower heat output since I intend to use this amp in the summer, and also for its simplicity. And it's a different implementation from other Sony VFET amplifiers on diyAudio. It's like playing with single ended triodes, which I did until I discovered SITs/VFETs. So my preference is actually single ended SITs/VFETs. I like to try different configurations. I started with common source and now I am at common drain.
I'm old and weak so 40 or 50 kg amp is too heavy for me. 
The Triad C56-U has much lower inductance than the Hammond 159ZC so the low frequency response would not be as good. It all depends on what you want from your amp. Perhaps Hammond are not readily available in Europe but the 159ZC is made in Canada and available for the equivalent of about 21 Euro, so not too expensive.
The Triad C56-U has much lower inductance than the Hammond 159ZC so the low frequency response would not be as good. It all depends on what you want from your amp. Perhaps Hammond are not readily available in Europe but the 159ZC is made in Canada and available for the equivalent of about 21 Euro, so not too expensive.
SIT FOOLLOWER
I know you'll not like it, it's too complicated
but I simply like an idea of having proper biasing mechanismus
besides, with minor changes (aux psu voltage) and resistor or two values, good for Tokin Biguns and eeeny Sony
in other words - any functional SIT will work, without worry about Tempco
took 2SK77B model simply because it's best model I have
choke value put arbitrary - 50m@0R47
CCS noted as expensive LT3092, aux PSU - small 3VA tranny (or aux winding on main Donut), few bobs and LM317 will do
U2 is not ultrxpensive as noted, regular LM385-1V2 is appropriate
I know you'll not like it, it's too complicated
but I simply like an idea of having proper biasing mechanismus
besides, with minor changes (aux psu voltage) and resistor or two values, good for Tokin Biguns and eeeny Sony
in other words - any functional SIT will work, without worry about Tempco
took 2SK77B model simply because it's best model I have
choke value put arbitrary - 50m@0R47
CCS noted as expensive LT3092, aux PSU - small 3VA tranny (or aux winding on main Donut), few bobs and LM317 will do
U2 is not ultrxpensive as noted, regular LM385-1V2 is appropriate
Attachments
Last edited:
You're right, it is too complicated for me. My brain can't comprehend it.
However I can see a use, especially since VFETs and Tokins are getting harder to find. A biasing circuit that can handle devices that have tempco issues that might otherwise make them unfit in a simple fixed bias circuit could then be used in an amplifier.
Fortunately I have had only one 2SK180 frustrate me because of temperature/biasing problems and I finally gave up on it and bought another to replace it. The 2SJ28s that I will be using have been proven in use to have no biasing issues.
However, looking at your schematic gave me the idea of putting a 12V bias supply on top of the V+ for a maximum of 34V instead of my previous scheme of using a 30V stand-alone bias supply.
However I can see a use, especially since VFETs and Tokins are getting harder to find. A biasing circuit that can handle devices that have tempco issues that might otherwise make them unfit in a simple fixed bias circuit could then be used in an amplifier.
Fortunately I have had only one 2SK180 frustrate me because of temperature/biasing problems and I finally gave up on it and bought another to replace it. The 2SJ28s that I will be using have been proven in use to have no biasing issues.
However, looking at your schematic gave me the idea of putting a 12V bias supply on top of the V+ for a maximum of 34V instead of my previous scheme of using a 30V stand-alone bias supply.
Attachments
well, as we are having more and more Wakootalians around, that just reminds us that all roads are leading to their Capital
as above
I'm just stubborn, always striving for complicated Swiss Knife solutions
Pa is always making simplest and sweetest, so what else is left for me ........
