Hi Ben and company,
I am getting my feet wet in the SIT world, drinking from the firehouse over the past 48hrs. I have mostly dealt in tubes, but would like to try my hand in single-ended SIT design and have been reading your thread with much interest. A "simple" low parts count design is up my alley for the first time.
I recently purchased THF-51S quad for Mr. Pass' recent announcement of upcoming project. I plan to save a pair for his design, but would like to experiment with my own, and perhaps even try other SIT, although it seems numbers are dwindling.
I have a few questions if you do not mind.
1) In substitution of the THF-51S for 2SK180 in common source configuration, given the input capacitance of 5000pF, are you still seeing a wide HF bandwidth with your Luminaria SIT preamp? I believe NP recently suggested common source configuration of THF-51S would necessitate output Z of 10ohm from the previous stage for wide bandwidth from his measurements.
2) I had also come across Russian 2P926B low power SIT. Similar to your thought to use 2P926B as voltage gain stage as input to a 2SK180 follower output, how about 2P926B as follower input buffer? Although from the design below, seems Ciss of 2P926B is 1600pF. Perhaps Sony 2SK60 as follower input buffer?
SIT-Preamp
Some of my tube designs below, my DIY credentials
I am getting my feet wet in the SIT world, drinking from the firehouse over the past 48hrs. I have mostly dealt in tubes, but would like to try my hand in single-ended SIT design and have been reading your thread with much interest. A "simple" low parts count design is up my alley for the first time.
I recently purchased THF-51S quad for Mr. Pass' recent announcement of upcoming project. I plan to save a pair for his design, but would like to experiment with my own, and perhaps even try other SIT, although it seems numbers are dwindling.
I have a few questions if you do not mind.
1) In substitution of the THF-51S for 2SK180 in common source configuration, given the input capacitance of 5000pF, are you still seeing a wide HF bandwidth with your Luminaria SIT preamp? I believe NP recently suggested common source configuration of THF-51S would necessitate output Z of 10ohm from the previous stage for wide bandwidth from his measurements.
2) I had also come across Russian 2P926B low power SIT. Similar to your thought to use 2P926B as voltage gain stage as input to a 2SK180 follower output, how about 2P926B as follower input buffer? Although from the design below, seems Ciss of 2P926B is 1600pF. Perhaps Sony 2SK60 as follower input buffer?
SIT-Preamp
Some of my tube designs below, my DIY credentials
Attachments
Good afternoon chaps, instead of a Greedy Boy, let me introduce myself as a naughty boy.
I'm coming from a similar place to L0rdGwyn - I've been building SE-OTL amps and DACs for a while. Here's my latest creation, an SE-OTL amp using one 13E1 tube per channel for a glorious one and bit Watts per channel - I use this with 15ohm Lowther BLH speakers.
I was planning a MoFo project but the recent furore about the VFET amp lottery got me thinking about the pair of THF-51S SITs I have. I was initially thinking of ZM's SissySIT but came across this thread, which, in it's follower incarnation is very similar to the MoFo so I think this is probably the way I'll proceed. I am working on a 2P29L-based preamp with close to 20dB of gain so a follower with just current gain should better fit my requirements.
My requirement for this amp is to power a high-efficiency full range speaker in a Nelson Pass SLOB project (I'll use a chip amp for the bass unit, crossing over at circa 150-180Hz. ideally I would like to run the full-range speaker without any crossover, instead bandwidth limiting the amplifier driving it by rolling off the response at the bottom end - I had a plan to do that with the MoFo by adjusting the hi-pass filter point via the input cap and running a smaller inductor. I have built up a single MoFo to test this aproach out and take some FR measurements, which I'll be able to do as soon as the new preamp is ready.
Comments/observations?
Cheers
I'm coming from a similar place to L0rdGwyn - I've been building SE-OTL amps and DACs for a while. Here's my latest creation, an SE-OTL amp using one 13E1 tube per channel for a glorious one and bit Watts per channel - I use this with 15ohm Lowther BLH speakers.
I was planning a MoFo project but the recent furore about the VFET amp lottery got me thinking about the pair of THF-51S SITs I have. I was initially thinking of ZM's SissySIT but came across this thread, which, in it's follower incarnation is very similar to the MoFo so I think this is probably the way I'll proceed. I am working on a 2P29L-based preamp with close to 20dB of gain so a follower with just current gain should better fit my requirements.
My requirement for this amp is to power a high-efficiency full range speaker in a Nelson Pass SLOB project (I'll use a chip amp for the bass unit, crossing over at circa 150-180Hz. ideally I would like to run the full-range speaker without any crossover, instead bandwidth limiting the amplifier driving it by rolling off the response at the bottom end - I had a plan to do that with the MoFo by adjusting the hi-pass filter point via the input cap and running a smaller inductor. I have built up a single MoFo to test this aproach out and take some FR measurements, which I'll be able to do as soon as the new preamp is ready.
Comments/observations?
Cheers
LOrdGwyn,
Beautiful tubes. I used to do single ended tubes until about three or four years ago. I must say though that my creations did not look as beautiful as yours.
The Luminaria has an output impedance of about 75 Ohms. The THF-51S in common source does roll off a bit. According to LTSpice simulation, it is about 2dB down at 20kHz. I have the BAF2015 THF-51S (Singing Bush without input buffer) and I also tried the THF-51S 193V common source configuration for a couple of days, and did not think that the high end was rolled off. However I may have to rethink that as I hear more high end with the Fokin. That is even though I can't hear anywhere close to 20KHz if I listened to a frequency scan.
However, since I prefer the follower configuration now the discussion is moot. The follower configuration has an extended high frequency response.
For the KP926B for voltage gain stage, I will play with it but probably will end up similar to the operating point of pinholer's 2SJ28 voltage gain stage (but V+ instead of V-). For input buffer I was thinking DN2540 as in the Luminaria. It's too bad that I haven't been able to find a Spice model of the KP926B though.
Beautiful tubes. I used to do single ended tubes until about three or four years ago. I must say though that my creations did not look as beautiful as yours.
The Luminaria has an output impedance of about 75 Ohms. The THF-51S in common source does roll off a bit. According to LTSpice simulation, it is about 2dB down at 20kHz. I have the BAF2015 THF-51S (Singing Bush without input buffer) and I also tried the THF-51S 193V common source configuration for a couple of days, and did not think that the high end was rolled off. However I may have to rethink that as I hear more high end with the Fokin. That is even though I can't hear anywhere close to 20KHz if I listened to a frequency scan.
However, since I prefer the follower configuration now the discussion is moot. The follower configuration has an extended high frequency response.
For the KP926B for voltage gain stage, I will play with it but probably will end up similar to the operating point of pinholer's 2SJ28 voltage gain stage (but V+ instead of V-). For input buffer I was thinking DN2540 as in the Luminaria. It's too bad that I haven't been able to find a Spice model of the KP926B though.
Attachments
Fokin Measurements Part 2
The next day I remembered that I have a HP 3310A function generator. I purchased it about 20 or so years ago as surplus equipment for about $100. It was about 30 years old when I purchased it. Brand new it was about $1100 in 1970 dollars, so it was an expensive piece of equipment at the time.
I hooked it up to my Focusrite/REW test rig and let it warm up, and it stabilized after about one and a half hours. I did some baseline measurements of the 3310A to determine its own distortion and it was fairly consistent throughout its voltage output range.
The maximum voltage output was 10.5 Vrms, so I was able to use it to determine the distortion of the amps up to 10W. The 3310A's distortion profile showed consistent distortion levels throughout its harmonic range. Since the Fokin at its lower power output did not have higher order harmonics in its distortion profile, I discarded the harmonics that were higher than the fourth order to present a more realistic picture of the amplifier's harmonic distortion profile, assuming that those discarded higher order harmonics were a reflection of the 3310A's distortion profile. So just close your right eye and only use your left eye to look at the FFT plots. I think it is close enough to locate the amplifier's distortion neighbourhood.
The measurements show that the 2SK180 used as a follower has quite low harmonic distortion. At 10W output the THD was less than 0.20%. The measurements in my previous post showed the Luminaria's distortion profile added to the Fokin's profile.
I really like the sound of the Luminaria/Fokin combination in my system. There is enough second and third harmonic distortion but not too much as to muddle the sound. Before I converted the 2SK180 amp to follower, I was very happy listening to the higher harmonic distortion of a single ended common source amplifier. But now I can hear the added clarity of the 2SK180 in single ended common drain mode and I am impressed. There seems to me to be more high frequency detail that I had not heard before. I said it previously but the sound has attained a higher level of clarity. It may be just my imagination, but the sound is closer to live when listening to good recordings. I'm not a critical listener who analyzes minute details in recordings. I just listen and differences in what I hear in familiar recordings leave certain impressions with me.
Now the next step is build a voltage gain stage for the Fokins. It can be transformer based or electronics based. Mark Johnson has designed several stages for the much in-demand DIY Sony VFET amplifier (Scourge, Bulwark, Marauder, Dreadnought "front end" cards for DIY VFET amp). Pinholer in his 2SJ28-2SK180 build is using a Sony 2SJ28 for his voltage gain stage (Mo' Chokes: Building 2SJ28 - 2SK180 Monoblocks).
I have two 2SJ28 that I can use but I want to use a N channel device so that I can also use the power supply for an input buffer. I have some Russian KP926 in transit and I will probably use them for a gain stage. For a buffer I am thinking of using a DN2540 as in the Luminaria. I don't need a voltage gain stage but it will be a good exercise, plus the amp will then be more universal and can be used with all preamps, and the full power potential of the amplifier can be utilized. Of course if you have a preamp that can output about 16Vrms with good distortion profile, then no voltage gain stage is needed to get the full power out of the amplifier.
Meanwhile, I'm enjoying the music.
The next day I remembered that I have a HP 3310A function generator. I purchased it about 20 or so years ago as surplus equipment for about $100. It was about 30 years old when I purchased it. Brand new it was about $1100 in 1970 dollars, so it was an expensive piece of equipment at the time.
I hooked it up to my Focusrite/REW test rig and let it warm up, and it stabilized after about one and a half hours. I did some baseline measurements of the 3310A to determine its own distortion and it was fairly consistent throughout its voltage output range.
The maximum voltage output was 10.5 Vrms, so I was able to use it to determine the distortion of the amps up to 10W. The 3310A's distortion profile showed consistent distortion levels throughout its harmonic range. Since the Fokin at its lower power output did not have higher order harmonics in its distortion profile, I discarded the harmonics that were higher than the fourth order to present a more realistic picture of the amplifier's harmonic distortion profile, assuming that those discarded higher order harmonics were a reflection of the 3310A's distortion profile. So just close your right eye and only use your left eye to look at the FFT plots. I think it is close enough to locate the amplifier's distortion neighbourhood.
The measurements show that the 2SK180 used as a follower has quite low harmonic distortion. At 10W output the THD was less than 0.20%. The measurements in my previous post showed the Luminaria's distortion profile added to the Fokin's profile.
I really like the sound of the Luminaria/Fokin combination in my system. There is enough second and third harmonic distortion but not too much as to muddle the sound. Before I converted the 2SK180 amp to follower, I was very happy listening to the higher harmonic distortion of a single ended common source amplifier. But now I can hear the added clarity of the 2SK180 in single ended common drain mode and I am impressed. There seems to me to be more high frequency detail that I had not heard before. I said it previously but the sound has attained a higher level of clarity. It may be just my imagination, but the sound is closer to live when listening to good recordings. I'm not a critical listener who analyzes minute details in recordings. I just listen and differences in what I hear in familiar recordings leave certain impressions with me.
Now the next step is build a voltage gain stage for the Fokins. It can be transformer based or electronics based. Mark Johnson has designed several stages for the much in-demand DIY Sony VFET amplifier (Scourge, Bulwark, Marauder, Dreadnought "front end" cards for DIY VFET amp). Pinholer in his 2SJ28-2SK180 build is using a Sony 2SJ28 for his voltage gain stage (Mo' Chokes: Building 2SJ28 - 2SK180 Monoblocks).
I have two 2SJ28 that I can use but I want to use a N channel device so that I can also use the power supply for an input buffer. I have some Russian KP926 in transit and I will probably use them for a gain stage. For a buffer I am thinking of using a DN2540 as in the Luminaria. I don't need a voltage gain stage but it will be a good exercise, plus the amp will then be more universal and can be used with all preamps, and the full power potential of the amplifier can be utilized. Of course if you have a preamp that can output about 16Vrms with good distortion profile, then no voltage gain stage is needed to get the full power out of the amplifier.
Meanwhile, I'm enjoying the music.
Attachments
-
2SK180 193V Follower HP3310A R Ch 5W.jpg -
2SK180 193V Follower HP3310A R Ch 1W.jpg -
2SK180 193V Follower HP3310A L Ch 8R 10W.jpg -
2SK180 193V Follower HP3310A L Ch 8R 5W.jpg -
2SK180 193V Follower HP3310A L Ch 8R 1W.jpg -
HP 3310A 10,0Vrms.jpg -
HP 3310A 3,0Vrms.jpg -
HP 3310A.jpg -
2SK180 193V Follower HP3310A R Ch 10W.jpg
Ben Mah: Thanks for the distortion measurements. I'm in the market for a new function generator and have been combing through all the models on ebay trying to find something with low numbers. That 3310A seems to measure a lot better than the other analog models. The Kenwood AG-204D is rated for 0.02% THD but it seems to be pretty rare. Anything lower than that is getting into digital territory.
Yes, I also noticed that when I switched to the 2SK82-2SJ28 source follower amplifier from a common source 2SJ28 amp that the midrange and highs were clearer. The high end sparkle was more evident and tones were more pure. My Luminaria preamp provides enough 2nd harmonic, so no need for too much more in the power amplifier.
That's a good start!
So just drop the THF-51S in place of the 2SK180 in your schemating on post #109;
https://www.diyaudio.com/forums/pas...-choke-loaded-2sk180-lamp-11.html#post6601437
What voltage are you running for the power supply now? I've picked up on the need to tweak the operating point for different SITs to optimise performance but haven't got my head around it properly yet.
Regarding the idea to bandwidth limit the amplifier to roll off the bass, I'll fire up a filter calculator later and see what I can make of adjusting the input filter (C1 & R2/P1/R3 in series). As I said earlier, I plan to test this with a MoFo that I've quickly put together, adjusting the input and using this inductor;
http://catalog.triadmagnetics.com/asset/c-56u.pdf
Thanks again.
I got the olive cutting board from an ebay seller here in the UK. It's about 18mm thick. The seller was realy helpful and selected a board the right shape and size for me to cut it down to the required panel size.
I plan to use the offcuts to make a headphone stand.
The finish is Liberon Finishing Oil, not sure if you have Liberon brand over there but it's a low viscosity oil, wipe on, wipe off, to gradually build the finish.
I plan to use the offcuts to make a headphone stand.
The finish is Liberon Finishing Oil, not sure if you have Liberon brand over there but it's a low viscosity oil, wipe on, wipe off, to gradually build the finish.
nautibuoy,
I have not changed the power supply. It is the same CLC and it outputs 38VDC +/- depending on the VAC from the wall socket. That is with the amp drawing 2.5A.
That voltage will give Vds=35.5V+/- which seems to work well with the Tokins. If you have a lab power supply along with an USB sound card you can experiment with operating points and see what makes you happy.
I have not changed the power supply. It is the same CLC and it outputs 38VDC +/- depending on the VAC from the wall socket. That is with the amp drawing 2.5A.
That voltage will give Vds=35.5V+/- which seems to work well with the Tokins. If you have a lab power supply along with an USB sound card you can experiment with operating points and see what makes you happy.
My 3310A measured 0.13% THD at 1kHz.
The official specification for Sine wave distortion for frequencies 10Hz to 50kHz on the 1K range is greater than 46 dB (0.5%).
So it seems my 50 year old 3310A is still well within specification.
It's not 0.02% though.
Thanks for the compliments Tungsten, ZM, and Ben. I am not ready to give up my tubes just yet, I need my exercise lugging iron around the house. I have upcoming 801A A2 design in progress, but the SIT bonanza has grabbed my attention...
It seems the THF-51S is preferred as a follower in many designs, and the 2SK180 too. Still, I think I would like to build a choke-loaded single-ended common source SIT amp similar to the original design of this thread, perhaps I will try it both ways and let my ears judge as Ben did. With that in mind, I suppose I must buy another SIT as the 2SK180 or 2SK182 will be preferred in common source configuration with lower input capacitance than THF-51S. Unless a different SIT comes along that is better for the job, I think that is my approach
It seems the THF-51S is preferred as a follower in many designs, and the 2SK180 too. Still, I think I would like to build a choke-loaded single-ended common source SIT amp similar to the original design of this thread, perhaps I will try it both ways and let my ears judge as Ben did. With that in mind, I suppose I must buy another SIT as the 2SK180 or 2SK182 will be preferred in common source configuration with lower input capacitance than THF-51S. Unless a different SIT comes along that is better for the job, I think that is my approach
Please keep in mind that I had problems with my 2SK180 in common source. That is why I changed to common drain and luckily discovered that I preferred it that way.
Since you are into tubes, I assume that you have a tube preamp. If it is single ended with gain, it probably has high output impedance and will only work with the Tokin common source amp, regardless of which Tokin you use, if there is a buffer in between.
Also do not discount the THF-51S in common source mode without trying it. If you build an amp, it would be very easy to switch between the various Tokins. However as mentioned, a preamp with low output impedance (75 Ohm or lower) or an added buffer is necessary.
Since you are into tubes, I assume that you have a tube preamp. If it is single ended with gain, it probably has high output impedance and will only work with the Tokin common source amp, regardless of which Tokin you use, if there is a buffer in between.
Also do not discount the THF-51S in common source mode without trying it. If you build an amp, it would be very easy to switch between the various Tokins. However as mentioned, a preamp with low output impedance (75 Ohm or lower) or an added buffer is necessary.