L0rdGwyn's DIY Audio

[Tjj226 Angel]

I have no update on the phono stage just yet, still planning to powder coat tomorrow, but I have some down time at work, something has been on my mind...

I am fan of Nelson Pass, I admire his balanced approach to designing his amplifiers that blends both objective measurements and subjective impressions. He has also been a huge supporter of the diyAudio community, and he is just a cool dude IMO.

Obviously Nelson deals in solid stage gear, and I am a self-proclaimed "tube guy", HOWEVER, I have been very intrigued by his First Watt line of amplifiers for some time as they incorporate many of the same philosophical design goals sought by tube SET nuts like myself, that being simplicity of circuitry, pure class A, and minimal NFB.

I'm sure some are already familiar with First Watt, but it is essentially Nelson's experimental playground where he designs what he wants outside of the Pass Labs brand. Of particular interest to me are his designs that utilize static induction transistors (SIT), also known as VFETs. What makes these devices very interesting is that their characteristic curves are much more similar to that of triodes than your typical transistor, whose curves are more similar to that of pentodes.

Why is this important? Because triodes / SIT transistors can be operated in class A without feedback at reasonable levels of distortion, whereas pentodes / JFETs / MOSFETs are much less linear and require NFB to tame their distortion to acceptable levels when used as voltage gain devices.

Below are the curves of one of Nelson's SIT transistors, as you can see, they are much like that of a triode.



Anyway, why am I ranting about all of this...an attractive project for me is to build a tube preamp and pair it with a First Watt SIT amplifier...except the First Watt SIT models are essentially unavailable as the SIT devices were made in limited quantity, and they went for around $4,000 a piece!!! BUT Nelson intermittently releases kits to the DIY community, and he happens to have one coming up that utilizes NOS VFETs. I am going to try and get my hands on one when they are released, and if so, I will build it and perhaps pair it with a tube preamp.

Long story short, I might build a DIY SS power amplifier, assuming I can get one of the kits. If not, perhaps I will build one of Nelson's other DIY designs, we will see.

Will hopefully have some powder coated phono stage chassis to show off tomorrow

I have looked at doing the same thing a few times. Something that you might want to look into are the Russian SIT/Vfets. I haven't looked to see which parts are good and which ones are bad, but they are generally cheaper and somewhat readily available.

Couple other things worth considering.

1: Super triodes. If you are going to make a hybrid amp anyways, take a look at super triodes. You can get tube curves, with SS power, for a fraction of the cost of an SIT part.

2: Feedback. There are supposedly different ways you can introduce local feedback to a mosfet to make it act like a triode.

3: Push pull with identical parts rather than complementary parts. If you take the time to design an amp with two identical output devices, the distortion characteristics will cancel more completely than if you had used an N and P channel complementary pair. This means that the majority of your distortion will come from your single ended driver ie your tube front end. Also worth noting is that with a bipolar PSU, feedback, and DC offset, you can eliminate the output capacitor. And trust me, a DC servo circuit is WAYYYYYYYYYYY cheaper than a high quality output capacitor.

-----------------------

A SIT amp would be cool. Any of the other options are cool too. Just want to give you some options before you see the price tag on the SIT parts : P

 

[Xcalibur255]

I wanted one of those SIT amps, but by the time I learned of them the window of opportunity had already opened and closed. What I do not understand is why nobody seems to have any interest in doing another run of the VFET chips or designing and producing an updated one. I don't expect any mainstream brand (which ironically would easily have the budget and manufacturing ability) to pursue it because they're all to busy re-selling the same aluminum jewelry they've been repackaging for decades now, but a boutique manufacturer would be wise to crack this nut and market it in a savvy way. It's essentially offering something similar to a tube experience sound-wise without any of the drawbacks or care and feeding issues of actually having tubes.

Jason Stoddard at Schiit mentioned being really into Nelson Pass a couple of times so I was holding my breath for a while that he would pick this ball up and run with it (IMO they're the perfect people to implement and market something like this), but I think if it was going to happen there would have been talk of it by now.

 

[L0rdGwyn]

I have looked at doing the same thing a few times. Something that you might want to look into are the Russian SIT/Vfets. I haven't looked to see which parts are good and which ones are bad, but they are generally cheaper and somewhat readily available.

Couple other things worth considering.

1: Super triodes. If you are going to make a hybrid amp anyways, take a look at super triodes. You can get tube curves, with SS power, for a fraction of the cost of an SIT part.

2: Feedback. There are supposedly different ways you can introduce local feedback to a mosfet to make it act like a triode.

3: Push pull with identical parts rather than complementary parts. If you take the time to design an amp with two identical output devices, the distortion characteristics will cancel more completely than if you had used an N and P channel complementary pair. This means that the majority of your distortion will come from your single ended driver ie your tube front end. Also worth noting is that with a bipolar PSU, feedback, and DC offset, you can eliminate the output capacitor. And trust me, a DC servo circuit is WAYYYYYYYYYYY cheaper than a high quality output capacitor.

-----------------------

A SIT amp would be cool. Any of the other options are cool too. Just want to give you some options before you see the price tag on the SIT parts : P

Thanks for the ideas, I am pretty keen on the VFET designs in particular. I know they are very rare and expensive now, but one thing about this Pass kit that is appealing is that he is providing the devices, unless builders already have them (some do). According to NP, he has about 200 N channel and 200 P channel devices on hand. Finding them on my own would be a big hassle, not sure I am ready to go down that road just yet

I wanted one of those SIT amps, but by the time I learned of them the window of opportunity had already opened and closed. What I do not understand is why nobody seems to have any interest in doing another run of the VFET chips or designing and producing an updated one. I don't expect any mainstream brand (which ironically would easily have the budget and manufacturing ability) to pursue it because they're all to busy re-selling the same aluminum jewelry they've been repackaging for decades now, but a boutique manufacturer would be wise to crack this nut and market it in a savvy way. It's essentially offering something similar to a tube experience sound-wise without any of the drawbacks or care and feeding issues of actually having tubes.

Jason Stoddard at Schiit mentioned being really into Nelson Pass a couple of times so I was holding my breath for a while that he would pick this ball up and run with it (IMO they're the perfect people to implement and market something like this), but I think if it was going to happen there would have been talk of it by now.

Right? You would think someone else would have picked this up. There is one company in Japan apparently that has current production SIT devices in their gear, Digital do MaiN, pretty boutique stuff from what I can tell. Doesn't sound like NP is ready to do another batch of them himself, from my reading, he paid a very pretty penny to have them made. I'd imagine part of the difficulty with marketing SIT amplifiers would be the power output.

 

[Xcalibur255]

Seeing how well stuff like the Schiit Aegir is doing is proof in my mind that people can be convinced that they don't actually need 250 watts per channel for their personal stereo systems, and that the quality of the first watt is what they should be caring about. The right company speaking to the right audience could make huge waves with VFET.

 

[Tjj226 Angel]

Thanks for the ideas, I am pretty keen on the VFET designs in particular. I know they are very rare and expensive now, but one thing about this Pass kit that is appealing is that he is providing the devices, unless builders already have them (some do). According to NP, he has about 200 N channel and 200 P channel devices on hand. Finding them on my own would be a big hassle, not sure I am ready to go down that road just yet



Right? You would think someone else would have picked this up. There is one company in Japan apparently that has current production SIT devices in their gear, Digital do MaiN, pretty boutique stuff from what I can tell. Doesn't sound like NP is ready to do another batch of them himself, from my reading, he paid a very pretty penny to have them made. I'd imagine part of the difficulty with marketing SIT amplifiers would be the power output.

I believe the Vfets are depletion mode devices. If I am correct, it is not that complicated to build a PP amp using two N channel Vfets. In fact the design looks a lot like a regular PP tube amp.

I HIGHLY advise using two identical devices in PP. It sounds MUCH better than using a N/P complimentary pair. In fact if you dig for it, NP even has a write up about this technique.

 

[L0rdGwyn]

Seeing how well stuff like the Schiit Aegir is doing is proof in my mind that people can be convinced that they don't actually need 250 watts per channel for their personal stereo systems, and that the quality of the first watt is what they should be caring about. The right company speaking to the right audience could make huge waves with VFET.

Good point, but Schiit are the masters of marketing they're going against the grain "we're not like the other guys" approach has worked very well for them so far. But I absolutely agree, marketed as a "tube sound without tubes" might go far, Pass didn't really take that approach from what I have seen, but then he has a reputation, anything he makes will sell!

I believe the Vfets are depletion mode devices. If I am correct, it is not that complicated to build a PP amp using two N channel Vfets. In fact the design looks a lot like a regular PP tube amp.

I HIGHLY advise using two identical devices in PP. It sounds MUCH better than using a N/P complimentary pair. In fact if you dig for it, NP even has a write up about this technique.

I'll look for that writeup, in fact I think I just read it today, but it is all a blur I'm trying to exert minimal effort on this one LOL I don't want to design my own VFET amplifier, purely looking to work on one of NPs designs, again why the upcoming kit I keep referencing is appealing.

I believe his plan is to release kits with the P channel version of the VFET first, then do a second release with the N channel, it is a single-ended amplifier with the VFET setup as a source follower with a constant current sink. The voltage gain front end is a step up transformer, however it sounds like there will be "options" for the front end if one doesn't want to use the traffo. These are going to sell like hot cakes so hopefully I can grab one

 

[Tjj226 Angel]

I wanted one of those SIT amps, but by the time I learned of them the window of opportunity had already opened and closed. What I do not understand is why nobody seems to have any interest in doing another run of the VFET chips or designing and producing an updated one. I don't expect any mainstream brand (which ironically would easily have the budget and manufacturing ability) to pursue it because they're all to busy re-selling the same aluminum jewelry they've been repackaging for decades now, but a boutique manufacturer would be wise to crack this nut and market it in a savvy way. It's essentially offering something similar to a tube experience sound-wise without any of the drawbacks or care and feeding issues of actually having tubes.

Jason Stoddard at Schiit mentioned being really into Nelson Pass a couple of times so I was holding my breath for a while that he would pick this ball up and run with it (IMO they're the perfect people to implement and market something like this), but I think if it was going to happen there would have been talk of it by now.

Tooling. It would be one thing if you only had to pay a company to make 1000 units, but in reality you are paying for all the tooling, labor, ect that goes into making that part. Its usually a 100K+ investment to get a few thousand devices.

Nelson pass did it because he didn't mind investing the time, energy, ect to make the gamble work.

However this sort of an investment is usually frowned upon in the business world. In general you want to avoid having merchandise on hand as much as possible. That way you are not paying for the overhead to just warehouse stuff and your insurance is lower. Blah blah blah.

The truth is that if you want to make these companies produce certain parts, you have to prove that they have a unique use in industry. For instance united SIC is making power Jfets again because they make for really good high power switches. If a vfet/SIT can somehow be used to replace mosfets and make a system more efficient, they will start making them in bulk for 10 cents a piece.

 

[Xcalibur255]

I'm actually fairly well versed on the subject too, and I agree with everything you just wrote. There are two things that hang me up really:

1) The tooling for the existing VFET NP used should exist still, if it was disposed of that was not a wise decision on somebody's part. The investment was already made on somebody's part, so the production capacity existed, yet nobody wanted to utilize it again. There must be more to the story we weren't told because that doesn't add up. Possibly the tooling and fixtures encountered an obsolescence issue.....?

2) There was supposedly a military application, or possibly a *potential* military application for these parts. This is, without question, the best way to guarantee that the tooling and capacity will be created and will continue to exist (at least in the USA it is), yet in this case it didn't work out that way.

We do custom PCBAs where I work and nearly all of the cost is in initial tooling investment and ICT fixtures. If nothing else it means the product needs to have a lifespan measured in years, not months, and some volume behind it to justify the cost of initial start-up, but I don't see this as a deterrent to any reasonably deep-pocketed audio company. The world of tubes and the tube sound is really becoming a "known" thing in the high-end mainstream and the interest in a solid state product that achieves a similar result will likely be strong.

 

[Tjj226 Angel]

We do custom PCBAs where I work

 

[Xcalibur255]

Your pic is broken so I'm missing out on what was potentially a good meme.

I should clarify that by "do custom" I don't mean manufacture I mean we have them manufactured for us for the product we make.

 

[johnjen]

Tjj's response made me laugh…
But then I like to laugh and will go out of my to do so, so there is that…

JJ

 

[Xcalibur255]

Personally I think a "doesn't everybody" meme would have hit harder. 7/10

 

[Tjj226 Angel]

Personally I think a "doesn't everybody" meme would have hit harder. 7/10

I thought you had basically outed yourself as a resource for the entire DIY community.

I figured I would point that part out so that you could edit it and hide it as quickly as possible lololol.

 

[Xcalibur255]

I thought you had basically outed yourself as a resource for the entire DIY community.

I figured I would point that part out so that you could edit it and hide it as quickly as possible lololol.

Ohhh. Yeah, my wording is dumb sometimes. The products we make require custom designed and fabricated PCBAs, so I have some familiarity with the logistics and processes of that world. I didn't mean to imply that we manufacture PCBs, though that would be an interesting job.

 

[L0rdGwyn]

Ohhh. Yeah, my wording is dumb sometimes. The products we make require custom designed and fabricated PCBAs, so I have some familiarity with the logistics and processes of that world. I didn't mean to imply that we manufacture PCBs, though that would be an interesting job.

Would have been disappointed if you were holding out on us this long @Xcalibur255 LOL I use JLCPCB, if you can learn to use EasyEDA (which is some sloppy software, IMO, but hey it's free!), getting some PCBs designed and fabricated is pretty simple.