Welcome to the SEWE300B Design Log.


This is a continuation of the the Western Electric to Manufacture Tubes in the USA thread found here:

https://www.decware.com/cgi-bin/yabb22/YaBB.pl?num=1648154321

With the war in Ukraine, Russian tubes are hard to get and Western Electric is stepping up to the plate to bring many of these tubes to market. Right now they have the original 300B tubes but are adding more.  To support this USA company I am making a DECWARE 300B amplifier specifically for the Western Electric 300B tubes that cost nearly half the price of the amplifier!  Obviously the resulting sound of a Decware 300B amp would have to be so over the top to justify such an expensive tube, but we are confident it is not only possible but likely.

So in the world of 300B amplifiers the $1500 price for a pair of tubes would only be small percentage of the amplifier cost... welcome to high end audio of which we completely are but aren't.  If you do the math based on my comments we should be able to hand build a 300B amplifier without tubes starting around 3K that can carry the Decware badge with honor.  

Below is the beginning of the design log for this new Decware amplifier.  I'm sure it will be interesting because they always are.  A tumultuous journey for myself as I try to rise to the occasion and make things happen... these real time journals are import so I don't forget any of the stops along the journey and it gets those who care inside my head so they know what to expect before they ever hear it.

---

DESIGN LOG - DECWARE 300B

So one of the things that has always given me some pause with respect to directly heated triodes like the 300B is the fact that if you run an AC heater, you have to use a hum balance pot on the cathode.  That would be a 2 watt or 5 watt wire wound linear 100 ohm pot where the center wiper becomes the cathode path to ground!  This means you would have to have an extremely high quality wire wound pot not to tank the sound right here.  This is the critical signal path where even the type of wire used will effect the signature of the amp.

So naturally this is why some of the more high end offerings use DC heater supplies for the tube.  That also worries me because we're not talking about a battery here, this would be a rectified AC supply with however much noise and sag you decide to allow.  Another scary prospect.  Even scarier are the solid state regulated supplies that promise perfection but I only see artifact.

Then I find from the WE spec sheet and literature that while they use the best coatings, DC on the heaters can still cause some stripping  and reduce the quality and life of the tube.  Obviously we're not going to do that.  So then besides the worry about the quality of the pot comes the worry about hum.  The other reason many hi end offerings go with DC on the heaters.

To do this right, the power transformer is going to have to be really good.  The extra trouble to wind it with reduced stray fields, and no primary capacitance coupling.  Not to mention since the cathode is the AC heater for the tube, and there are two tubes, there should be two separate heater windings, one for each.  And then a 3rd for the input tubes and a 4th for the rectifier.

I figured out that all the pot is doing (other than balancing AC for zero hum) is acting like the center tap on the heater winding passing the signal from  ground to either side.  That means that if a heater supply is center tapped, you should be able to use that for the cathode resistor instead of the center wiper of the balance pot.

To make the balance pot work, the bypass cap for the cathode resistor can be placed on the wiper since we're only dealing with a low frequency (50/60Hz).

With the cathode current and signal no longer going through the hum balance pot or it's wiper, the amp stands a chance at being a Zen Triode.

Below is a diagram of how I'm doing it.

 


The right side seems to be the standard method.  The left side would be superior on multiple levels because now the only thing the pot does is balance AC to the ZERO point while the DC current and signal do not pass through it.


So this forms the basis for my plan, along with a custom transformer.  The output transformers will be the same UFO transformers used in the SE34I.5 now SE34I.6.  The driver stage will also be identical but with different tubes.  At least at this point in the development that is the plan.


At this stage I am worried about the giant power supply transformer, worried about possible hum, and worried about too much heat.  I'm not going to lie, it's stressful like driving around with your muffler dragging the ground.  For this reason I am not wasting any time on this project.  It's some new territory for me and let's be honest, the pressure is high.  Not only would everyone expect it to be better than other 300B amps because it's a Decware but its performance will be measured against the 25th Anniversary Zen Triode amp which has so far been completely unbeatable.  This would be a lot of pressure for an amp builder who only dealt with DHT tubes, but for me, of course it's over the top so as I said no pressure here!



Steve

I look forward to the “Decware 300B amp development” thread.

Yes, I see. I was thinking of this amp being an alternative to the sound of your existing amps, for the novelty of variety. A richer sound or a sweeter sound or what ever it is that 300B offers.

I had not thought of the transformer taking away a voicing tool. That of course makes good sense.

Brian

This should be an interesting thread.  Looking forward to how the story unfolds.

Wishing you the best of luck.  I understand that this is WAY out of your comfort zone.  You will figure out how to make it reach your high expectations. Looking forward to seeing it at Decfest 2022. I hope the Audio Gods drop in an show you the way.

Instincts can save hours, days, even months of trial and error.  Instincts come from the Audio Gods, the trick is to listen past your own ego and follow them.  If not for this simple truth a Decware 300B amp would not be possible.  I would have to build at least 20 different amps each with two or three different front end variations to figure out by process of elimination the best possible sound.  It would take years just like it did with the original Zen amp and look how simple that was!  I don't have years to dedicate to this project.

The way I look at it is simple.  In less than a month if I didn't do anything else to it, it's already clearly a gift from the Audio Gods to have it sound this good already.  There's just no way I can take credit for it.  My skill comes from being able to hear what isn't there and then recognize it when it is there.  I use simple tenacity and instincts to create the bridge between the two, not brain power.

Anyway back to the amp.  I am happy to report that my fears about it running hot did not manifest and it actually runs the perfect temperature, slightly cooler than most of my stuff.  Also the fear about hum was also for not, as the amp is quiet.  The AC heaters worked great.

The only question is gain at this point.  The tube is a bit harder to drive than an EL34 or KT88 so the input sensitivity is a bit on the low side.  No problem if you have a preamp so I am now spending time with it without a preamp to see how the gain works out.  

I have been using the 25th Anniversary CSP325 to drive this amplifier.  It has been my observation that when used with any of our amps, things get a bit denser and have a touch better tone.  It always makes things a notch better.  So I expected some of the sound I am hearing from this amp was due to the preamp driving it, so tonight I realize while listening with no preamp that I can't even tell I took the preamp out.  Wasn't expecting that.

Quote:


Damn it! Talk about good news/bad news . Now I have to decide if I want to change my order from a pair of UFO25s to a pair of these. Good thing I have plenty of time.

Never mind, I saw where you said 6.5w

NM.

FWIW, Craig Uthus of Eddie Current (he's retired now so not producing anything) used an unusual method to heat the 300B/PX4 tubes in his Balancing Act headphone amp. As I understand it, he designed a power amp that produced a 42K Hz AC signal to heat the tubes. His power supply was separate and as large as the amp itself.

I had one for a time (shoulda' kept it) and can concur that the soundstage was, as one reviewer described it, otherworldly. Most credit that characteristic to the heater circuit.

can you tell us which input tubes you will be using for the 300B amp?

I'm curious too, my guess is 12AX7.

Quote:


The 6N2P and 12AX7 have been tried already and didn't suck but the reality is that this tube needs current.  I think the original 6N6P is going to be the tube to beat.  I haven't found anything I like better yet. But God, there's so many ways (circuits and values) to dial a tube in.  The possibilities are overwhelming. Nevertheless that is what I am using at this stage and already know there's not going to be any way to beat it in this circuit... a circuit that is still evolving btw.

Here is a shot of the amp as I was finishing more updates just before I buttoned it up for tonights listen.


(click to enlarge)

It looks like a mad science experiment because it has been changed so many times...

Trust me at this stage of the process we are in the heat of the race -- you don't care what it looks like until you finish the sound.  Then you will simply build a new one a re-lay things out. See it's hard going into a layout when you don't actually know what the circuit is going to be : )

So if you recall I was digging the sound, but worried about the gain. It had none. Sounded great with a preamp, or a ZBIT or a ZROCK2 or a ZSTAGE, or high output DAC. But that was with the volume turned all the way up. I could have easily lived with the sound, it was really good.  But now we're locking you into a preamp or higher gain source.  Unless that higher gain source is a ZBIT, it is adding another tube stage, capacitor and resistor or two or three... So by making it require a preamp we eliminate the opportunity to hear a single capacitor and two resistors which is the essence of a Zen Triode amplifier.

This morning when I came into work I was pondering adding another tube for more gain and ruin my simple signal path... and tonight without adding anything I have twice as much gain as I could possibly use! And I wish I could describe what I'm hearing but my face melted off on the first song.

I had an idea to use a voltage regulator tube as part of the plate resistor one day a long time ago and tonight that very thing came to pass by accident. Well, what happened was I created two different power supplies for the tube, 150 volts apart and then proceeded to hook them up in reverse. This created a nearly perfect square wave which was a little distressing. Then I correct that mistake and forgot to move the capacitor which put the voltage regulator tube in series with the plate.  Once I discovered this  I knew to leave that way because -- by now if you don't know I'm not going to tell you.

So this is tonights mad science experiment and the way it is currently dialed in the resulting sound has taken presence to a whole new universe. It's like listening to watch ticking on the other side of the room vs. being shrunk to the size of an ant and being inside the watch. I can't even process this. We might be onto something serious here.

Well there is no doubt the tube is happy, both of them : ).  This is such a trip.  If people could experience the wild swings in sound that happen during this process they would understand why I freeze when asked what parts sound best, be it tubes or otherwise because I know that the parts and or tubes have precious little to do with it compared to the circuit and execution.

All I can say about the sound right now is it feels like the James Webb Telescope. It's like you can hear 10x more information than there was before. Where the hell did it come from?  This is a completely new sound.  I think it might be time to order the WE300B tubes.  I think we're about to land.  I've actually never heard anything like this.  I've heard quite a few 300B amps.  



Steve

So what will this amp be called?

Z300B?

Or Z300B SFM (Super Face Melter)

This is so exciting. If I had speakers that were more sensitive (88db), I'd strongly consider switching my existing order from the MKV. I know they are VERY different amps so you can't really compare, but the notion of having the WE tubes is very appealing.

Love this thread Steve, thanks for letting us into the mind of a mad scientist.

I believe "Steve" as a kind of signature is already out there. He should continue using that!

Keep your orders going, then sell me which ever one you decide to let go, later. I will take care of it for us.


I know we are all jumping the gun, here, but it's fun to imagine ...

Quote:


This thread, while quite interesting, doesn't move me a bit.

For me, there's zero question that I'm dying for my Torii MK 5-25th order. About 4 months and counting. I'll take its versatility, tuneability & power 24-7.

Hats off to Steve though for pursuing this wonderful and challenging path.

- Brad

It won't be much longer piezoman, hang in there buddy.

When you order, Piezoman? I got mine in shortly after Christmas.

Dang, you'll be kicking out some sweet tunes well before I can! I can't wait for a full report on your MKV! I've yet to read one. Any ETA for you?

Any difference between working with a "true" triode and a triode-strapped pentode?

The prototype under test today.

I only have a couple of questions.

Is it possible to have the new SE300UFO equipped with a UX4 rectifier socket? As an option it would be nice to be able to roll 274a, Globe 80, or 5z3 rectifiers without an adapter.

Are you going to try out oil filled can capacitors in the power supply? That is of course after you have nailed down the circuit values with your usual (and awesome) arsenal of caps.

The Rachael is one of the coolest amps that I ever owned. Now that I have built the DNA2, I would like to try her out again.
Steve, I am enticed by the 'what if' factor in making just about anything. In this case, I had to know, what if, I build a class D power amp and used a tube preamp to drive it. As time went by, I found myself fiddling more than anything. I won't say that I was disappointed, but I simply missed the sound that I had before.
I am still using a directly heated triode for the preamp part, and then using yet another class A power amp. I seem to like the 20-40 watt region of power despite some drawbacks.
What I am getting here is the beauty of being able to start, either by yourself, or with a bunch of help (like I did), a project that came partly from some sort of inspiration. It doesn't always end up where you would like or as planned but the trip is worthwhile.

BIG UPDATE:  Sunday Results

Since my last post, the main curiosity I had was about the VR tube connected to the plate resistor vs. connecting to a capacitor and the plate resistor at that node.  I had to wire a switch so I could have the amp warmed up and then switch the capacitor in and out.  I started in the afternoon yesterday and streamed classic rock which sounds like plastic when anything is even slightly off...  I did this because I am interested to hear the amp sound better with each change, so play material the sounds almost like crap, make the change and then see if it sounds less like crap.  This is easier than playing reference recordings and trying to hear which one sounds better.

I started with the capacitor in the circuit, which is the normal way and the way I have always done it so far as I can remember.  After about 3 hours of listening to it sound borderline like crap one particularly bad sounding piece came on and I walked in there and flipped the switch.  I was stunned.  It got so much better.

So I proceeded to listen to it that way the rest of evening. Not in the sweet spot but from my work bench.  Sweet spot too distracting when trying to hear what is wrong.

By evening I had concluded that the present circuit wasn't going to fly.  The amplifier sounded like I went to a retail hi-fi shop and purchased a retail tube amp. Couldn't find anything wrong with it, nor measure anything wrong with it, and it had it's moments, but it just didn't move me.  Not at all, in fact it made me understand that I'm really going to have to work  for it.

CIRCUIT

The original circuit was the exact SRPP stage used in the SE34i.5.  It sounded really good, but as you recall, couldn't drive the tube to the same output level resulting in the amp having too little gain.

The best sounding tube in that circuit was the 6N6P.  6N6P is a really bomb proof high current tube of which the original ones sound pretty good. We'll the amp still didn't have any more gain than before, it just sounded better. So I decided to optimize the circuit for the 6N6P and created a really rather high current direct coupled gain stage where the output tube was driven of the plate of the second stage. This got complicated really fast because the circuit needed another voltage, hence filter section. Prior to this I had added bypass caps on everything trying to get more life out of the top end and more speed. (That proved to be stupid because that is an improper use of the technology). Before you knew it, we had an ultra complicated driver stage with 4 or 5  times as many parts and that was rubbing me in the wrong way because I'm watching the price of the amplifier gradually almost double due to the complexity of the build and extra caps.  

Anyway this was some pretty textbook engineering except for the SG5B regulator tube trick, and even with that the sound was average. It contained some veil, some hardness, some plastic aftertaste, some poor low level performance, a forward sound without the hallmark DECWARE touch despite my efforts it's still not I want. I will recognize it when I hear it even if I've never heard it.

When I went to bed I knew without any doubt that the problem was I needed to be listening to a cathode in the driver stage, not the plate. I just knew it because I know the sound signatures within this gain structure by memory. Never in this situation does the plate sound better. The cathode always sounds better. More musical, more spacious, fuller, warmer... you get it. But you get zero gain in most cases, the exception being a SRPP stage where both triodes are in series and you drive the output from the center.

I could have pretty easily modified the existing circuit with new values and moved the output to the cathode but so far this driver circuit hasn't done anything to suggest I'm going to love it, so I just removed the entire thing and all the extra bypass and filter caps and created my own SRPP stage from scratch which was basically like the original SE34I.5 circuit that I started with, but now with slightly different values and two wicked little twists, one I'll keep to myself, the other I will say was to connect the SG5B output directly the plate of the SRPP stage.

I plugged in a 6N5P, the lower gain sister of the 6N1P because from a gain standpoint that would be a worse case scenario and kept my fingers crossed as I popped it on the analyzer that it would have the same input sensitivity as our other amps which is about 1.5 volts and that's exactly what it came out to.  

The distortion signature, and power and gain are all perfect. The power comes in a 6 watts.  

I took it into the listening room and hooked it up and began to raise the right volume control just in time to catch the last note in a song as it trailed off to silence. About 1 second long, a note/sound I am well familiar with and instantly I felt it. It was right. I nailed it. I could tell from just that 1/2 second of sound on one channel.

So I raised both dials to the half way point and waited for the next track to start. It was eerily silent and then suddenly music just flowed into the space like Holy water. Yes, this was the magic sound I was waiting to hear, and as is usually the case in these situations, even better than the sound I heard in my mind. Boy this is going to be fun!

Warm and full with good dynamics at low volume, like a Zen amp with that insane transparency and layering that we can't live without.

At risk of jinxing it I am going to say the amp is done and it is back to it eloquent simple self again. I am still listening the the 6N5P and trilled it has enough gain to bring the amp to past 6 watts with 2 volts and the sound.  The 6N5P sounds so good right now that I swear if I tried to remove it to hear a different tube -- lightning would strike me dead on the spot just for being stupid.  It's simple, the rule has always been that when it sounds this good you don't jack with it.

This is definitely the sound I want my name on for the 300B tube. It's just like listening to the smaller amps but with that 300B top to bottom magic. So far in this adventure I have created and listened to 6 different ways (driver stages) to waste the sonic potential of this tube and the sad thing is they all look fantastic on paper, and would be considered by any non-DECWARE customer to sound great.

And the weird circle of driver stage design produced something I would have never come up otherwise. Using a VR tube instead of a plate resistor has just throttled the transparency through the roof, and the SRPP stage that can be found in several Decware products proves once again to be unbeatable.  

Can this new SRPP stage design trickle back to the other products? Not exactly like it did here, no. It's about the whole of all the variables unique to this amplifier, the voltages and the 300B tube itself.

That's cool because that would justify the amplifier as something beyond a 'popular demand' design into something that pushes Decware fidelity into new territory.  At this point I will listen to it all day all week on streaming music off the algorithm so that great and non great recordings are intermixed and in the evenings listen to reference material.

One thing about this tube that is different in this particular amplifier is that the distortion comes on in a more linear fashion than even our other amplifiers and that will make it more tolerant of being pushed. For example if you want to take it to 10% distortion you can get around 10/11 watts out of it. Now obviously no-one is going to turn an amp up to an RMS level with 10% distortion, but what it will do is increase the instantaneous peak power so that when the amp is pushed it's headroom is better managed.

So what makes this design unique, is so far:

A) The driver stage - there is probably no other one like it on the planet. The driver stage IS the amp.  
B) Cathode circuit for the 300B is separated from the AC hum balance by using the center tape of the transformer instead of the wiper on the hum balance pot keeping AC and DC paths separate and quality factor of the cathode circuit high.
C) UFO output transformers.
D) One capacitor, 2 resistors in the whole signal path with no preamp required or even recommended.

And this is perhaps the most striking part of this sound relating to the 300B tube. Zen Triodes often sound better with a preamp because it makes the source a little juicer and adds some weight. As I often site it is a tool to make your source better. So now we have an amplifier that sounds exactly like it has our best preamp in front of it and it is hooked directly to a DAC via the ZBIT set to 2 volts. The way I see it, this saved the audiophile the price of a preamp because for some fascinating reason all the density and bloom a preamp adds is already here. Another way to express it is to say that based on what I am hearing now, it really can't get any better than this so a preamp probably isn't going to help you with a killer DAC like the HOLO Audio Kitsune editions.  

So what does a 300B tube bring to the Decware experience... density and more vidid colors.  Everything else is now the same but the density and tone a preamp adds is already there. Transparency is now on par with the 25th Anniversary Benchmark and that was my biggest worry as I went into this especially when I kept hearing less than stellar sound from each of these driver stages as we went along.

The 25th Anniversary Zen Triode amplifier made this design possible by kicking it's ass until the very end.