[baronbeehive]

Some of the main differences I noted in various bias points, when listening, were more of image distance and liveliness, rather than distortion.

Sometimes a livelier tube bias settling, was also a more forward sound, so selecting bias points always needed listening observations.

Yeah, I know you looked into this at he time. I expect distortion might be masked by the fact that we have a slightly slower, more tubey sounding amp than the Oblivion, which can still sound pleasant under different operating conditions, so long as the harmonics are lower order ones.

I just wondered if there were tube varieties we haven't tried yet that could fit the bill.

This amp runs such a low voltage and current, it needs a bit of help. I suppose that's the downside for having an OTL amp. Having said that though it's still TOTL for me.

What tubes are you using now?
I tried so many, I lost count.

Haha, I would love to see your stash lol.

I'm back with the TS6SL7's, can't beat 'em. Its less of a problem now though, the amp sounds so good whatever I've got in, also different power tubes don't seem to matter now.

My tube of choice was also limited to the fact I did not use CCS.

If I were to change tubes now I would have to recalculate resistor values for the CCS. I'm very happy with it in though.

Anyways, my pick ended up being the 6SL7 variety, not the 6SN7.
My fav being the "coke bottle" version, with top anode(6C8G):

I like the look of them, I bet they sound good. but you know me, I'm a skinflint so could not justify splashing out that sort of money for tubes...having said that I will have to justify spending out for some OCC silver to complete the silver wires inside soon... I believe I can get a slightly edgier sound closer to the APPJ.

I heard you were finishing the upgrades soon...
.

 

[baronbeehive]

Having looked again at the 6SL7 operating point tests I did on page 230 I realised that it might have been difficult to understand to anyone looking at it... if there was anyone there, so I will just summarize my findings so that anyone can easily see what to do to make the changes.

This graph shows Sonic's, in brown, Maxx's, in red, and stock, in black, operating points

This graph shows the points I picked, in green to test.

Below were my comments:

To summarize my results there was one stand out position which hit me the moment I started listening, the sound was full, palpably real with good clarity and texture which you could almost feel. The sound had a presence, transparency, resolution and dynamism. This for me was the obvious setting to use. I was a bit surprised that my proposed intermediate settings in between Sonic and Maxx's settings were not my favourite, probably because it fell in the middle it was neither one thing nor the other. The best operating point was with an anode of 69K and cathode of 380R, point 7 in the graph. I will scan in my notes page later so you can see in more detail.

33K - good baseline score, with cathode of 380R, as above, however at 2 of the cathode settings, 1K and 1.5K, I felt that the sound was somewhat flat and lifeless. High gain sounded better, slightly more transparent, with both 6SL7 tubes.

46K - again, good baseline score or thereabouts, with cathode of 380R, but with the GE tube with scores of 1 for bass, mids, treble, and soundstage I felt that this justified a second place position for SQ. Very good transparency and detail. With the 1.5K cathode the scores fell to -1 with a flat rather lifeless sound.

69K - here with all 3 cathodes I felt was the best sound with all 3 cathodes the sound was considerably above baseline. With the 1.5K cathode I felt that with the GE tube it justified second equal position in SQ. With the 1K cathode and with both 6SL7's I felt that second equal position was justified with full bodied, incisive, real and transparent sound. However as soon as I started listening with the 380K cathode in I knew straight away that I would be very happy with this setup. The sound was full bodied, had a real presence, and had the most texture to the transients and treble which gave it a dynamism. For example piano had weight and attack, guitar notes stood out, vocals sounded real, bass had weight and definition and drums had attack, you could almost feel the kick drums.

Below is my recommendation:

I would recommend that overall it would be best not to use the 33K anode setting, or the 1.5K cathode settings and go for the first place or one of the second place settings.

So, the 69K anode and 380R cathode are the best values to use, when using 6SL7 tubes. But, interestingly the 69K anode, and the 1.5K cathode, point 9 on the graph, is the position with the most voltage swing so, that might be worth a try also, however in my tests I put it in second equal position, most likely because with that amount of voltage swing the load line entered into the nonlinear region of the grid.

 

[coinmaster]

Oh I forgot about this thread, Sup losers.

Below is my recommendation:

I would recommend that overall it would be best not to use the 33K anode setting, or the 1.5K cathode settings and go for the first place or one of the second place settings.

So, the 69K anode and 380R cathode are the best values to use, when using 6SL7 tubes. But, interestingly the 69K anode, and the 1.5K cathode, point 9 on the graph, is the position with the most voltage swing so, that might be worth a try also, however in my tests I put it in second equal position, most likely because with that amount of voltage swing the load line entered into the nonlinear region of the grid.

Just use a horizontal or vertical load line. No need to muck around if linearity is the goal.
You should probably keep in mind the extreme lack of voltage swing required for headphones too, you're not gunna need 200v of swing. Even the worst offenders don't use more than like 30v p-p at ear splitting volumes. My HD800s only use a few volts.

 

[baronbeehive]

Oh I forgot about this thread, Sup losers.

Hey!!

Is that THE coin?

Welcome back to the losers thread.

Does Frankenstein live?

Just use a horizontal or vertical load line. No need to muck around if linearity is the goal.
You should probably keep in mind the extreme lack of voltage swing required for headphones too, you're not gunna need 200v of swing. Even the worst offenders don't use more than like 30v p-p at ear splitting volumes. My HD800s only use a few volts.

We can't use horizontal load lines without grid drivers, and probably gyrators and a couple of other things to provide current in that area of the grid. I seem to remember you banging on about this way back at the start of the thread... when nobody was listening lol! Don't forget this amp is old school. It's sounding great though, I know you liked yours.

I must say the point I chose with 40V of swing was very linear and sounded good, probably better than the grid area with 110V swing which may have started to distort.

Stick around!
.

 

[baronbeehive]

This site has gone bananas, no wordwrap, or something…, can't get anything down properly.

That was supposed to say I seem to remember you banging on about this way back at the start of the thread... when nobody was listening lol!

Don't forget this amp is old school, its sounding great though, I know you liked yours.

 

[baronbeehive]

The website maintenance appears to have introduced a bug that only affects me, most of my post is cut off at the right hand side of the page which is pretty weird!

 

[coinmaster]

Hey!!

Is that THE coin?

The one and only

We can't use horizontal load lines without grid drivers

I think it was the other guy that was insisting on grid drive. The grids of the following stage will draw some current but we're talkin like micro amps, the load line will still be effectively horizontal for all intents and purposes.
You could use a mu follower gyrator to drive all of the current in the next stage or apply my leamskie follower mod to ensure the next stage requires practically no current to drive and doesn't have frequency dependent current sharing on the push pull operation like the stock design does but those are a bit more complex to implement than just applying a simple current source.
A vertical load line is another option and it will greatly increase the gain of the first stage which can allow for extra feedback and will have no issues driving the next stage since it functions purely on the current axis. Implementation involves just a simple source follower between the tube and the plate resistor.
There's a zillion mods that can be done here, I design vastly superior amps at this point so the MK6 is just a paper weight for me nowadays.

 

[baronbeehive]

the MK6 is just a paper weight for me nowadays.

I hear you can get cheaper paperweights than that!

The grids of the following stage will draw some current but we're talkin like micro amps, the load line will still be effectively horizontal for all intents and purposes.
You could use a mu follower gyrator to drive all of the current in the next stage or apply my leamskie follower mod to ensure the next stage requires practically no current to drive and doesn't have frequency dependent current sharing on the push pull operation like the stock design does but those are a bit more complex to implement than just applying a simple current source.

Could you demonstrate a horizontal load line, I only know the old style ones
that go from 0mA/max B+ to desired plate current.

What is the Leamskie follower?

A vertical load line is another option and it will greatly increase the gain of the first stage which can allow for extra feedback and will have no issues driving the next stage since it functions purely on the current axis. Implementation involves just a simple source follower between the tube and the plate resistor.

The nearest I've got to the vertical load line is the brown line on the graph
in the previous post that Sonic used. But even then there is the same problem
right near to 0mA, are you saying the line stops before you get to this region?
And how do you get voltage swing on a vertical line?

I design vastly superior amps at this point

So your Frankenstein monster IS alive.. and getting ready to get off of your workbench...

 

[coinmaster]

I hear you can get cheaper paperweights than that!

Yeah I need to sell it at some point.

Could you demonstrate a horizontal load line, I only know the old style ones
that go from 0mA/max B+ to desired plate current.

A horizontal load line is another way of saying constant current.
A vertical load line is another way of saying constant voltage.

What is the Leamskie follower?

It's a more elegant name for the improved WCF design that broskie helped me complete when I first started designing amps.
https://www.tubecad.com/2016/02/blog0339.htmIt's essentially the idealized version of the MK6. I believe I mentioned my impressions after implementing it into the mk6 quite a while back in this thread somewhere.

I also designed a balanced input stage that uses one tube to cancel the distortion of the other tube using anti phase currents across the same load line, 4 triodes total for balanced, but it sounded just as good as my good SS circuits so I decided against using it due to cost and complexity of implementation into a finished design when a much cheaper and smaller SS circuit sounds just as perfect.

I was building a balanced version of the follower using 6c33c tubes years ago but couldn't finish it due to lack of funds. Coincidentally I'm going to finish it soon because I found a way to finish it without spending money.
It's more of a novelty project now since I can design way better stuff but it uses 8 6c33c tubes so it looks cool in any case.
It will have an adjustable output impedance and a feedback configuration that allows a pot to let it adjust between from 100% tube sound to 100% SS sound, the SS portion of the dial will cause the amp to produce unmeasurable distortion.

And how do you get voltage swing on a vertical line?

You don't, as far as the tube is concerned. You put a source follower on the plate of the tube so the tube just sees a DC voltage. Then you put the "plate" resistor on the drain of the source follower. The current of the vertical load line flows right through the source follower into the plate resistor.
This means that the gain of the circuit is set by the plate resistor resistance and the current swing of the tube rather than the gain of the tube.

So your Frankenstein monster IS alive.. and getting ready to get off of your workbench...

No that stuff is garbage compared to what I make now.
I've been working for a while on an amp that has distortion well below -300db and never produces any measurable distortion under any load until clipping.
I designed a PCB for it that can pump out 25W into 8 ohms, True class A, balanced,can fit in your hand, and would sell for roughly $150.
Need to get a reputable third party measurement lab to "officially" verify the specs first before I can go to investors, that is a cost I can't incur any time soon so I'm kind of sitting on it at the moment.

That's just the tip of the iceberg though, I've developed headphones that can loudly and powerfully produce a 20hz signal and make the HD800 sound like toy in comparison and has insanely larger soundstage.
I also have a slew of other equally cool circuits that are tube based and SS based but they will never leave prototype stage until I can fund them somehow.
I tried fishing for investors on DIYaudio but god most of those guys are insufferable and mostly a colossal waste of my time. So much ego, emotion, lack of reading comprehension, and a painful lack of interest in real innovation going on over there it truly boggles the mind.

 

[baronbeehive]

Hats off for sticking with it.

I will check out all that later, I remember you doing a lot of that stuff a while back.

My opinion, for what it's worth, why don't you take a few of your best ideas and build a complete amp, and bring it to one of the threads with a following and watch it take off from there. You may have already done that, but that would be a way of getting going without needing a lot of funds. And it would demonstrate the potential of your amps.

Re: the LD, the upgrades we have done have transformed it totally. Despite the pushpull - WCF design, which is rather clever actually, it's really good now, very transparent sounding with a lot of textural detail, but still euphoric. I probably wouldn't want to change it as it stands now. As far as I can make out it doesn't have any harmonic distortion, certainly not the unpleasant kind but difficult to say, anyway it sounds good!

 

[coinmaster]

My opinion, for what it's worth, why don't you take a few of your best ideas and build a complete amp, and bring it to one of the threads with a following and watch it take off from there. You may have already done that, but that would be a way of getting going without needing a lot of funds. And it would demonstrate the potential of your amps.

Not a whole lot I can effectively demonstrate in a thread without giving too much away.

My opinion, for what it's worth, why don't you take a few of your best ideas and build a complete amp, and bring it to one of the threads with a following and watch it take off from there. You may have already done that, but that would be a way of getting going without needing a lot of funds. And it would demonstrate the potential of your amps.

The WCF has many intrinsic flaws that are only solved in the leamskie follower version.
I agree that the amp is transformed with the mods on this thread as I also went overboard with the modding if you recall. However, it simply can't compete with the stuff I currently make. Not even close.
The main issue is the output stage, a tube can be made sonically transparent in the input stage without too much problem but using tubes in the output stage in the way the MK6 does can only go so far sonically.

 

[baronbeehive]

Not a whole lot I can effectively demonstrate in a thread without giving too much away.

No, but if you had a working version and got it reviewed the message would get out.

The main issue is the output stage, a tube can be made sonically transparent in the input stage without too much problem but using tubes in the output stage in the way the MK6 does can only go so far sonically.

The CCS and filter I put on mine went a long way to cleaning up the sound, and I don't have to worry about achieving balance anymore.

 

[coinmaster]

No, but if you had a working version and got it reviewed the message would get out.

I'm meeting Maxx today. You can ask for his opinion.

 

[baronbeehive]

It will have an adjustable output impedance and a feedback configuration that allows a pot to let it adjust between from 100% tube sound to 100% SS sound, the SS portion of the dial will cause the amp to produce unmeasurable distortion.

I'm impressed that you managed to get good distortion free SS sound.

You don't, as far as the tube is concerned. You put a source follower on the plate of the tube so the tube just sees a DC voltage. Then you put the "plate" resistor on the drain of the source follower. The current of the vertical load line flows right through the source follower into the plate resistor.
This means that the gain of the circuit is set by the plate resistor resistance and the current swing of the tube rather than the gain of the tube.

Did you implement this? It could overcome the problems of the nonlinear part of the grid curves.

That's just the tip of the iceberg though, I've developed headphones that can loudly and powerfully produce a 20hz signal and make the HD800 sound like toy in comparison and has insanely larger soundstage.

20hz, wow, I want these when you've built them... while I can still hear down there... if I can still hear down there.

... So much ego, emotion, lack of reading comprehension, and a painful lack of interest in real innovation going on over there it truly boggles the mind.

...tell me about it... I hate all that.

 

[coinmaster]

Did you implement this?

Never got around to it.

It could overcome the problems of the nonlinear part of the grid curves.

So would a horizontal load line. A horizontal load line would likely give lower distortion than a vertical one on most tubes.
A vertical load line would allow more feedback and provide a bit more linear current response into the non linear input capacitance of the next stage assuming that is an audible issue.

20hz, wow, I want these when you've built them... while I can still hear down there... if I can still hear down there.

You'd be surprised how audible 20hz is.
99% of speakers can't play 20hz well which is why it sounds so low or inaudible if you try. Technically you can hear way below 20hz. The commonly known theory that a person cannot hear below 20hz is only because nothing can typically play that low without impractically large drivers. Mine doesn't have that issue.
I got it measured yesterday. The LF response is essentially flat and begins to rise as it approaches 0hz. So in other words it's loud all the way to 0hz.
The rest of the FR is similar to other high end headphones. The sound stage is in another league from the HD800 though and the sound quality makes the HD800 remind me of a phone speaker.
The prototype is essentially a first blind attempt, further testing revealed that it can be optimized much much further than it is now. But alas I have no monies it's probably gunna sit there for quite a while before I can get back to it.