[johnjen]

So here are the latest updates to the layout and power supply circuit.
45 amp layout-13.pdf
and
45 PSU.pdf

JJ

 

[johnjen]

Only to be superseded by the next iteration.

45 amp layout-14.pdf
45 PSU-2.pdf

And I've torn out a bunch of the innerds to rewire it and deal with a few layout issues.
As well as work on a few more design related issues and plan for their implementation.

The saga continues.

JJ

 

[johnjen]

So here are 2 pics, the 1st of which is analog section when the amp was first run.




The 2nd, is part way thru the rebuild.


Nothing is soldered in place yet, the ground buss has yet to go in (the black wire running down the middle), and I'm moving most of the xfmrs around just a bit to get a bit more room in places for the filament cable runs and a few other 'refinements' to the layout.

And we are still figur'n on how to lower the line voltage to fix and stabilize the internal operating voltages and allow for dc heating of the tubes, if need be.

To that end our current count of 14 xfmrs, could rise to 19, which follows the tube DIY'rs dictum of 'you can solve any problem with moar iiiiirrrrrrnnnn'.
(note, ya gotta say iiiiirrrrrrnnnn with a texas drawl, as in shoot'n iiiiirrrrrrnnnn, as there seemed to be some confusion as to the proper pronunciation, and intended inner meaning. hahahahahahahahahaha).

JJ

 

[johnjen]

So today I made up the filament cables and figured out how I can use the scant space to wire up the filament outputs and run them to the tubes.
And the tricky part was to be able to accommodate both ac and dc voltages and currents setups, to feed the tubes, all from the same space.

Going with heaters that are fed dc requires transformers, filter caps, diode bridges, and chokes (if I can fit all of this in the ≈4.5"x4.5" back corner of the chassis.
Whereas ac heaters only needs the output of the filament xfmrs
The secret to my success is going vertical, since I do have 4" of height to use.
It looks like a telephone pole in miniature with the filter caps hanging down from the tag boards which are 'elevated' well above the diode bridges which are stuck on the top plate and using it as a heat sink (not that this is even gunna get hot. Each filament runs 6.3vx0.3a = 1.89watts for the 6J5 driver tube and 2.5vx1.5a = 3.75watts for the 45 tubes).

The space was tight and I had to move the mains and primary choke over an inch, as well as the secondary and plate chokes inwards to accommodate the re-routing of the filament wiring.
I'm hoping all of this helps balance out and reduce the hum issue, which is a common struggle with tube gear especially when using ac for the DHT (Directly Heated Triode) tubes.
To that end I have made matched, shielded, twisted pair, heater cables (the yellow cables resting on top).

And I still have to rebuild the diode bridge or replace it with an 'all in one' unit.

I'll add the pictures tomorrow and run down a few technical solutions we are working on.

JJ

 

[johnjen]

So my line voltage swings from ≈121Vac to 125+Vac which is a tad bit high.
I have contacted City Light to see what can be done, and we have figured that using a multitap isolation line voltage xfmr would sufficiently drop the voltage as well.
And we might just go for a custom line voltage xfmr with multiple taps to be able to adjust the voltage back down to a more suitable range (115 to 118Vac).

This may seem like an exercise in picking at nits, but in this case every 1 volt increase in line voltage, results in an increase of ≈3Vac in the B+ voltage (300Vdc).
Which further means the jump from 115Vac (the specified input voltage for the mains xfmr) to 125Vac means a 30Vdc increase in the B+ voltage (330Vdc).

This line voltage increase is sufficient to screw with the bias voltages and current flow in the tubes and chokes, and for the filament voltages as well.
And because this is an old skewl design there are no voltage regulators, on purpose and by design.

Once we achieve a 'natural' balance with the 'correct' voltages and current flows, which will optimize all of the operating parameters, all based upon the line level voltage, the entire amp will 'self regulate' in a manner of speaking and there will be no need for internal regulators of any kind.

This is the essence of the old skewl design approach,
the system, while operating, allows for and accommodates it's own type of self regulation because that is it's 'natural' way of operating.
This is the masterful design niftyness of the old skewl approach, and as long as the line level voltage matches up with the mains xfmr's designed for input voltage, and the rest of the design is dialed in as well, the overall entire circuit is really very simple and basic.

We have come up with a couple of work arounds, as well as seeing if we can get city light to step in and hopefully lower the mains voltages as well.

It will be interesting to hear if my main system changes due to this line voltage reduction as well.

And tonight I built v.2 of the B+ diode bridge using 1200v rated Hex-FRED diodes (super duper audio gud stuff).

here is a pic or 2+ of the latest efforts.


This shows where the filament supply is located and those yellow cables are the shielded twisted pair feeds for the 4 tubes.
Believe it or not I can fit 4) xfmrs, (2 go on top), 2) diode bridges, 2) 10,000µfd filter caps, the ground buss, and all 6 cables, that originate in this corner of the amp, all together in that corner on the lower right hand side.



Here is an over all shot of the progress.



Here is a shot of v.1 and v.2 of the diode bridge.
Yes I did scavenge the heat sinks and used them in v.2



Again I'm using the vertical height available and those 'gold' colored holes in the center of the tag boards are the perfect connection point for the 14g wire I'm using in the entire power supply.

And there is yet way more fun ahead, as I install the cables, v.2 of the diode bridge, and modify the input stage of the amp as I add line level input xfmrs that can handle balanced and single ended input sources and convert/hand them over to single ended driver tube.

More later.

JJ

 

[gefski]

Let us know how this goes, especially how much it costs to correct City LIght's problem.

 

[johnjen]

Say there gefski.
It shouldn’t cost much at all, really as their internal documentation states it shouldn’t exceed 124vac with an extreme limit of 126vac, which I am approaching. And really it should (hopefully) be only changing a tap on the step down xfmr in the alley.

But we’ll see soon enough.

JJ

 

[gefski]

Say there gefski.
It should cost much at all, really as their internal documentation states it shouldn’t exceed 124vac with an extreme limit of 126vac, which I am approaching. And really it should (hopefully) be only changing a tap on the step down xfmr in the alley.

But we’ll see soon enough.

JJ

I guess that when the "heavy hitters" in Seattle (Boeing and JJ) say jump, City Light says "how high?".

 

[johnjen]

City light and I have this ‘relationship’ that goes way back, even before they knew we had a ‘relationship’.

They even said I should come work for them, which I ‘tactfully’ declined.

And they really ****ed up when the installed their LED street light replacement program, much to comcast’s consternation.
They used SMPS’s to run their LED arrays and used no shielding, neither light barrier nor electrical, which really screwed with upload stability and speed, and light trespass.

In other words it was a slap-dash implementation, all to save money, which thus far I have not seen reflected in the my electrical bill, go figure...
Not to mention their ‘smart metering’ program, which is another time bomb just waiting to go off.

JJ

 

[derekphone]

The HD-800 a headphone ahead of its time.
And the ‘need’ for modifications, common assumptions, and my different approach and results.

When the 800’s were first released it was a love/hate relationship with many users.
Some couldn’t stand them as they were to ‘bright’ or had a ‘hard edge’ or the top end was ‘strident’ etc.

And at the time the 6.5KHz spike in FR was pointed out as the culprit for these problems.
And so those who figured they could ‘tame’ these problems went about devising several generations of modifications to address these issues.

Along the way I tried them all and after much playing around and trying various different materials and techniques I was sorta satisfied with the results.

That is until I found SAA (Stephan Audio Arts) and subsequently had my 800’s modified and hardwired.
This whole approach opened my eyes to what was and wasn’t going on.
That isn’t to say it became completely clear what was going on, but it is to say, at least for me, it was obvious that the single issue of that 6.5KHz ’spike’ wasn’t the real source of the problem.
Because the results were so far superior to any of the other attempts I had made, there was no turning back.

And along the way I discovered that listener fatigue (I call it tLFF or theListenerFatigueFactor) WAS a contributor to all of this.
And in addition, how important the upstream gear was, and how well setup it was, were also key elements, in not just taming the 800’s, but in allowing them to ‘sing’ as it were.
They are after all only going to create the signal based upon what is fed to them.
And with the tweaks I am using, the upstream system is even more critical in achieving my goal of just listening to music and for ‘the system’ itself to disappear, or as I like to put it, “get the system out of its own way”.

But there were a couple of issues that 800’s have, that could be addressed based upon all of the work I had performed from those previous generations of mods that were, at the time, the mainstay of the available solutions to the 800.
And a significant contributor learned from those previous mods, to improving the 800’s, was dealing with the ‘trapazoid’ and it’s surrounding ‘wings’.

But the SAA mods also showed me a few other related physical modifications that were also very helpful.
Such as removing the ‘protective screen’ that is on the outside of the ear cups that covers the drivers.

And then there are those ‘special’ connectors Sennheiser uses.
Ever since I first examined my 800’s and saw those tiny pins that were used in the connectors for each driver, I knew I was going to hardwire them, which also made a significant improvement as well.
Of course choosing a suitable replacement cable was another contributor to getting the most out of my 800’s.

In short, in my experiments and research, the much maligned 6.5KHz spike wasn’t the real problem after all.

Because you see, the SAA mods had NO dampening material in the earcups and the degree of listener fatigue was reduced WAY down, all based upon how the drivers were secured to the ear cup assemblies, along with the removal of the external protective screen, and being hardwired.

This got me to wondering if these changes would be revealed using any of the available testing procedures.

So I contacted Tyll at InnerFidelity to see if he would measure them, and he agreed.
This stared my investigations in earnest as I now had measurements of the changes from stock with which to compare to my modded pair of 800’s.

And indeed I did find one specific test that connected up the dots I was looking at, and
that was related to listener fatigue, and,
that ‘opened up’ the 800’s, as was revealed by the SAA mods.

That test was the 20 or 30Hz square wave test.
I could have used the impulse test, but the square wave test was more revealing and useful, since it showed the Initial Leading Edge Response (ILER), that was of particular interest to me.

What I began to realize was,
the overshoot of that leading edge WAS the primary source of the listener fatigue,
and the LEDI Leading Edge Dynamic Impact, was the source of the improvement that the SAA mod made, that was so compelling.

So as I continued to experiment and make other modifications to my 800’s, I began to zero in on optimizing that ILER, because I realized that the use of leading edge overshoot, which is very common, was not optimal.

And as I further refined my approach to dialing in my mod, the better my 800’s sounded, and,
the closer to ideal that was to the original signal, and,
the more I didn’t want nor need to make any further changes.

All I wanted to do was to just listen to music.

To me, this was a VERY good sign, that I was on the right track.
Not to mention a major step in achieving my goal of listening to music and not ‘the system’ itself.
And this pattern was repeated several times and the ‘need’ to further modify (other than by sheer curiosity) grew weaker after each new generation was achieved.
I took this as an even better sign, that I’m on the right track, and I’m further dialing in my approach.

So my conclusion is the ‘solution’ to the maximizing the 800’s involves the entire system that drives them, which isn’t exactly a new idea by itself.
But tweaking the ILER waveform to more closely match the step response input, helps optimize these tweaks to the 800’s with wonderful and surprising results.

So I kept diving deeper and deeper down this rabbit hole and when I purchased a set of ears (3dio head simulator) my efforts sped up considerably as I didn’t have to wait for measurements, I could make a change and then run those mods and see how the variables I was juggling told me whether what I did brought me closer in achieving my goal of dialing in that ILER, or not.

So the full VMC mod (Vibration Management Control) involves the 3 aforementioned tweaks (hardwired cable, external screen removal, trapazoid treatment), plus what I have come up with as my tweak (VMC), which together all make up this mod.

And in principle it’s amazingly easy to describe what my VMC tweak involves, at least in general terms.

Take the metal inner mounting 'clamp ring' (my terminology) off, coat the metal side with dampening material, then mount it in reverse, with the new dampening material ‘down’ into the driver and it’s mounting structure.

Put it all back together and then listen.
Give them at least 1 hour of operation to settle in before any critical evaluation, after all you just disassembled the entire driver mounting assembly and then changed it.
And depending upon the state of tweak of your system, as mentioned above, this initial period of improvement may only last for several hours before morphing, only to return as they continue to settle in.

And as the saying goes, ‘the devil is in the details’, and this project has plenty of them buried in the many many small little details in the actual implementation of this tweak.
But those details are the very same ones it has taken so long to ferret out, coupled with learning what aspects of re-assembly are important etc.
Not to mention the various types of dampening material etc.

Like I said, in principle it’s easy to describe, but the doing and achieving of optimal results, is much more involved.
That isn’t to say an experienced DIY’r can’t implement all 4 of these tweaks in this mod, its not rocket science.
And let me add, this change to the 'clamp ring', can be easily reverted back to stock by simply reversing the 'clamp ring' again, as in a before and after test of sorts.

And as I posted, this final tweak to my 800’s, the Gen-5 level of performance, is settling in nicely.
So in Part 2 of this report I am going to release more of the details I use to achieve my results, along with the measurements and a few pics of the tweaks themselves.

It will take a bit of time, because 1st I want to fully mod my 2nd pair of 800’s, my ‘B’ set, so I will have 2 fully modded pairs of 800’s to measure and compare to each other.


JJ

Thanks for replying. Wow you’ve really put your heart into this and I think it’s inspiring. I’m really excited to get into this.

I will say that as a noob to headphones and new to DIY I’m pretty intimidated. Can I even pull any of this off? The Popular SDR Mod (the little donut resonator) definitely seems easy enough. But wiring and the other stuff that I don’t even know what those words mean ....maybe I’m in way over my head.

You sent me to this post for a list of the mods. But I don’t see the list. How about instructions on how to do all of this. Tutorials?

I’m beginning to think this is for experienced DIYers. For a first time DIYer like myself (with headphones anyway, I have made some sweet stereo cables before with great results) are there any resources for the mods you’ve made, and are there some mods that you suggest I don’t touch. Eg. I’ve never soldered before.

What do you think of this guys mods...

https://www.head-fi.org/threads/senn-hd800-ultimate-s-mod.879024/

Also, it would be great to know what each mod does to the sound, more than just very tech talk, but also what it does to the sound signature, for us average hobbyists that know the sound we want so we don’t end up making a bunch of mods that move us away from what we personally want out of sound.

Thank you so much. And this work is awesome. Such a great community. I would love to do this all myself because I think it would be fun, and really fulfilling to know I did it.

 

[johnjen]

Say there derekphone.
I figured your questions and my answers might be better served if made more public as others may have similar questions as well.

I have been working on these mods for my 800's for several years now and overall my efforts have evolved and morphed into a multifaceted approach involving not just making modifications to the 800's themselves but also the signal that is fed to them.
This involves using DSP to compensate for the 'natural' FR of the 800's (especially from ≈100Hz down to ≈10Hz) as well as a few physical mods to the 800's themselves.

And as I wrote "So the full VMC mod (Vibration Management Control) involves the 3 aforementioned tweaks (hardwired cable, external screen removal, trapazoid treatment), plus what I have come up with as my tweak (VMC), which together all make up this mod."

I have since re-named the mod the HD800-Jmod or some such.

So here are a few pics of the cable I used in comparison to the stock cable connectors.
Yes this is about as big a set of wires as I'd be able to 'easily' fit in the existing space.


and here is the cable installed and wired.


And here is a pic of what my treatment to the trapazoid looks like.


and another of the outside of the 800's with the external screen removed.


As you can see these mods are not for the newb modifier as they are extensive and fully warranty obliterating, as is my usual modus operandi.

The EQ mods I have added in DSP are outlined starting in this set of posts.
https://www.head-fi.org/threads/the-diyrs-cookbook.781268/page-15#post-12300653

This should give you enough reading/cogitating/head scratching time to ponder if this approach is one you want to pursue.

As for the other approaches to modding the 800's that have been taken, including the one you cited, I can't say I have tried them to know if they would meet my expectations, or not. I have used elements of them (the outer screen removal and trapezoid treatment) along the way and have settled upon this set of mods as my final set of mods, and I don't see any need to pursue this any further as the results satisfy what I was seeking and so much more.

I hope this helps answer your questions, but do feel free to ask more if questions remain.

Thanks JJ

 

[johnjen]

So the answer back from Igor the City Light Voltage Control Specialist (anyone else catch the humor potential here?) is no, there is nothing they can do, as there are no additional taps on the step down xfmr on the pole in the alley.
And they certainly aren't going to adjust the 2400Vac that feeds that pole mounted step down xfmr.
In fact he was saying 124Vac that's good right the more the voltage the better, right?
It's the Uhmerikan weigh, MOAR is ALWAYS better, kind of approach.

So on to step 2, aka Plan B, the option of adding even MOAR BIG iiiiirrrrrrnnnnn to the mix.

I have acquired a Hammond 169J line voltage isolation xfmr with multiple taps on the secondary to compensate for the 122-124Vac (most of the time) voltage I normally see.
https://www.hammfg.com/electronics/transformers/line/169

This puppy is huge and weighs 22lbs all by itself, and I swear it must be warping gravity because it feels like it weighs 80lbs.

So as soon as I complete the re-wire job on my my BIG eyerrrrrrrnnnnnn amp I'll use it to drop the primary input voltage down to the 118 to 116Vac range, and then feed that to the amp and see what effect that has on all of the internal voltages and currents.

The saga continues…

JJ

 

[derekphone]

Say there derekphone.
I figured your questions and my answers might be better served if made more public as others may have similar questions as well.

I have been working on these mods for my 800's for several years now and overall my efforts have evolved and morphed into a multifaceted approach involving not just making modifications to the 800's themselves but also the signal that is fed to them.
This involves using DSP to compensate for the 'natural' FR of the 800's (especially from ≈100Hz down to ≈10Hz) as well as a few physical mods to the 800's themselves.

And as I wrote "So the full VMC mod (Vibration Management Control) involves the 3 aforementioned tweaks (hardwired cable, external screen removal, trapazoid treatment), plus what I have come up with as my tweak (VMC), which together all make up this mod."

I have since re-named the mod the HD800-Jmod or some such.

So here are a few pics of the cable I used in comparison to the stock cable connectors.
Yes this is about as big a set of wires as I'd be able to 'easily' fit in the existing space.


and here is the cable installed and wired.


And here is a pic of what my treatment to the trapazoid looks like.


and another of the outside of the 800's with the external screen removed.


As you can see these mods are not for the newb modifier as they are extensive and fully warranty obliterating, as is my usual modus operandi.

The EQ mods I have added in DSP are outlined starting in this set of posts.
https://www.head-fi.org/threads/the-diyrs-cookbook.781268/page-15#post-12300653

This should give you enough reading/cogitating/head scratching time to ponder if this approach is one you want to pursue.

As for the other approaches to modding the 800's that have been taken, including the one you cited, I can't say I have tried them to know if they would meet my expectations, or not. I have used elements of them (the outer screen removal and trapezoid treatment) along the way and have settled upon this set of mods as my final set of mods, and I don't see any need to pursue this any further as the results satisfy what I was seeking and so much more.

I hope this helps answer your questions, but do feel free to ask more if questions remain.

Thanks JJ

Say there derekphone.
I figured your questions and my answers might be better served if made more public as others may have similar questions as well.

I have been working on these mods for my 800's for several years now and overall my efforts have evolved and morphed into a multifaceted approach involving not just making modifications to the 800's themselves but also the signal that is fed to them.
This involves using DSP to compensate for the 'natural' FR of the 800's (especially from ≈100Hz down to ≈10Hz) as well as a few physical mods to the 800's themselves.

And as I wrote "So the full VMC mod (Vibration Management Control) involves the 3 aforementioned tweaks (hardwired cable, external screen removal, trapazoid treatment), plus what I have come up with as my tweak (VMC), which together all make up this mod."

I have since re-named the mod the HD800-Jmod or some such.

So here are a few pics of the cable I used in comparison to the stock cable connectors.
Yes this is about as big a set of wires as I'd be able to 'easily' fit in the existing space.


and here is the cable installed and wired.


And here is a pic of what my treatment to the trapazoid looks like.


and another of the outside of the 800's with the external screen removed.


As you can see these mods are not for the newb modifier as they are extensive and fully warranty obliterating, as is my usual modus operandi.

The EQ mods I have added in DSP are outlined starting in this set of posts.
https://www.head-fi.org/threads/the-diyrs-cookbook.781268/page-15#post-12300653

This should give you enough reading/cogitating/head scratching time to ponder if this approach is one you want to pursue.

As for the other approaches to modding the 800's that have been taken, including the one you cited, I can't say I have tried them to know if they would meet my expectations, or not. I have used elements of them (the outer screen removal and trapezoid treatment) along the way and have settled upon this set of mods as my final set of mods, and I don't see any need to pursue this any further as the results satisfy what I was seeking and so much more.

I hope this helps answer your questions, but do feel free to ask more if questions remain.

Thanks JJ

Thanks for the pictures. It perhaps is a bit out of my league as a first time diy project. If they were cheap headphones that would be different. Another issue I’m seein* as I’m going knee deep in research on these mods is that everyone has different taste and goals. We can all agree on the 6kh peak yes. So not being really sure what each mod is doing to the sound makes things impossible to decide if it’s worth implementing. I confess, maybe the sound characteristics have been explained super tech terms, but it has its been out of my understanding.

This, combined with the difficulty level of the mods, have me slowly taking quiet steps backward so maybe nobody would notice me. I kid.

I thought that the HD800 modded would definitely be my route but I’m now considering the 650 moddd because from what I read I think the sound signature is more to my liking. The question is whether the mods would have it compete with a modded 800. There are few comparisons online and of them , only of certainmods. Certainly not your mod. So I’m left with a lot of unknowns, on top of my own personal tastes.

2 things:

Has anyone compared HD650 modded vs. HD800 modded

Say there derekphone.
I figured your questions and my answers might be better served if made more public as others may have similar questions as well.

I have been working on these mods for my 800's for several years now and overall my efforts have evolved and morphed into a multifaceted approach involving not just making modifications to the 800's themselves but also the signal that is fed to them.
This involves using DSP to compensate for the 'natural' FR of the 800's (especially from ≈100Hz down to ≈10Hz) as well as a few physical mods to the 800's themselves.

And as I wrote "So the full VMC mod (Vibration Management Control) involves the 3 aforementioned tweaks (hardwired cable, external screen removal, trapazoid treatment), plus what I have come up with as my tweak (VMC), which together all make up this mod."

I have since re-named the mod the HD800-Jmod or some such.

So here are a few pics of the cable I used in comparison to the stock cable connectors.
Yes this is about as big a set of wires as I'd be able to 'easily' fit in the existing space.


and here is the cable installed and wired.


And here is a pic of what my treatment to the trapazoid looks like.


and another of the outside of the 800's with the external screen removed.


As you can see these mods are not for the newb modifier as they are extensive and fully warranty obliterating, as is my usual modus operandi.

The EQ mods I have added in DSP are outlined starting in this set of posts.
https://www.head-fi.org/threads/the-diyrs-cookbook.781268/page-15#post-12300653

This should give you enough reading/cogitating/head scratching time to ponder if this approach is one you want to pursue.

As for the other approaches to modding the 800's that have been taken, including the one you cited, I can't say I have tried them to know if they would meet my expectations, or not. I have used elements of them (the outer screen removal and trapezoid treatment) along the way and have settled upon this set of mods as my final set of mods, and I don't see any need to pursue this any further as the results satisfy what I was seeking and so much more.

I hope this helps answer your questions, but do feel free to ask more if questions remain.

Thanks JJ

I had a long response and my browser messed up. Ugh.

Basically.... thanks for your help and images. I’m left feeling like this is too much for a first timer to wisely take on. Secondly I still don’t know what each mod really does to the sound quality and signature. On top of that , what I want out of sound may be different than what you want. There are obvious improvements most can agree on, like the 6kh peak. So maybe I just go with the top few obvious tweaks and that are doable for a first timer. Which would you suggest? (I’ll add that I’m a tube, analoge, natural , magical vocals, kind of guy, with as much punch to drums etc without taking away from the other stuff.)

That being said, as I’ve explored the HD650 mods I’m thinking perhaps this headphone might be worth a look for me. I’m going to test the 650 vs the 800S at the store to see which I like better. I did test both but I gave the 650 like 1min of time because the comfort of the 800S was better and that trumps everything for me. But I do remember thinking damn those were nice! And thinking they just didn’t have the detail. But it had that, “I want to listen to these” feeling. Maybe that’s the tonality and midrange beauty that people talk about.

Has anyone tested modded 650 vs modded 800? And what mods?

 

[derekphone]

Also, there is a jmod 650. Is this by you JohnJen?

 

[johnjen]

No, I've never had a pair of any of the 6xx series.
And the ones I have tried were moderately uncomfortable for long duration listening.

But one of my next projects will be to mod the HD700's using the same techniques I used on the 800's.

I have worked on a set and can see some easy mods that would mimic the same type of modifications to the overshoot, applied to the 700's.
But don't hold your breath, my amp project is on the front burner and it be cook'n…

JJ