Bravo / Indeed Headphone amplifiers

[Deleted]

@ Alex,

I know... I was just having a laugh. ;D
Can't hear beyond 18 kHz myself.

For the non technical readers I will explain the input-filter Alex mentioned.
(skip it if you don't want to get bored)

The input filters in SS amps are to prevent the weakest parts (slowest amplifying parts in an amplification stage with feedback) of the used opamps/transistors from saturating causing distortion in the audible range.
This happens because it takes some time for saturated transistors (also transistors inside an opamp) to recover catch up with the faster parts and function properly again.
When this happens the amplification stage cannot 'do' anything else.
Since the open loop gain of MANY of these amps falls in the audible band....

This only happens when the rise/fall times of the source exceed the rise/fall times of the amplification circuit.
In other words the output of the amp can't 'follow' the input signal anymore and internally 'clips'.
The faster the op-amps the higher these filters cut-off frequencies can be constructed.
A lot of manufacturers (even some high-end) don't even apply such a filter... stupid IMO
Just a few parts and cost near to nothing !

The input tube of these little amps doesn't have this affliction (no overall feedback and the only part that amplifies) and the FET behind it is faster and alos in source follower mode so no filter needed.
The 10kHz square wave in the article looks quite exemplary IMO
Even the 100kHz is not even that bad..

Just look at the squarewaves from a lot of opamp manufacturers and you will see nasty edges (especially on the corners) and ringing (damped oscillation) or overshoot (a peak higher then the original signal that should not be there) at high frequency square waves.

Fortunately these high frequencies are not present in audiosignals (unless you are using cheap and low bandwith opamps) so they are easy to filter at the input preventing the opamp/amp to ever reach these conditions.
These high freq components can, however, couple in or seep through filters from sources so amps that do not have this input filter still could 'suffer' because of this..

Hence.. use a filter on the input.
Oh gosh ... me and my filters...that's probably why Mike put 'Filtration, filtration, Filtration the 3F's' under my name.
Demonstrates the importance of filters (be it on the power input, output of amps (boucherot filter) or input filters.

[Deleted]

The input tube of these little amps doesn't have this affliction (no overall feedback and the only part that amplifies) and the FET behind it is faster and alos in source follower mode so no filter needed. The 10kHz square wave in the article looks quite exemplary IMO

To put that into context, attached are a couple of screen grabs from 2 different solid state headphone amplifiers WITH input filters, AND using negative feedback. One is the G.S. Novo, and the other is the SC HA using a 50K Alps pot.

[Deleted]

Simple SS designs can easily outperform simple tube-amp designs .. no doubt about it.
For a cheap tube amp with no input filter (modded amp that is) the 10kHz SqW is just as good as Alex's right scope picture above.

SS amps will have no compression effects and smaller THD.
If you look carefully in the article's scope pictures and know how a sinewave should look like you can see compression taking place (in the bottom half of the sinewave) at full output power.

The left picture resembles the unmodded Miri more closely (limited bandwith compared to the right scope picture)

The pictures below shows what happens when you use the venerable OPA627/637 and you do not use an input filter on it. (left above is what I was talking about)
Right below shows clear overshoot.
Mind you, the PCB layout and decoupling of the power rail can evoke similar effects even in op-amps that do not exhibit this behavior under the right circumstances.
(excerpt from the original datasheet)
Talking about MHz's bandwith here, much more then the tube amp .. but do you really need this bandwith (MHz's)?
This is what I meant..



Well designed amplifiers will perform like the examples from Alex of coarse.

[PinkFloyd]

Jun 12, 2010 9:23:49 GMT @sandyk said:

The input tube of these little amps doesn't have this affliction (no overall feedback and the only part that amplifies) and the FET behind it is faster and alos in source follower mode so no filter needed. The 10kHz square wave in the article looks quite exemplary IMO

To put that into context, attached are a couple of screen grabs from 2 different solid state headphone amplifiers WITH input filters, AND using negative feedback. One is the G.S. Novo, and the other is the SC HA using a 50K Alps pot.

Which one is the GRAHAM SLEE NoVo?..... left or right?

[Deleted]

Which one is the GRAHAM SLEE NoVo?..... left or right?

Mike
Do you REALLY need to ask ?
I think that Frans comments about before and after of the Miri are a pretty good guide, and for similar reasons .

The left picture resembles the unmodded Miri more closely (limited bandwith compared to the right scope picture)

Alex

[PinkFloyd]

Jun 12, 2010 22:11:19 GMT @sandyk said:

Which one is the GRAHAM SLEE NoVo?..... left or right?

Mike
Do you REALLY need to ask ?

NO...... "but" I don't want to be the one that the son of god is pointing his finger at..... he thinks he's the messiah of headphone amps and takes badly to "real world" criticism..... I don't want to upset him, he might start another Volcano eruption or an oil spill

[Deleted]

The 'messiah' made the one on the left. (pretty sure of it..)
With a similar bandwith as the original Miri, bravo, indeed.
acc to your 'friends' specs (he calls it flat.. the charlatan !, I start to dislike him too now Mike )
Flat: 43Hz to 16kHz (-1dB), 22Hz to 30kHz (-3dB)
Original Miri: 1.2Hz to 15kHz (-1dB), 0.9Hz to 32kHz (-3dB)
I call it: 'mimicing the bandwith of poor designed tube amps in order to let it sound similar'.
He (the Messiah) goes even further and limits the lower frequency band also to emulate an output transformers limited bandwith (the b@5#&~d)
The modded Miri comes closer (in bandwith) to the SC HA.

Funny to see you can make a SS amp sound tube amp-alike (acc. to the Messiah and the 'revaleting press') by limiting the bandwith.

@ Alex,

What the hell are are you doing with a NOVO...
I thought you'd like your bandwiths MUCH wider then that ?

[Spirit]

Jun 13, 2010 7:02:27 GMT @solderdude said:

@ Alex,

What the hell are are you doing with a NOVO...
I thought you'd like your bandwiths MUCH wider then that ?

I'm pretty sure that was from when one of Alex's mates was part of the Novo auditions...?
(Sample unit(s) was/were sent round Oz and a few other places)

[Deleted]

Jun 13, 2010 8:26:12 GMT Spirit said:

Jun 13, 2010 7:02:27 GMT @solderdude said:

@ Alex,

What the hell are are you doing with a NOVO...
I thought you'd like your bandwiths MUCH wider then that ?

I'm pretty sure that was from when one of Alex's mates was part of the Novo auditions...?
(Sample unit(s) was/were sent round Oz and a few other places)

Phil
Yes, a Sydney RG member and myself auditioned one.
I found the bottom end "warm" and lacking impact.
The attached photos taken by Dave Mitchell from the U.K.will show the rolled off LF response used to help achieve that.
Alex

URL=http://img28.imageshack.us/i/100hzsqcomparison.jpg/][/URL]

[PinkFloyd]

Jun 12, 2010 9:23:49 GMT @sandyk said:

The input tube of these little amps doesn't have this affliction (no overall feedback and the only part that amplifies) and the FET behind it is faster and alos in source follower mode so no filter needed. The 10kHz square wave in the article looks quite exemplary IMO

To put that into context, attached are a couple of screen grabs from 2 different solid state headphone amplifiers WITH input filters, AND using negative feedback. One is the G.S. Novo, and the other is the SC HA using a 50K Alps pot.

This is interesting stuff and good material for a seperate thread.... what do you think Alex?

[Deleted]

Mike
Unless Frans wishes to further expand on the subject of square wave testing, and how deficiences, deliberate or otherwise, can affect the resulting sound, even to the extent of giving a potentially very neutral sounding SS amplifier, many of the characteristics of a cheap valve design, then perhaps we should leave it at this, in case we are accused of G.S. bashing ?

Alex

[Deleted]

Nahhh.

Me thinks enough is said and done on this subject.

a SS vs Tube thread is likely to end up the same as Vinyl vs CD.

A thread for testing amps with specialised equipment is another subject.
Hard to do/understand if you don't have enough knowledge and equipment.
Most RG members/guests are non technical music/gear loving people, a few technical members aside, and not able to do these tests nor (fully) comprehend the significance of the results so don't see the point of a separate thread for these reasons as it would only be interesting for a happy few.

< edit >

O.K. maybe there are more intersted then I thought

couldn't resist also (yes resistance was futile... again)

the schematics for a G.S. Novo emulator.
Put it between a rather low Ohmic source (HP output of source) and a HP amplifier with a large bandwith (R in bigger then 10k!) and you will be auditing a G.S. Novo.

[Seamus]

Let me say as someone who isn't very technically minded, that I'm still interested in how and why things work and have enjoyed reading the posts of those who do understand and can demonstrate these things. I'd be very happy for more technical discussions like this, just wouldn't be able to contribute much

[Deleted]

Jun 14, 2010 13:39:01 GMT Seamus said:

Let me say as someone who isn't very technically minded, that I'm still interested in how and why things work and have enjoyed reading the posts of those who do understand and can demonstrate these things. I'd be very happy for more technical discussions like this, just wouldn't be able to contribute much

I'm with Seamus on this - I don't understand much of it at all but find it interesting and isn't that how we learn. This of course is dependant on the technical guys being willing to answer questions from numpties like me in layman's terms.
If you don't want it, don't visit the thread - simples! .
Dave.

[PinkFloyd]

Yes, I'm with seamus too! Nothing to do with Graham Slee bashing.... these guys put these amps on the market making the claim they are the best things since sliced bread.... it's interesting to see just "how" substantial these claims are

Look at the Bravo / Indeed (for example) totally misleading / inaccurate "specs".

[leo]

A lot of commercial amps are set to limit bandwidth , going by the shite stuff fed into the amps it probably helps reduce hassle

Its quite easy to higher the bandwidth of some amps by changing a few parts, as long as the amp doesn't take off that is

[Deleted]

True..

But the Novo's bandwith is made that way deliberately.
(quite sure of that)
As is probably also the Indeed and alikes.

This is to 'emulate' tube sound (make 'm sound warmer (by slightly damping the top end) and give tighter bass (by removing the bottom, in case of the Novo)).
The crap you are fed by telling you special this and that is used is .. (not gonna say it)
See how many peeps like the Novo and other GS amps.
See how many people like the unmodded Indeed and alikes.
See how many people like certain other brands..
I am sure more brands use the same or similar 'tricks'.
That is for a reason...

It's not fooling people... it's giving them what they like/want/prefer and suit their personal taste !
Doesn't mean you can't improve/alter any of these amps.
They will just sound more like all other amps which is NOT what 'they' want.
'They' want to stand out and be different the rest.
Have something that has an edge.

It's for the users (with their ears) to find what they really like.

Me likes them straight.. not everyone does.

Rather pointless hauling them all through the testbench to see how they do.
Just enjoy the things as they are or buy another one (or mod them if you want to improve them)

That's why the IRF630 sounds uhmmm.. warm (rolled off treble)
The IRL/IRF530 sounds full (-0.5dB @ 20kHz is found to be near ideal for most)
The IRL/IRF510 sounds sparkly. (frequency range expands beyond the audible range i.e. straight)

That's really the biggest part there is to it.
Yess. there is distortion, odd and even harmonics, crossfeed or no crossfeed (for HP's) and other small things but the frequency range tells the biggest story.

[Deleted]

Thanks Mike..

Yep.. I feel sorry I joined too late to catch Rick in a better condition.
He was well liked here and after he died I read back a lot of his posts.
He was a knowledgeable guy for sure.
Can see why he was well liked.

Me telling it like I see it can be a bit ..uhm .. confronting (and always too serious too)

[rmouser]

:oHello Gents,

I'm a noobie (Ron) to the forum from across the 'pond', Colorado. I have been monitoring this thread for a while. I have done the IRF510, crosstalk, power switch, linear PSU and different tubes to my Bravo 6N11 and 12AU7. I got them on eBay when they could be had for $40 USD including shipping.

I will probably stop the mods here since they both sound simply incredible with mods done so far, thanks to your many pearls of wisdom!

Edit: OK, I think I know how to add the smilies now

Apart from the Bravos, most of my experience has been op amps with buf634, irf510 buffers and 3 channel active ground ; 16 builds so far including ones I am selling catering to those with reasonable hearing impairments (those amps have EQ and balance controls).

Your forum is the best I have encountered and I look forward to exchanging info across the pond.

Cheers!

[Deleted]

Jun 14, 2010 20:55:16 GMT @solderdude said:

True..

But the Novo's bandwith is made that way deliberately.
(quite sure of that)
As is probably also the Indeed and alikes.

This is to 'emulate' tube sound (make 'm sound warmer (by slightly damping the top end) and give tighter bass (by removing the bottom, in case of the Novo)).
The crap you are fed by telling you special this and that is used is .. (not gonna say it)
See how many peeps like the Novo and other GS amps.
See how many people like the unmodded Indeed and alikes.
See how many people like certain other brands..
I am sure more brands use the same or similar 'tricks'.
That is for a reason...

It's not fooling people... it's giving them what they like/want/prefer and suit their personal taste !
Doesn't mean you can't improve/alter any of these amps.
They will just sound more like all other amps which is NOT what 'they' want.
'They' want to stand out and be different the rest.
Have something that has an edge.

It's for the users (with their ears) to find what they really like.

Me likes them straight.. not everyone does.

Rather pointless hauling them all through the testbench to see how they do.
Just enjoy the things as they are or buy another one (or mod them if you want to improve them)

That's why the IRF630 sounds uhmmm.. warm (rolled off treble)
The IRL/IRF530 sounds full (-0.5dB @ 20kHz is found to be near ideal for most)
The IRL/IRF510 sounds sparkly. (frequency range expands beyond the audible range i.e. straight)

That's really the biggest part there is to it.
Yess. there is distortion, odd and even harmonics, crossfeed or no crossfeed (for HP's) and other small things but the frequency range tells the biggest story.

Frans, That was really inspiring writing. Thanks for that.

[PinkFloyd]

X2

[Deleted]

A last photo to show how LF tailoring helps achieve the sound some people find pleasing.The RHS photo was taken when there were 4.7uF input capacitors fitted.

URL=http://img3.imageshack.us/i/18hzsq.jpg/][/URL]

[Deleted]

Seriously mate...... just take a back seat

Mike

I would presume that Leo's demotion from Global Moderator was a glitch in the system, that I have now corrected.

Alex

[Deleted]

Jun 11, 2010 21:20:22 GMT @solderdude said:

Here is a write-up I did about modding Mick's Miridiy amp.
A technical article for those that are interested in the workings of this amp. (2MB pdf file)
You can download the file from this location:

www.mediafire.com/file/vjmjjmukwj2/modding the Miridiy amplifier.pdf

<clip>

Frans,

Do you mind elaborating a bit more on the mod involving installation of the TracoPower TSR1-2465 switch-mode voltage regulator, and cutting of traces? It would be most useful if you would be willing to indicate this mod on some photos, showing which traces exactly need to be cut. I think I can trace the heater power supply to pin 4 OK, but I am not too clear about the heater power supply to pin 5. I am worried that with my limited experience I may cut the wrong traces on the PCB.

I have successfully performed some of the other mods (IRL510, LT317A, cathode bypass caps, photoflash cap in parallel with the power cap, gold plated stereo jack on the way), and my G2 is much improved as a result, powered by a regulated SMPS. All this, thanks to your unselfish willingness to share your expert knowledge freely!

Thank you in advance for your continued guidance offered to some of us amateurs! You are one of a kind! And, this is perhaps the best audio forum on the globe!

[Deleted]

I will when I have some time for it.
This can ONLY be done if you move the LM317's to a large cooling surface.
I do this by mounting an aluminum plate underneath the amp.
This mod can't be done with the LM317's on the small cooling fins.