Bravo / Indeed Headphone amplifiers

[PinkFloyd]

May 31, 2010 22:26:06 GMT Judge Buff said:

No one has taken up the question that mrb has posed above and it is to the detriment of the entire community of Indeed/Bravo amp owners that hasn't made this mod.

It's covered in the G2 thread judge.... Leo mentioned it quite some time ago rockgrotto.proboards.com/index.cgi?board=b&action=display&thread=5213&page=14

Mike.

[Judge Buff]

Jun 1, 2010 10:11:34 GMT PinkFloyd said:

Hi Judge,

I notice you have used Nichicon MUSE 1000uF (NON POLAR) coupling caps.... have you checked the voltage across these caps?

Nope... I had 16V Silmic II 470s in earlier for 4 months with no probs, but I swapped in the 1000uF Muses when I changed the bias caps to 1000uF about 7 weeks ago. The sound is amazing for the pittance I have in it...

[Judge Buff]

Jun 1, 2010 6:49:40 GMT @solderdude said:

Hi Judge Buff,
Welcome aboard..

I respectfully disagree with some of your views.
All these amps have about 27dB amplification.
The specs that come with these amps have NO truth in them at all.
They are 'invented out of thin air' so to speak.
Do NOT trust the specs from all these different branded vendors.
Don't believe me ? ... measure for your pleasure... tone generator and oscilloscope needed
Especially the frequency range and amplification are complete nonsense.
They are all basically the same design.
Differences come from amplification factors of the tubes but most of all from the volume potmeters.
The ones that are thought to have such high gain simply have linear vol pots.
The ones with 'lesser' gain have log vol pots.
So the reason for different settings of the vol pot does NOT originate mostly from amplification factors but vol pot characteristics.

pots with an A on it are log type and have better travel, pots with a B on it are lin type and already start playing loud when you barely touch the vol pot making you think they have higher gain.

The only way to really check total gain is by comparing them with the vol pots at maximum setting and use the volume control of the source

SandyK is right on my account regarding the cathode caps IMO.
I'm a gonna do experiments and measurements on this subject on HH's amp.
As I like to validate all my mods (I know a stupid non-audiophile reason)

All these amps have crosstalk when unmodded. (it's because of the design.. it has the filament as a shared component, the audiosignals of BOTH channels pass through it and crosstalks to the other channel via the LM317 current source... that causes it.)
Simply pull one of the input plugs and listen to the channel that is supposed to have no sound on it.
Now cranck up the volume !!
Then tell me there is no nasty sounding crosstalk above a certain level.
The nasty sounding crosstalk above a certain (high level) point is caused by the LM317 not being in it's minimal required operating point.
IRL MOSFETS put this point further then IRF types, so higher levels are needed to reach the point of horrible crosstalk.
In normal operation the output signals are in phase so it isn't that bad in practice but there are certain passages (with very stereo signals) that won't sound anything near optimal

@ Imagemaker:

I still suspect your powersupply !
Them is cheap crappy chinese PS.

Set the bias level around 13 ... 13.5 Volt.
It's the optimal setting for the tube if you want symmetrical clipping.
This will give an output voltage of around 12Vpp (so around 4.2Veff)
Ofcoarse this is when loaded with 300Ohms.
With 32 Ohms the output voltage will have dropped to around 2Veff.

Use lineair regulated power supplies with these amps.
Highly recommended as these amps are very sensitive to the noises on the powersupply rails.

Hey solderdude,

The only thing I'll respond to concerns the "role" of the bias caps. Anyone with a multimeter knows that the caps shunt or bypass AC to ground, but that isn't the primary role here. They are meant to provide a constant cathode bias voltage by either charging via the tube or discharging across the parallel resistor. This is the "reservoir" that I was talking about. The other extraneous effects are noted, but they are just that in this part of the circuit: extraneous. That's the way I learned it 33 years ago as a Navy ET.

I've obviously had some career changes over the years, and these days I just tinker around with this stuff. It's not like I can tell what a system sounds like from just looking at the a scope or specs such as you but I do know how this amp sounds in regards to the mods I've made. If the %#$@ thing explodes tomorrow, it surely sounds awesome tonight with the Siemens Gray Shield I have in it at the moment.

I hope it doesn't explode, but if it does, I have two G2s and a Bravo V2 on the way, right now. I gotta go, as I just won six 5814s on the Bay... Cheers, Gents!

[Judge Buff]

Jun 1, 2010 10:35:31 GMT PinkFloyd said:

May 31, 2010 22:26:06 GMT Judge Buff said:

No one has taken up the question that mrb has posed above and it is to the detriment of the entire community of Indeed/Bravo amp owners that hasn't made this mod.

It's covered in the G2 thread judge.... Leo mentioned it quite some time ago rockgrotto.proboards.com/index.cgi?board=b&action=display&thread=5213&page=14

Mike.

Thanks, Mike... I saw it after I posted.

[Deleted]

Jun 1, 2010 6:49:40 GMT @solderdude said:

<snip>

@ Imagemaker:

I still suspect your powersupply !
Them is cheap crappy chinese PS.

Set the bias level around 13 ... 13.5 Volt.
It's the optimal setting for the tube if you want symmetrical clipping.
This will give an output voltage of around 12Vpp (so around 4.2Veff)
Ofcoarse this is when loaded with 300Ohms.
With 32 Ohms the output voltage will have dropped to around 2Veff.

Use lineair regulated power supplies with these amps.
Highly recommended as these amps are very sensitive to the noises on the powersupply rails.

Hello again, Frans

Thanks for your response to my messages. And, I think you are absolutely right about the PS being the culprit to the amp cutting out. Since I posted last, I received two regulated power supplies - one is a 24V/2A, and the other 24V/6A. When I would connect the 2A PS to the modded G1 or the G2 amps it would immediately cut out in a pulse-like pattern, indicating that the amp is pulling more than 2A of current , but when I would connect them to the 6A PS, all was well , and the amps sounded fantastic , with an amazingly deep bass - probably because of the 10000uF + 80uF photoflash caps installed in parallel.

About the bias - I still have it at 15V on both the G1 and the G2, but for the reasons you indicated, I will turn it down to 13.5V as soon as I get to it. I have not been comfortable keeping it at 15V - the trim pots are almost at their end position of lowest resistance. I do, however remember reading an early post by Pink Floyd (Mike), recommending the 16.5V bias , and I don't know how he was able to achieve this, for I tried, but reached the end of the trim pot travel at a little over 15V. I do recall that he was talking about the Bravo amp (the one that has the trim pots), of which I have two of the 12AU7 (no trim pots) version now, with one more on the way, which I in an auction 'dirt cheap' - $21.50US at eBay, neither of which have been modded yet, though I will, when I get around to it.

Thanks again!
imagemaker

[Deleted]

It's Solderdude.. as in soldering-iron holding dude, not a soldier dude although sometimes I am thought to be at war with certain Grotto members which, for me is, not the case as I am a pacifist...and just a bit ... too direct perhaps.. which is mistaken for something else.

The current in the anode circuit is constant because of the CCS (constant current source made up of the transistor, emitter resistor and red LED) as soon as the cathodes have their proper temperature/emission.
The cathode resistor is an autobias resistor here.
a more stable voltage across it (reservoir cap theory) does not make the anode current more constant as it already is ALWAYS very constant, regardless of the output swing on the anode.

The anode has a CCS instead of a resistor, as is mostly used in this configuration, but HAS to be used in order to get it to perform well enough because of the low voltages used.
If a resistor were used the current would vary too much and since it needs to amplify an input voltage and the gain is determined by S (mA/V) and the output swing is very close to the supply voltage this is the only proper way to go if you want to minimise distortion as the tubes output is mA not volts. To get volts you need to have a constant current which a resistor does not provide when a different voltage falls over it. (i.e. sound good)

So.. the cathode resistor biasses the output voltage of the anode so you can get it in it's proper output voltage operating point which is technically spoken between 13 and 13.5V for these amps for symmetrical clipping (when using IRL)
When using IRF with this anode voltage you will see the DC output change when you apply a big sinewave because the LM317 is out of it's operating point and the filament does NOT get enough voltage/current resulting in drifting of the DC voltage.
That's why the manufacturer says 12V operating point.
This way the glowing current is set a bit too low and musical contents let the DC bias vary under signal so that the average output voltage will rise some (slowly as the heater reacts slowly) moving it toward the same DC value as when an IRL were used (around 13-14V when the amp needs to deliver voltage).

With IRL these problems do only occur at almost maximum output power (your ears won't like it).
More stable DC as a result, less distortion in the negative half of the sine wave, the positive side of a sine wave is never distorted.

Anyway... the cathode bypass caps... they do only that.. bypassing AC in this amp.
NOT reservoiring for a constant current as it IS already always constant.
If it were needed for that purpose the manufacturers would have included it in their design or are they totally stupid ?
In any case they don't know how to measure their own specs so... they might be.... ;D

the gain of the amp, as there is no overall feedback, is determined by the tube (12.5 mA/V).
the input voltage across the grid and cathode resistor determine the gain (local feedback be it of very small importance in this case.)
When you bypass the cathode resistor the gain is determined by the voltage across the grid (tube) alone.

The gain of the amp = 27dB that's about 22x (most HP amps are between 5 and 10 times) with a max output voltage of 4 V the input voltage = 0.18V to achieve this with the vol pot at max.

the cathode resistor is about 1k (pot = 2k and usually in middle setting)
The current source = 0.4mA and the voltage across it about 11V so 27k resistance.
so if the gain determined by the anode/cathode resistor the gain would be (when S > say.. 50mA/V) 27 times because of the local feedback that is created and would be around 30dB in this case.
BUT S = 12.5mA/V so the gain is 'limited' by the tube's gain (about 22 times) and not by the resistors.
Now there will only be a slight change in the gain (about +1dB) when the cathode resistors (actually pots) are bypassed.
As this is very small difference this 'extra gain' is often perceived as 'something else'.

So, whatever you learned is not applicable in this particular situation regardless of how you may feel the amp works.
These are the facts as any engineer/designer can tell you.

Ofcoarse I am used to things IRL (in audioland) to be different so when you apply the caps there will be 'improvement' the amp will sound.. uhmm... better without a doubt.

Personally (from a technical point of view) I would use polarised caps in the output as there is 8 to 10V DC over these caps and polarised caps handle that better then non-polarised (designed for non DC carrying things)
Ofcoarse again, non polarised sound better... matter of taste I presume.
You be the Judge...

I know... there will always be discussions between audiophiles and techs (sound vs measurements)
My technical expose's here are often not (fully) understood as most guys here are audiophiles with no or little technical background.
There are some members that do have technical background.

Please don't let me discourage you posting here as any member here is valuable.

End of technical shit...
Let's get back to listening again ..

[Deleted]

Jun 2, 2010 8:23:47 GMT @solderdude said:

I am Solderdude as in soldering-iron holding dude, not a soldier dude although I sometimes told I am at war with certain Grotto members which, for me is, not the case as I am a pacifist... just a bit ... too direct perhaps.. which is often mistaken for something else.

We loves ya really Frans.

That first sentence sounds like the start of a film ......

'I am John Connor. We are at war with the machines.' (Terminator)

You are scarily good with a soldering iron though.

[Deleted]

@ imagemaker

The amp itself draws only about 350mA (0.35A) so if the 2A PS cuts out there is probably something wrong in this power supply or something is drawing huge currents (2A @ 24V = 50Watts)

SMPS usually don't particularly like to 'see' huge capacitors.
In fact often a maximum is specified.
See it as an enormous flywheel that has to receive huge blows to get it to move and if you have to change its speeds huge amounts of energy are needed again.
Smaller caps are equivalent to a light flywheel.
It will still spin but takes less force to start it or adjust it's speed.

If you have the opportunity see how much the amp draws.

The 6A PS is virtually 'unloaded' with 0.35A so it should be very easy for it and not even get warm.

See mods for the 12AU7 version in the 12AU7 thread in the DIY section.
Easy to get rid of the crosstalk in these amps and to adjust the bias by strategically placing pots.
All other mods (IRL etc) are valid for these amps too.
Had been thinking about buying one myself but already have plenty stuff lying around.

[Deleted]

You are scarily good with a soldering iron though.

Its great reading your technical stuff Frans, be even better if i understood it more, you certainly have great understanding and knowledge of Electronics my friend

[Deleted]

Jun 2, 2010 19:44:06 GMT @solderdude said:

@ imagemaker

The amp itself draws only about 350mA (0.35A) so if the 2A PS cuts out there is probably something wrong in this power supply or something is drawing huge currents (2A @ 24V = 50Watts)

SMPS usually don't particularly like to 'see' huge capacitors.
In fact often a maximum is specified.
See it as an enormous flywheel that has to receive huge blows to get it to move and if you have to change its speeds huge amounts of energy are needed again.
Smaller caps are equivalent to a light flywheel.
It will still spin but takes less force to start it or adjust it's speed.

If you have the opportunity see how much the amp draws.

The 6A PS is virtually 'unloaded' with 0.35A so it should be very easy for it and not even get warm.

Frans,

I did measure the current that the G2 draws -- after the initial momentary surge of 900mA upon power-up, it settles down to a steady 570mA, and that is with the 10000uF + 1000uF + 80uF Photoflash caps in parallel installed, and with the bias at 13.5V. With the 6A PS there is no problem, but it seems that the protective circuitry of the smaller 2A PS is set somewhat low. The latter works fine with the unmodded Bravo with its 6800uF cap, so apparently it draws less current, which I didn't measure.

See mods for the 12AU7 version in the 12AU7 thread in the DIY section.
Easy to get rid of the crosstalk in these amps and to adjust the bias by strategically placing pots.
All other mods (IRL etc) are valid for these amps too.
Had been thinking about buying one myself but already have plenty stuff lying around.

I have yet to get around to modding the Bravos I have, but I will soon. I have the parts now; it is just a matter of finding the time to do it.

As always, your valued expertise is most welcome!

[Deleted]

Don't know the exact schematics of the G2 (I expect the amp part to be the same) the relay shouldn't draw that much extra.. but might account for the extra mA if it is 'on' together with the raised power resistor.

The old versions draws around 350mA (that's why the old version works well on the PS probably)

< edit >

O.K. clear to me now. (thanks to pictures Mike took a while back)
The G2 draws (acc. to calculations) around 590 mA.
The current is higher because the output stages of the G2 run at twice the current of the first design.
This is accomplished by the big resistor in parallel to the filament that dissipates the extra current into heat.
downside: the LM317's not having had a bigger cooling fin as the MOSFET's have is getting even hotter then before.
MUCH hotter when an IRL is used !
I expect it to run at 100oC easily !

Here I smell a modding opportunity to get rid of the crosstalk !!
Very simple and SHOULD do the trick... but someone has to try.
Underneath the big (lifted) power resistor there is a wirebridge (at least I expect it to be that or a very low-ohmic resistor.. can't see).
Just solder it out or cut it.
Then check for proper operation and if the crosstalk is gone.
IF it isn't a wire but a (not so low ohmic) resistor crosstalk may already be greatly reduced and this might be also why the G2 sounds better then the previous version.

Please report back if someone has had the nerve to try this 'hunch'.
great to see if the crosstalk ads or deteriorates the sound in some way.
a daredevil me even solder a switch in place of the wire and switch to see (actually hear) what happens.

[mrarroyo]

Have you all seen the Bravo V3?

[Deleted]

Yep,

Man227 (Chaz) spotted it in April.
rockgrotto.proboards.com/index.cgi?board=b&action=display&thread=5289

It's a V1 with a 3 band tone control next to it.
The specs that come with it cannot possibly come from this amp.

I'd spend my money on a G2 over this one unless you like/need some tone control.

[Deleted]

Here I smell a modding opportunity to get rid of the crosstalk !!Very simple and SHOULD do the trick... but someone has to try.Underneath the big (lifted) power resistor there is a wirebridge (at least I expect it to be that or a very low-ohmic resistor.. can't see).Just solder it out or cut it.Then check for proper operation and if the crosstalk is gone.IF it isn't a wire but a (not so low ohmic) resistor crosstalk may already be greatly reduced and this might be also why the G2 sounds better then the previous version.Please report back if someone has had the nerve to try this 'hunch'.great to see if the crosstalk ads or deteriorates the sound in some way.a daredevil me even solder a switch in place of the wire and switch to see (actually hear) what happens.

Hi Frans,
the resistor under the raised one on the G2 that you want removing looks to be the same size as the one on top.
Is it still ok to remove it

[Deleted]

Hi Mick,

thanks for checking.
Maybe you can do the following to make things clearer to me.

1: Check if there IS crosstalk in these amps (by pulling one input plug and listening to see if the channel that is supposed to be silent is truly silent.

2: Measure the resistance of the top and bottom resistor in the Ohms setting of a multimeter with NO power on the amp and disconnected from the power supply for at least 1 minute.
The tube MUST be out of it's socket for this measurement as the filament is parallel to the upper resistor and measurements would not be accurate.

I expect the upper resistor to be around 22 Ohms and the bottom one only a few Ohms.
Deleting the bottom resistor or not depends on the findings.

thanks for assisting so far,

Frans

[Deleted]

Hi Frans,

They are both 34 Ohms, and cross talk does occur but only at high(12oclock and beyond, to loud for normal listening) volume.

[Deleted]

O.K. thanks Mick !

Will change the schematics.
Crosstalk should be far less compared to the old versions.

You cannot take the bottom one out just like that in this case.

This can only be done (bottom one that causes the crosstalk) when the top resistor is replaced by a 18 Ohms (3W - 5W) resistor.
In this case there is no more crosstalk at all.
Both channels have completely separated channels if this is done (like any other HP amp).

Still curious about one thing...

When the amp is working I am interested in 2 voltages.
The (DC) voltage from 1 leg of the bottom resistor to ground and the other side of the bottom resistor to ground.
ground being the shield of the input cinch connectors.
The amp has to be playing... be carefull not to short things.

[Deleted]

ground being the shield of the input cinch connectors.The amp has to be playing... be carefull not to short things.

Not to sure what input cinch connectors are Frans.

presume you meant RCA grounding.

Looking front on with Amp standing as normal , right hand leg on both upper and lower resistor is reading 6.56 volts.
Left hand leg on both is reading 0.02mv.

Readings taken at normal listening levels, source Cowon mp3 in 3.5 socket, HP`S Senn 250`s.

[Deleted]

Cinch connectors are also known as RCA phonos, the twin I/P and O/P on the G2.

[Deleted]

Yes thought that to be the case Chris, cheers.

[Deleted]

Chris and Mick, thanks for the input.

Here is the deal with the 2 resistors in the G2 amp.
They are actually a nifty circuit.
As most tube rollers know there are 6V and 12V tubes.
These 12V tubes can also be used on 6V by using a center-tap pin.
For this purpose the 6-12V switch is the solution.

When a glowing VOLTAGE is used the drawn current is determined by the voltage and resistance of the filament.
No troubles with tube rolling in this configuration.
Most amps have a 6.3V AC (or DC) heater voltage but this amp uses a heater current.
Actually the class-A currents in the end stage (MOSFET/LM317) determine the glowing current.
This is where the troubles start when tube rolling.

As most tube rollers will be aware of there are tubes that draw 300mA and 365mA at 6.3V
When a current supplies the filament this would mean that either the 300mA would be powered correctly and the 365mA would be 20% underpowered OR the 365mA would be powered normally and the 300mA would glow far too bright (20% over voltage) and have a shorter lifespan.
There is always the compromise... 330mA.. still the 300mA would receive 10% overvoltage and the 365mA 10% too low.

This is where the 2 power resistors do their thing.
The way the resistors are combined they actually divide the heater current so that the + and - 10% over and undervoltage that would occur come closer to their desired levels,
a 365 mA tube will receive about 350mA (so only 5% undervoltage) and a 300 mA tube will receive about 315mA (so only 5% overvoltage) while still maintaining a constant and equal class A current in both end stages.
Furthermore the crosstalk is much less then the older versions.
This is perhaps most why these amps sound better then the older versions.

Ofcoarse every upside has it's downside... that's life..

What's the downside ... that is (when using IRF MOSFETS) the LM317 in one channel is already almost in it's minimal required operating point giving distortion in 1 channel at lowering listening levels then the other channel and the 'nasty' crossover from one channel to the other starts earlier then the other way around.

Can we get rid of the crosstalk and bring the distortion down..
YES, but at a price... the price being the tube rolling cababilities will be less as you can either use only 300mA tubes or 365 mA tubes and there have to be things modded.

Using IRL510 instead of IRF510 helps already in the distortion part.
adjusting the anode voltage to 13-14V too (LM317's will get very hot... yes every upside has it's downside)

The bottom resistors has to go and the upper one has to be replaced by a 22 Ohm resistor for 300mA tubes and 18 Ohms resistor when using 365mA tubes. (3W resistors, will get hot)
When using 300mA tubes the LM317 resistors can stay in their place but when using 365mA tubes the free spaces on the PCB next to the LM317 resistors have to be filled with a 22 Ohm resistor.

Not for those that roll tubes regularly but the people that don't roll tubes and want lower distortion and no crosstalk this might be a good mod.
Otherwise leave it as it is and replace the IRF630 wit IRL510 or IRL530

[PinkFloyd]

Frans, just to confirm, they are both 34 ohm as Mick says (well, I measured 34.2 ohm)...... I haven't looked at the G2 for months now, it's sitting out in the workshop gathering dust.... in fact, there was a spider's web on top of it when I checked the value of the resistors

As you are actively helping people with their Indeed amps (and don't own one yourself) I am going to send mine over to you as a gift Frans, please PM me your details and I'll ship her over on the next flight out.

There is a limit to the amount of headphone amps any one man can listen to and I have reached that limit! I'm, literally, tripping over these nasty Chinese SMPS power supplies (have boxes full of them) and have way too many unused headamps kicking about the place....

I used to be SO interested in headphone amps when I didn't have one.... now they are just getting on my nerves.... over exposure to "anything" isn't good...

The main reason I'm taking a break from the forum, I've lost interest (due to being over exposed)..... time to sit back from it all and have a rest.... when one headamp sounds the same as another that's a message that you are over indulging / have become complacent......

I've got to get my "hunger" back and the only way to do that is to starve myself for a few months and then come back to the table with a fresh "hungry" pair of ears.

It also doesn't help matters that I have got back into loudspeakers in a big way.....

Anyways, Frans, PM me your details and I will get the G2 over to you ASAP.

Mike.

[Deleted]

Frans,

Of course you know we'll all be waiting for a detailed run down on how to mod a G2 to soopa doopa status!

No pressure

[Judge Buff]

Jun 2, 2010 8:23:47 GMT @solderdude said:

It's Solderdude.. as in soldering-iron holding dude, not a soldier dude although sometimes I am thought to be at war with certain Grotto members which, for me is, not the case as I am a pacifist...and just a bit ... too direct perhaps.. which is mistaken for something else.

End of technical shit...
Let's get back to listening again ..

Solderdude... now that does make more sense. I gotta change this theme so I can read easier.

As to the rest of the $%&# : I stand corrected. The Muses do sound better than the Silmic IIs did.

[Deleted]

Hi Mike,
Sorry to read of your over exposure - hope your appetite returns soon. At least the timing is a good as such a thing can be - plenty of other things to do on these long warm days. Let's hope normality returns in time for those long dark winter evenings .
Dave.