[NeilR]

Where exactly is the Millet's closed feedback loop? I must have missed it when I reviewed the schematic

 

[rickcr42]

looking at the schemo right now (alweays a good idea

) and dammit,seems you are right !

I can't see any connection from the output back and so it would seem unless you are paralleling buffers the resistor can be tossed without ill effect.
sorry 'bout the screwup guys.Made an assumtion and you know what that usually means (Made an A*S out of U and ME with this case being me

)

 

[00940]

Rick : we're mixing way too many different things here.

- RFI : best taken care of with a Zobel.

- capacitive loading : not a problem normally with buf634 but the output resistor can be useful indeed. Bandwith limiting could help too in case of instability.

- reflections on transmission lines : at audio frequencies and with the lengths of wire involved, a non issue.

- back emf : not sure a simple resistor will do much.

 

[NeilR]

You have no idea how much satisfaction a peon like me gets by tripping up one of the grizzled vets here

It almost makes this hobby worth while.

Seriously though, the Millet is probably the only totally feedbackless amp circuit floating around here. It is easy to forget. It should have a logo with circle and bar or something like that.

I run mine with the discrete bufers and I was pondering taking them out, just for some entertainment value and a change in pace. I agree though, that with an average of 0.25% to 0.50% THD and IMD, it kind of takes the edge off many of the tweaks. It's like the good old days when men were men and amps put out a little distortion and that's just the way it was!

 

[Voodoochile]

Well, the discrete buffer does not have current protection built in, as the BUF634 does. So you might want to leave them... or at lease be cautious in seating the HP plug all the way.

Do you have a decent resistor in that location?

As for the distortion, rerview the article for that reason also. Pete talks a fair amount about gain and distortion, relative to bias. The better-sounding bias point often has more distortion in it, but it falls in a different harmonic than that of the lower distortion point of operation.
(that's a horrible sentence... good luck interpreting it)

 

[NeilR]

Voodoochile,

I know what you are saying.... 2nd order harmonics are "ok", higher order is baaaaaaaaad. Also regarding the buffers, that is what has kept me from doing it. I would hate to have to write an operating manual for a Millet with HA3-5002 or discrete buffers and no output resistor. It would go something like this:

"Always unplug the headphones before turning the amp on and off because dangerous DC trancients and annoying thumps may be experienced. Failure to heed this warning may damage your expensive headphones"

"Never unplug your headphones while the amp is turned on. Otherwise, you may blow up your output buffers"

Kind of a catch-22, I guess. Pick which thing you want to blow up. Of course, in the real world, no one is complaining about either blowing up. I was "trained" on the Headroom amps (over 10 years ago) so I am used to the idea of being very careful about removing headphones while the amp is on, or at least the volume is turned all the way down. Interestingly, their Micro amp cannot be powered off with a switch while it is connected to AC. The Millet is just the opposite and it is hard to switch back and forth on the shut down procedure. I think Headroom still makes amps with unprotected HA3-5002 buffers. I wonder what their service rate is for blown buffers.

Regards,
Neil

 

[amb]

The diamond buffer uses two pairs of MJE243/253 or BD139/140 transistors which are rated high enough that the polyfuse should trip long before the transistors suffer any ill effect (not to mention that there are emitter resistors to help too). Not to say that the output resistor isn't additional peace of mind, but the likelihood of transistor damage should be very small.

 

[rickcr42]

Quote:

Rick : we're mixing way too many different things here.

I am in the middle of something "off site" so will have to get back to you with the documentation for what I was trying to explain.All there and very clearly layed out in plain english.a zoebal is seriously overkill when a simple resistor will cure the possible ill.Kinda like fishing with TNT.you will catch fish but likely mangle them in the process

Quote:

You have no idea how much satisfaction a peon like me gets by tripping up one of the grizzled vets here

luckily I can take it as well as dish it

then again "no where to hide" means just that

Quote:

Seriously though, the Millet is probably the only totally feedbackless amp circuit floating around here. It is easy to forget. It should have a logo with circle and bar or something like that.

You are right about that.I have been SE/Open loop for a while but my amps are hardly DIY project material.The Millett is most likely the absolute biggest bang for the buck amp done as a straight build with no fancy parts yet having enough wiggle room to make it "Mongo Amp' to suit any taste.Pretty cool really.

Quote:

I run mine with the discrete bufers and I was pondering taking them out, just for some entertainment value and a change in pace. I agree though, that with an average of 0.25% to 0.50% THD and IMD, it kind of takes the edge off many of the tweaks. It's like the good old days when men were men and amps put out a little distortion and that's just the way it was!

it has been proven beyond any shadow of doubt that an amp speccing in at 0.001 %THD (TL072 opamp for one) can sound way worse than one having even 1% distortion and the human ear has a way of liking things test equipment does not and dispising things loved by the bench technician who is all about numbers and comparing specs but never actually listens.

always wondered why some feel they should be ashamed somehow for liking the "sound" of something because it does not rip to shreds other amps that sound worse in specs.I am pretty sure my main power amp rolls off around 30hz and highly doubt it is dead flat at the top yet is sounds better than any other amp I have ever owned bar none so I have no desire to replace it until I can find a way to get the SAME sound in a higher power amp (mine is only 8 watts

)

Quote:

The better-sounding bias point often has more distortion in it, but it falls in a different harmonic than that of the lower distortion point of operation.

The old "lean and mean" or "ballsy" from the guitar amp biasing sector.Same amp,same parts,nothing but a bias change,two entirely different creatures.Works for guitar amps because a musician is all about expressing what THEY want you to hear,a home amp you try for the romantic side of nutral because you are not trying to change as much as enjoy what is already recorded.
to make matters even worse there is no "one size fits all" since we all have varying grades of hearing loss,different taste,different headphone/amp balance requirements.all about locking in the sound so it sounds "pretty good" with everything thrown at it instead of it being a "rock" or "classical" or "jazz" centric amp which usually is a prime identifier of extreme coloring and so not close to natural.find the happy medium then close that sucker up tight !

 

[NeilR]

Quote:

Originally Posted by Voodoochile Do you have a decent resistor in that location?

I used the 1/4 watt VD resistors spec'd in the various BOMs. I didn't consider the issue when I ordered my parts. If I make another I will upgrade that part. Good point.

 

[McRat]

What an interesting read my little thread turned out to be.

But it does seem that there hasn't been too much conversation about the output resistor, so maybe it's about time.

I'm all for the great sound instead of great measurements and this little amp seems to underline that point.

 

[nikongod]

the output resistor is totally optional.

it is easy enough to jumper after you build if you should decide that you want to change.

i built it with a resistor, listened for a while, jumpered, and then UNJUMPERED!!! i just liked the sound better with the resistor inline.

 

[NeilR]

Am I the only one who has used the HA3-5002 buffers? Just curious since all the discussion has centered around BUF634 and the discrete buffer.

This is an interesting amp and a unique tweak opportunity since there are four different possible buffer options built into the board. (hint to the OP)

 

[McRat]

Quote:

Originally Posted by NeilR This is an interesting amp and a unique tweak opportunity since there are four different possible buffer options built into the board. (hint to the OP)

Thanks for the hint. I do know that the amp will accept those HA3-5002 buffers. I have also been reading about the diamond buffers, but at the time I only have the BUF634 and OPA551 at hand. I don't think I will be building the DB's but I haven't ruled anything out.

 

[NeilR]

If you do try the HA3-5002, be aware that the buffer will instantaneously self destruct if you short the bias test points. This is different than a headphone output short because it is behind the coupling capacitor and sitting at 12-15VDC.

If you do not have dedicated test point sockets, use one of the ground vias around the edge of the board for your black test probe. It is soooo easy to short the buffer out if you try to use both test points adjacent to each other.

 

[Voodoochile]

Quote:

Originally Posted by amb The diamond buffer uses two pairs of MJE243/253 or BD139/140 transistors which are rated high enough that the polyfuse should trip long before the transistors suffer any ill effect (not to mention that there are emitter resistors to help too).

How true! I really like the way Pete did that whole power entry arrangement. There was another thread where someone was concerned about excessive draw due to another potential malfunction, and I pointed out the same. It's a nice arrangement.

And as for hot-plugging headphones with either the HA3-5002 or diamond buffer module in general, if you're reasonably careful, then it should not matter anyway. I've never blown any buffers of either flavor with my PPAs. Just have to pay attention when plugging and unplugging.

As for the test points, I have some tip jacks I plan to install in the back panel if I ever case my millet. On the first one I built a few years ago, I added header pins, and then made a little adapter from a motherboard LED plug to fit it, with a pair of alligator clips on the other end. That worked well for an amp "cased" as Pete did in the original article- wide open.