How to attenuate source so amp has more volume play??

[n_maher]

Glad they worked out for you. I've been playing with numbers based on Justin's simple resistor-divider suggestion and wondering if something like it wouldn't be a great way to add "gain" adjustment to something like the beta22. It wouldn't be all that difficult to make it up on protoboard and wire it to a multi-position switch. It would make it a lot easier to build the beta so that it would be far more suitable for both speaker (high gain) and headphone (low gain) use.

[Beefy]

OK I just got these goldenjacks, -14 dB version, so far they seem to do the trick.

I'm glad my suggestion wasn't a dud

Just read this post now, WoW they do look good.

Could probably whip something up DIY but this looks a lot neater

Yeah, there is actually not much to them..... but getting the resistances right is the key. If you're really keen though, these plugs from DIY Fidelity don't look that different and you may be able to jam a couple of resistors in there - no promises though! http://www.diyfidelity.com.au/index.php?cPath=_34

Biggest/only problem with the Goldenjacks is their bulk. They add a LOT to the connection, and you don't want to go knocking them.

Glad they worked out for you. I've been playing with numbers based on Justin's simple resistor-divider suggestion and wondering if something like it wouldn't be a great way to add "gain" adjustment to something like the beta22. It wouldn't be all that difficult to make it up on protoboard and wire it to a multi-position switch. It would make it a lot easier to build the beta so that it would be far more suitable for both speaker (high gain) and headphone (low gain) use.

Are you thinking of this on the input or output?

[grawk]

Are you thinking of this on the input or output?

Pretty sure you'd want to attenuate the output, not the input. You want the amp to have as much to work with as possible, without overloading. Since overloading the B22 input is pretty unlikely, you'd want to do the attenuation later.

[JimP]

Beefy, thanks a lot for your recommendation!!!

So far, for headphones, these seem to serve its purpose (I cannot tell you how annoying being restricted to 7 - 8 o'clock is, in day-to-day use). Next test is full speakers - I have an Omega Super 5 + DeepHemp Cube Sub on order that I will also run off my Almarro. If the attenuation is too much, I can separately run an XLR to RCA (these are custom Premium 'Aniston' special cables) to the second set of inputs on the Almarro. That is, goldenjacks in the chain for my K1000/DT880-600, and straight up input for the Omegas. Will see how that works out.

[Beefy]

Pretty sure you'd want to attenuate the output, not the input. You want the amp to have as much to work with as possible, without overloading. Since overloading the B22 input is pretty unlikely, you'd want to do the attenuation later.

I would think so too, but some people seriously crap their pants at the thought of any sort of output attenuation.

[n_maher]

Are you thinking of this on the input or output?

I'd only do it on the input. From where I sit the output would be a very bad idea since you'd be destroying the output impedance of the amp.

I would think so too, but some people seriously crap their pants at the thought of any sort of output attenuation.

No pants crapping here, but it does sound like a shitty idea.

[grawk]

I'd only do it on the input. From where I sit the output would be a very bad idea since you'd be destroying the output impedance of the amp.

No pants crapping here, but it does sound like a shitty idea.

The volume knob is just a resistance attenuator, tho. What's the difference?

[Duggeh]

I've tried a few sets of in-line RCA attenuators, all of them destroyed the high end. Couldn't you just fit a bigger volume knob?

[n_maher]

The volume knob is just a resistance attenuator, tho. What's the difference?

No difference, but the pot is always on the input.

Rough wiring diagram:

source ---> attenuator ---> amp ---> headphones

[grawk]

Oh, ok, then I was mistaken. My suggestion is then just use a bigger ladder for your attenuator, or a better volume knob

[JimP]

all of them destroyed the high end

I'm using K1000 and DT880-600, both known for it's high end. I've gone back and forth with and without the goldenjacks. So far, I can't tell any difference. Again, this is with Almarro A205AmkII amp and 840C source. Maybe my ears suck (not unlikely given all the rock concerts I've attended to date; now I'm grown up -- I attend operas and classical concerts )

[Pars]

Since the attenuator (pot, stepped, whatever) is a voltage divider, it is the ratio of input to amp and amp to ground impedance that determines the volume. Therefore, it doesn't matter how big of pot you use in terms of volume; they are all the same. The different sizes will effect noise, etc. however. Tube stuff typically uses the larger values whereas SS stays at or below 50K (typ. 10K).

For the B22, can you not play with the feedback resistors (looks like R4 / R3) to control gain?

[Beefy]

I'd only do it on the input. From where I sit the output would be a very bad idea since you'd be destroying the output impedance of the amp.

No pants crapping here, but it does sound like a shitty idea.

I'm pretty sure you could find resistor pairs that perfectly match the impedance of a given set of phones, and still divide the driving voltage between the phones and paralleled resistor.

So far, I can't tell any difference.

There shouldn't be. Voltage dividers are good because they can be designed to attenuate without changing the impedance seen by the source device. Inline resistors suck because they do change the impedance, regardless, which can kill the dynamics.

I think

For the B22, can you not play with the feedback resistors (looks like R4 / R3) to control gain?

Yes, but there are also specific value compensation caps (C2-C5) that need to be used for a specific gain setting.

[n_maher]

For the B22, can you not play with the feedback resistors (looks like R4 / R3) to control gain?

The gain of the the amp is also tuned with compensation caps which should be changed for different gain configurations. So the switch required to change all of that for even two channels would be mind-numbingly complex and the wiring would be begging for noise.

And it would have to be a pretty crappy in-line attenuator to roll off any portion of the signal now that I look into it further. I'm sure there are crappy ones out there, but provided you choose wisely you shouldn't see any real signal degradation. If a source has north of a 2V output you can probably safely cut that in half (6db reduction) and still have more than enough signal for the amp to work with.

[Beefy]

I'd only do it on the input. From where I sit the output would be a very bad idea since you'd be destroying the output impedance of the amp.

No pants crapping here, but it does sound like a shitty idea.

I worked one out for 32ohm drivers as an example.

Sig --16R----------

| |

32R Driver

| |

Gnd ---------------

Amp still sees a total of 32ohm impedance, driver only sees half the voltage, and it is only a single resistor in the signal path. I'm not seeing why this is such a terrible idea

[n_maher]

I'm pretty sure you could find resistor pairs that perfectly match the impedance of a given set of phones, and still divide the driving voltage between the phones and paralleled resistor.

And by doing so complicate the heck out of the arrangement and also have to deal with a lot more power dissipation over the resistors.

[Beefy]

And by doing so complicate the heck out of the arrangement and also have to deal with a lot more power dissipation over the resistors.

I didn't say it would be practical. But that doesn't mean it is a 'shitty idea'.

[grawk]

I dunno, impractical = shitty to me

[justin]

I worked one out for 32ohm drivers as an example.

Sig --16R----------

| |

32R Driver

| |

Gnd ---------------

Amp still sees a total of 32ohm impedance, driver only sees half the voltage, and it is only a single resistor in the signal path. I'm not seeing why this is such a terrible idea

You'd have a much lower damping factor resulting in the amp having less control over the headphone, most noticeable in the bass. You may prefer this, as many OTL tube amps have practically no damping factor and the headphones are free to resonate.

[Beefy]

You'd have a much lower damping factor resulting in the amp having less control over the headphone, most noticeable in the bass. You may prefer this, as many OTL tube amps have practically no damping factor and the headphones are free to resonate.

Ah, see, now THAT might make it a shitty idea

Still, if your amp is just simply too loud with low impedance phones, I'd imagine that it would better than an in-line attenuator.......?

[Pars]

The gain of the the amp is also tuned with compensation caps which should be changed for different gain configurations. So the switch required to change all of that for even two channels would be mind-numbingly complex and the wiring would be begging for noise.

The compensation cap is in parallel with R4, so if you tuned gain by varying R4, you could simply have appropriate comp caps in parallel with those = mind-numbingly complex. Of course, you are correct regarding the wiring, etc., particularly if more than one choice is desired. If you did it with relays on a small PCB you could keep it quite close to the area on the B22 board.

Another possibility would be switchable series resistors on the input in front of the pot, with a short being one choice...

[n_maher]

I didn't say it would be practical. But that doesn't mean it is a 'shitty idea'.

I beg to differ and justin explained it better than I could have

Still, if your amp is just simply too loud with low impedance phones, I'd imagine that it would better than an in-line attenuator.......?

Why would you imagine that? You can select tightly matched, high-quality resistors that are impedance appropriate. In that scenario this first attenuation stage would probably be doing less harm to the signal than the variable attenuator that you're using as a volume pot.

[Beefy]

I beg to differ and justin explained it better than I could have

Don't give yourself too much credit; I don't think you had any idea why it was a shitty idea. At least I was still asking a question.

Why would you imagine that? You can select tightly matched, high-quality resistors that are impedance appropriate. In that scenario this first attenuation stage would probably be doing less harm to the signal than the variable attenuator that you're using as a volume pot.

I was still talking about the output side. If you had to attenuate at the output I'd imagine a voltage divider is still better than an inline attenuator, that are just series resistance - like you yourself said, screwing up the output impedance.

[grawk]

Dude, this is a great way to end up on everyone's asshat list.

Don't give yourself too much credit; I don't think you had any idea why it was a shitty idea. At least I was still asking a question.

I was still talking about the output side. If you had to attenuate at the output I'd imagine a voltage divider is still better than an inline attenuator, that are just series resistance - like you yourself said, screwing up the output impedance.

[Beefy]

Dude, this is a great way to end up on everyone's asshat list.

I understand and entirely appreciate n_maher's efforts to the community. But credit where credit is due - not when somebody is being elitist. I'm just trying to help, discuss and learn here, and getting slammed for no apparently good reason.

I really don't understand this place......