Analog Output Signal to Amp Input Sensitivity Calculation

[tomb]

Yeah, 0.25 is what will allow maximum output, but maximum output is rarely what we're after, so it could be unimportant.
A possible issue with using a higher voltage to feed the amp is that you might reach high distortion earlier (at a lower listening level into a given load/headphone). I say this in this so very conditional way because it is.
1/ tube amp lean into distortion in a softer, more progressive way than SS one.
2/ when they do, for many tube amp users, that's when they finally get that tube sound they like
3/ the headphone is an important variable as it impacts both impedance as a load for the amp and the necessary voltage for your preferred listening level. As a rule of thumb, a lower impedance, lower sensi headphone would be more likely to cause distortion/clipping issues when used.
4/ the amp design also has a big say in this.

You certainly can try and check if you notice differences. For example, with a lower input you will apply more gain on the amp to get the same listening level, that could increase the background noise. Or the other way around, you might find that the distortion level (texture, warmth or however that manifests itself) is different. If you struggle to notice a difference, then most likely you're nowhere near the max of what your amp can do, and you're not clipping the signal, so whatever you use is likely still working in a pretty nominal way.
About life expectancy of the tubes or the amp itself and possible damage, I have no idea. Maybe someone like @tomb has some advice on this? Not sure how expert he is as he discusses stuff I usually know nothing about, but I know he's one willing to help a fellow Headfier in need anytime he sees one on the side of the audio road in the help section or DIY areas.

Wow! Thank you for those very kind remarks!

I can't speak to the details of this discussion because there's not really enough specific info available. I can state a few generalities, though:

• Equation above - That S, R, I equation up there fails dimensional analysis, because the ohms have no cancelling term. It could be the poster simply left that off of one of the terms, though.
• Gain vs Attenuation - Technically speaking, "gain" is set at a constant by the amplifier circuit design. The ratio of that final voltage or current level to the input level is the "gain." The volume control has no effect on this - it only attenuates the entering signal. The amp will still amplify it by the same "gain." Of course, this is just semantics in the context of this conversation. We all know the volume control will reduce the overall output of the amplifier.
• Signal Strength vs Amplifier Optimization - The question of optimizing amplifier performance relative to signal input is a good one. It's answered in almost every case by the industry standard of 2V RMS output for signal sources, along with a volume control. What is in play here is that if the signal is too small, the amplifier amplifies some of its own noise floor along with the signal. Conversely, If the signal is too strong, the amplifier can clip. Think about phono preamps. There the signal is so small a pre-amp is absolutely required. Why? Because an amp designed for a 2V RMS input will not only not be loud enough, it will be amplifying mostly noise.
• Where is the Volume Control? - For a signal too strong, you might say, "But the volume control will prevent clipping by attenuating the input signal." Well, yes and no. It depends on where in the circuit you place the volume control. NWAVGUY notably placed the volume control of his O2 amplifier between the signal stage and the amplifier stage. This meant that if a source produced a signal too strong, the O2 would clip - and it didn't matter where you adjusted the volume knob! Thankfully, that was very unique in the case of the O2 and you probably won't find it anywhere else. That said, there are audio companies who like to "play" with the signal output, letting it go a bit higher than 2V RMS. This is an old trick and used to be done back in the day when auditioning speakers: what sounds loudest at first hearing, sounds the best.
• Tubes and Distortion - Tubes are OK driven to distortion ... for a shorter life. It's how guitar amplifiers work. You drive the tubes to distortion to get all those great sound effects. They certainly don't fail right away, but you'd have to ask a guitar player how long they last. Generally speaking, a stronger input signal is not going to damage a tube, even if it clips. You'd have to run the tube that way for a long time and assuming it got hotter than it should be doing that, it might have a shorter life. But then, you'd be adjusting the volume knob so it wouldn't do that, right?
• Guitar Amps are Different - Guitar amps drive a tube to distortion by using a much higher voltage on the plates. This is much more severe than the distortion by running a tube in clipping mode. If anything, most tube headphone amplifiers run with lower voltages on the plates, making them safer to operate and assemble. Famous DIY designs by Pete Millett and Alex Cavalli were designed this way: 100, 150, 200V tubes running on 24 or 48V. The tubes will become less linear in frequency response, but traditional distortion is still kept low - and they are safer to build.

Hope some of that helps ...

 

[WoodyLuvr]

I sincerely appreciate everyone's input and help on this. I found everyone's input quite beneficial in clearing the cobwebs. I am on the right track now thank you. Cheers and respects. Back to lurking and learning. Cheers,

 

[lantian]

My suggestion, would be to match as closely as possible the dac output, to be exactly what the amp want s for max output, any more is just useless bordering on damage/clipping territory. So if it says full output power at 250mv which is standard, than that's what you should try and givi it. By giving it the input it is rated for you will be able to use the whole pot and use it in its sweet spot, 11 till 13 o clock.
If your amp gets plenty loud while 9 o clock or god forbid sooner you are giving it way to much voltage and the gain is way to much for your volume control device whatever it might be.
Myself built resistor voltage divider for my luxman input that way always giving it what it wants and having sweet spot on pot at around 11 o clock.

 

[WoodyLuvr]

Thank you all, especially @tomb @lantian @71 dB @castleofargh I appreciate the in-depth troubleshooting you all provided me.

No, I never have heard any clipping or gross "unwanted" distortion occurring even with the maximum output (3.3V) from my DAC using the unbalanced RCA outputs.

I was mainly concerned about both the noise floor (though I could not detect any unwanted noise) and most especially running my amp (and tubes) with too much voltage (risking damage or in the very least speeding up the lifespan of my gear). My custom headphone tube amp's volume was always well under the 9 o'clock position even in "Low Gain" mode so thus my concern that my source signal might have been too hot.

As everyone suggested I experimented around and have now permanently switched my iFi Audio ZEN DAC Signature V2 (w/ current firmware v7.4b) from "Fixed" to "Variable" output mode with the analogue volume control knob on DAC now set at around the 12 o'clock position which I believe is outputting a signal of ~1.65V approximately 50% of 3.3V as this analogue volume control pot (on the DAC) starts at 7 o'clock and ends at the 5 o'clock position. EDIT: I am being silly; please see posts below.

My custom headphone tube amp's volume is now sitting at a much more logical position, sitting around the 11 to 12 o'clock position depending on the track I am listening too.

An interesting note: again, when using the unbalanced RCA outputs with my iFi Audio ZEN DAC Signature V2 (w/ current firmware v7.4b) set on "Variable" output mode and with the analogue volume control knob set to maximum (@ 5 o'clock position), which should be an output of 3.3V, the output is considerably softer sounding than when in the "Fixed" 2.1V setting which is so much louder that I must reach quickly for the volume knob on my amp to turn it down. Even now when I switch over from "Variable" at a 50% setting (~1.65V) to the "Fixed" output mode (2.1V) it is still noticeably much louder sounding. I can not fathom why this would be. This behavior was also confirmed by another owner of the same DAC model over on the friends audio forum. Maybe the folks over at iFi Audio can offer a logical explanation @iFi audio?

 

[castleofargh]

Not sure that half the course of a knob is 50% of max voltage? It doesn't feel very likely to me.
For a fast ballpark idea, you can take a cellphone with a free decibel meter app and check the values coming out of your headphone while you play some noise/test tone, and turn the volume knob (obviously caution is needed to avoid getting too loud for your poor headphone). Even with no proper calibration, those apps tend to be reliable when it comes to the variation between 2 readings.
And here you can convert db to voltage ratio, or you can enter the max voltage of your DAC, the voltage you want, and calculate the variation in dB. Then the decibel meter app on your phone can be used to target that voltage.
http://www.sengpielaudio.com/calculator-gainloss.htm

 

[gregorio]

So if it says full output power at 250mv which is standard, than that's what you should try and givi it.

This is singularly poor advice. Firstly, if you’re going to give the amp the same input power as its output power, why do you even need an amp in the first place? Secondly, this advice can, and sometimes does, lead to significant gain-staging problems. I’ve seen audiophile systems where the amp input (DAC output) is so low, because the amp output is too high for the transducers, that even the great SNR of digital audio becomes audible.

My custom headphone tube amp's volume was always well under the 9 o'clock position even in "Low Gain" mode so thus my concern that my source signal might have been too hot.

Your concern should have been that your amp output was too high for your transducers. IE. It was the wrong amp for the job, you needed a less powerful amp. 2.1V isn’t too hot, it’s about right.

As everyone suggested I experimented around and have now permanently switched my iFi Audio ZEN DAC Signature V2 (w/ current firmware v7.4b) from "Fixed" to "Variable" output mode with the analogue volume control knob on DAC now set at around the 12 o'clock position

It’s hard to know what’s going on here. Your DAC appears to be just a DAC (in “fixed” mode) or a DAC plus amp (in “variable” mode).

If we’re just talking about a DAC then the only output volume adjustment is digital level and this is where we could run into issues with the digital noise floor.

In “variable” mode you’ve now also got an analogue amp after the DAC and here is where it’s difficult to know exactly what’s going on. From what you describe, my guess is that the output is 2.1V which you’re attenuating with the volume control and then feeding it to your tube amp which is amplifying it again. This is inherently poor gain-staging, although in practice you might just be effectively applying two stages of attenuation. Whether or not this poor gain-staging has any consequences that are actually audible is hard to say. We don’t seem to be talking about particularly large amounts of attenuation/gain stages so I wouldn’t expect audible consequences but with audiophile gear (especially tube stuff) there’s no telling, there could be some JND.

G

 

[WoodyLuvr]

Not sure that half the course of a knob is 50% of max voltage? It doesn't feel very likely to me.
For a fast ballpark idea, you can take a cellphone with a free decibel meter app and check the values coming out of your headphone while you play some noise/test tone, and turn the volume knob (obviously caution is needed to avoid getting too loud for your poor headphone). Even with no proper calibration, those apps tend to be reliable when it comes to the variation between 2 readings.
And here you can convert db to voltage ratio, or you can enter the max voltage of your DAC, the voltage you want, and calculate the variation in dB. Then the decibel meter app on your phone can be used to target that voltage.
http://www.sengpielaudio.com/calculator-gainloss.htm

Concur, I was being a dummy... it is not linear. Although digital volume it still would be logarithmic.

As you suggested, I tested via my Decibel X meter app and an 1kHz -5dB Test Tone for Setting Amp Gains and used the linked calculator that you provided. I am getting a voltage ratio gain of ~1.9 V at 12 o'clock (~56dB) which does seem to be more correct. The voltage and gain calculator at the bottom also provides a -21.9dB required gain loss to reach 0.25 V (~56dB) from 3.3V (77.9dB) which falls in line with the former calculated values, as well as with the DAC volume knob falling in the exact same spot at 12 o'clock (or very nearly so).

So the 12 o'clock position on my DAC's analogue volume control (when set on "Variable" output mode) should be ~0.25V (or very close to it) and not the ~1.65 V I foolishly proposed before! LOL!

 

[WoodyLuvr]

Your concern should have been that your amp output was too high for your transducers. IE. It was the wrong amp for the job, you needed a less powerful amp. 2.1V isn’t too hot, it’s about right.

It’s hard to know what’s going on here. Your DAC appears to be just a DAC (in “fixed” mode) or a DAC plus amp (in “variable” mode).

If we’re just talking about a DAC then the only output volume adjustment is digital level and this is where we could run into issues with the digital noise floor.

In “variable” mode you’ve now also got an analogue amp after the DAC and here is where it’s difficult to know exactly what’s going on. From what you describe, my guess is that the output is 2.1V which you’re attenuating with the volume control and then feeding it to your tube amp which is amplifying it again. This is inherently poor gain-staging, although in practice you might just be effectively applying two stages of attenuation. Whether or not this poor gain-staging has any consequences that are actually audible is hard to say. We don’t seem to be talking about particularly large amounts of attenuation/gain stages so I wouldn’t expect audible consequences but with audiophile gear (especially tube stuff) there’s no telling, there could be some JND.

G

This was custom designed for higher impedance earphones and headphones 300-600Ω but it also can handle as low as 64Ω (even down to 16-32Ω with certain headphones/earphones). It has a switchable low and high impedance mode, as well as a switchable low and high gain mode. You can see more specs and pics of the amp if you click on "GEAR" down in my signature.

With this particular DAC the "Fixed" mode as per iFi Audio: "this will bypass the ZEN DAC Signature V2 as if there was a ‘straight wire’ from the digital source to the analogue amplifier" which is now making me think I should indeed use this mode instead. Where as, the "Variable" mode turns it into a DAC/pre-amplifier device, which can also regulate volume of any connected power amps as per iFi Audio: "variable setting applies volume control to the audio signal, enabling the ZEN DAC Signature to act as a preamp to connect directly to a power amp." Also, please note that this DAC uses a 16-core XMOS chip which does not currently accept Microsoft Windows' digital volume controls so the volume control slider in Windows does absolutely nothing.

 

[castleofargh]

Concur, I was being a dummy... it is not linear. Although digital volume it still would be logarithmic.

As you suggested, I tested via my Decibel X meter app and an 1kHz -5dB Test Tone for Setting Amp Gains and used the linked calculator that you provided. I am getting a voltage ratio gain of ~1.9 V at 12 o'clock (~56dB) which does seem to be more correct. The voltage and gain calculator at the bottom also provides a -21.9dB required gain loss to reach 0.25 V (~56dB) from 3.3V (77.9dB) which falls in line with the former calculated values, as well as with the DAC volume knob falling in the exact same spot at 12 o'clock (or very nearly so).

So the 12 o'clock position on my DAC's analogue volume control (when set on "Variable" output mode) should be ~0.25V (or very close to it) and not the ~1.65 V I foolishly proposed before! LOL!

Nothing foolish here, we can't think of everything all the time. You tried checking by yourself, got a more reliable set of data to make your decision. Which, as I often say, is all too rare in this hobby. IMO, anybody who works his way through a test method to get some answers should be proud of himself. I know I am when I do it^_^.

 

[gregorio]

With this particular DAC the "Fixed" mode as per iFi Audio: "this will bypass the ZEN DAC Signature V2 as if there was a ‘straight wire’ from the digital source to the analogue amplifier" which is now making me think I should indeed use this mode instead.

In theory you should. The DAC’s 2.1V output is slightly hot but not excessively so. In practice though, particularly with your tube amp which appears to want significantly lower than consumer line level input (which is a bit strange), it could be a balancing act. You stated the volume knob/pot was below nine o’clock even on low gain, which I’m presuming is pretty close to its minimum setting. Very low settings with some/many amps can causes issues, channel imbalances quite commonly for example. Therefore, it might be worth trading improper gain-staging, having your DAC in variable mode and lowering the output, in order to have a more optimal volume setting on your tube amp.

However this is something only you can figure out as it depends on if your tube amp actually does perform poorly with a very low volume setting or if there really are any audible consequences of the poor gain-staging and if “yes” to both of these, which one would you rather live with? My guess is there’s no audible consequences of the improper gain-staging and probably that’s the best solution in this case (except for getting a more appropriate amp), but this is just a guess.

G

 

[iFi audio]

With this particular DAC the "Fixed" mode as per iFi Audio: "this will bypass the ZEN DAC Signature V2 as if there was a ‘straight wire’ from the digital source to the analogue amplifier" which is now making me think I should indeed use this mode instead. Where as, the "Variable" mode turns it into a DAC/pre-amplifier device, which can also regulate volume of any connected power amps as per iFi Audio: "variable setting applies volume control to the audio signal, enabling the ZEN DAC Signature to act as a preamp to connect directly to a power amp." Also, please note that this DAC uses a 16-core XMOS chip which does not currently accept Microsoft Windows' digital volume controls so the volume control slider in Windows does absolutely nothing.

That's correct. The 'fixed' mode disengages ZEN DAC Signature V2's volume control circuit, so the product acts like a pure DAC that outputs line level signal. Thanks!