Matching devices - impedance etc

[spaace]

Hi,

I have a newbie doubt on how to go about matching devices.

I recently had my chinese DAC die on me due to what i think might be a power fluctuation and i went back to my old Schiit modi multibit, feeding it directly into my tired old receiver being used as the 2.1 amp.

Well, the setup with Schiit did sound Schiit. However i wasn't sure if it was due to the Schiit DA functionality being bad or if it was due to a mismatch between the amplifier and the DAC.

I have this doubt since -

1. . The chinese DAC was quite louder than Schiit to begin with. I have to turn up the reciever volume much higher to get the same audio levels when using Shiit and i wasnt sure if the extra amplification level was causing a distortion.
2. Level matched, a 20$ DAC and 500$ DAC arent supposed to be audibly different as per some folks. (But in this case me, my wife and kid can hear the difference, when the setup changes even without AB-ing them - mainly with the bass levels being quite low and overall music being grating)

So my doubt is this -

1. Can impedance mismatch cause audio degradation ?
2. What are the electronic compatibility factors to be looked for when connecting a dac to preamp or preamp to amplifier & why do they matter?
3. Do DAC's always produce the output with the same signal strength or is it determined by its input signal ? is there a limit here? What is the spec that allows me to find this in source documentation

Thanks
Arun


Modi 2 multibit

Output Impedance: 75 ohms
Maximum Output: 2V RMS


chinese dac - xiangsheng 05B

Output Level: 2V
Output Impedance: 600 ohms

Yamaha DSP A2070

Input impedance - 150mV


ps : To be fair the source changed too from a hifiberry dac digi+ with RCA cable to to an allo digione with its own supplied BNC cable - so iam not sure if thats a factor too - the allo output went kaput too in the same event - so need to test all of these but i still wasnt sure about the impedance factor& hence the question. .

 

[sergeauckland]

In general terms, as long as the load impedance is at least 10x the source impedance, there is no likelihood of distortion. There is the possibility that with an AC coupled output, the output capacitor is too small, so bass gets rolled off unless the load impedance is higher. This is rare, but did, for example affect the Quad FM3 tuner unless it fed at least a 50-100k input impedance.

It is also possible that with some valved (tubed) outputs, the output impedance is so high that not only does it need a very high load impedance, but also short low capacitance cables to avoid hf rolloff.

In your case, and with most modern SS equipment, impedance is a non-issue.


Edit:-
In answer to Question 3, a DAC's output is entirely dependent on the digital data it's fed with, not the amplitude of that data. Whatever the input level to the DAC in volts, provided it's enough for the DAC to lock to, it will produce the same output voltage as that depends on the digital data, not voltage.

Most DACs will lock perfectly well on any sensible digital input level. The S-PDIF spec is quite wide, 0.5-0.6V as the nominal level, but should work correctly down to 0.2 v peak to peak. There's no upper voltage level specified, but AES balanced can go up to 7v PP and all S/PDIF inputs I've tried will cope with an AES-EBU balanced input if just one leg used, so up to 3.5v.

S

 

[Tom C]

We need a true expert to jump in here, but my experience is that there are two first-line considerations when feeding an active component.

One is matching the impedance of a component’s output to the input of the next component. A minimum of ten to one is often encountered as a manufacturer recommendation. So if your Modi has an output impedance of 75 ohm, you would want the input impedance of the Yamaha DSP to be at least 750 ohms. If it were less, too much of the output of the Modi would be dropped across the Modi’s own internal output stage and the Yamaha input would be poorly driven. If the input impedance of the Yamaha is over 750 ohms, even way over, it shouldn’t be a problem.

You cite the input impedance of the Yamaha as 150 mV, which I’m guessing is actually a rating of the input sensitivity (input and output impedance are expressed in ohms). Which brings up another consideration when matching components. If you overdrive the input, distortion can result. I would expect that if the Modi put out, say, 1000 mV, that would overdrive the Yamaha input. 15 mV out of the Modi would result in loudness being too low.

Personally, I think the impedance matching needs to be respected (especially when running near the cut off of acceptable parameters, such as when matching speakers to tube amps), but matching output levels to input sensitivity is more important when you’re trying to match two active components, and you have to choose one over the other.

 

[spaace]

We need a true expert to jump in here, but my experience is that there are two first-line considerations when feeding an active component.

You cite the input impedance of the Yamaha as 150 mV, which I’m guessing is actually a rating of the input sensitivity (input and output impedance are expressed in ohms). .

The documentation listed this data as "Input Sensitivity/Impedance" - And the number is listed as 150mv/47K ohm. I misread the title to mean that the sensitivity is the same as impedance. (else i would have asked about sensitivity too - Guess i have lots of googling to do.

So in this case the impedance is 47K ohm for the Yamaha. Thanks for catching that.

 

[spaace]

In your case, and with most modern SS equipment, impedance is a non-issue.

Edit:-
In answer to Question 3, a DAC's output is entirely dependent on the digital data it's fed with, not the amplitude of that data. Whatever the input level to the DAC in volts, provided it's enough for the DAC to lock to, it will produce the same output voltage as that depends on the digital data, not voltage.

S

Thanks for replying Serge. If the input signal to the DAC does not matter & neither does the impedance, then i guess this simply has to do with the inherent "output level" of the DAC ? (whats the term here? I meant the strength of the signal produced by the DAC)

I found it quite advantageous if the DAC outputs a stronger signal to begin with, since that seems to have a multiplying effect on the amplification. (My amp volume knob is it lower levels) I would assume people would love DACs that present a stronger signal for amplification to the amp. Is this a valid assumption? Which parameter in the specs would describe the output signal strength ?

 

[RayDunzl]

Which parameter in the specs would describe the output signal strength ?

Dacs that plug into a wall outlet generally aim for a 2Vrms maximum output on an unbalanced (RCA) connector.
2Vrms = 2.828Vpeak = 5.656Vpeak-peak

The value may vary.

Example, where a comprehensive description is supplied:

Dacs that run off USB may provide less. USB provides a 5Vdc source voltage. The DAC has to come up with both positive and negative voltages, so it depends on what it (its power circuitry) does with the 5Vdc it has available to work with.

 

[RayDunzl]

I would assume people would love DACs that present a stronger signal for amplification to the amp. Is this a valid assumption?

Only up to the point that the output voltage of the DAC exceeds the sensitivity of the input to the amp/preamp.

The Volume knob on an analog amplifier attenuates the signal from the DAC.

If the amplifier has 1V sensitivity, and the DAC has a 2V output, the DAC can overload (clip or otherwise distort) the input/output of the amp.

The volume knob attenuates the input voltage to a comfortable level, 1V (full output) or, usually less, as full output is usually "too loud".

 

[sergeauckland]

Thanks for replying Serge. If the input signal to the DAC does not matter & neither does the impedance, then i guess this simply has to do with the inherent "output level" of the DAC ? (whats the term here? I meant the strength of the signal produced by the DAC)

I found it quite advantageous if the DAC outputs a stronger signal to begin with, since that seems to have a multiplying effect on the amplification. (My amp volume knob is it lower levels) I would assume people would love DACs that present a stronger signal for amplification to the amp. Is this a valid assumption? Which parameter in the specs would describe the output signal strength ?

The DAC's 'inherent' output voltage means the voltage when given a 0dBFS signal, and that's the most the DAC will output (It is possible for a DAC to output more than the 'inherent' output voltage, due to intersample overs, but those are rare in real music recordings, and anyway, a good DAC shouldn't just clip at 0dBFS, it should have some headroom).

Generally, yes, if a DAC outputs a stronger signal, it allows the volume control to be at a lower level, and that reduces noise, BUT!! Firstly, volume controls tend to lose balance as they get turned down, so the stereo image might not stay centred. Secondly, volume controls get coarse twards the bottom, where a small rotation makes a big difference to volume, so operating the volume control well down is clumsy. Thirdly, if the volume control comes after an active buffer stage, then it's possible for the strong DAC signal to overload the input buffer causing distortion. Fourthly, the benefit of noise reduction with a strong signal is pretty minimal as any half-decent DAC will have sufficiently low noise. Ideally, you would operate the volume control such that it's around 80% of full rotation at your loudest listening level. This will give you a bit of headroom for quiet recordings, yet not risk any overload, and keep a decent balance between the channels. Unfortunately, most amplifiers have excessive sensitivity, as people tend to judge how powerful the amplifier is by how far they rotate the volume control.....this amp's so powerful it blows the windows out at only quarter volume....

If your DAC has its own volume control, then you can set the amp's volume to around 80% and adjust the volume on the DAC so it's just too loud on most recordings. That'll give you be best balance between sensitivity, balance, avoiding overload and noise.

S.

 

[andreasmaaan]

1. Level matched, a 20$ DAC and 500$ DAC arent supposed to be audibly different as per some folks. (But in this case me, my wife and kid can hear the difference, when the setup changes even without AB-ing them - mainly with the bass levels being quite low and overall music being grating)

The Schiit Modi Multibit is one of those DACs that performs poorly enough that it's likely to have its own sound, because it produces a lot of higher-order harmonics, which tend not to be masked by the auditory system. It's unlikely the bass output of the Schiit actually is lower, because measurements of it I've seen show a flat frequency response, but the presence of high-ish levels of intermodulation distortion and higher-order harmonics, which are likely to be perceived as adding harshness or shrillness to the sound, are a possible explanation for the sense of less bass.

 

[spaace]

The DAC's 'inherent' output voltage means the voltage when given a 0dBFS signal, and that's the most the DAC will output (It is possible for a DAC to output more than the 'inherent' output voltage, due to intersample overs, but those are rare in real music recordings, and anyway, a good DAC shouldn't just clip at 0dBFS, it should have some headroom).

Generally, yes, if a DAC outputs a stronger signal, it allows the volume control to be at a lower level, and that reduces noise, BUT!! Firstly, volume controls tend to lose balance as they get turned down, so the stereo image might not stay centred. Secondly, volume controls get coarse twards the bottom, where a small rotation makes a big difference to volume, so operating the volume control well down is clumsy. Thirdly, if the volume control comes after an active buffer stage, then it's possible for the strong DAC signal to overload the input buffer causing distortion. Fourthly, the benefit of noise reduction with a strong signal is pretty minimal as any half-decent DAC will have sufficiently low noise. Ideally, you would operate the volume control such that it's around 80% of full rotation at your loudest listening level. This will give you a bit of headroom for quiet recordings, yet not risk any overload, and keep a decent balance between the channels. Unfortunately, most amplifiers have excessive sensitivity, as people tend to judge how powerful the amplifier is by how far they rotate the volume control.....this amp's so powerful it blows the windows out at only quarter volume....

If your DAC has its own volume control, then you can set the amp's volume to around 80% and adjust the volume on the DAC so it's just too loud on most recordings. That'll give you be best balance between sensitivity, balance, avoiding overload and noise.

S.

Awesome. Thank you.

 

[spaace]

The Schiit Modi Multibit is one of those DACs that performs poorly enough that it's likely to have its own sound, because it produces a lot of higher-order harmonics, which tend not to be masked by the auditory system. It's unlikely the bass output of the Schiit actually is lower, because measurements of it I've seen show a flat frequency response, but the presence of high-ish levels of intermodulation distortion and higher-order harmonics, which are likely to be perceived as adding harshness or shrillness to the sound, are a possible explanation for the sense of less bass.

oh !! Thats good to know. Upto what levels are these audible? Amongst all the bad measurements a DAC can have (jitter/distortion/low SNR etc) which is absolutely the worst? Is there a checklist somewhere that can be used to check off a new DAC that says atleast x amount of SNR, jitter to be lower than 120 db etc?

 

[solderdude]

https://www.audiosciencereview.com/...ad-distortion-comparison-graph-for-dacs.4814/

IMHO anything with a SINAD > 80 is probably low enough in distortion for most folks. This single number does not tell it all... it's just a number.
Above SINAD 100 you are safe when it comes to ... well distortion at 1kHz and and noise/hum levels.
You also need to take the FR response, multitone response, jitter, linearity and other measurable quantities into account.

Impedance matching is ONLY needed for transmission of digital signals (SPDIF via Coax, I2S, AES/EBU) and video signals.
The in and output impedances usually are between 50 and 75 Ohm but for USB for instance it is around 110 Ohm (balanced connection).
Normally this is already taken care of IN the used equipment and you don't need to worry about this. Just select the correct cables for the job.

For analog audio like vinyl (MC and MM) you need to apply the correct LOAD, this is NOT the same as impedance matching.
Input resistance and input capacitance are important and should comply to the used cart+wiring.

For speakers and headphones there is also no need for impedance matching (there is nothing to match) but for these devices the amplifiers must be able to 'properly drive' the load impedance and be able to supply enough voltage and current.

For analog inputs you can use the rule: Input resistance should be at least 10x higher than the output R. So 100x is even better.
And it is no law just a guideline.
In some cases the input resistance can even be smaller than a factor 10 without degradation of sound.
No need to use 'special' cables here.

There are some folks that claim speaker cables need to be '8 Ohm nominal' as well and use high capacitance and 'filtered' cables or use conductive isolators. Also there are folks that claim/want interlinks to be impedance matched (so input, output and cable impedance) to be the same (around 150 Ohm or so) but most equipment won't like this.

The biggest issue one comes across is voltage level differences. Whether or not this could be problematic (clipping/distortion) or not totally depends on how the incoming signals are handled. Some 150mV inputs clip when 1V is applied, some even lower but many have no problems handling 10V. Depends on the used circuits. The volpot position may become an issue (have to turn it up or down too much).
When it does not distort/sound harsh .... you're fine.

 

[Tom C]

Hi,

I have a newbie doubt on how to go about matching devices.

I recently had my chinese DAC die on me due to what i think might be a power fluctuation and i went back to my old Schiit modi multibit, feeding it directly into my tired old receiver being used as the 2.1 amp.

Well, the setup with Schiit did sound Schiit. However i wasn't sure if it was due to the Schiit DA functionality being bad or if it was due to a mismatch between the amplifier and the DAC.

I have this doubt since -

1. . The chinese DAC was quite louder than Schiit to begin with. I have to turn up the reciever volume much higher to get the same audio levels when using Shiit and i wasnt sure if the extra amplification level was causing a distortion.
2. Level matched, a 20$ DAC and 500$ DAC arent supposed to be audibly different as per some folks. (But in this case me, my wife and kid can hear the difference, when the setup changes even without AB-ing them - mainly with the bass levels being quite low and overall music being grating)

So my doubt is this -

1. Can impedance mismatch cause audio degradation ?
2. What are the electronic compatibility factors to be looked for when connecting a dac to preamp or preamp to amplifier & why do they matter?
3. Do DAC's always produce the output with the same signal strength or is it determined by its input signal ? is there a limit here? What is the spec that allows me to find this in source documentation

Thanks
Arun


Modi 2 multibit

Output Impedance: 75 ohms
Maximum Output: 2V RMS


chinese dac - xiangsheng 05B

Output Level: 2V
Output Impedance: 600 ohms

Yamaha DSP A2070

Input impedance - 150mV


ps : To be fair the source changed too from a hifiberry dac digi+ with RCA cable to to an allo digione with its own supplied BNC cable - so iam not sure if thats a factor too - the allo output went kaput too in the same event - so need to test all of these but i still wasnt sure about the impedance factor& hence the question. .

Were you able to sort out what the trouble was?

 

[spaace]

Not yet - I need to try and bring up my old source which also died in the mishap to try and see if thats what is causing it ....the cables are also different ....bnc vs coaxial ..iam beginning to suspect the source could be it since music played through my alternate source ( a google audio chromecast) through its toslink output is not too bad though the volume levels are indeed lower compared to before

But the problem is that unless i can bring up old DAC soon, i feel I will get adjusted to the new norm and forget what the old sound was like as iam going by the ear and not measurements ...

Will post back here once i have it sorted out. Thanks for following up.

 

[spaace]

Happy to report my original dac is back ...and i played it with the same source cable etc and it sounds as gorgeous as i remember it.

If my cheapo chinese dac can make so much of a difference i can't imagine how it would sound with a better amp and premp and dac. Wow ....

Thanks for all your advice folks.