Puffin: dB Gain to Voltage Calculation?

[watchnerd]

I finally got around to hooking up the Parks Audio Puffin on kind loan from @BDWoody

While I'm still learning some of the features of the Puffin, gain staging of the cartridge is the question I have for now.

The cartridge in question is a new (maybe 7 hours of use so far) Audio Technica ART-7:

Output: 0.12mV (very low!)
DC Resistance: 12 ohm
Inductance: 8 µH (1 kHz)

The current configuration is:

TT -> Puffin -> Devialet line-in

Puffin Gain: 72 dB (max)
Devialet input voltage (configurable): 2V

[At some point I'll try other configurations, like using with my SUT or comparing the DSP tube simulation to real tubes, but we're keeping it simple for now].

Given that both the Puffin and Devialet have configurable gain, what are the rules of thumb for gain staging the phono stage vs the preamp input?

Would I be better off dropping the gain on the Puffin and increasing it on the Devialet?

I suppose the answer is hard to know without measuring noise. Also, FWIW, I believe both the Puffin and Devialet are conducting gain in the digital domain after ADC.

 

[watchnerd]

Okay, doing a little calculator math:

Input: .00012V
Gain: 72 dB
Output: 0.48V

At a minimum, it sounds like I might want to drop the Devialet input voltage down from 2V to something lower, like 1V?

 

[scott wurcer]

I
Given that both the Puffin and Devialet have configurable gain, what are the rules of thumb for gain staging the phono stage vs the preamp input?

Would I be better off dropping the gain on the Puffin and increasing it on the Devialet?

I suppose the answer is hard to know without measuring noise. Also, FWIW, I believe both the Puffin and Devialet are conducting gain in the digital domain after ADC.

Big difference between standard RIAA and DSP based. The NJM op-amp they use is not going to give you the lowest possible noise with that cart. IME you are close at the max gain, maybe a few dB less will be optimal. ChannelD has an AES paper on headroom with flat pre-amps and DSP RIAA, they tested a lot of LP's.

 

[watchnerd]

Big difference between standard RIAA and DSP based. The NJM op-amp they use is not going to give you the lowest possible noise with that cart. IME you are close at the max gain, maybe a few dB less will be optimal. ChannelD has an AES paper on headroom with flat pre-amps and DSP RIAA, they tested a lot of LP's.

Who is the 'they' in this case?

Parks?

NVM, 'they' in this case is obviously referring to the Puffin.

 

[Sir Sanders Zingmore]

out of interest, how do you find the Puffin compared to the Devialet's own phono stage?

 

[watchnerd]

out of interest, how do you find the Puffin compared to the Devialet's own phono stage?

Sonically, too early to say, I'm still getting the settings dialed in.

Feature-wise:

--The Devialet has far more loading options for both MM and MC. Example: the Puffin only allows for 200 ohm and 47k ohm resistance loading

--The Puffin has a few more bells and whistles, in terms of tilt control and DSP transfer function simulations, like 2H (tube) and 3H (tape)

I played with the tube simulator. It was interesting, but I don't think it sounds much like my tube RIAA stage (Fi Yph).

 

[Soniclife]

Sonically, too early to say, I'm still getting the settings dialed in.

Feature-wise:

--The Devialet has far more loading options for both MM and MC. Example: the Puffin only allows for 200 ohm and 47k ohm resistance loading

--The Puffin has more bells and whistles, in terms of tilt control and DSP transfer function simulations, like 2H (tube) and 3H (tape)

I played with the tube simulator. It was interesting, but I don't think it sounds much like my tube RIAA stage (Fi Yph).

I'd really like to know how they compare for hiss when you get it configured well

 

[Soniclife]

I suppose the answer is hard to know without measuring noise. Also, FWIW, I believe both the Puffin and Devialet are conducting gain in the digital domain after ADC.

That sounds wrong to me, they would risk overload into the ADC line stage if true in the devialet.

 

[Soniclife]

The cartridge in question is a new (maybe 7 hours of use so far) Audio Technica ART-7:

Output: 0.12mV (very low!)
DC Resistance: 12 ohm
Inductance: 8 µH (1 kHz)

Interesting cart, suicidal output voltage. Good job you have a SUT.

 

[watchnerd]

Interesting cart, suicidal output voltage. Good job you have a SUT.

Yeah, although I'm not using the SUT with the Puffin.

 

[watchnerd]

That sounds wrong to me, they would risk overload into the ADC line stage if true in the devialet.

Yes, I believe I was incorrect.

 

[Blumlein 88]

What you are really asking is which combination gives you the best overall EIN for the setup. Usually mic pres and phono pres give better EIN with larger amounts of gain vs smaller. So I'd concur with what Scott Wurcer suggested.

If they are relying mostly on digital gain it won't matter either way. If there is some variable analog gain it would matter and a bit more is probably better on the phono.

 

[watchnerd]

What you are really asking is which combination gives you the best overall EIN for the setup. Usually mic pres and phono pres give better EIN with larger amounts of gain vs smaller. So I'd concur with what Scott Wurcer suggested.

If they are relying mostly on digital gain it won't matter either way. If there is some variable analog gain it would matter and a bit more is probably better on the phono.

Reading more about the Puffin, my remark about digital gain was probably wrong.

There seems to be an analog stage before the ADC.

 

[watchnerd]

Big difference between standard RIAA and DSP based. The NJM op-amp they use is not going to give you the lowest possible noise with that cart. IME you are close at the max gain, maybe a few dB less will be optimal. ChannelD has an AES paper on headroom with flat pre-amps and DSP RIAA, they tested a lot of LP's.

What you are really asking is which combination gives you the best overall EIN for the setup. Usually mic pres and phono pres give better EIN with larger amounts of gain vs smaller. So I'd concur with what Scott Wurcer suggested.

If they are relying mostly on digital gain it won't matter either way. If there is some variable analog gain it would matter and a bit more is probably better on the phono.

I dropped the gain on the Puffin to 68 dB (it only moves in 4 dB steps, at least at the high gain end of the range).

Doing basic calculations:

Input: .00012V
Gain: 68 dB
Output: 0.30V

Devialet input set to 1.0V

 

[watchnerd]

I'd really like to know how they compare for hiss when you get it configured well

Interesting cart, suicidal output voltage. Good job you have a SUT.

With the ART-7 (0.12mV) torture test:

I didn't do a back to back direct A/B comparison, and didn't have the SUT in the chain, but based upon memory, and with the volume cranked up to -0 dB:

--The Puffin definitely has hiss, but it might be a bit less than straight into the Devialet.

But...

--Hum. At first there was blatant hum, depending on orientation and if I was touching things. I added a 2nd ground wire, which helped.

--Line noise...or something? There is a motorboating noise at 0 dB gain that is barely audible from a distance, audible if I put my ears to the cone, and I could feel the cone vibrating even with no music. It felt like ~120 Hz-180 Hz or so, trying to count the beats I was feeling in my fingers.

I normally listen in the -20 dB to - 15 dB range, so these effects are not obvious at the level, and with music there is masking. But it might erode detail.

 

[watchnerd]

Update on going SUT-less:

Warning: I'm about to delve into subjective foo foo.

Noise aside, my subjective impression is that, without the SUT in the picture, the ART-7 straight into the Puffin lacks macro-dynamic capacity.

It's perplexing, as I can hear very soft sounds and detail is good.

But the loud passages don't seem as loud as they should be.

They don't sound like they're clipping, just a bit flat, as if double fortes are single fortes.

I can think of two semi-plausible explanations:

1. It's a subjective normalization problem -- because I can hear lower-level details, I'm able to listen at lower level volumes, and thus just not using as much master volume gain as I normally would.

2. 68 dB of gain on the Puffin is still too close to its max gain of 72 dB, and may be lacking headroom.

As an aside, the Puffin documentation refers to supporting MC carts in the 0.25mV range, and the ART-7 is half that....

Any thoughts from the academy?

 

[MRC01]

I used to have an Ortofon MC-30 Super II which was rated at 0.18 mV, only about 3.5 dB higher than your ART-7. At the time, trying a few different head amps, I found the DACT CT-100 to have the best cleanest sound especially at these high gain settings. It was dip-switch selectable for loading & gain from 40 to 80 dB. In my case, 68 dB of gain seemed to be the sweet spot for this cartridge. That's a voltage ratio of 2512:1, which takes 0.18 mV to 452 mV. This line level voltage is on the low side, but worked perfectly well even with just a passive attenuator between it and my power amp.

 

[watchnerd]

I used to have an Ortofon MC-30 Super II which was rated at 0.18 mV, only about 3.5 dB higher than your ART-7. At the time, trying a few different head amps, I found the DACT CT-100 to have the best cleanest sound especially at these high gain settings. It was dip-switch selectable for loading & gain from 40 to 80 dB. In my case, 68 dB of gain seemed to be the sweet spot for this cartridge. That's a voltage ratio of 2512:1, which takes 0.18 mV to 452 mV. This line level voltage is on the low side, but worked perfectly well even with just a passive attenuator between it and my power amp.

Thanks, although given the Puffin is a very different device, I'm not sure what to make of your findings vis-a-vis the Puffin.

It sounds like you were running it 12 dB below it's max gain...

Which I guess would argue for running the Puffin at 60 dB gain (72 - 12)?

 

[scott wurcer]

2. 68 dB of gain on the Puffin is still too close to its max gain of 72 dB, and may be lacking headroom.

As an aside, the Puffin documentation refers to supporting MC carts in the 0.25mV range, and the ART-7 is half that....

Any thoughts from the academy?

A true lack of headroom would be closer to clipping, you describe more something like compression. I suppose you don't have a way of looking at actual waveforms? There are some free sound card scopes.

 

[watchnerd]

A true lack of headroom would be closer to clipping, you describe more something like compression. I suppose you don't have a way of looking at actual waveforms? There are some free sound card scopes.

My scope died a few years ago and I haven't replaced it.

And, yes, it sounds more like compression than clipping -- perhaps the cart is too new and still stiff.

But I don't know if ASR believes in cartridge break-in.