Phono preamp headroom - why?

[JP]

Not for me. Until recently, the pre I was using was the cheapest I'd been able to get at a local store - Project Phono Box.

In July I got the waxwing - and even with the "Magic" feature turned off noise (particularly pop/crackle) is (seems) much lower - or at least much less intrusive. To the point after 6 sides of vinyl last night, I hadn't noticed any outside of the lead out.

Now this is perceptive and sighted listening - Id have to try to measure it to be totally certain there is a real difference.

That's not exactly what I meant. I don't think anyone is contesting that a phono stage that behaves well with high-amplitude input (ticks/pops) will make them sound less intrusive than ones that don't. Generally speaking the latter can take many cycles to recover which makes the event stand out. I have contested that overhead is the only solve.

I was speaking directly about the notion that most ticks, etc. are a product of the phono stage and not the vinyl. In my experience I've not seen a phono stage that does this - nearly every artifact I've seen is in the vinyl and shows up on every playback with any phono stage. That's not to say that every phone stage does an equal job of avoiding clipping, or recovery, or how the event may sound. Just that it's there if I look for it.

IIRC, the disclosure of a setup known to exhibit the fabrication of ticks has been requested, but none given.

 

[levimax]

That's not exactly what I meant. I don't think anyone is contesting that a phono stage that behaves well with high-amplitude input (ticks/pops) will make them sound less intrusive than ones that don't. Generally speaking the latter can take many cycles to recover which makes the event stand out. I have contested that overhead is the only solve.

I was speaking directly about the notion that most ticks, etc. are a product of the phono stage and not the vinyl. In my experience I've not seen a phono stage that does this - nearly every artifact I've seen is in the vinyl and shows up on every playback with any phono stage. That's not to say that every phone stage does an equal job of avoiding clipping, or recovery, or how the event may sound. Just that it's there if I look for it.

IIRC, the disclosure of a setup known to exhibit the fabrication of ticks has been requested, but none given.

I am confident that a well behaved phono stage can help minimize the intrusiveness of pops and clicks. I also have never experienced pop and click "fabrication" but it is an interesting theory. What I can't wrap my head around is "20 dB of headroom". First off I have never seen a record damaged so bad that it can create a signal more than about 4 dB above the peak signal. Second how is it helpful to send huge signals to the following gain stages if the signal will clip that stage? I must be missing something as some smart engineers have designed phono stages with huge amounts of headroom. The only thing I can think of is that the "headroom" is measured at 1 Khz and that the "similar to square wave" nature of pops and clicks creates relatively large amounts of HF energy which is where the overload / bad behavior occurs. To get 5 dB of headroom @ 20 Khz maybe you need 20 dB at 1 Khz? A HP filter has been mentioned as being helpful, maybe for this reason? What about using a 44.1 Khz sample rate on an ADC to capture the cart signal and then do RIAA digitally?

 

[restorer-john]

I also have never experienced pop and click "fabrication" but it is an interesting theory.

JP invented the click fabrication theory. I see nobody else in this thread mentioning clicks materializing from the phono stages themselves.

That said, I have had some faulty phono stages to repair with noisy transistors that made their own snap, crackles and pops.

What I can't wrap my head around is "20 dB of headroom". First off I have never seen a record damaged so bad that it can create a signal more than about 4 dB above the peak signal. Second how is it helpful to send huge signals to the following gain stages if the signal will clip that stage? I must be missing something as some smart engineers have designed phono stages with huge amounts of headroom. The only thing I can think of is that the "headroom" is measured at 1 Khz and that the "similar to square wave" nature of pops and clicks creates relatively large amounts of HF energy which is where the overload / bad behavior occurs. To get 5 dB of headroom @ 20 Khz maybe you need 20 dB at 1 Khz?

There's no cut or boost at 1kHz, whereas there is 20dB of cut at 20kHz in the RIAA EQ. And it doesn't stop there. The curve continues after 20kHz. Many stages spec RIAA deviation out to 100kHz.

The stage also needs approximately 35-40dB of voltage gain (MM) at 1kHz to bring a 2.5-5mV nominal output cartridge up to line (ish) level. Trackable content can be >16dB above nominal ([email protected]/3.9cm/S).

The RIAA can be done in either the NFB loop (most common) or passive in between stages or even in two (or more) stages. So, you are getting rid of 20dB at one end of the spectrum and boosting by 20dB at the other.

You need a lot of available voltage in your PSU rails for the phono stage to cater for all that. Then along comes an almighty crack or pop. So, you create a stage that really can't overload or distort and then attenuate that signal to the level you want at the next stage. A bit like so-called 32bit floating point recording.

 

[JP]

JP invented the click fabrication theory. I see nobody else in this thread mentioning clicks materializing from the phono stages themselves.

I invented something I’m arguing against? Very clearly in this same thread:

Phono preamp headroom - why?

Amir often gives extra points to phono preamps that can cleanly and linearly pass big input spikes from pops on the disk as big output spikes. It's been a while but I recall one of them could output 20 V spikes. And I thought: Why is clipping in a subsequent stage in the system better then...

www.audiosciencereview.com

 

[restorer-john]

I invented something I’m arguing against? Very clearly in this same thread:

Phono preamp headroom - why?

Amir often gives extra points to phono preamps that can cleanly and linearly pass big input spikes from pops on the disk as big output spikes. It's been a while but I recall one of them could output 20 V spikes. And I thought: Why is clipping in a subsequent stage in the system better then...

www.audiosciencereview.com

You took an overload situation supposedly caused by excitation/stimulation of a system and rebranded it as "fabrication".

A very long bow to draw, IMO. But it could be your and my definition of fabrication in the context used, differs somewhat. All good as they say.

 

[Brian Hall]

Don't think so far as there's no 'dust-adherence-adding-component‘ involved.

It is called static electricity.

 

[JP]

You took an overload situation supposedly caused by excitation/stimulation of a system and rebranded it as "fabrication".

A very long bow to draw, IMO. But it could be your and my definition of fabrication in the context used, differs somewhat. All good as they say.

If the click is from this theorized overload event caused by an electrical oscillation and not from what the stylus directly scanned, then the phono stage did materialize it.

 

[antcollinet]

IIRC, the disclosure of a setup known to exhibit the fabrication of ticks has been requested, but none given.

Sorry - I haven't seen (or just missed) the claim that a phone stage created clicks.

 

[markstein]

I am starting to think the 20 dB headroom requirement is a myth. Looking at the worst of the worst of my second hand records it is rare to see a click or pop louder than the music peaks and in a few instances I have seen it is more like 2 to 3 dB above peak music and these are extremly damaged records. I think the issue is more complicated, see a plausable explanation from a member https://www.audiosciencereview.com/...lain-the-vinyl-renaissance.32420/post-1332275

Edit
I come to the conclusion - based on measurements with some of my records - that my preamp should support velocities of up to 73 cm/s to avoid clipping.

Parks Audio Waxwing measurements

Here is an additional measurement of mm mode THD with 23dB of attenuation added to the RME output (with it's level adjusted up by 23dB), to minimise the RME noise contribution. Measurement range a bit different from the original (20-20k vs 10-22.4k). Noise improvement really just due to the...

www.audiosciencereview.com

 

[atmasphere]

Ok I'm fine with that.

Clean is really good. I have tried the manual method according Neil Antin just to get the acidic step, but the lack of Alconox and Citranox made me try the method of "liquid wash power sensitive" + "distilled white vinegar". I just wanted to try the acidic step to remove any potential salts embedded in the groove. However, I actually found that my method with multiple rounds of Humminguru ultrasonics was better for noisy records; it removes/reduce clicks and pops for every round. So my suspicions is that there are things really hard stuck dirt in the grooves in some records that is not easily removed neither manually or by machines.

As mentioned, if you are using rumble filter and HP filter, you will probably reduce level of pops and clicks since they are broadband and may contain out-of-phase components. It may have a bigger impact of the clicks and pops than "headroom".

I should point out that with a MM cartridge the peak is often just above the audio band and it can be up to 20dB (caused by the inductance of the cartridge in parallel with the capacitance of the tonearm cable, plus the input capacitance of the phono section). There's plenty of tracking noise and harmonics just above 20KHz so there's plenty up there to mess with a phono section, and not really all that broadband either.

Sorry - I haven't seen (or just missed) the claim that a phone stage created clicks.

I've stated this many times so you can blame it on me. I first discovered this was a phenomena about 33 years ago when an employee complained of ticks and pops in his system that I wasn't getting. We narrowed down the possible causes and eventually he brought in his (Toshiba) preamp and there were the ticks and pops. His preamp was opamp based and it really wasn't that hard to overload it. Apparently the designer didn't take into account the significance of what happens when an inductor is in parallel with a capacitance (Electronics 101, first week stuff). So all that ultrasonic tracking noise and harmonics was taking the phono section to the cleaners.

With LOMC cartridges the resonance is much higher frequency and the peak is higher and tighter owing to the much higher Q of the coil in the cartridge. In this case you have a different problem where the resonance, when in excitation, acts as RFI injected directly into the input of the phono section. It can be up to 30dB higher level than the signal. If you place a 'cartridge loading' resistor across the input of the phono stage, you can detune the resonance, which has led to the mistaken idea that LOMC cartridges need to be 'loaded' to sound right. But if you measure one on the bench you'll find their inductance is so low that loading does nothing to them at audio frequencies unless the load is such a low value that the output of the cartridge actually goes down. But no-one loads their cartridge that low. At any rate if the phono section is RFI immune you won't need the loading resistor. Of course most people don't take this into account but if you place a lower resistance across the output of the cartridge you will cause it to do more work- so the cantilever will become stiffer as a result. This can mess with the mechanical resonance and the ability of the cartridge to track high frequencies.

 

[Thomas_A]

Below is what I get with test signal (=5 cm/s peak mono, or 3.54 cm/s peak L or R) vs. music cut.

Reference signal at -22-23 dB RMS and -19 peak

Music example: About -22 to -23 dB average RMS power, -12 to -13 dB max RMS within 10 ms window, and about -3 dB peak.

So it might depend on what reference is used (which vary with standard) but 20 dB headroom is not enormous.

 

[restorer-john]

So it might depend on what reference is used (which vary with standard) but 20 dB headroom is not enormous.

Play a hot 12" 45RPM single and it'll be several dB higher again.

 

[Thomas_A]

Play a hot 12" 45RPM single and it'll be several dB higher again.

Sure, agreed. I was just giving an example of a normally cut 33 RPM. Headroom is needed for the music content.

 

[markstein]

Sure, agreed. I was just giving an example of a normally cut 33 RPM. Headroom is needed for the music content.

As stated some posts above, I come to the conclusion - playing some of my 33 and 45 rpm records on my system - that peak velocities up to 73cm/s occur.

Taking 5cm/s as reference point, this would translate into >23dB of "headroom" required.

I think we should focus on peak values, not RMS, as the intention of sufficient headroom should be to avoid clipping of pops and ticks.

 

[Thomas_A]

So handling 100 mV at input is not unreasonable? It would not apply to many preamps reviewed here.

 

[sergeauckland]

So handling 100 mV at input is not unreasonable? It would not apply to many preamps reviewed here.

And that's much of the problem with many of today's phono preamps. Noise levels are much lower than in the '70s, but at the expense of headroom.
20dB (i.e. 50mV on a nominal 5mV input) is about the minimum for a decent phono preamp. 23dB, i.e. 100mV is that bit better, but bear in mind that many of today's cartridges output more like 3mV, (especially 300uV MC cartridges through a 10:1 SUT) that one gets more than 20dB overload in actual use.

Noise levels are important, but considering what is the background noise inherent with vinyl, chasing -75dB or more noise levels is unnecessary, but makes for a nice easy-to-understand specification, whereas headroom is more complicated to understand.

S.

 

[markstein]

I think there are two options for a "good" phono preamp with regard to headroom:

Either provide an extraordinarily huge headroom to cover all kind of cartridges with varying sensitivity,

Or - even better - to provide sufficiently adjustable input gain not to "waste" S/N ratio for cartridges with less sensitivity.

I guess headroom should also be frequency independent. Have not measured the spectrum of a tick or pop, but I assume there is considerable high frequency content.

Regarding noise level in vinyl: I like intelligent solutions reducing low frequency noise level. In my view, a good phono preamp should also offer a well implemented rumble filter and a way to activate mono bass below a certain frequency (say, 120-200Hz).

 

[Multicore]

I am glad I asked. Very interesting thread. I want to try to distill some lessons of what I think I learned.

First, that my OP reflects simplistic thinking about gain staging by assuming all is linear and time-independent until clipping and then clipping is simple. Apparently it isn't always so with analog phono preamp circuits. Some have dynamic behavior at or close to overload that's worse than just clipping. For example there can be oscillation that takes time to recover.

Iiuc the main point @restorer-john (among others) has been making: with all the gain and eq required, good circuits must have lots of headroom to ensure they stay in their well-behaved range even under the biggest pop inputs. This suggests that Amir's "extra points for headroom" is maybe a good heuristic even though it doesn't actually measure a device's dynamic behavior under overload.

And iiuc @atmasphere adds to this that some circuits won't necessarily show their worst dynamic behavior without a realistic input RLC network. There are examples of phono preamp circuits that can turn normal vinyl tracking noise into big nasty transients. Such examples may be exceptional but it raises the question about test fixtures if we want to see dynamic behavior at or close to overload.

And finally, since I was never an analog circuit designer, my inclination is: carefully digitize ASAP with all the headroom you'll need and take it from there, as we already discussed on page 1 of this thread. (I'd prefer if professionals did this without needing to print signals to plastic disks for us to digitize at home but that's another discussion.)

In any case, I now feel my one and only contribution to digitizing vinyl for the world is not up to snuff.

 

[markstein]

The two most interesting phono preamps checking all these (my) boxes are the MM Pro / MC Pro by Michael Fidler (completely analog design) and the Waxwing / Puffin by Parks Audio (DSP based design). Would really like to test both, especially the Waxwing which seems to have all the useful functionality you can imagine in one box.

I am currently running my turntable into a mic preamp and doing all the RIAA and DSP on my PC, which gives me all the flexibility I want, however, booting a PC in order to listen to vinyl is kind of strange ;-)

 

[Multicore]

I think there are two options for a "good" phono preamp with regard to headroom:

Either provide an extraordinarily huge headroom to cover all kind of cartridges with varying sensitivity,

Or - even better - to provide sufficiently adjustable input gain not to "waste" S/N ratio for cartridges with less sensitivity.

I guess headroom should also be frequency independent. Have not measured the spectrum of a tick or pop, but I assume there is considerable high frequency content.

Regarding noise level in vinyl: I like intelligent solutions reducing low frequency noise level. In my view, a good phono preamp should also offer a well implemented rumble filter and a way to activate mono bass below a certain frequency (say, 120-200Hz).

see @sergeauckland post here.