No less important is that they have an extreme amplitude. The main reason for phono preamplifier overloads.
Red - cartridge output. Blue - MM phono preamp output. Vinyl groove surface click.
This shows why it’s so important to ensure good phono preamp overload performance. This is an extreme case, but even with a more normal pop or click, it’s easy to see why 9 dB is just not enough.
Please do not react angrily, but I do not follow this simple reasoning.
I suggest looking at the clicks regularly - that is, in a wave editor, after sampling and digitizing. At the same time, you should ask yourself how a short impulse is converted by the mechanical sensor and then what the impulse response to a first-order low-pass filter or PI element with f-3=2kHz should theoretically look like. The 1.2V peak shown by PMA should not pose a problem for a correctly dimensioned RIAA EQ on the input side. Just think of the short and powerful mechanical nudge, a jump ... A low-pass or band-limited system reacts to jumps with a delay, integrating.
If the first stage of a multi-stage EQ process is only a P element or does not include the 75µs time constant, then we have a problem, namely a very short overload.
Let's take a two-stage circuit as an example: +30dB / passive RIAA low-pass / +30dB. If the first stage can follow the signal rise 1 to 1, then we only don't bang the RAILS if they are greater than +/- 34Vdc.
I think we should leave the issue of clipping alone. Especially when I look at PMA's picture a second time and realize that the pickup system only delivers about 0.06Vpeak, we can now also roughly estimate the excursion of the pickup.
1200mV / 60mV = 20
---> +26dB
40-26
= 14
From the point of view of the RIAA low-pass filter, one could now claim that it is a (one period) single, simple 10kHz sine wave.
😉
Above we see the slight drop before the inlet groove. Below we see a classic click at the start of the outlet groove. Everything with 44k1Hz and 16Bit ...
21 samples indicate the 75µsec pole.
The triggered system therefore reacts with an oscillation, ergo f=2122k1Hz ...
or do I now have a plank in front of my head?
😕
or do I now have a plank in front of my head?
😕
Below is DSO scope photo that I took circa 2010 when I was looking at phono preamp design and overload issues. It shows the raw cartridge output in purple and the RIAA equalized output from a V15 Type V in yellow when exposed to a natural, but fairly large-amplitude, tick from a real record.
The vertical scale is 20 mV/div for the cartridge output and 500 mV/div for the RIAA preamplifier output. The peak-to-peak amplitude at the cartridge is about 70 mV and the time scale is 100 us per division. It appears to show fairly wide bandwidth with virtually no ringing. The leading edge of the impulse appears to have a risetime of about 40 uS, while the falling edge is about 30 us. Perhaps the bandwidth of the cartridge can be inferred from this.
The DSO is a fast scope, so what I assume to be the surface noise as seen at the cartridge output is notable. That "surface noise" is highly attenuated after the RIAA EQ. Due to the wide bandwidth of the scope, I cannot rule out that what I am suggesting is surface noise is some EMI. The cartridge is terminated in 47k and about 150 pF, including the turntable interconnect.
Cheers,
Bob
The vertical scale is 20 mV/div for the cartridge output and 500 mV/div for the RIAA preamplifier output. The peak-to-peak amplitude at the cartridge is about 70 mV and the time scale is 100 us per division. It appears to show fairly wide bandwidth with virtually no ringing. The leading edge of the impulse appears to have a risetime of about 40 uS, while the falling edge is about 30 us. Perhaps the bandwidth of the cartridge can be inferred from this.
The DSO is a fast scope, so what I assume to be the surface noise as seen at the cartridge output is notable. That "surface noise" is highly attenuated after the RIAA EQ. Due to the wide bandwidth of the scope, I cannot rule out that what I am suggesting is surface noise is some EMI. The cartridge is terminated in 47k and about 150 pF, including the turntable interconnect.
Cheers,
Bob
Sorry in advance for the short off-topic.
A few corrections:
In the drop in the upper picture of posting #233 we can of course see the swing-out with the entire mechanical resonance of the combination of tonearm and cartridge, the classic, here with 10Hz.
And for mechanical excitation by scratches, the simple band limitation by 1/(2*PI*75µsec) is always present.
The whole scenario is therefore directional.
Very interesting Bob, thank you.
3.5 * 20mV & 3.5 * 500mV
175 / 7 -> 28dB
40 - 28 = 12 we look into |H(jOmega)| and find f_riaaa pproximately >8kHz; shorter 125µsec, your deflection coefficient is 100µsec per div.
If we want to look at it very simply, we can see exactly the decay behavior. The absolute classic ...
always short and higher frequent - no real overload-tendency !
3.5 * 20mV & 3.5 * 500mV
175 / 7 -> 28dB
40 - 28 = 12 we look into |H(jOmega)| and find f_riaaa pproximately >8kHz; shorter 125µsec, your deflection coefficient is 100µsec per div.
If we want to look at it very simply, we can see exactly the decay behavior. The absolute classic ...
always short and higher frequent - no real overload-tendency !
The RIAA slope above 2kHz is -20 dB/decade so if you have only HF click noise it’s going to be attenuated. The problem arises when these events are riding on other HF material.
This Tomlinson Holman’s 1976 graph of peak signals from across a number of records. If 5mV/cm is used as the midband reference (typical cart output), the signal level out of the cart can be read directly in mV. The red trace is typical RIAA amp response in dB also read off the y-axis.
This Tomlinson Holman’s 1976 graph of peak signals from across a number of records. If 5mV/cm is used as the midband reference (typical cart output), the signal level out of the cart can be read directly in mV. The red trace is typical RIAA amp response in dB also read off the y-axis.
Last edited:
This is the same click in higher time resolution:
and I agree that for a competent phono preamp design it should be no problem. Integrating character of the RIAA EQ response helps a lot and the click should have been much broader to send the preamp to output saturation or internal saturation.
Frequency representation does not help much (IMO) to evaluate the click or mistracking effect.
You would have to do a FFT analysis of the click to pull out the frequency content and amplitudes, but it is just easier and in my opinion better to settle on all-active RIAA for opamp level supply rails. All of the overload issues being discussed here are then resolved.
Demonstration of the operation of an analog record click suppressor developed by the laboratory of A.S. Bogatyrev at the Museum of Magnetic Recording Research Institute (Kiev) - [
It was done with a predictable result, but again, it is difficult and time consuming to get meaningful result from frequency domain analysis of a single impulse. Time domain is much better for this task. Frequency domain is good for periodic or quasi periodic signals.
Let us assume that it is possible with the storage medium to display a reasonable sine wave with f=10kHz and a nominal output voltage of 2Vrms (!) and that this is vectorially added to the interfering pulse (it interferes and irritates the mm transducer in the first instance) at the peak, then the output of our electronics controls to 10V_peak to peak_. If one of the first internal stages is not already stuck in a saturation phenomenon (for a very, very short time), and this will not be the case (unless we want to open the big barrel of negative feedback systems now), then the problems that are so often cited simply do not exist at all. With +/- 12Vdc supply voltages and the usual allInOne RIAA EQ ..! I myself define 24V ergo +12V and -12V rails as the minimum. However, you can also work extremely borderline and work with +/-9Vdc, as is the case with the OREAD and Supra24 MM-compatible equalizers.
With regard to this topic, we should not unsettle any DIY (to which we also belong) - any more.
I am also familiar with this work.
For example,
if we look at the published specifications of your great equalizer Andrew - they speak to us as follows:
"There are no problems, I have an irrepressible overdrive reserve ..."
kindly,
HBt.
+1
complete agreement,
all of these points are undisputed,
but I don't understand why they are always being emphasized.
HBt.
Could we please discuss this exciting topic in full in another thread - and especially in English. Thank you in advance. The many YT insertions from time to time are more distracting than helpful.
kindly,
HBt.
because deep scratches on a gramophone record are very annoying to everyone, especially listeners of incorrectly designed phono preamplifiers.
fancy
peak velocity at 6k32Hz with 80cm/sec
This immediately results in a peak deflection of 2µm.
Even with very sensitive MM systems, this is not a real problem. Our grandfathers did their homework and found technical and physical solutions.