Because I try to make 20-20k flat, 2Hz is -3db on my builds. That's fine for digital. My phono stage makes subsonic rumble 5-10Hz depending on the record and playback speed. Furthermore, any high energy hit (bass drum, snare drum) makes a wiggle.
I'm rubbish at filters, but I think a good LF filter that passes 20Hz and above while being 4th order below that?. I don't know the math to design one... Any pointers?
OTOH, it doesn't effect the sound at all. Just seems like a waste of power.
I'm rubbish at filters, but I think a good LF filter that passes 20Hz and above while being 4th order below that?. I don't know the math to design one... Any pointers?
OTOH, it doesn't effect the sound at all. Just seems like a waste of power.
Last edited:
If you are listening to old recordings be sure to set the filter pole near below or above bbc dip. Otherwise you miss out on alot of action.
For example I use 2240hz for 2250hz. This is opposed to a typical filter pole set at 2500hz or dare say 3000hz
Don't alter it too much off or you will get phase error multiplication.
For example I use 2240hz for 2250hz. This is opposed to a typical filter pole set at 2500hz or dare say 3000hz
Don't alter it too much off or you will get phase error multiplication.
Last edited:
There are possible IM distortion products from the infrasonic noise combining with in audio band info. Also, the infrasonic noise can cause O/P trafo core saturation, in designs with a GNFB loop.
Initially, I suggest you try a simple 1 pole RC filter whose F3 is in the 16-18 Hz. range, right at the power amp's I/P.
For something sophisticated, check out the offerings of Marchand Electronics. You should be able to latch on something that's 4th order Linkwitz-Riley, which exhibits no phase shift at the "corner" frequency.
Initially, I suggest you try a simple 1 pole RC filter whose F3 is in the 16-18 Hz. range, right at the power amp's I/P.
For something sophisticated, check out the offerings of Marchand Electronics. You should be able to latch on something that's 4th order Linkwitz-Riley, which exhibits no phase shift at the "corner" frequency.
There is a good HP filter circuit in this manual's schematic.
APT Corporation Holman - Manual - Stereo Pre Amplifier - HiFi Engine
APT Corporation Holman - Manual - Stereo Pre Amplifier - HiFi Engine
Thanks, guys. I'll have a look and see.
I had the APT Holman Power Amplifier 1... It was an excellent amp but I sold it to a friend, and now a Yorkville AP1200 sits in it's place (my "summer" amp).
And then it dawned on me. I own TCJ Filter Designer and it's now working on Wine 5 staging!
I had the APT Holman Power Amplifier 1... It was an excellent amp but I sold it to a friend, and now a Yorkville AP1200 sits in it's place (my "summer" amp).
And then it dawned on me. I own TCJ Filter Designer and it's now working on Wine 5 staging!
Last edited:
I added an input filter on my build of Eli's RCA phono mod
circuit. I used 0.1uF into 1meg resistor to ground, for 1.59 hz
cutoff. Cuts out some of the record warp frequency, but not all
possible rumble frequencies, but that's not an issue with my
setup. It seems to reduce some of the low frequency groove
noise. I would think too much subsonics would tend to upset the
bias of the input tube also.
The main reason I put the filter in was I was getting a low
frequency constant hummmmmmm with my Grado gold (not a 60hz or
120hz, just low hum) and the turntable was elminated as the
culprit. I also lifted the input tubes circuits from circuit
ground with 15 ohms each. I had the same problem with a previous
12ax7 input preamp. One or the other change fixed the problem, I
suspect normal grid current was enough to cause the hum. I've
never seen any reference to this source of hum, which was a
little surprising.
circuit. I used 0.1uF into 1meg resistor to ground, for 1.59 hz
cutoff. Cuts out some of the record warp frequency, but not all
possible rumble frequencies, but that's not an issue with my
setup. It seems to reduce some of the low frequency groove
noise. I would think too much subsonics would tend to upset the
bias of the input tube also.
The main reason I put the filter in was I was getting a low
frequency constant hummmmmmm with my Grado gold (not a 60hz or
120hz, just low hum) and the turntable was elminated as the
culprit. I also lifted the input tubes circuits from circuit
ground with 15 ohms each. I had the same problem with a previous
12ax7 input preamp. One or the other change fixed the problem, I
suspect normal grid current was enough to cause the hum. I've
never seen any reference to this source of hum, which was a
little surprising.
I can point you at this, which allows you design a filter which I've used a few times.
3rd order Sallen-Key High-pass Filter Design Tool
The op-amp can be a simple emitter follower or a cathode follower such as a 12AX7. The input impedance needs to be much lower than R2 I think.
3rd order Sallen-Key High-pass Filter Design Tool
The op-amp can be a simple emitter follower or a cathode follower such as a 12AX7. The input impedance needs to be much lower than R2 I think.
Yep that's done it nicely. A third order filter is good enough in which case you can do that with a single stage.
If you just want a high-pass filter, I'd go for Butterworth rather than Linkwitz-Riley.
A Linkwitz-Riley has a curved phase response, as all lumped analogue filters have to some extent. The peculiarity of Linkwitz-Riley is that the difference in phase response between the low-pass and the matching high-pass filter is a constant multiple of 180 degrees, while the sum of the magnitudes of their responses is constant. That makes them very useful for loudspeaker crossover filters, but I don't see why that would be an advantage when you only want a high-pass filter.
However, there are special filters that suppress rumble that is in antiphase between left and right, see The ultimate rumble filter - far more effective than just a high pass filter! From post #85 onwards, tubewaller explains how Linkwitz-Riley filters can be used in such an application.
I guess it works similar to how stereo -> mono gets rid of the scratch noise since they tend to be out of phase. That's how I record mono 78 RPM records... No noise removal, just summing the channels.
Third order in one stage puts a lot more demands on component tolerances though - two 2nd-order stages will be more tolerant as well as having an extra pole.
Good rumble filters are hard BTW - phase distortion in the subbass represents significant time delays and waveform distortion, suggesting a Bessel filter might be least problematic in the time-domain, but they are very poor roll-off and you want a nice solid drop to the stopband by 10Hz and a Bessel simply won't give you that.
Butterworth has good amplitude response but the phase response isn't great near the roll off point - this suggests using a high order Butterworth but with a roll off definitely below 20Hz, say 15Hz.
Linear phase equiripple filters are a hybrid of Bessel and Butterworth and can be tuned by a phase-ripple parameter, but they don't fare much better than Bessel.
Some good comparisons can be made from the graphs here:
https://www.analog.com/media/en/training-seminars/design-handbooks/Basic-Linear-Design/Chapter8.pdf
I'd suggest 4th order Butterworth high-pass, perhaps cutoff in 16 -- 18Hz region. Two Sallen Key 2nd order stages.
Bessel high-pass filters have a poor phase response, because the low-pass to high-pass transfer messes it up. Only Bessel low-pass, narrow-band band-pass and complex filters have a good phase response.
Considering I've already built it with tubes? Yup 😀
Since the phase shift will happen at such a low frequency, and the filter is not inside the global feedback loop, I'm not that concerned with it.
Those two filters are the most complex part of the whole integrated amp LOL
Koda
Just a follow up. The circuit I posted works a treat on the output of the aikidoesque phono stage I use. All the LF rumble from the record is gone, and the insertion loss is welcome as the phono was hot anyway (about 48db instead of the normal 40db).
This is actually not true, the best way to play or record mono vinyl is with a mono cartridge. There is no out of phase. In a mono cut 78 there is only a single lateral grove.
This article from Sound Matters sums it up quite nicely:
A mono cartridge is by far the best way to experience mono records. You could take the easy option and use a mono switch, but this will not fix the errors caused by a stereo cartridge never quite producing the same signal in both channels. In other words, playing a mono record with a stereo cartridge will not achieve the same signal in both the left and right channels. It will not be “true mono”. Imperfections such as crosstalk, phase errors, and tracking errors will result in a degradation of sound quality. Using a mono cartridge eliminates these problems by only producing one signal that is then distributed to both speakers on your stereo system. The signal in both your left and right speakers will be identical.
Using a mono cartridge can also reduce the response to dust and dirt substantially as any vertical element will not be reproduced. Often, the final result is a much cleaner reproduction of a mono record. Again, a mono switch cannot remove this noise—once the vertical element is tracked, it’s tracked, and this signal now has to be processed by the system whether you mono the signal retrospectively or not.
Do You Need a Mono Cartridge to Play Mono Records? - Sound Matters
I think electrically, it's the same thing. Any noise cancellation in a mono cartridge is probably done by summing two coils that are 90 degrees to each other. How else would you do it? So a stereo cartridge can be used as a mono cartridge by summing (shorting) L and R. Perhaps I'm wrong, but I see no other electromechanical way of doing it.
Last edited: