Subsonic Filter
Printed From: Graham Slee Hifi System Components
Category: DIY AUDIO
Forum Name: Owner's Hot-Mods
Forum Description: Tell us how and why you hot-modded your audio gear
URL: https://www.hifisystemcomponents.com/forum/forum_posts.asp?TID=3005
Printed Date: 24 Apr 2024 at 3:52am
Software Version: Web Wiz Forums 12.01 - http://www.webwizforums.com
Topic: Subsonic Filter
Posted By: GrahamD
Subject: Subsonic Filter
Date Posted: 19 Aug 2016 at 12:02am
|
I know Graham Slee may never forgive me but I have added a DIY subsonic filter to the output of my Gram Amp 2 !!!!! I have 3 or 4 records in my collection that produce excessive cone pumping on the first 2 or 3 tracks which while not audible are not good for the speakers. ( Monitor Audio 6 ). I have whipped up a little circuit which is plugged in between the preamp and power amp. It is nothing more than a 0.1uF cap in series and a 75k ohm resistor across the output of each channel which provides attenuation below 20k hz. A switch allows for it to be bypassed when not needed. It works fine and reduces the cone pumping to a slight tremble with no ill effect to the sound. Cheers Graham |
Replies:
Posted By: Graham Slee
Date Posted: 19 Aug 2016 at 9:43am
|
Why should I be upset with you for doing this? The values give a high pass filter which turns over at 21.22Hz (-3dB) , but when the load of the amp it is driving is taken into consideration, the 75k will not be 75k, it will be less. By how much depends on the input impedance of the amp (or preamp etc). If a 1Meg valve amp input the parallel combination will be 70k, and the turnover frequency only slightly higher at 22.74Hz. At the other extreme of say a 22k input solid state, the parallel combination will be 17k, and the turnover frequency much higher at 93.6Hz. Another much more complex way of doing it would be to use the channel subtract and sum method which has been doing the rounds since 1979 (see: http://www.douglas-self.com/ampins/wwarchive/wwarchive.htm). This cross feeds the bass from around 150Hz and below and cancels the lateral differences from the cartridge where the rumble, we are told, comes from. Purists may argue that it makes the lows mono, which isn't a good thing, but we are told that records are cut with mono bass below 300Hz anyway. I beg to differ on that, because some early stereo records have the bassists over to one side or another. Anyway, as bass is less directional and the "cross feed" circuit doesn't result in phase differences within audibility (which your filter does), then it looks like a better idea. However, to get the cross feed a large number of op-amp stages are required to do the subtraction and summing and so the resulting sound might not be as good. Cone flap should not damage well made speakers, and if so it would only be of visual annoyance. Replacing the grilles would remove that annoyance, and not cost anything. Just my take. ------------- That none should be able to buy or sell without a smartphone and the knowledge in how to use apps |
Posted By: GrahamD
Date Posted: 19 Aug 2016 at 10:31am
|
Hi Graham. Just kidding about you not forgiving me, but I know you are not keen on subsonic filters. The amp is a Yarland Pro 88 SE valve amp with a stated input impedance of 100k ohms @ 1kHz. Not sure what that does to the crossover frequency, maybe I could use higher value resistors to fine tune it lower. I didn't want a complicated circuit, just something simple to use when required to stop me worrying about the cone movement on the few records that seem to have low frequency rubbish recorded on them. Cheers Graham |
Posted By: Chris Firth
Date Posted: 19 Aug 2016 at 11:15am
... and therein lies your problem. Cone flap occurs with bass reflex speakers, and isn't as much of an issue if your amplifier has a high electrical damping factor. MA tend to use big solid state power amps with high damping factors when they design their speaker models. Your amp effectively has sod all electrical damping factor, which in turn means that the back emf from your speaker drivers will have an effect on the signal. The cure is to get a loudspeaker that is mechanically well damped, and has a relatively benign load. |
GrahamD wrote:Posted By: BackinBlack
Date Posted: 19 Aug 2016 at 2:05pm
Another cure - 2 x Proprius to keep them cones under control  ------------- Just listen, if it sounds good to you, enjoy it. |
Posted By: Graham Slee
Date Posted: 19 Aug 2016 at 3:12pm
|
-3dB turnover frequency 37Hz, 0.1u into 43k (75k || 100k). Slope of filter 6dB/octave. Agreed, damping factor (or should we call it driving impedance?) isn't much from an output transformer, and as such "sod all" is appropriate wording. Reflex "loading" allows the driver to resonate (pump) at its resonant frequency and a complete lack of damping will allow large cone excursions, but I would have thought the manufacturer would have taken that into consideration so as not to "bottom out" on its suspension. Then again, Chris says these are designed using amps with a stiff driving impedance. It's always a good idea to do a lot of research before buying, but some manufacturers don't give the full details these days. And therefore I should remark that the Proprius also has sod-all damping factor  Based on common knowledge it is around 4 at 20Hz. I say common knowledge because some rubbish the whole idea of damping factor. Back emf is the release of inductor energy once the drive has been removed, but as the amplifier is now driving in the opposite direction the maths will be a little more complex, and a factor of 4 might be a long way off. I once connected a DC coupled amp to a drive unit and tapped the cone and then repeated it after disconnecting the amp. My ears heard no difference. No change in cone excursion. Which suggests that simple calculation is not indicative of the amplifier's control over the cone. Also being old enough to have only had analogue tape or vinyl sources, cone flap was not considered out of the ordinary. In fact one expected to see it and when not there we thought something must be wrong ------------- That none should be able to buy or sell without a smartphone and the knowledge in how to use apps |
Posted By: Richardl60
Date Posted: 19 Aug 2016 at 3:13pm
| Another angle. If this relates to a small number of records are they warped? If they are can they be flattened to reduce this issue? |
Posted By: GrahamD
Date Posted: 19 Aug 2016 at 10:22pm
|
Thanks for the input guys. I must admit I never considered the lack of damping provided by valve amp transformers. What value resistor do you suggest Graham to bring the turn over closer to 20Hz ? The records in question are not warped, in fact my Oracle Delphi has a record clamp, it seems to be recorded into the 'groove'. One of the culprits is a brand new pressing of Dave Gilmours Rattle That Lock. The first track on side one really gets the cones flapping. Cheers Graham |
Posted By: GrahamD
Date Posted: 25 Aug 2016 at 2:06am
|
Well, in the absence of further advice I went ahead and changed the resistor value to bring the turnover closer to 20Hz. I had a couple of 270k resistors available and according to the advice given these in parallel with the amps 100k should result in 73k which with the 0.1uF capacitors results in 21.8Hz with a 6db/octave slope. It certainly sounds fine to my old ears as I can't detect any difference in the sound switching it on or off but a noticeable reduction in cone movement at these low frequencies. I am very pleased with outcome as I only wanted a very simple and non-intrusive filter that I could activate when required with little or no impact on the music. Cheers Graham 
|
Posted By: Graham Slee
Date Posted: 25 Aug 2016 at 3:38am
|
Well done! ------------- That none should be able to buy or sell without a smartphone and the knowledge in how to use apps |