Jump to content
Why become a member? ×

High Pass Filter - Simple New Design.


Stub Mandrel
 Share

Recommended Posts

I've been looking into high pass filter design.

I'm looking at a basic Sallen & Key format with two two-pole stages switchable for 12dB an 24dB per octave.

The advantage of this is simplicity in design and  also it's easy to make one stage variable frequency to tune the response (e.g. set one stage to ~30Hz and the other variable, say 20-50Hz).

A two-stage butterworth or chebyshev filter would give a sharper 'knee'. This does mean looking at gain greater than 1.

My thought is that the 'cost' of going to chebyshev is even higher gain and a less flat curve for the pass band.

While 'gain is good', ideally a filter like  this should have gain close to 1 for the pass band so what comes out is essentially what goes in. Too high a gain means it could run out of headroom on transients, which defeats the object of the exercise.

I plan to use a charge pump to get +/- 9V to get plenty of headroom.

My thought is the butterworth arrangement (very flat passband, relatively sharp knee) is ideal, and tune it to have knee at about 30Hz by tuning for -3dB in the mid-low twenties.

 

Edited by Stub Mandrel
Link to comment
Share on other sites

I don't know much about filter designs, but if you use a charge pump, watch out for noise. Most of the ones I've encountered produce a high pitch whistling that can leak into the audio path. e.g. a big culprit was the RMI Dual Band Compressor - it was fine on isolated power, but stick it on a daisy chain and it was pretty audible. I put it down to the charge pump having heard the same result from using 9V-18V adaptor tips in the past.

  • Thanks 1
Link to comment
Share on other sites

4 minutes ago, dannybuoy said:

I don't know much about filter designs, but if you use a charge pump, watch out for noise. Most of the ones I've encountered produce a high pitch whistling that can leak into the audio path. e.g. a big culprit was the RMI Dual Band Compressor - it was fine on isolated power, but stick it on a daisy chain and it was pretty audible. I put it down to the charge pump having heard the same result from using 9V-18V adaptor tips in the past.

Good point; I suspect that this happens when people decouple the switched -ve supply but don't decouple the +ve supply. This is what the datasheet shows, trying to minimise component count, but if the pump is at the end of a daisy chain some decoupling at the converter is going to be needed.

Link to comment
Share on other sites

If you are going to crossover at 30Hz I don't think you need to agonise over Chebyshev v's Butterworth. 

We can't really hear anything below 50Hz in normal circumstances. Depending upon the exact frequencies your sensitivity going to be 18-24dB below the rest of the sound the band are making, including the upper frequencies of your bass, Because of the positioning of the pups there's at least another 12dB attenuation of output of those frequencies from the bass. My experience is that anything below 50Hz can be cut without anyone noticing and I believe most recording studios do this as a matter of course. 

Most ported bass speakers are tuned around the 50Hz mark and few bass guitar drivers have resonant frequencies much below this Even a few Hz below resonance the excursion of a cone rises dramatically and the coil will spend a lot of time outside the magnetic field endangering the driver and distorting the signal, the whole point of an HPF. If you go on the Eminence website you'll see they recommend HPF for most of their compact instrument designs. 40Hz makes more sense to me and even then I doubt you'd hear a difference playing an actual bass rather than test tones other than the filter would mean less distortion as the coil would stay in the linear part of the magnetic field

Link to comment
Share on other sites

32 minutes ago, Stub Mandrel said:

Frequency wise, I'm unashamedly copying the Thumpinator.

thumpinator_freq.gif

The HPF-Pre V3 is pretty much exactly what I had in mind, having one stage at 30Hz and the second stage 30 and up.

 

 

So, not being a negative Nancy here but isn't that minimal change/improvement? Or is this view misleading and it actually has a more dramatic effect audibly?

 

Link to comment
Share on other sites

49 minutes ago, visog said:

So, not being a negative Nancy here but isn't that minimal change/improvement? Or is this view misleading and it actually has a more dramatic effect audibly?

 

The idea is you don't hear any change - but 'sub-harmonics' and thumps get cut giving you more headroom and a cleaner, clearer sound.

Link to comment
Share on other sites

1 hour ago, Stub Mandrel said:

Frequency wise, I'm unashamedly copying the Thumpinator.

thumpinator_freq.gif

The HPF-Pre V3 is pretty much exactly what I had in mind, having one stage at 30Hz and the second stage 30 and up.

 

 

I'm not sure how this was all measured and there is no scale on the vertical axis but all I see here is a filter cutting in at around 25Hz plus a little bit of non linearity between 25 and 100Hz due to an underdamped circuit. I'm not sure I understand what you are saying about stages. If the vertical divisions are 10db then this looks to me (I've done no proper sums though) like a 4th order Chebyshev at around 25Hz. That's going to remove some subsonics but leave the audio signal almost untouched. People speak well of the Thumpinator so copying it is a good idea, saves a lot of experimentation. As a speaker designer however I can say you could get more protection for your speakers by filtering at a higher frequency and I believe it would also sound better if you did that. I can see the logic of following an established design though. I'll follow this project with interest.

Link to comment
Share on other sites

First shot at a design - 'rough and ready'  chebyschev , hoping that a 3:1tuning ratio for the second stage won't have too extreme an effect. Bodge from Art oif Electronics ideas plus HPF-pre warning C3 is connected the wrong way round in this schematic - I've corrected the simplified one).

Gumpinator.thumb.png.207bbc4fb7d07a31667eb9cba776e4b1.png

 

Super simple design with -3dB at 30Hz, using Analog Devices online designer which chooses clever resistor values and lets you have a zero gain design so no volume control needed. Inclined to think the simpler the better (edit gone back to E24 to make it more practical):

 

684658598_SimpleGUmpinator.thumb.png.7052fd53170a67db8ede06aa1c0eb5ea.png

image.thumb.png.88c19a4513a9754244262ef2dbb4e5c1.png

Edited by Stub Mandrel
  • Thanks 1
Link to comment
Share on other sites

By the way, as I have that HPF2, I would modify it to be a tool adjustable unit. The pots tend to move while carrying the stuff around, and the settings are of set-and-forget type. Smaller box would be easier to add on top of other equipment.

Waiting to see your solution, and maybe few extra PCBs, too...

Edited by itu
Link to comment
Share on other sites

1 hour ago, itu said:

By the way, as I have that HPF2, I would modify it to be a tool adjustable unit. The pots tend to move while carrying the stuff around, and the settings are of set-and-forget type. Smaller box would be easier to add on top of other equipment.

Waiting to see your solution, and maybe few extra PCBs, too...

Will definitely make the simple version, may try the more complex one. Good suggestion to use presets - it should be set and forget.

Link to comment
Share on other sites

  • 2 weeks later...
On 13/09/2019 at 20:20, visog said:

So, not being a negative Nancy here but isn't that minimal change/improvement? Or is this view misleading and it actually has a more dramatic effect audibly?

 

From a playing perspective, having owned a fair number of different effects pedals, a Micro Thumpinator hpf had the single most significant impact on my sound. In a loud rock band context fx pedals often (while sounding great at home) diminish bass in the mix. The simple addition of a Micro Thump gave me lots more headroom and an ability to cut through without the speakers moving alarmingly. I'd recommend that every bassist who gigs give one a try. 

Link to comment
Share on other sites

Ok, I've finally got round to etching the PCB, populating its and then testing the high pass filter.

The 50mm square enclosure was challenging - even fitting the two jack sockets was tricky, spacers on the outside. The tall caps and voltage doubler IC (for +/- 9v) fit at bottom right. All caps are either 10u or 100n for simplicity!

DSCN3179.thumb.JPG.2a71b90b18455fd98a9d32a1a3a6c395.JPG

The PCB had to be carefully hacksawed(!) right to the edges of the tracks, I used soft faced vice jaws as a guide for the saw. Note one corner was too tight so I had to 'divert' the track:

DSCN3186.JPG.14a1182b545fcd3c171869cd1cd46e53.JPG

Power LED just hard wired to the power connector and switched -ve via the ring of the input jack, so not on PCB. Need to put on a proper decoration - and the right way up!

DSCN3185.thumb.JPG.3bc7190254108cf3d9a5dc8f58e12656.JPG

First stick in a bottom B from my home-made frequency generator:

DSCN3180.JPG.568aab7781a366203006ef89f6b68f3a.JPG

4V P-P, (actually no visible difference between in and out signals up to 6KHz - I didn't bother going higher):

DSCN3181.JPG.3afab31a2931d3b5a035b03c633413c2.JPG

As hoped, gain rapidly drops off below 31Hz.

DSCN3182.JPG.f1ac8a772b66c687936e567fe99812c6.JPG

This is 0.4V at 22Hz:

DSCN3184.JPG.11540e8eec52bce73963cbcb35f59a45.JPG

4.0 top 0.4 is gain of 0.1 that's exactly 20dB cut at 22Hz, which is pretty much excatly what the modelling software predicted. Kudos to Analog Devices!

image.png

 

The nay-sayers will say I should have made the cutoff higher, but it's (supposedly) 'infrasonic' signals below about 16Hz that are the main problem.

 

 

Link to comment
Share on other sites

Update: used in anger a faint, high pitched  whistle was evident.

Back at the lab this was rapidly traced to the voltage pump, which was wired to the default operating frequency of 10KHz, and it appeared to be the 5KHz a sub-harmonic of this that was audible, despite being a very faint signal.

Larger supply capacitors could have cured this but not fitted a small board. The cure was simple - bridging the unconnected 'boost' pin to the V+ terminal which raises the switching frequency to 45KHz.

This is slightly less efficient but raises any 'whistle' out of the audible band and also increases the filtering effect of the output capacitors four-fold. The modified box is silent and sounds 'transparent' although if you 'pat' the strings with the palm of your hand it clearly attenuates the 'thump'.

I'll replace the circuit diagram and board layout in my previous post.

 

  • Like 1
Link to comment
Share on other sites

Gosh. I actually understand what you've made and what it does. Everything beyond that, with the exception of the bit about hacksawing, raised the technical frequency to 45KHz, or beyond my thinking range. 

Well done and I love the name. Let me know when you're ready to sell them. 

  • Like 1
Link to comment
Share on other sites

45 kHz is related to that power section. The voltage is regulated by a switcher (voltage pump) that makes the audible noise. When the frequency is far over the audible band (20 kHz), it just vanishes from your attention. It is there but you can not hear it. I suppose that other parts in the signal chain will also limit the tweeting.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...