Spækerbœx: 27.2Hz High-pass filter

[EricssonB]

His filter circuit is nearly 100% there already. It will also be noise/distortion free when he makes a couple of modifications.

To make your filter operate properly with the negative supply grounded, you will need to apply a Vbias to the grounded ends of R3 and R6 on the above filter schematic.

Use the Vbias circuit from the opamp buffer circuit in post #41 (R1 R2 and C1). Then lift the grounded ends of the resistors of R3 and R6 in the filter schematic in post #53 and connect them to the Vbias divider.
Got it. I plugged it in BEFORE I made the changes. Whoa boy does that make this a hot circuit. Ear splitting highs and mids.

Then measure the DC voltage of each opamp output to ground for one half of the supply voltage.
Got it. Good: half and half.

-Frank

I should probably pay you. If I'm not mistaken, you sent some advice on my January project.

Congratulations, now you have a fuzz circuit.

Anyways, I got this thing running. I think it works.

Yeah. I think it does. It's transparent when in or out of the loop -- trouble with an HPF is that it's difficult to objectively test it. I turned off the amp's compressor and I think the cone moves less than it did before, but I'm not sure.
GC
Anybody have any empirical tests?

Next day update: the folks in the 24-hour office have an oscilloscope. If it has a tone generator, I'll check the frequency cornering. Or rig up some iPhone-base creation at home.

 

[EricssonB]

Mute switch with LED, dark when no input cable.

I have common cathode RGB LEDs. I think if I switch to common anode, then I'll be able to get the blue one to light up when there's no cable plugged in. Maybe I can with CC, but the late hour is killing my critical thinking.

 

[EricssonB]

Tested with a 20Hz tone. Works well. It's not 100%, but the amount of cone movement with it on is significantly lower than it is without. The orange LED on the amp is also less energetic.

I'd improvise an oscilloscope and plot the reduction of frequencies, but my interest in 1 testing it any further and B opening this thing back up to change anything is waning quickly. Closing this out. I'll update the original post with a complete schematic.

Conversely/regrettably, the flag waves its freedom considerably less with this circuit on.

 

[EricssonB]

Tested through DAW peak graphs:
- flat response ~60Hz and up
- slight reduction between 30-60Hz
- dramatic drop starting at 30Hz
- 10dB loss by 20Hz
- ...tone generator doesn't go below 20Hz

Since I removed the 220k ohm resistor from ground to add the Vbias, would I have lost that range (~7Hz) of high pass? Sure, it's ultra low, but that's what I'm trying to gut.

Should I add another 220k to ground?

 

[boomertech]

From your observations and measurements it appears that your filter is working perfectly. The 47uF capacitor on the Vbias circuit attached to the two 220K resistors essentially appears as a direct connection to ground for AC signals. Do not add another resistor to ground!

Look at figure 2 from that ESP website for the filter response. Figure 2 is for the 18.1Hz filter configuration, but just visually slide the red graph to the right by approx 10Hz to see what your 27.2Hz response should look like. Approx 10dB down at 20Hz for your filter is about right.

If you want a steeper slope then build a second filter and run one into the other… you would have a total of 72dB/oct filter.

-Frank

 

[EricssonB]

It's official: I'm tired of this thread, and I'm sure you are, too.

Closing it out with this final layout. Seems accurate. It's ugly -- I know. I combined the two boards but didn't make it any more efficient on space, but it's not too large to begin with.

Note: power section ripped out from original HPF schematic. Uses +V and +Vbias from Buffer section.

BOOMERTECH THANK YOU FOR ALL OF THE HELP YOU'RE AWESOME I SHOULD SEND YOU SOME MONEY

 

[boomertech]

I’m happy to help and really glad to see your project working.



-Frank