Blown Tweeter in an Acme?

[martens-koop]

hi Tim x

so then with my rudeimentary math skills, I am finding that I cannot make a 28hz cutoff point work with a .47uF cap because I would need to make the total resistance be 12K which is not possible by putting any amount of resistors in parallel... my math is as follows:

1/ (1/10,000ohms) + (1/1,000,000ohms) = 9900 ohms

basically, no matter what resistor value I use, I can't get the total above the 10,000 ohms of the input impedance...

but If I go to a 4700nF with a 1.4K resistor or a 47000nF capacitor with a 120ohm resistor, I get the desired Fo results..

(with the 47000nF cap), the 120 ohm resistor gets is close enough to the desired R value that the 10K input impedance does not seem to effect it much..) with the 4700nF cap, my math indicates that I need a 1.4K resistor to offset the 10K impedance... does this make sense?

is it because the 47000nF and 120 ohm are so far below the input inmpedance as to not effect things much?

if thats the case, will there be a huge audible difference? would it be negligent to try and build both to A/B them? is there a chance of actually wrecking anything?

I don't think that I will ever change power amps in the foreseeable future... I like what I've got, and can't afford anything new!!!

33hz just seems like too high of a cutoff point, because at that level, 31hz will be way too quiet... I'm actually somewhat worried that at a 28hz cutoff, 31 hz will still be too quiet...

sorry to be so long winded, but until now I've only been able to talk about this with my wife and 2 yr old daughter, and all they hear is "blah blah blah ohms farads and widgets" its nice to hear from someone who doesn't think I'm spouting nonsense!!!

jmk!!!!

 

[Tim__x]

120 ohms is a pretty hard load for a pre-amp to drive.
If 33.87hrtz sounds too high it would be easiest to lower it by changing the cap's value. For example if you parallel two caps, 0.22uf + 0.47uf = 0.69uf, you'll get a 23.06hrtz cutoff which will remove some handling noise but not any actual bass sound.

These are just 6db an octave filters so I suggest you buy both caps and try it first with just the 0.47uf and then with both, you may be suprised by how much lows you still have.

You could probably fit it all in the jack on a patch cable.

 

[martens-koop]

well, I just bought a box, and tested it out... it sure made a HUGE difference to how much the cones move when I pull hard on the G string.. but it really didn't cut that much into the low b... out of some old copper-clad board I made permenant traces and soldering spots for the various components... I also put a bypass switch in so that a person can play with or without the highpass.

on monday I'm going to get a bunch more capacitors and resistors to see if I can bump it up a few notches... (I was hoping that this box would be so aggressive that I would actually notice a drop in volume on the low B...)

all the hard work is now done... from here on in, its just a matter of swapping our capacitors and resistors...

you were right about how low 33hz is... I'm going to recalculate and see what happens if I centre it around 40 hz or so... then if I need to, I can always boost the bass on my preamp...



thanks SOOOO much for your help!!! it has been invaluable!!!

jmk

 

[Petebass]

I see a problem. The reason your low B is still moving around too much is because you're combining a shallow rolloff with a very low cuttoff point.

You shouldn't worry about setting your cuttoff above 31Hz if it's only a 6dB/ octave filter. At a 33Hz, there should be no audible difference to the 31Hz signal, simply because a 1 or 2 dB cut isn't recognisable to the human ear, certainly not at gig volume. You could even set your cutoff at 40Hz and you're still only 3dB down at 30Hz.

If you were using a 12dB/octave rolloff or higher, I certainly wouldn't recommend going as high as 40Hz. But with only 6dB/octave, you have 2 choices - higher cuttoff, or steeper slope.

Let us know how your experimentation goes.

 

[Tim__x]

I see a problem. The reason your low B is still moving around too much is because you're combining a shallow rolloff with a very low cuttoff point.

You shouldn't worry about setting your cuttoff above 31Hz if it's only a 6dB/ octave filter. At a 33Hz, there should be no audible difference to the 31Hz signal, simply because a 1 or 2 dB cut isn't recognisable to the human ear, certainly not at gig volume. You could even set your cutoff at 40Hz and you're still only 3dB down at 30Hz.

If you were using a 12dB/octave rolloff or higher, I certainly wouldn't recommend going as high as 40Hz. But with only 6dB/octave, you have 2 choices - higher cuttoff, or steeper slope.

Let us know how your experimentation goes.

Actualy the 33hrtz point in the filter I posted is already down 3db.

I do agree, these 6db/octave filters are fairly subtle, but the problem is that to do a higher order filter requires either inductors (big, expensive, hard to find) or active filters.

Martens-koop, if you do want to try an active filter I suggest this one Link Removed just use the high-pass section of it and leave out the right channel and the low-passes.
I suggest C=270nf R=13k 2R=26k (two 13ks in series), these are standard values and will give a cut-off of around 32hrtz but this time it'll be 24db/octave (four times as steep).
If you want 12db/octave I suggest C=270nf and R=18K this will give you 32.5hrtz.

 

[martens-koop]

wow man! that active crossover is crazy technical!!! those triangly things... I assume they are Transistors? (for some reason my gut tells me they might be diodes...)

i think that I'm going to have to think a whole before trying to put something like that together... since i've got my box and a hard copper board to monkey around with, I'll try different combos of resistors and caps and see if that gets me anywhere.... if it doesn't, I'll have to try the active one...


I found a website that has a java program that tells you what value caps and resistors to use for a specific frequency cutoff... so figuring out which resistor to actually get is just a matter of figuring in the input impedance... unfortunately, I just upgraded my os to win xp, aand I heard a rumour that the java virtual machine is no longer included with internet explorer... oh well... can't win em all!!!




jmk

 

[Passinwind]

wow man! that active crossover is crazy technical!!! those triangly things... I assume they are Transistors? (for some reason my gut tells me they might be diodes...)

Those are op amp integrated circuits. The ESP site has a primer that explains stuff like that somewhere, I believe.