Not sure which forum this belongs in, could be amplifier, speaker, line level...?
The mission: Use tube amp as full range amplifier, have amplifier2 driven by tube amp with RC low pass filters of at least 2nd order, preferably 4th order. Amp2 is driving a 15" woofer in an OB speaker. Experimentation is required.
Now all logic prevents us from going RC passive filtering past 1st, or perhaps 2nd order. The signal loss is just too much, and each section must handle the loading of the next, etc.
But in this case I have a large signal out of the tube amplifier needing to be attenuated approx 30dB before hitting the inputs of the amp2. The amplifier can drive low impedance values the first stage of the RC filter will present, so I can start off with a low value. I would like to load the tube amp as little as possible tho, so an input impedance of sa 25-50ohms is preferred.
One little challenge which probably limits the number of cascaded sections is my amp2 has input resistance of only 1.8kohms. (The amp uses mixed feedback voltage¤t, and to limit offset the input resistance equals the feedback networks resistance to ground).
But I have no notes on cascading RC filters... Cannot find much out on the W-cube either.
Any suggestions for calculating LinkRiley 2nd, 3rd, and 4th cascaded RC filters?
If this venture is promising, I will likely go for a LC version later. I can do inductors at work, if I ever get the time...But right now I just want to get going for the weekends test run.
Thnx
The mission: Use tube amp as full range amplifier, have amplifier2 driven by tube amp with RC low pass filters of at least 2nd order, preferably 4th order. Amp2 is driving a 15" woofer in an OB speaker. Experimentation is required.
Now all logic prevents us from going RC passive filtering past 1st, or perhaps 2nd order. The signal loss is just too much, and each section must handle the loading of the next, etc.
But in this case I have a large signal out of the tube amplifier needing to be attenuated approx 30dB before hitting the inputs of the amp2. The amplifier can drive low impedance values the first stage of the RC filter will present, so I can start off with a low value. I would like to load the tube amp as little as possible tho, so an input impedance of sa 25-50ohms is preferred.
One little challenge which probably limits the number of cascaded sections is my amp2 has input resistance of only 1.8kohms. (The amp uses mixed feedback voltage¤t, and to limit offset the input resistance equals the feedback networks resistance to ground).
But I have no notes on cascading RC filters... Cannot find much out on the W-cube either.
Any suggestions for calculating LinkRiley 2nd, 3rd, and 4th cascaded RC filters?
If this venture is promising, I will likely go for a LC version later. I can do inductors at work, if I ever get the time...But right now I just want to get going for the weekends test run.
Thnx
Cascaded R-C filters have a very gradual "cut off". Almost useless for protecting drivers or reducing IMD.
L-C filters can work. If you can find/make the values. If your slide-rule is well oiled.
A simple 30:1 passive pad, with a buffer after, and a buffer before your 1.8k amp, allows simple cookbook active filtering.
L-C filters can work. If you can find/make the values. If your slide-rule is well oiled.
A simple 30:1 passive pad, with a buffer after, and a buffer before your 1.8k amp, allows simple cookbook active filtering.
I've grown lazy over the years, and sure, while it's good to know the basic behavior of typical filters and alignments, one can also be a slob like myself and just lean on LTspice to solve for the particulars. Cobble up some Rs and Cs and see what they do... cascade 'til you're tired and see how that works out. Yes, it's intellectually lazy, so having a bit of theory to add to the "hunt and peck" design process can help out, but just the same, an LTspice .ac analysis will be happy to show you the response of whatever nonsense you hook up, without having to build anything, run equations, or break out the test set.
It's physically impossible to make a Linkwitz-Riley filter with cascaded passive RC sections, but you can approximate second-order Linkwitz-Riley quite well when the source impedance is much smaller than the load impedance, as in your case.
The simplest approach is to just take two RC sections with equal cut-off frequencies and give the second a much higher impedance than the first, so it doesn't load the first much. If you want to be perfectionistic about it, you have to give them slightly different RC values, that allows you to approximate Linkwitz-Riley just a little bit better.
Higher order Linkwitz-Riley filters can't even be approximated this way.
The simplest approach is to just take two RC sections with equal cut-off frequencies and give the second a much higher impedance than the first, so it doesn't load the first much. If you want to be perfectionistic about it, you have to give them slightly different RC values, that allows you to approximate Linkwitz-Riley just a little bit better.
Higher order Linkwitz-Riley filters can't even be approximated this way.