Hi,
Is it feasible to construct a line-level passive circuit consisting of three cascaded RC filters?
I guess the showstopper, if any, would be insertion loss.
The amp is a Crown K2 with 10Kohm input impedance.
Could someone tell me where to look for design formulae?
Background:
I have two ported subs using Avalanche 12, tuned to 18 Hz.
I have occasional bottoming on aggressive soundtracks, which I'd like to fix with a 3rd-order (is Bessel the recommended type?) HP filter at, say, 16 Hz.
Thanks
Noah
Is it feasible to construct a line-level passive circuit consisting of three cascaded RC filters?
I guess the showstopper, if any, would be insertion loss.
The amp is a Crown K2 with 10Kohm input impedance.
Could someone tell me where to look for design formulae?
Background:
I have two ported subs using Avalanche 12, tuned to 18 Hz.
I have occasional bottoming on aggressive soundtracks, which I'd like to fix with a 3rd-order (is Bessel the recommended type?) HP filter at, say, 16 Hz.
Thanks
Noah
Thanks for that new link, Tweeker.
That one doesn't have second order, and doesn't seem to account for the amp impedance the way the first one does.
Speaking of which, can I just use the 10K amp impedance as R2 and dispense with the extra resistor?
Thanks
"AFAIK the K2 already has a 3rd order HPF. If you are not afraid of tweaking that you could try to change its cutoff-frequency."
Right, at 8 Hz, but it's a lot easier to add an outboard filter than to dig around in the amp.
That one doesn't have second order, and doesn't seem to account for the amp impedance the way the first one does.
Speaking of which, can I just use the 10K amp impedance as R2 and dispense with the extra resistor?
Thanks
"AFAIK the K2 already has a 3rd order HPF. If you are not afraid of tweaking that you could try to change its cutoff-frequency."
Right, at 8 Hz, but it's a lot easier to add an outboard filter than to dig around in the amp.
You cannot add a 3rd-order passive filter to that amp because the amp input impedance is quite low at 10k. Even 2nd-order would be a struggle. And you would need a strong source.
The filter will be very droopy and affect the passband.
You need to build an active filter, which will give far superior performance in all respects, and can be very small. I think I might have a pair of 3rd-order active filter PCBs lying around at home, or could find a layout
The filter will be very droopy and affect the passband.
You need to build an active filter, which will give far superior performance in all respects, and can be very small. I think I might have a pair of 3rd-order active filter PCBs lying around at home, or could find a layout
If you could you could do it on the first order filter also.
This second order filter is just a cascade of two first order filters. Use the equations from t-linespeakers.org or Art Ludwig.
To use the java tool, you need to add in the amp impedance, it lies in parallel with R2.
The Java tool aggrees with Richie, effect on passband and high insertion loss.
Tweeker,
"If you could you could do it on the first order filter also."
I don't understand.
Looking at the schematic for 2nd-order HP at t-linespeakers.org, the amp impedance can be merged with R2, but there's nothing directly in parallel with R1 to merge it with.
Richie00boy,
"The filter will be very droopy and affect the passband."
Yikes, if R1 = R2/10, doesn't that automatically make the insertion loss 20 dB?
But why can't R2 be, say, R2/4?
"I think I might have a pair of 3rd-order active filter PCBs lying around at home"
Sounds interesting. Does it include a PS, run off a wall wart, or ?
I just need one BTW.
You can email me details at noah.katz at lmco.com
Or is there something I can just buy ready to go?
Thanks
"If you could you could do it on the first order filter also."
I don't understand.
Looking at the schematic for 2nd-order HP at t-linespeakers.org, the amp impedance can be merged with R2, but there's nothing directly in parallel with R1 to merge it with.
Richie00boy,
"The filter will be very droopy and affect the passband."
Yikes, if R1 = R2/10, doesn't that automatically make the insertion loss 20 dB?
But why can't R2 be, say, R2/4?
"I think I might have a pair of 3rd-order active filter PCBs lying around at home"
Sounds interesting. Does it include a PS, run off a wall wart, or ?
I just need one BTW.
You can email me details at noah.katz at lmco.com
Or is there something I can just buy ready to go?
Thanks
I have just looked and the PCB is a 3rd-order filter with input buffer and output drive capability. It's a board I made a while ago, so you would have to be able to follow a hand drawn component overlay on a seperate sheet of paper as there is none on the board. A PSU is needed, but this could be very simple.
Probably a bit ineffective to send to the US, but you're welcome to it if you pay postage.
Probably a bit ineffective to send to the US, but you're welcome to it if you pay postage.
filter
I think you may find it hard to achieve much attenuation of 16Hz if you want to operate at 18-20Hz. I think you are looking at a clocked digital filter. I seem to remember using a very simple design using an LTC clocked filter device. They are easy to use and could be battery powered. You can obtain a 7th order design quite easily. Worth taking a look at the LTC website (Linear Technologies) I think the part I used previously was an LTC 1083 or something similar. I used this as a sub filter and it worked well.
regards,
Steve
I think you may find it hard to achieve much attenuation of 16Hz if you want to operate at 18-20Hz. I think you are looking at a clocked digital filter. I seem to remember using a very simple design using an LTC clocked filter device. They are easy to use and could be battery powered. You can obtain a 7th order design quite easily. Worth taking a look at the LTC website (Linear Technologies) I think the part I used previously was an LTC 1083 or something similar. I used this as a sub filter and it worked well.
regards,
Steve
Replying to your edits:
The passband will be droopy. This means that the roll-off curve will be relatively gentle because the Q is very low, with the effect that the passband will see some attenuation as you get nearer to the actual cut-off frequency.
You can reduce insertion loss by making the impedance differences smaller, but then the filter sections interact. Going active is really the only sensible thing to do with 2nd-order or above.
By all means try digital/chips, but a 3rd-order filter with Q around 0.8 might be just fine and is simple.
For best integration you would build a stereo infra-sonic filter and run it before your main amp, e.g. through a tape loop. That way the group delay/phase effects happen to all speakers so it's not so noticeable.
The passband will be droopy. This means that the roll-off curve will be relatively gentle because the Q is very low, with the effect that the passband will see some attenuation as you get nearer to the actual cut-off frequency.
You can reduce insertion loss by making the impedance differences smaller, but then the filter sections interact. Going active is really the only sensible thing to do with 2nd-order or above.
By all means try digital/chips, but a 3rd-order filter with Q around 0.8 might be just fine and is simple.
For best integration you would build a stereo infra-sonic filter and run it before your main amp, e.g. through a tape loop. That way the group delay/phase effects happen to all speakers so it's not so noticeable.
filter
I'm pretty sure there is no low frequence limit and I apologise for the incorrect part number. Yes you are right about the external Rs and Cs. It's worth a go. As I said, I have used them as low pass filters for sub use, set as 50Hz high pass without problems. Some people may say they are a little noisy, but at these very low frequencies I doubt that it would be noticable.
I'm pretty sure there is no low frequence limit and I apologise for the incorrect part number. Yes you are right about the external Rs and Cs. It's worth a go. As I said, I have used them as low pass filters for sub use, set as 50Hz high pass without problems. Some people may say they are a little noisy, but at these very low frequencies I doubt that it would be noticable.
Richie,
"By all means try digital/chips, but a 3rd-order filter with Q around 0.8 might be just fine and is simple."
I'm confused by the "but"; you mean a 3rd-order w/the digital chip, right?
Steve,
What did you do for PS?
A wall wart on its own would have 60/120 Hz hum, right? I guess I could just put a big filter cap on it.
Thanks guys
"By all means try digital/chips, but a 3rd-order filter with Q around 0.8 might be just fine and is simple."
I'm confused by the "but"; you mean a 3rd-order w/the digital chip, right?
Steve,
What did you do for PS?
A wall wart on its own would have 60/120 Hz hum, right? I guess I could just put a big filter cap on it.
Thanks guys
No I meant go digital if you want with whatever filter order you want,,,,,,,,but,,,,,,,the analogue filter you can build which is 3rd-order is simple and easy to build and will probably do the job fine.
Actually that assumption is not correct in theory. Precisely because the sub should only emit low frequencies makes the (higher frequency content of) noise even more audible, because it is not masked.
I think you misunderstand how to build a PSU. You can't just use an AC wall wart 'as is' -- it needs to be rectified and filtered to get the DC. A very nice yet still fairly simple one is on my website if you can make your own PCBs. Or maybe I can find a spare board for that as well Note though that you can get away with a simpler PSU if needs be for this particular application.
Actually that assumption is not correct in theory. Precisely because the sub should only emit low frequencies makes the (higher frequency content of) noise even more audible, because it is not masked.
I think you misunderstand how to build a PSU. You can't just use an AC wall wart 'as is' -- it needs to be rectified and filtered to get the DC. A very nice yet still fairly simple one is on my website if you can make your own PCBs. Or maybe I can find a spare board for that as well
Note though that you can get away with a simpler PSU if needs be for this particular application.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.