Finally my plans are crystallizing. I'm going to an amplifier for each channel (L,R) each with active crossover inside. Each amp will drive a small 2way speaker (passive crossover) and a larger sealed bass box.
My assumptions (please correct me) are that a sealed 2way going down to say 250Hz, and a bass box 20Hz-250Hz will each be relatively easy to design.
What are the things I need to think about to get the 2 boxes working together nicely? Or is a 4th order crossover, parametric equalisers, and adjustable gain all I need (eg MOXlite).
Thanks
Jim
My assumptions (please correct me) are that a sealed 2way going down to say 250Hz, and a bass box 20Hz-250Hz will each be relatively easy to design.
What are the things I need to think about to get the 2 boxes working together nicely? Or is a 4th order crossover, parametric equalisers, and adjustable gain all I need (eg MOXlite).
Thanks
Jim
Hi,
I have a suggestion- if you're using stereo amps, use one amp for bass, and one amp for treble. I know it's attractive to put the crossover inside the amp case, but consider this:
Most stereo amps include a spec on "stereo seperation". It improves imaging to have good stereo seperation. If you use a stereo amp for tweeter and bass, your "stereo seperation" figure becomes a bleeding of highs to the woofer and bass to the tweeter.
Although this figure is generally better than 60db separation, I still recommend that one amp is dedicated to bass, one to treble.
I have a suggestion- if you're using stereo amps, use one amp for bass, and one amp for treble. I know it's attractive to put the crossover inside the amp case, but consider this:
Most stereo amps include a spec on "stereo seperation". It improves imaging to have good stereo seperation. If you use a stereo amp for tweeter and bass, your "stereo seperation" figure becomes a bleeding of highs to the woofer and bass to the tweeter.
Although this figure is generally better than 60db separation, I still recommend that one amp is dedicated to bass, one to treble.
I see what you're saying (although it took me a while to see how stereo separation could be a problem if LR are in separate boxes).
My first post is missing the word BUILD. The two channels in each box will be LM3886 chip amps.
Cross channel interference wasn't one of my concerns, but if you have any tips on what causes it / minimising it ?
My first post is missing the word BUILD. The two channels in each box will be LM3886 chip amps.
Cross channel interference wasn't one of my concerns, but if you have any tips on what causes it / minimising it ?
There are lots of good designs for small two ways - see www.zaphaudio.com for a number of good ones as well as data on drivers to help you select your own. Of course there are other sites, too. Recommendations will depend on your goals and budget.
If you're going to the trouble of building an active XO for the bass-mid, why not go fully active? Just add another filter section and another 3886 for the tweeter. I find active XOs easier to tweak - both physically and psychologically, since the changed out parts usually only cost pennies, unlike quality caps and coils.
If you're going to the trouble of building an active XO for the bass-mid, why not go fully active? Just add another filter section and another 3886 for the tweeter. I find active XOs easier to tweak - both physically and psychologically, since the changed out parts usually only cost pennies, unlike quality caps and coils.
Fully active was my first thought (this was some time ago when I first looked into all this) then I got scared by tweeter protection issues, and thought maybe it was overkill as the biggest benefit comes from the active woofer/mid crossover ( so most people say).
Plus the small 2way is a standalone item that can function independently, and maybe I will just find/buy a suitable new or 2nd hand speaker.
However my plans have already changed 50 times so once more won't hurt. Adding another chipamp is about $50 total, which must compare favorably to the passive crossover.
Plus the small 2way is a standalone item that can function independently, and maybe I will just find/buy a suitable new or 2nd hand speaker.
However my plans have already changed 50 times so once more won't hurt. Adding another chipamp is about $50 total, which must compare favorably to the passive crossover.
If you just look at the damping factor advantage, then active XOs on the mid-woofer get you your biggest bang for the buck - especially if you are copying an existing M-T design.
For designing my own systems, I like the ease of altering the XO to suit the drivers - for pennies I can change the filter response shape and/or frequency, or for a couple of bucks I can add baffle step correction, a notch, a boost or shelving EQ to compensate for less than perfect driver response. Of course, this is only an advantage, if like me you have lots of amps handy, or can build them cheaply.
A pair of Brian Bell's LM4780 kits is $35 more than his dual monoblock LM3886 kit. You wouldn't need significantly more heatsink or transformer, since the tweeter power requirement is low. If you just want to dead-bug another 3886 on the heat sink you have, you can do it for about $10 per channel.
You would need a bit of perfboard and another opamp and a few bucks in caps and resistors to make the XO. Assuming that you are putting all amps and active XO in one box that's all the extra stuff you need. You can run the XO through regulators (LM7815/7915) and maybe a dropping resistor right from your main rails.
I use a Dayton 25uf cap for tweeter protection in my active systems (3 so far). I can't notice a difference between them and some similar value arrays of SCR and AXON caps I have on hand. Haven't listened to any real exotic types, though.
Last night I worked up a ROM price on a system for a buddy and your $50 is about right for a two way 3rd order passive XO using Dayton caps, Jantzen coils (air core except for BSC). Add another $10-15 for tweeter padding resistors and woofer zoebel. More exotic parts raise that cost.
If your woofer is less than 8 ohms, I strongly suggest either using an LM4780 with both channels in parallel or probably better, a pair of 3886s in parallel. You don't want the chips to get anywhere near kicking in SPIKE protection. The 3886 is big enough to build an amp dead bug style. The 4780 is doable, but much harder, making boards like Brian's a bargain.
** disclaimer ** I have no connection with Brian, but have heard good things about his boards, kits and service here.
For designing my own systems, I like the ease of altering the XO to suit the drivers - for pennies I can change the filter response shape and/or frequency, or for a couple of bucks I can add baffle step correction, a notch, a boost or shelving EQ to compensate for less than perfect driver response. Of course, this is only an advantage, if like me you have lots of amps handy, or can build them cheaply.
A pair of Brian Bell's LM4780 kits is $35 more than his dual monoblock LM3886 kit. You wouldn't need significantly more heatsink or transformer, since the tweeter power requirement is low. If you just want to dead-bug another 3886 on the heat sink you have, you can do it for about $10 per channel.
You would need a bit of perfboard and another opamp and a few bucks in caps and resistors to make the XO. Assuming that you are putting all amps and active XO in one box that's all the extra stuff you need. You can run the XO through regulators (LM7815/7915) and maybe a dropping resistor right from your main rails.
I use a Dayton 25uf cap for tweeter protection in my active systems (3 so far). I can't notice a difference between them and some similar value arrays of SCR and AXON caps I have on hand. Haven't listened to any real exotic types, though.
Last night I worked up a ROM price on a system for a buddy and your $50 is about right for a two way 3rd order passive XO using Dayton caps, Jantzen coils (air core except for BSC). Add another $10-15 for tweeter padding resistors and woofer zoebel. More exotic parts raise that cost.
If your woofer is less than 8 ohms, I strongly suggest either using an LM4780 with both channels in parallel or probably better, a pair of 3886s in parallel. You don't want the chips to get anywhere near kicking in SPIKE protection. The 3886 is big enough to build an amp dead bug style. The 4780 is doable, but much harder, making boards like Brian's a bargain.
** disclaimer ** I have no connection with Brian, but have heard good things about his boards, kits and service here.
Hi,
here's my two pence worth since I do not agree entirely with the earlier postings.
Run one stereo amp at each mid /treble speaker.
Either keep the passive crossover and use the two channels in bi-amp mode. OR build the mid treble as an active set up with active crossovers inside the power amp. Run two rectifiers and two sets of smoothing caps to the chip amps. The treble amp could use smaller high quality caps since the lowest frequency is determined by the high pass crossover. 200uF to 400uF may be enough and much of this could be film rather than low esr electrolytic.
Build a pair of monoblock chip amp power amps for the active bass speakers. Build in lots of current delivery into both the PSU and the output stage.
A cheaper alternative would be a pair of three channel amps with three rectifiers etc.
Select speakers that have a smooth response right out to one octave beyond each crossover frequency. Much has been said about Linkwitz Reilly at least give this a try. Keep some flexibility in the crossover boards for this. You may want to experiment by using a temporary DCX2496 to develop the crossovers and equalisation and then use the Freq & Qs from DCX to build the discrete crossovers.
here's my two pence worth since I do not agree entirely with the earlier postings.
Run one stereo amp at each mid /treble speaker.
Either keep the passive crossover and use the two channels in bi-amp mode. OR build the mid treble as an active set up with active crossovers inside the power amp. Run two rectifiers and two sets of smoothing caps to the chip amps. The treble amp could use smaller high quality caps since the lowest frequency is determined by the high pass crossover. 200uF to 400uF may be enough and much of this could be film rather than low esr electrolytic.
Build a pair of monoblock chip amp power amps for the active bass speakers. Build in lots of current delivery into both the PSU and the output stage.
A cheaper alternative would be a pair of three channel amps with three rectifiers etc.
Select speakers that have a smooth response right out to one octave beyond each crossover frequency. Much has been said about Linkwitz Reilly at least give this a try. Keep some flexibility in the crossover boards for this. You may want to experiment by using a temporary DCX2496 to develop the crossovers and equalisation and then use the Freq & Qs from DCX to build the discrete crossovers.
Thanks for everyone's inputs. I've been around the options so many times now.
I've just spotted the FR125S which could be the answer, you certainly can't get simpler than no crossover.
Anyway, my original question was about crossing to the woofer. If both extend beyond the crossover point sufficiently is it as simple as a 4th order crossover, or is there more to it?
I've just spotted the FR125S which could be the answer, you certainly can't get simpler than no crossover.
Anyway, my original question was about crossing to the woofer. If both extend beyond the crossover point sufficiently is it as simple as a 4th order crossover, or is there more to it?
I fancy making spherical enclosures, but I couldn't find any conclusions in the discussions about internal design/resonances etc.
Modelling the fr125s in winISDpro shows that it does easily go down to 120Hz in a sealed enclosure. What difference does volume make for sealed enclosures, is it just bass extension and therefore no problem as I'm cutting the low end off anyway? A 10" sphere size to match my woofers will give 3or4 litres.
PS. how do you model 1sub +2woofers +2fr125s ? Do you double the power & volume & no. of drivers for the woofers and fr125s?
Modelling the fr125s in winISDpro shows that it does easily go down to 120Hz in a sealed enclosure. What difference does volume make for sealed enclosures, is it just bass extension and therefore no problem as I'm cutting the low end off anyway? A 10" sphere size to match my woofers will give 3or4 litres.
PS. how do you model 1sub +2woofers +2fr125s ? Do you double the power & volume & no. of drivers for the woofers and fr125s?
baffle step
How do I calculate the reinforcement from the rear wall? ie how much less than 6db baffle step compensation I will need.
My FR125S will be in a pseudo spherical cabinet and will be about 500mm from the the rear wall.
Or should I just build a plain crossover and measure the results before adding BSC?
How do I calculate the reinforcement from the rear wall? ie how much less than 6db baffle step compensation I will need.
My FR125S will be in a pseudo spherical cabinet and will be about 500mm from the the rear wall.
Or should I just build a plain crossover and measure the results before adding BSC?
my active crossover design - help needed
Summary: My bi-amp system will be FR125S in 6litre sealed enclosure, sitting above 10" driver in 40litre sealed box. Each driver will be driven by a chip amp. The active crossover will be in the amplifier case.
I will build a board per amp, and think I need the following:
Mid/High:
Input Buffer
Adjustable Gain
BSC
HP - 4th order at ~160Hz
Low
Input Buffer
Adjustable Gain(?)
Linkwitz Transform
LP - 4th order at ~160Hz
HP - ?order @ ?10Hz
I've not looked at BSC circuits yet, but saw mentioned somewhere that BSC can be incorporated into the input buffer (I assume that means without additional op-amps). Can I do buffer, gain and BSC all with 1 opamp?
It strikes me that the BSC and LT are doing the same thing, so are the circuits similar? can I use one circuit for both jobs?
For ease of layout on stripboard, should I be looking at single or dual op-amps?
It may be that I can answer these myself with further investigation, but any help to speed me down the right direction is appreciated.
Summary: My bi-amp system will be FR125S in 6litre sealed enclosure, sitting above 10" driver in 40litre sealed box. Each driver will be driven by a chip amp. The active crossover will be in the amplifier case.
I will build a board per amp, and think I need the following:
Mid/High:
Input Buffer
Adjustable Gain
BSC
HP - 4th order at ~160Hz
Low
Input Buffer
Adjustable Gain(?)
Linkwitz Transform
LP - 4th order at ~160Hz
HP - ?order @ ?10Hz
I've not looked at BSC circuits yet, but saw mentioned somewhere that BSC can be incorporated into the input buffer (I assume that means without additional op-amps). Can I do buffer, gain and BSC all with 1 opamp?
It strikes me that the BSC and LT are doing the same thing, so are the circuits similar? can I use one circuit for both jobs?
For ease of layout on stripboard, should I be looking at single or dual op-amps?
It may be that I can answer these myself with further investigation, but any help to speed me down the right direction is appreciated.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.