A DIY Sub Crossover -what would you like to see?

[Amphiprion]

Ok guys, I am taking it upon myself to design a generic two-way subwoofer crossover. My basic idea is to make PCB's available at cost to folks here. Now, it's my design, so I retain total authority over what's involved, but I'm soliciting suggestions anyway.

Basic idea: symmetrical, 4th order LR type. Stereo full range L/R inputs, highpass L/R outputs, mono subwoofer output. Additional 2nd order highpass rumble filter on the subwoofer that can be defeated. Also, MAYBE an incorporated Linkwitz Transform circuit, useable with or without the rumble filter to kill out unecessary <20Hz gain of the LT ckt (also defeatable). Phase delay compensation on sub out.

Places where I need help? Making this thing easily adjustable in terms of crossover frequency. I don't want folks to have to pop open the box and swap out resistors or caps to change out the xo freq. I want discrete, fraction-of-an-octave spaced selectable xo frequencies. I don't know how to do this - any suggestions are very much welcome! I don't know if I could do it with resistor arrays and relays to set resistor values in the sallen-key filters or what. I don't know what type of relays are good (low noise) for audio.

So, Discuss! Let me know what recommendations you might have. I could probably do these at Advanced Ckts for the 3 for $33 deal they have, or I could do it with their Bare Bones PCB service (though I really hate not having a solder mask and ref designators on a board).

Also, I would probably make it buildable into one of these:

http://www.hammondmfg.com/1455ptblb.htm

I've used them at work and they are nice. They also have plastic front and back plate options for those not wanting to drill aluminum end plates.

[mazurek]

http://www.linkwitzlab.com/Pluto/woofer-asp.htm

[chasw98]

How will you power the device? And while you are at it put in balanced output for the sub so that it will feed the Behringer line of pro devices that people use for equalization. Have you thought about adding 2 sections of parametric EQ to the <20 Hz region? Maybe something along the lines of this?




Chuck

[Amphiprion]

Power will be a small onboard linear regulated supply, a small Amveco PCB mount toroid with some LM317 regs for +/-15V. I'd prefer not to have any parametric EQ on board, I personally don't think it's of much value <20Hz and even though lots want it due to the limits of the Behringer equipment it would complicate things some. I had thought of a simple -10dBV/+4dBU switch with RCA and 1/4" TRS outputs, but hadn't considered a true balanced connection, considering the runs to some sub amps that might be a good addition. Are there converter chips that handle that or can I find some simple opamp ckt to do it?

[Amphiprion]

Oh, and are people OK with Molex connectors or prefer Phoenix screw-down types?

[cjd]

Ok guys, I am taking it upon myself to design a generic two-way subwoofer crossover. My basic idea is to make PCB's available at cost to folks here. Now, it's my design, so I retain total authority over what's involved, but I'm soliciting suggestions anyway.

Normal, etc. Sounds like a plan.

Basic idea: symmetrical, 4th order LR type. Stereo full range L/R inputs, highpass L/R outputs, mono subwoofer output. Additional 2nd order highpass rumble filter on the subwoofer that can be defeated. Also, MAYBE an incorporated Linkwitz Transform circuit, useable with or without the rumble filter to kill out unecessary <20Hz gain of the LT ckt (also defeatable). Phase delay compensation on sub out.

Symmetrical slopes are quite limiting. I'd like to see at the very least the ability to mix and match 2nd order/4th order (so I can do the Dolby thing and combine 2nd order with a sealed enclosure main and have a net 4th order response on my mains to match a 4th order slope on the sub(s). If you want to make it gravy, let me choose knee point on each leg. Perhaps 2 2nd order on the high pass, 3 2nd order on the low pass (that is how active filtering cascades, right? Or is my memory of these things totally gone?). With the option of assembling any combination of such filters as wanted.

I'd like to see stereo OR mono output options. And an optional LFE sum would be excellent. Mains signal in, LFE signal in, cross mains, send low frequency mains signal + full LFE signal (i.e. all the way p to 200Hz) to sub(s).

LT is good. Very good.

I'd like to see ANY rumble filter be variable, at least resistor options to make it so. 20Hz doesn't cut it for me. 10Hz is probably ok, though 5-8Hz is preferable. Yes, I know the risks. This is DIY though.

I would like to see some filter options to give at least some ability to deal with room nodes. 2 or 3 probably. Yeah, that's parametric EQ in all liklihood.

And balanced signal options would be great. There are some relatively simple options for this. Twisted Pear has some great response on the simple little circuit they have using the THS4131 (differential driver) along with the LM4562MA (op-amp).

Places where I need help? Making this thing easily adjustable in terms of crossover frequency. I don't want folks to have to pop open the box and swap out resistors or caps to change out the xo freq. I want discrete, fraction-of-an-octave spaced selectable xo frequencies. I don't know how to do this - any suggestions are very much welcome! I don't know if I could do it with resistor arrays and relays to set resistor values in the sallen-key filters or what. I don't know what type of relays are good (low noise) for audio.

I can probably dig in again. I would love such a flexible circuit. Jon likely has the answers tucked away just outside of his normal brain functions like breathing and gawking at shapely women.

What PCB design software would you be using? I have Eagle (with only partial luck using it) and the PCB Express software (I can do some pretty fun stuff with this, and their service isn't bad - maybe a bit more costly than ideal but whatever). Would be happy to help out if I can.

edit: connector-wise, I like having a choice to hard-wire vs. using any kind of connector. So as long as that's there... The Phoenix stuff makes it easy though.

C

[chasw98]

Power will be a small onboard linear regulated supply, a small Amveco PCB mount toroid with some LM317 regs for +/-15V.

PM me with some more info about this please. Thanks.

I'd prefer not to have any parametric EQ on board, I personally don't think it's of much value <20Hz and even though lots want it due to the limits of the Behringer equipment it would complicate things some.

This I understand completely but I am in the process of breadboarding and testing it as we speak just for fun.

I had thought of a simple -10dBV/+4dBU switch with RCA and 1/4" TRS outputs, but hadn't considered a true balanced connection, considering the runs to some sub amps that might be a good addition. Are there converter chips that handle that or can I find some simple opamp ckt to do it?

Yes, see the attached PDF. It is what used in the box above . Works very well with only 4 components, 2 caps and 2 resistors.

drv134.pdf

This will allow you to offer both -10 and +4 at the same time with no switch needed. TRS is not the best idea. A TRS connection was meant to be plugged and unplugged often. An XLR connection is meant to be connected and left over a longer period of time compared to a TRS.

Chuck

[Amphiprion]

I have Mentor Graphics PADS available to me via work (otherwise is like a 3K license). It's not my favorite (the autorouter sometimes does stupid things) but it's what we started with years ago and have stuck with, no real reason to switch. I'm also a lot better with it than Eagle, and we have tons of pre-defined parts libraries that make it easy for me to use. Plus the boss encourages me to use work stuff for hobby stuff, so no getting in trouble for that

I've got to run to catch an afternoon poker game, but will post back with more comments tonight.

[cobbpa]

I definitely don't have the knowledge of Chris or Chuck, so take my thoughts as a sort of outsider's perspective. (and the words of a college student doing anything to avoid studying)

I would definitely enjoy having a unit like this. In my thoughts, I could probably determine where I wanted this box to start cutting out LFE before building. Or at least, test different components during the build to see what seemed to balance sub protection vs. LFE grunt best. After that, I probably wouldn't vary it until my room or subwoofer changed. But for those with different scenarios, on the fly adjustability might be handy.

cjd
let me choose knee point on each leg. Perhaps 2 2nd order on the high pass, 3 2nd order on the low pass
--
Mains signal in, LFE signal in, cross mains, send low frequency mains signal + full LFE signal (i.e. all the way p to 200Hz) to sub(s).

I read these ideas and thought, "oh how cool!" Just seconding these, really.

My concerns, and again these may be stupid questions: If I don't want to build 100% of the unit, is it possible to do it partially? I.e., if your schematic includes an LT circuit, could I choose not to include that? Or if there was some input / output I won't use, would I still have to buy & include those parts? Another question that popped in my head was, "If I had this, what functions would I have my receiver handle?" I would probably have it do a lot, but some with more advanced equipment may only want this for its rumble filter. I guess this goes back to building options & overall design goals.

I'm thinking that when I get around to building a subwoofer this summer, I'll have to include this in the project. Sounds like a great idea :T Now I'll step back and let those who know provide more worthwhile input, but this is definitely a thread I'll keep watching!

[cjd]

Yes, see the attached PDF. It is what used in the box above . Works very well with only 4 components, 2 caps and 2 resistors.

DRV134 works but isn't the best sounding device out there.

[Jonathan DA]

こんにちはHayenga サン

I don't think I got that right, but it was worth a try

Anyway, you know I'll build one of these regardless, but I'll second the comments about offering stereo output as well as summed mono output. And I'd like the choice to have 2nd order highpass or 4th order highpass.

Any particular reason behind the onboard power supply instead of a wall wart?
EDIT: Nevermind, it's for your +/-15v rails.

[chasw98]

DRV134 works but isn't the best sounding device out there.

In the range of <80 Hz, I have found the specs to be very acceptable. I have isolated it and measured, S/N ratio and distortion. For full range I would definitely build a better circuit along the lines of a Neve output circuit that would give gain up to +24 db. :W Plus it is so easy to incorporate and doesn't draw much current.

Chuck

[cjd]

Wall warts suck anyhow. Miserable to fit into plugs or power strips or whatever, though some are better than others now about allowing room.

C

[dlneubec]

Hi Mark,

I think it's a great idea! :T I can't help, but can give an opinion.

I'd like to see the LT circuit included for sealed sub users.

I'd also suggest that the sub output be stereo or mono, in case you are running stereo subs. This is the type of circuit John K designed and built for my NaO Mini's to the stereo subs that I used as bass bins (and stands) for the 2-way MTM's. I will need exactly this type of crossover for my Mentor Omni's when I finally pull the sub drivers, amps and active crossover to make the NaO's opreational again.

I'd be happy to test anything you come up with and might be your first customer!

[cjd]

Chuck, good point about frequency range. But as an example, a 2nd order slope will mean these are still present at 360Hz

C

[chasw98]

Wall warts suck anyhow. Miserable to fit into plugs or power strips or whatever, though some are better than others now about allowing room.

C

It is very hard to find a bipolar wall wart. The problem I have found when trying to integrate +-15v supplies into projects is that the cost of the transformers drives the price up. You can spend as much on the transformer as you can on all the other parts combined. Many times it is easier to tap off an existing piece of equipment and use that unit's supply to power the project.This supply is one of the best and most affordable around. It is excellent for testing. I am trying to come up with a toroid based supply that will fit in a 1 unit 19 inch rack mount case for use with little projects like this.

[ThomasW]

Although we've been waiting for Jon to do the PCB layout for the box shown on Chuck's first post, I suppose it's time to take the lid off this project since it has some of the things Mark is talking about doing.

So drum roll please ........ introducing the H@LF box

[Habs4life]

Places where I need help? Making this thing easily adjustable in terms of crossover frequency. I don't want folks to have to pop open the box and swap out resistors or caps to change out the xo freq. I want discrete, fraction-of-an-octave spaced selectable xo frequencies.

For the low pass adjustment you could use a pair of high quality dual ganged 50k? linear pots or you could have fixed resistors switched by a multi position rotary four gang switch or two separate dual ganged switches.

Having the two switches or pots would give the added flexability of setting each 12db LPF at a different frequency which may aid in the blending of the subs to the mains.

Ive built several projects into those Hammond cases and they are perfect for the purpose and look good.

[chasw98]

Although we've been waiting for Jon to do the PCB layout for the box shown on Chuck's first post, I suppose it's time to take the lid off this project since it has some of the things Mark is talking about doing.

So drum roll please ........ introducing the H@LF box

I started to build this originally as a High Pass filter for a sonotube when I found out you cannot get HP filters at low frequencies very easily. After that I decided I wanted more of a selection of options to augment a BFD. I had read about a lot of people spending $50 to $75 on a clean box just to up the level from -10 unbalanced to +4 balanced to feed a pro amp. I also decided that phase control could be a handy tool to have, especially when calibrating a sub with REW and trying to get it to integrate with main speakers in a troublesome location in a room.

Thomas and I started going back and forth about what to put in it and how to construct it. In the end you will be able to use all parts of the circuit or go 'a la carte' for whatever section you need. You will be able to mount it inside an existing piece of equipment that has +-15 volts available or put it in a case of your choice. You will be able to set it up by using jumpers or by using switches and pots mounted in an enclosure. We expect the cost for all parts and the PCB to be less than $50.00. This does not include enclosure or power supply.

The HP filter is adjustable from 8 HZ to 20 Hz in 2 Hz increments. It is set at 24 db per octave (4th order LR). The HP filter also has adjustable gain at the filter point. The phase is adjustable from 0 - 180 degrees with a 180 degree switch that gives you a full range of 360 degrees. The output includes both
-10 RCA out and +4 db XLR balanced output.

In the prototype all 1% metal film resistors, 1% poly capacitors, and OPA series op amps have been used. In addition gold plated DIP sockets were incorporated. There will be a pictorial on how to assemble it for the 'electronically' challenged. As you can see the web page is still under construction, but I have had 2 of the units in my system at my house working well for over a month now.

Chuck

[digital desire]

.This supply is one of the best and most affordable around. It is excellent for testing. I.

That is an awesome price for that. I was about to suggest the kit available from marchand.com, which is perfect little supply, and easy kit, but 3X the price.
Actually, marchand has some great crossover electronics (and the basis! :T ), very well engineered kits.

[BobEllis]

For the low pass adjustment you could use a pair of high quality dual ganged 50k? linear pots or you could have fixed resistors switched by a multi position rotary four gang switch or two separate dual ganged switches.

Having the two switches or pots would give the added flexability of setting each 12db LPF at a different frequency which may aid in the blending of the subs to the mains.

Ive built several projects into those Hammond cases and they are perfect for the purpose and look good.

Ganged pots generally do not track well enough for use in the typical Sallen-Key filter topology. Multi-feedback filters are less sensitive to component variations, but are slightly more complex to lay out and would require an additional variable resistance.

The best solution in the analog domain is to use a multi pole switch to either select the appropriate resistor or shunt out a part of a resistor string. Separate switches for each pole would give maximum flexibility at increased cost, front panel space and longer leads possibly leading to stray inductance/capacitance problems.

Another solution would be sockets for changing either individual resistors or groups. Once I get my XOs dialed in I don't mess with them. OK, so they live with their tops not screwed on for a few months, but I don't have SAF to worry about.

I suggest using Sallen-Key equal component value topology. This will give you a larger pool of parts when matching capacitors (1-2% for accurately building what you calculate) Another advantage is that you can vary the Q of each pair of poles - so your rumble filter can also be your EQ. That's not as good as a Linkwitz Transform in the group delay department, but I am happy with it.

Not to put a damper on Mark's project, but there's a board that already meets most of what is being discussed here. Over at diyaudio, Jens Rasmussen designed an XO board and PSU for a series of group buys at that I ran. Each 80x100 mm board has an input buffer, shelving low pass (BSC) HI/LO or bandpass 2nd or 4th order filters, two peak/notch EQ sections, a Linkwitz Transform, a first order all-pass section and buffered/variable attenuation outputs. Outputs are single ended, so those seeking balanced outs would need to add a small balanced driver board. Two required for stereo and mono/stereo sub. If you want a rumble filter on the sub channel(s) you'll need to add another board per channel.

The circuit is set up so that you can vary the Q (low end boost) without affecting the overall gain if you replace one of the Q setting resistors with a pot. In this case ganged pots probably track closely enough.

Advanced Circuits has the gerber files, tied to my account. I don't have the time to run group buys anymore but if someone would do the order taking and shipping, I could buy the boards. It would help if that someone was a DIYaudio member, since there are quite a few people there looking for both filter boards and PSUs. Some at the Orion owners group have used the psu and are very pleased with the results.

More info on both filter and psu boards can be found here: http://mysite.verizon.net/vze3xvxs/audio/id20.html I AM NOT SOLICITING ORDERS. I have a filter calculation spreadsheet that may be useful posted there as well as a guide to using the filter boards with some design examples.

The pricing was $5-6 per board if ordering 100 pieces last time. I don't know if that has changed. The four rounds of group buys totaled over 1200 filter boards and 500 psu boards. The only problems seem to have been people who liked the basic idea but tried to change it to something slightly different. Super fast opamps like the LM6172 need careful attention to bypassing, but the OPA2134 and slower are fine (and sound good to my ears)

I might be able to convince my son to pack up PSU kits if there is enough interest to justify a group parts buy. He did a good job on it before.

[Amphiprion]

Symmetrical slopes are quite limiting. I'd like to see at the very least the ability to mix and match 2nd order/4th order (so I can do the Dolby thing and combine 2nd order with a sealed enclosure main and have a net 4th order response on my mains to match a 4th order slope on the sub(s). If you want to make it gravy, let me choose knee point on each leg. Perhaps 2 2nd order on the high pass, 3 2nd order on the low pass (that is how active filtering cascades, right? Or is my memory of these things totally gone?). With the option of assembling any combination of such filters as wanted.

This is all do-able. Filter block sections will be Sallen-Key equal component types, but you can pop in any R and C you want in each 2nd order filter section and get whatever 2nd order response you want out of it, and cascade that with another section. Adding a third section for 6th order, maybe, I'm not convinced of the usefulness of 6th order filters for subwoofers unless crossing >80 or 100 Hz. I think you are right to include the 4th/2nd order combo filter option and will incorporate that. I will think about the 6th order idea.

I'd like to see stereo OR mono output options. And an optional LFE sum would be excellent. Mains signal in, LFE signal in, cross mains, send low frequency mains signal + full LFE signal (i.e. all the way p to 200Hz) to sub(s).

Stereo and mono outputs are easy easy to do and I will include. It will take some convincing for me to do the LFE sum. Most HT receivers will already do this and it seems like mostly and HT application type thing, that and I don't really understand the reasoning behind it. For me subs are to alleviate the bass duties of full range speakers, even good 3-ways can have some serious distortion issues <50Hz (and two ways really start to suck wind <100Hz).

I'd like to see ANY rumble filter be variable, at least resistor options to make it so. 20Hz doesn't cut it for me. 10Hz is probably ok, though 5-8Hz is preferable. Yes, I know the risks. This is DIY though.

All doable. I can make it adjustable.

I would like to see some filter options to give at least some ability to deal with room nodes. 2 or 3 probably. Yeah, that's parametric EQ in all liklihood.

I'm going to nix the idea of any PEQ on this board, I think that is much much better handled by things like the DSP1124P and a few bands on this board, while maybe correcting a major problem or two, doesn't do near what one of those little $100 Behringers can do.

And balanced signal options would be great. There are some relatively simple options for this. Twisted Pear has some great response on the simple little circuit they have using the THS4131 (differential driver) along with the LM4562MA (op-amp).

The availability of easy to use little DIP balanced line drivers will make this easy to include. Consider it included.

edit: connector-wise, I like having a choice to hard-wire vs. using any kind of connector. So as long as that's there... The Phoenix stuff makes it easy though.

We use the Phoenix stuff at work. The only thing I don't like about them is on the 2 and 3 terminal blocks, when tightening down the screws, they tend to rotate (the plastic bodies that is, not the solder joints ). Does anyone have any alternative similar screw down terminal blocks that they like? I've seen some blue ones and grey ones used on different projects but I don't know brands or if they are quality.

PM me with some more info about this please. Thanks.

This Q was about the power supply. I'll go ahead and post it here since others will be interested. If you look at the PS that Marchand offers, it will basically be a smaller unit than that. A ~16 dollar PCB mount Amveco 30VCT, some LM317 regs (SOT-223 if I can get away with them, otherwise TO-220 with some cheap clip on heatsinks). The Amveco's are at Digikey, just do a search for toroidal transformers, then in the selection tool choose PCB and Amveco. They are the 70000 series PCB mount toroids, and look pretty decent. The smaller units would fit in a 1RU chassis. I posted a PS schematic here a while ago, it would be very similar but a little simplified (might eliminate the adjustability pots and just use a single resistor instead, and my bypassing was way overkill - will just use some 1206 SMT .1uf NP0 ceramics). It will have a bridge rectifier probably using 1N4000 series diodes, 1N4002 at a minumum. There are some nice Schottky rectifier diodes with lower forward voltage drop but PIV seems to be much lower, probably too low for this application.

This I understand completely but I am in the process of breadboarding and testing it as we speak just for fun.

Some people definitely want it, I will consider adding it just because the Behringers won't do it - but I doubt I will. If anything, your box will become vastly more popular

I'm breaking up this reply into two posts, so please wait for me to respond again before posting any comments.

[Amphiprion]

My concerns, and again these may be stupid questions: If I don't want to build 100% of the unit, is it possible to do it partially? I.e., if your schematic includes an LT circuit, could I choose not to include that? Or if there was some input / output I won't use, would I still have to buy & include those parts? Another question that popped in my head was, "If I had this, what functions would I have my receiver handle?" I would probably have it do a lot, but some with more advanced equipment may only want this for its rumble filter. I guess this goes back to building options & overall design goals.

Yes, no, no, and yeah you'll have to figure that out If someone only wants a rumbly filter they would be way better off with other options. Even just a protoboard, a pair of 9V recharchables, and a simple op-amp ckt. That would be way cheaper.

I'm thinking that when I get around to building a subwoofer this summer, I'll have to include this in the project. Sounds like a great idea Now I'll step back and let those who know provide more worthwhile input, but this is definitely a thread I'll keep watching!

Timelines are something I can't promise - I never make them unless my salary or job is at stake This is just a weekend project for me. I can post schematics as I develop the ckt and get feedback and error-checking, that would certainly help and would ease the initial prototyping and debugging that I'll have to do before releasing the final design. Late summer may be a possibilty depending on my schedule, but again no commitments here on timing. On the plus side, schematic and PCB design is probably my favorite part of my job, so I'm likely to work on it even if it gets frustrating.

I don't think I got that right, but it was worth a try

Pretty good, but I would prefer to have seen it in kanji Also I think it is supposed to be wa (character) instead of ha (pronounced as wa) but honestly I can't remember.

Have you gotten to use my two most important phrases while over there? Remember, it's yooshiki no toire wa doko desu ka? and and fashion hoteru wa doko desuka? (not that I've ever been or that you should go with Jennifer stateside ). But still, renting one for a brief period might just be worth it to see all the crazy shit they have. A discrete concierge might be able to help you out, or might kick you out of your hotel, so I dunno how to handle finding one.

Anyway, you know I'll build one of these regardless, but I'll second the comments about offering stereo output as well as summed mono output. And I'd like the choice to have 2nd order highpass or 4th order highpass.

Yep, will be done.

Any particular reason behind the onboard power supply instead of a wall wart?

Getting +/-15VDC out of a wall wart isn't hard, either just an AC wallwart with everthing else on board, or a simple DC wallwart with a switching DC/DC converter on board. TDK makes some decent ones for 5V applications but I don't particularly like the ripple specs on the dual +/- 15V supply ones. I get about 100mV p-p ripple with them, which I could just hit with a linear regulator, but that seems kind of silly. Besides, people building their own PS is educational if a bit more expensive. And wall-warts? Jeez, I thought you were high-end

In the range of <80 Hz, I have found the specs to be very acceptable. I have isolated it and measured, S/N ratio and distortion. For full range I would definitely build a better circuit along the lines of a Neve output circuit that would give gain up to +24 db. Plus it is so easy to incorporate and doesn't draw much current.

I like the ease of use of the DRV chip, and if you say it's good < 80Hz that's good enough for me. Of course I am also open to other solutions. I might consult you for app info on the DRV chip.

I'd be happy to test anything you come up with and might be your first customer!

You really want to be a guinea pig? First run boards always have something wrong with then, need jumper wires, crap like that At least at work they do, since we're on such tight time schedules and things just have to get made to get tested (and a few jumpers in a proto to correct errors is no biggie). But being a tester on this would cost you some significant additional cash. Still, if you are up for it, I would be happy to include you :T

For the low pass adjustment you could use a pair of high quality dual ganged 50k? linear pots or you could have fixed resistors switched by a multi position rotary four gang switch or two separate dual ganged switches.

Me no like ganged pots. 1% metal films all the way, switched.

Ive built several projects into those Hammond cases and they are perfect for the purpose and look good

I agree, those Hammond extruded aluminum boxes are *awesome* for the price. I used one at work to house a PC/104 single board computer and a power supply and was amazed at what I got for $25.

The best solution in the analog domain is to use a multi pole switch to either select the appropriate resistor or shunt out a part of a resistor string. Separate switches for each pole would give maximum flexibility at increased cost, front panel space and longer leads possibly leading to stray inductance/capacitance problems.

What I was considering was a set of resistors and a DIP switch (are those acceptable for analog signal passing?). The DIP would be connected such that binary inputs on 3 pins would allow you to select 8 resistances by connecting/disconnecting resistors in parallel while letting the unused ones float disconnected at one end. A 6 position DIP switch should handle a single 2nd order filter block section, or I could use an 8 position switch to allow for bypass/buffer through a single 2nd order section as well. If any of this makes sense.

Another solution would be sockets for changing either individual resistors or groups. Once I get my XOs dialed in I don't mess with them. OK, so they live with their tops not screwed on for a few months, but I don't have SAF to worry about.

Me no likey the socket idea, I'd prefer something someone can just throw a couple of DIP switches on and get a bunch of different xo frequencies. Again if that idea sounds like an OK one.

To everyone that has mentioned it:

Yeah, I know there are probably a lot of different options and existing boards out there for this type of thing - I'm just trying to build something I've wanted to design/build and throw in some options that others would find useful that wouldn't be too hard to add on (and may in fact improve the useability for myself). I don't expect it to be cost-competitive with others, just well-engineered, well documented, and made in such a way that laypeople and new types can understand how each part works. Other projects will have other features mine doesn't, and mine will have features theirs don't, but it's all good in the end. It's the whole make some people happy some of the time type thing and whatnot. Having this a project done at a small board like this inceases the likelihood of support, ease of communication, tailoring to what most/at least some want, and things like that.

Again, if I've nixed or expressed doubt about any of your ideas, PLEASE feel free to try and persuade me - I might even nix it for now in the initial stages but add it in at a later time, so your efforts may not be in vain.

OK, that's it for now, ready for the next onslaught of comments

[cjd]

Ok. I can respond now I think! :P

If I am going to get a Behringer (for EQ), I'm just going to get a DCX and forget about this. The point is, most applications need 2 or 3 nodes of EQ and that's usually sufficient. I want a one-stop solution. I don't want to do this and then add something else. I don't even know what my signal source might be - perhaps trying analog output from the HTPC. It's doubtful I'll ever jump to a pre/pro. Yeah, the computer has some ability to do this signal management too sometimes. But I've found it is finicky. And I'm trying to move away from Windoze entirely which makes it trickier. Plus I'd love to be able to add a sub into my main setup. Which would require this handle balanced signal input not to mention output.

And balanced output... you planning on only offering balanced out for the sub? Otherwise, my concerns re: DRV134 remain.

I may consider the concept that the option of testing with pots, measuring resulting values, and soldering in the correct value resistors. Yeah, a little headache. Of course, I think there would be nothing to prevent me from doing that no matter what solution you go with.

C

[Amphiprion]

Mkay, you have me considering PEQ. Any schematics you could point me to?

[chasw98]

Mkay, you have me considering PEQ. Any schematics you could point me to?

Try looking at the pe17sch.pdf . They offer response down to 10 Hz.

[Habs4life]

Mkay, you have me considering PEQ. Any schematics you could point me to?

he may not want his circuit copied but for reference purposes there is this.http://sound.westhost.com/project28.htm
It is not as functional as a full PEQ but it should be effective enough for low frequency use and not take up as much PCB realestate and front panel space as a normal 3 band PEQ.It could be modded so as all the controls affect the same 35-150hz range.

[Amphiprion]

That looks simpler than anything I've found (in the Rane PEQ app note) (ETA: not the link just posted above, I haven't reviewed that one yet). I could easily adapt that. Would anyone mind me not giving them front panel controls, and just using multiturn cermet trim pots inside? PEQ seems to be something someone sets once for a given room and sub location and then doesn't want anyone fiddling with the knobs on again in my mind. Keeps the front panel looking cleaner too And I guess if someone really wanted panel mount adjustability, they could always just wire their panel mount pots to the pads meant for the trim pots...

[Jonathan DA]

Jeez, I thought you were high-end

Oh yeah, what was I thinking? With an onboard power supply I can go buy one of those $1000 power cables, plug it into your xo and make it sound waaay better than you ever could with your fancy engineering. I scoff in your general direction!

Sign me up as a guinea pig too. I ain't afraid of no jumper wires!

Count one more vote for PEQ too!

[Amphiprion]

Hey J,

PM me your new house line, or your SO's cell. The 670 number ain't workin no more.

[Amphiprion]

Man, that Zaph discussion sure killed off this thread! Anyone with anymore thoughts before I start designing this weekend?

[kingpin]

Anyone with anymore thoughts before I start designing this weekend?

Made this real quick.

Image not available

Mike

[Amphiprion]

So you guys want a manual?

[kingpin]

So you guys want a manual?

No no Mark.

I don't think anyone would expect you to write a manual for something this specific or advanced. I don't think this is something for the novices.

I just like to take potshots at you guys(in a joking matter) who are clearly at a different level when it comes to this hobby. :T :T

Call it a jealously that resides in me 'cause for the life of me I can't understand what you guys are talking about.

I think what you are doing is an extremely nice gesture to the folks at this forum. ;x(

Mike

[Habs4life]

Mark, maybe I missed it but what are your plans for phase adjustment,variable 0-180,or 0-360 or a just a 0-180 switch?

[Amphiprion]

I'm thinking about that. Very hard, actually. Read Rod Elliot's article here. I'm torn between 0/180, variable 0-180, or an outright delay. An outright delay would require me doing some digital work on the board, which I would like to avoid. But I'm considering it.

[BobEllis]

The manual I wrote for the boards I mentioned earlier might offer a good start for the basics. It covers most of what Mark plans to offer and is written for "dummies with a bit of a clue." The spreadsheet allows you to skip the math completely. http://mysite.verizon.net/vze3xvxs/audio/id20.html

[ThomasW]

is written for "dummies with a bit of a clue."

isn't that an oxymoron.... :B

The manual is a very well written document for the novice... :T

[kingpin]

The manual I wrote for the boards I mentioned earlier might offer a good start for the basics. It covers most of what Mark plans to offer and is written for "dummies with a bit of a clue." The spreadsheet allows you to skip the math completely. http://mysite.verizon.net/vze3xvxs/audio/id20.html

Thanks Mr. Ellis, I will give it a shot. :T

Mike

[Habs4life]

Here is a couple of projects using the nice little Hammond cases.The black one contains the xover and PS boards that Bob mentioned.

[Amphiprion]

Hey Habs,

Any problem with putting an analog signal through a DIP switch or rotary DIP switch? (contact corrosion, etc). I notice they list some as 'sealed'....

[chasw98]

I'm thinking about that. Very hard, actually. Read Rod Elliot's article here. I'm torn between 0/180, variable 0-180, or an outright delay. An outright delay would require me doing some digital work on the board, which I would like to avoid. But I'm considering it.

I used that circuit as a starting point in the project I built. I had to modify part of it to get a full 360 degrees of shift. In order to get a full 360 degrees a 0/180 switch must be used. I also found that you must test it at all frequencies to see the effect compared to the 'non phase adjusted' signal. Where you place it in the circuit can have different effects also. That is why I built it as a seperate module so that it could be used independently or placed in the signal chain at a different point. When I actually put it in my system at home, I found it helped relieve a null that was not there when measuring the sub alone, but did appear when I measured the sub and the rest of the system. I was able to dial in the phase to eliminate the null.

[cjd]

So, analog signal through a switch. Yes, there can be issues over time - how often do we find switches that crackle and pop in older equipment, if not worse? You can also pick up "dirt" just soldering stuff together. Lots of things. That doesn't mean you shouldn't use them.

If you're looking at a (relatively) small number of connections, consider a jumper block.

And, I think variable phase (0-180 + 0/180 switch) would really be superb.

I should go research this stuff myself, start getting an understanding of what we need to put together, and maybe start tossing some ideas down as far as PCB layout. I like puzzles, and PCB layout is a puzzle. So much to do, so little time.

C

[Amphiprion]

The ones I'm looking at are gold plated and sealed, and a bit more expensive than standard DIP's (they are more than twice the cost in single units). They're quite nice, I have a few here, I will probably take a dremel to one and see what they are like inside. But they are sealed from dust and I want to make sure the internals are plated gold and not just the soldering pins.

The Digikey P/N is 563-1002-5-ND for anyone wanting to look them up.

ETA: Actually they are not sealed, bummer, b/c they are nice. Anyway, the whole reason for this DIP switch thing will soon become apparent. It's kind of nifty actually.

[Habs4life]

Those DIP switches should be ok but I would prefer they only be in the LP section.The less switch contacts in the full range signal path the better.
(I assume it is for frequency selection).
If you want to do a full 0-360 variable phase adjustment I can provide a schematic.