Hi
Thanks Jacco, but making a dual layer PCB is relaxing comparede to the 12 layer stuff I do at work
Nice one, can you add some more comments?
Have you reviewed the pcb files yet?
\Jens
jacco vermeulen said:Jens,
please stay healthy
Looks like there is room for a single to dual opamp doubledecker on the board, very nice. (i happen to have a tube of filthy decadent single opamps lying here.)
Thanks Jacco, but making a dual layer PCB is relaxing comparede to the 12 layer stuff I do at work
BobEllis said:
Nice one, can you add some more comments?
Have you reviewed the pcb files yet?
\Jens
Jens, You're the man. Could I get the gerbers for filter 4 to do a group buy (assuming enough interest)? I'd give up the regulators for an all pass section or two 
Here is an updated version of the spreadsheet. it includes more instruction, shading to indicate entered and calculated values and schematics pulled from Jens' manuals and (Linkwitz for the all pass) with credits given.
Edit: Preliminary interest check before starting a GB thread. Would potential buyers prefer the most recent version or would you like to add an all pass section?
Thinking out loud, would jumpers between each 2nd order section be worthwhile? If you only need 2nd order electrical to achieve the desired response you could do it without cutting traces or bending pins
Here is an updated version of the spreadsheet. it includes more instruction, shading to indicate entered and calculated values and schematics pulled from Jens' manuals and (Linkwitz for the all pass) with credits given.
Edit: Preliminary interest check before starting a GB thread. Would potential buyers prefer the most recent version or would you like to add an all pass section?
Thinking out loud, would jumpers between each 2nd order section be worthwhile? If you only need 2nd order electrical to achieve the desired response you could do it without cutting traces or bending pins
Forgive me for being dense, but when I look at the schematic this looks like a bandpass-filter with two EQ's? Wouldn't the board be more versatile if it could be configured as a highpass + EQ and lowpass + EQ as well?
If I could do a two-way, 4th-order LR-filter on a single board I know I'd buy a pile of them
/U.
If I could do a two-way, 4th-order LR-filter on a single board I know I'd buy a pile of them
/U.
That is the current state of affairs. If we add jumpers at the output at each stage it can be configurable as 4th order LP/HP/BP or 2 x 2nd order HP/LP/BP with the EQ's on any section. My goal is something like a 2 section MOXLite with the addition of all pass.
Adding jumpers might take up the space Jacco was looking at for his single-dual op amp boards, though.
EDIT: Interest check WIKI here: http://www.diyaudio.com/wiki/index.php?page=Active+Filter+Boards+
Adding jumpers might take up the space Jacco was looking at for his single-dual op amp boards, though.
EDIT: Interest check WIKI here: http://www.diyaudio.com/wiki/index.php?page=Active+Filter+Boards+
BobEllis said:
Hi, ever since I read chipco3434 excitement w/ his mox I have been doing a crash learning course on active crossovers and MOX to see if these boards would be a great way to go for a future project. I learned (I hope) that the earlier boards ‘mox’ were for prototyping and using w/ the advantage of changing values at will and ‘Lite’ which were boards used after the mox prototyping or straight after calculations (no prototyping) for a smaller footprint.
I was wondering if these new boards are configurable w/ jumpers to provide multiple combinations of whatever your needs. For example, would I be able to configure a two or three way/ 2nd order with an additional third order filter for subwoofer with these new boards? Or should I stay and explore the original mox boards which I believe are not obtainable anymore.
Are these the new generation configuerable 'Lites' for multiple combinations or is this a new 4th order filter capability only to use with the previous boards?
I'm confused as usual but interested in the GB if configuerable to multiple needs.
Thanx - Stan
Edit:
Sorry, this should have been posted in the new thread...
Hi guys:
Here is my two cents.
First off, I'm glad that the MOX still has quite a following.
I don't kow the extent of your skills, abilities and experience in designing active systems but I think you should bear in mind that the theoretically calculated parameters of activer crossovers will rarely correspond to the actual requirements of your individual systems. As I'm sure you all know, the speaker response is in no way ideal, which renders the ideal-case calculated crossovers rather useless. Crossover values should be decided in situ. So, if the MOX with all its discrete amps and options is too complex, the MOX Lite is IMO a bit too simplistic and inflexible. If anything, you should be able to switch the crossover frequency, Q factor and the gain quickly and easily. In practice, this translates as only a couple more resistors and caps in the new design for a small additional cost to you.
Speaking from my own experience, I believe one should always test a few options in order to be able to choose the best-sounding one in the context of one's own system. Now that you are considering a new design, why not consider a 'medium' version that would allow you to fine tune the system? For those of you who think this might be a good idea, one such Mid-MOX design is to be found on my website.
Regards,
Milan
Here is my two cents.
First off, I'm glad that the MOX still has quite a following.
I don't kow the extent of your skills, abilities and experience in designing active systems but I think you should bear in mind that the theoretically calculated parameters of activer crossovers will rarely correspond to the actual requirements of your individual systems. As I'm sure you all know, the speaker response is in no way ideal, which renders the ideal-case calculated crossovers rather useless. Crossover values should be decided in situ. So, if the MOX with all its discrete amps and options is too complex, the MOX Lite is IMO a bit too simplistic and inflexible. If anything, you should be able to switch the crossover frequency, Q factor and the gain quickly and easily. In practice, this translates as only a couple more resistors and caps in the new design for a small additional cost to you.
Speaking from my own experience, I believe one should always test a few options in order to be able to choose the best-sounding one in the context of one's own system. Now that you are considering a new design, why not consider a 'medium' version that would allow you to fine tune the system? For those of you who think this might be a good idea, one such Mid-MOX design is to be found on my website.
Regards,
Milan
On my website, it's under the heading MOX. The website version uses IC opamps and has a much more compact design than the original MOX. I called the version Mid-MOX just for the purposes of this thread.
Anyways, I don't think you should copy my design as is but use it as perhaps a staring point/inspiration in designing an improved Mox Lite version, and as an illustration of the points I made in my previous post.
Regards,
Milan
Anyways, I don't think you should copy my design as is but use it as perhaps a staring point/inspiration in designing an improved Mox Lite version, and as an illustration of the points I made in my previous post.
Regards,
Milan
I just rx'ed a pair of MOXlites (Thanks Ed) and was looking for a BOM of sorts to populate the boards. A "generic" parts list.. ie. r1-40, c1-25 and so forth to start calculating my required components and to ensure I don't "miss" any.
I looked through the thread at high speed and did not see mention of suggested op-amps (I am not going discrete).
Would the 5532(?) work or is everybody cringing on that question? I will most likely use sockets to try various ones, but to get started I always go cheap.
Jason, if you find yours I'd still be interested in them.
I am also on the wiki for the crossovers on the "other" thread.
I looked through the thread at high speed and did not see mention of suggested op-amps (I am not going discrete).
Would the 5532(?) work or is everybody cringing on that question? I will most likely use sockets to try various ones, but to get started I always go cheap.
Jason, if you find yours I'd still be interested in them.
I am also on the wiki for the crossovers on the "other" thread.
Troy -
I've used 5532s to date. Chances are you'll be happy with them. I have OPA2134s and LM6172s for evaluation when I get my moxlites up and running. Linkwitz uses the 2134s, although I have heard that some prefer the 5532 for bass sections.
The spreadsheet I posted earlier will calculate values for you. Use your favorite 100 nf to bypass the ICs. Other than that, there's just the power supply caps, diodes and regs. I think that the negative portion of the power supply won't give -15V (I could be wrong, nobody responded when I raised it before), so I cut a trace or two and used my own which required two separate DC supplies.
The SMD parts were fun for my old eyes and fat fingers. High power magnification and tweezers required.
I've attached the schematic to save a little searching.
I've used 5532s to date. Chances are you'll be happy with them. I have OPA2134s and LM6172s for evaluation when I get my moxlites up and running. Linkwitz uses the 2134s, although I have heard that some prefer the 5532 for bass sections.
The spreadsheet I posted earlier will calculate values for you. Use your favorite 100 nf to bypass the ICs. Other than that, there's just the power supply caps, diodes and regs. I think that the negative portion of the power supply won't give -15V (I could be wrong, nobody responded when I raised it before), so I cut a trace or two and used my own which required two separate DC supplies.
The SMD parts were fun for my old eyes and fat fingers. High power magnification and tweezers required.
I've attached the schematic to save a little searching.
Hi Bob,
Some weeks after you mentioned that regulator problem I went over the PCB and found that the PCB is correct. It is the schematic that is wrong. It calls up 78L15 for both + & - supplies.
Use 7915 for the -ve regulator (U21).
I have been pretty thorough in my checking, but it would be safer if someone else can confirm.
Some weeks after you mentioned that regulator problem I went over the PCB and found that the PCB is correct. It is the schematic that is wrong. It calls up 78L15 for both + & - supplies.
Use 7915 for the -ve regulator (U21).
I have been pretty thorough in my checking, but it would be safer if someone else can confirm.
As far as I can tell, the board follows the schematic. V- connects to the reference pin of the reg and the input pin is grounded. Not quite right for a 7915 either.
I think the layout started out as a conventional 7815/7915 layout and got halfway changed to use 7815s for both positive and negative, per the app. note. (7815s have lower noise and better regulation than 7915s) Remember that Tyler had a HD crash in the middle of development.
You could swap the connections for pins 1 and 2 of U21, but since I already had 7815s I used the app note version. I lifted the input of U21 and the + side of C27, connected them together and use that as the + input for the negative supply (separate winding and bridge for both polarities). U21's Pin 3 gets grounded, and the trace leading to C26 gets cut after D21. Then connect the negative pin of C26 to U21 pin 2. Made for an ugly rat's nest on the bottom of the board, especially since I needed to use through hole type 100 nf caps on the negative side. with these changes.
Moamps - I agree that you can rarely get away with a textbook filter, although you can get closer with textbook active filters than textbook passives. However, I disagree that the MOXlite is too simplistic (except for lack of all pass) With a MOX you can find the transfer function needed, and transfer it to MOXlite. It might be easier if the topologies were both the same with resistive Q adjustment, but it can be done (in a small form factor).
Once I find the "correct" transfer function, I'm looking for something small enough to put in the amp case so I can make active XO transparent to the end user - one in and one 4 conductor speaker cable with Neutrik connectors. Moxlite fits the bill, and so will the new boards for higher order filters (assuming that we can talk Jens into doing a little more on the layout for us)
I think the layout started out as a conventional 7815/7915 layout and got halfway changed to use 7815s for both positive and negative, per the app. note. (7815s have lower noise and better regulation than 7915s) Remember that Tyler had a HD crash in the middle of development.
You could swap the connections for pins 1 and 2 of U21, but since I already had 7815s I used the app note version. I lifted the input of U21 and the + side of C27, connected them together and use that as the + input for the negative supply (separate winding and bridge for both polarities). U21's Pin 3 gets grounded, and the trace leading to C26 gets cut after D21. Then connect the negative pin of C26 to U21 pin 2. Made for an ugly rat's nest on the bottom of the board, especially since I needed to use through hole type 100 nf caps on the negative side. with these changes.
Moamps - I agree that you can rarely get away with a textbook filter, although you can get closer with textbook active filters than textbook passives. However, I disagree that the MOXlite is too simplistic (except for lack of all pass) With a MOX you can find the transfer function needed, and transfer it to MOXlite. It might be easier if the topologies were both the same with resistive Q adjustment, but it can be done (in a small form factor).
Once I find the "correct" transfer function, I'm looking for something small enough to put in the amp case so I can make active XO transparent to the end user - one in and one 4 conductor speaker cable with Neutrik connectors. Moxlite fits the bill, and so will the new boards for higher order filters (assuming that we can talk Jens into doing a little more on the layout for us)
BobEllis said:
Hi Bob:
I was under the impression that you and others were making a "stand-alone" active crossover design that would not lean on the original MOX in order for it to work.
I'm not sure if you are aware of the fact that many people here quit or never even started building the original MOX design precisely because of its perceived complexity. How can we then expect people to use the MOX to find the 'correct' transfer function if the MOX itself is a problem most of them can't get past? That's why I suggested a medium MOX version "with everything on" that would be much more user friendly and would certainly make more sense.
If you think about the operating principles of professional active speakers (e.g. Genelecs), you'll remember they have a number of switches for changing transfer functions and other parameters. That's because the 'end user' must have or wants to have a certain degree of control over the operation of his active speakers. And that's what an upgraded MOX should be all about, IMO.
Anyways, that is just my opinion.
Regards,
Milan
Too bad so many people couldn't figure out the MOX - it is one of those great tools that makes rapid protoyping fun and easy to do.
I repeat my request - anyone have a pair or two of MOX boards they want to give up? That way I won't have to breadboard them.
My active speaker targets are fairly simple - mains for a friend's HT and the same for my fiance. Both places I have access to and will tune them for the installation. If they get moved, I'll be around to re-tune them. I want the simple hookup so I don't have to reinstall if they get unhooked for cleaning. No plans to go commercial or need to satisfy pro installers.
I know what I want, hopefully the others who express interest will understand what they are getting. One way to get flexibility is to use pluggable resistor modules, like Marchand does. It wouldn't be too hard to add headers to the board. Another way would be to use pots or switched resistor arrays - although this would take a lot of front panel space and adds cost.
I repeat my request - anyone have a pair or two of MOX boards they want to give up? That way I won't have to breadboard them.
My active speaker targets are fairly simple - mains for a friend's HT and the same for my fiance. Both places I have access to and will tune them for the installation. If they get moved, I'll be around to re-tune them. I want the simple hookup so I don't have to reinstall if they get unhooked for cleaning. No plans to go commercial or need to satisfy pro installers.
I know what I want, hopefully the others who express interest will understand what they are getting. One way to get flexibility is to use pluggable resistor modules, like Marchand does. It wouldn't be too hard to add headers to the board. Another way would be to use pots or switched resistor arrays - although this would take a lot of front panel space and adds cost.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.