Active Crossover Board Group Buy

[BobEllis]

Mouser sells the Wima FKP2s, there must be another distributor on your side of the world.

I like the sound of polyprops better than polyester, not that I have done a direct comparison. I've heard that the FKP (polypropylene film and foil) sound better than the MKP (metalized polypropylene film), but I haven't used them yet.

Polystyrene and Teflon are supposedly the holy grail. Mouser has some polystyrene caps that I used in a crossover. They are cheap, have 2 mm lead spacing and are only 10% so it was buy a big batch and sort. I was surprised how tightly they grouped, mostly within 1% of each other, but 5% off specified capacitance. This is my best sounding crossover to date, but I went from ~LR4 with notch for woofer to 8th order initial rolloff Cauer at a lower frequency. Again, no direct comparison to the previous polypropylene version.

F1 Fan, thanks for the suggestion.

 

[F1 FAN]

Your welcome Bob,I should mention I used 24nf and 47nf from the 417 series for the filter sections.The 100nf has bigger spacing but the 416 series 100nf will fit.

I used the 100nf Panasonic ECQ stacked films for the bypass caps.Digikey part P4525-ND

 

[tcpip]

what use have you in mind that requires Q<0.5?

Basically, I've done paper designs of many active xo's by taking their passive xo's and trying to convert them to active with exactly the same phase and SPL curves. I have found that this can be done, but after optimising the curves, I get filter stages with very strange Q sometimes. Omitting those stages is not an option; I get clearly unacceptable levels of phase errors.

During these experiments, I've seen that Q < 0.5 often occurs in one of the stages. That's why I was asking.

For a report of one such experiment, see here. In this example, there was no low-Q stage, but in other cases, it's happened.

 

[tcpip]

Hi tcpip,I used the BC components MKP 417 series 2% polyprop
available at digikey.They fit the board perfectly.

Yes, those were the closest fit among the ones I found. Thanks for the confirmation. 🙂

Mouser sells the Wima FKP2s, there must be another distributor on your side of the world.

Thanks. I'll check Mouser. Let's hope their Website will be easier to use than the nighmarish www.mcminone.com --- that's easily the worst online retailing electronics components Website I've seen.

I like the sound of polyprops better than polyester, not that I have done a direct comparison....

Thanks a lot. I'll keep these in mind. Of course, actually selecting caps for any project may be influenced more by size than type, given this tight-fit board. 🙂

BTW, how do you design your active xo's? Do you use SoundEasy, LSPCad or some similar package?

 

[BobEllis]

When it comes to hobby software, I am a tightwad. I use Speaker Workshop to measure in the cabinets. Then I use an eyeball approach to add a little higher Q to compensate for a droop near the crossover and vice versa. I get the impression that Linkwitz does it this way. So I figure with careful measurement and willingness to go a few iterations, I can get pretty good results, too.

Notches and shelving are done the same way. Baffle step is almost a textbook thing, although I tend to reduce it if the speakers will be close to the wall. I use "The Edge" for predicting baffle step response.

I build the XO, measure and tweak from there. Tweaking will be a lot easier now that I have a few of Jens' active filter 2 boards running. Nicely adjustable, with a prototype area that I have populated with sockets for adjustable notch filters.

SWCad from Linear is a nice free program that allows me to verify my spreadsheet calculations. It is particularly handy when doing Cauer filters.

It would be nice to have software to predict off axis response, but measure/tweak/repeat gives satisfactory results. Besides, for me much of the fun is the building. Wouldn't want that to go too fast. 😉 In some ways I get more satisfaction out of an unsuccessful experiment - it makes it easy to justify the next project. Thank goodness I only have my bank account to answer to. 😎

 

[Primetime]

Hi,
I'm more than a bit late in joining this GB... I would be greatful if any of you have a board or two to spare... Pls let me know..
I'm not sure if it is too much to ask, but if Jen or Bob could be kind enough to send me a pdf of the pcb layout, I could try to etch one myself 😉 though I would much prefer to have a professionally made board.

 

[frez]

Hi,

I'm, also interrested in a pair of PCBs. Any chance there's someone out there with a soft spot for a fellow DIYer?

THNX
/Frez

 

[spookydd]

May I revive this old dead thread a bit?
There used to be some interest in those xover boards, and I suspect some were left hanging with nothing at some point.
Perhaps some are still searching for such xover board projects...
I've been doing a lot of thinking over many years about making a 4 way xover filter for my system, and over time, many aspects of what I'm looking for have been tweaked.
If there is interest in this kind of project, I could share the results here.

Here are some of the things I definitely would want in such a project:

- being a diy project, absolutely NO SMT parts to solder anywhere (I can't handle those)
- single board design, with no external wiring (plugs right on the pcb, and the psu+power plug/module-fused-filtered-switched)
- 4 ways, with perhaps an easy way to swap out small boards with the parts needed to select the frequencies
- no switches, pots or whatever. The filters would be located near the speakers and amp racks, with the shortest speaker cables possible from the amps
- signal detect to turn on the filter. filter goes in standby mode after a while when no signal comes in
- 4th order filters only (linkwitz-riley)
- xlr inputs, balanced (plugs on pcb). balanced inputs is a must, with a long link to bring the signal in and short links to the amps from the outputs

those are the highlights. some things can evolve...

 

[Variac]

The diyAudio store is planning to have some active crossover kits within 6 months

 

[spookydd]

The diyAudio store is planning to have some active crossover kits within 6 months

That's cool and likely useful for many diyers out there, but that is very unlikely to be of much use to me for my much larger project.

So regardless, I will be working on a design of my own. Besides, it's more fun and rewarding to do things like this instead of simply using a ready made kit.

I have done some simulations, using the widely used and regarded salen-key topology and linkwitz-riley curves, but now I think I'll go for the state variable topology instead, still using the same linkwitz-riley curve...
What I'm after is a solid 4 way filter, with the best performance that can be achieved. And eventually along with the filter, several other features will be added, such as balanced input for example, and I want to figure out a way to have a few frequencies at each xover point to choose from when doing the tuning, so the best xover frequencies can be found for the speakers used. I'm probably going to have several sets of parts selectable in some way to change the frequencies, maybe with simple jumpers, or perhaps some type of plug-in card system like the short DIMMs in computers...
None of those features I think would be in a kit.