Testing my crossover wiring

[colin_young]

I'm building Paul Carmody's Overnight Sensation speakers and I've completed wiring up the crossovers. Is there any procedure involving a digital multimeter that I can follow in order to test that I've done the wiring and soldering correctly? Other than the obvious make sure paths that are supposed to be connected are connected and those that shouldn't be aren't?

I don't mean testing the crossover frequencies, etc. because I'm assuming he got that right. I mean making sure some idiot (me) didn't hook things up incorrectly. And considering the person who will be interpreting the circuit diagram for testing is the same person who interpreted it to wire it up, there is plenty of room for error. I mean, I consider myself pretty good at reading drawings and visualizing things physically, but I also have years of experiencing suggesting that I don't always get things right the first time. And it's going to be a lot easier to fix any mistakes before they are installed inside the box than after.

[billfitzmaurice]

A DMM can test continuity and DCR, but not Le or C. It can't even test DCR if there's a cap in the path, as with a high pass filter. One way to test it would be to hook the input to an amplified pink noise source, then touch the output leads of each filter one at a time to a woofer. Then you can hear the effect of each filter, without the potential for toasting a midrange or tweeter with low frequencies.

[fpitas]

You can double check you get the polarity right at the driver terminals, I guess. Other than that, an on-axis SPL measurement is the most sure-fire way of knowing the job has been done right. Barring that, you can use a signal generator or a tone test CD to make sure the signal at crossover isn't nulled from incorrect polarity etc.

Just saw Bill's answer. That's a good approach to test the crossover alone.

[colin_young]

Thanks. I was working under the assumption that I only had a multimeter and my ears for testing with, and no fancy signal generators, so I was going to say that I couldn't follow Bill's suggestion until I realized that I'm typing this on a computer and I can probably find a pink-noise generator for that, and I have the amplifier I'm planning to hook the speaker to.

That's a good tip about the tweeter. I had no idea that could be an issue (ruining it with low frequencies), but after a tiny bit of research, it makes complete sense.

[billfitzmaurice]

Software wise you can get anything and everything for testing as freeware. It's hardware you have to pay for.

[djg]

I generally use alligator clip test leads to hook it all up outside of the cabinet, play something through it. Has worked on numerous projects over 30 years.

[colin_young]

I generally use alligator clip test leads to hook it all up outside of the cabinet, play something through it. Has worked on numerous projects over 30 years.

That's exactly what I ended up doing. There is definitely a difference in how the tweeter and woofer outputs sound, and both crossovers sound different in exactly the same way, so I figure I either got it right or managed to do them both wrong in exactly the same way. The latter seems less likely.

Thanks for the help! Now I feel confident in going ahead and figuring out how I'm going to mount these things in the box. With a bit of luck I might get things assembled by the end of the weekend.

[Dave Bullet]

To force the brain to "see what's in front and not what it remembered" - draw a diagram "backwards" from the way you physically wired it up. You could also draw a diagram of the way you wired it up and ensure both diagrams match. I use the same technique when entering for example bank account number someone has told me. I read it backwards to them.

An impedance sweep is the best solution. You can do this with REW, as long as you have a computer with a sound card (line in and out - not just mic in) and hack a loopback as shown in the diagram with a load (Rsense) resistor. you don't even need an amplifier as the output from the soundcard is enough to create a sweep through the hooked up completed speaker. (Infact avoiding an amplifier is good as it eliminates any frying your soundcard risk)
https://www.roomeqwizard.com/help/he...asurement.html

[JensToft91]

Isn't it possible to place a phone with an spl meter about 5 inches away from the speaker, directly between the tweeter and woofer and play a 1kHz tone, and then a tone at the crossover region to see, if it has nearly the same output? If the second one is way down, the phase must be wrong.

[JensToft91]

Or run a sweep with any microphone, reverse the polarity of the woofer, and run it again.

[fpitas]

Yes, and yes.