I wanted to share a very strange experience I had with lower midrange distortion caused by a passive xover component.
The symptom was fuzzy sounding distortion in the lower mids when playing the speaker pretty loud, maybe 100 - 105 dB. It sounded like the amp was running out of power right around the impedance minimum of the LF driver. Sometimes it sounded like a VC rubbing or an enclosure panel buzzing. The speaker in question was a 2 way MTM with 2 x 8" Eminence Beta 8A and a 1" PRV compression driver on a B&C ME20 horn crossed at around 2k LR2.
Here's what I found - The LF low pass inductor (1.5mH 14 awg Solen Perfect Lay - see below) had a chunk of loose windings that were vibrating around. The coil was glued down with RTV to a piece of baltic birch plywood and the crossover itself was mounted onto the enclosure back wall, maybe 10 inches from the LF driver magnets. When I swept the speaker with a loud sinewave, I could feel the inductor rattling around and vibrating by reaching through the port and pushing on it. The proximity of the inductor to the speaker magnets probably provided enough magnetic field to force the inductor windings to rattle at higher playback levels.
The fix to the problem was a healthy dose of super glue across all of the inductor surfaces and fresh zip ties. Funny thing is when I first opened the shipping box containing the inductors, they were flying around loosely due to poor packaging. I think all that banging around in the box allowed the coils to break apart internally. I won't name the company who shipped me this stuff, but you can bet they're going to hear from me. I've gotten so many badly packed shipments from various companies. Problem is most of them try to blame it on the carriers and you get stuck in the loop of trying to resolve a damage claim. Not fun.
Anyways, I wanted to tell everyone about this experience because its potentially a hard thing to figure out. I definitely won't be trusting another one of these "baked" coils again straight out of the box. Another one of the inductors in this parts shipment just blew apart in my hands from merely handling it. The zip ties holding the windings just broke apart and it all uncoiled into a slinky - instant 40 bucks down the drain. There was no way no how I could get that thing back together after totally coming unraveled that way. After that I should have suspected the other inductors weren't in much better shape internally. Lesson learned.
Time to build a coil winder and start winding my own inductors....
Picture of the culprit -
The symptom was fuzzy sounding distortion in the lower mids when playing the speaker pretty loud, maybe 100 - 105 dB. It sounded like the amp was running out of power right around the impedance minimum of the LF driver. Sometimes it sounded like a VC rubbing or an enclosure panel buzzing. The speaker in question was a 2 way MTM with 2 x 8" Eminence Beta 8A and a 1" PRV compression driver on a B&C ME20 horn crossed at around 2k LR2.
Here's what I found - The LF low pass inductor (1.5mH 14 awg Solen Perfect Lay - see below) had a chunk of loose windings that were vibrating around. The coil was glued down with RTV to a piece of baltic birch plywood and the crossover itself was mounted onto the enclosure back wall, maybe 10 inches from the LF driver magnets. When I swept the speaker with a loud sinewave, I could feel the inductor rattling around and vibrating by reaching through the port and pushing on it. The proximity of the inductor to the speaker magnets probably provided enough magnetic field to force the inductor windings to rattle at higher playback levels.
The fix to the problem was a healthy dose of super glue across all of the inductor surfaces and fresh zip ties. Funny thing is when I first opened the shipping box containing the inductors, they were flying around loosely due to poor packaging. I think all that banging around in the box allowed the coils to break apart internally. I won't name the company who shipped me this stuff, but you can bet they're going to hear from me. I've gotten so many badly packed shipments from various companies. Problem is most of them try to blame it on the carriers and you get stuck in the loop of trying to resolve a damage claim. Not fun.
Anyways, I wanted to tell everyone about this experience because its potentially a hard thing to figure out. I definitely won't be trusting another one of these "baked" coils again straight out of the box. Another one of the inductors in this parts shipment just blew apart in my hands from merely handling it. The zip ties holding the windings just broke apart and it all uncoiled into a slinky - instant 40 bucks down the drain. There was no way no how I could get that thing back together after totally coming unraveled that way. After that I should have suspected the other inductors weren't in much better shape internally. Lesson learned.
Time to build a coil winder and start winding my own inductors....
Picture of the culprit -
Attachments
Last edited:
I have an affinity for dipping coils in gloss polyurethane. Since i wind my own, this is pretty much my standard practice.
Wolf
Wolf
What type / brand of poly do you use? I was going to use 2 part casting resin but I don't know if that's a compatible with the wire enamel.
My next step is winding my own coils. I bought some supplies already.
What's your coil winding setup look like?
My next step is winding my own coils. I bought some supplies already.
What's your coil winding setup look like?
https://www.diyaudio.com/community/threads/anyone-custom-wind-inductors.339405/#post-5831177
I currently do them by hand, but I want to enlist a hand-crank at some point. Keeping the tension on the bobbin is the most important and yet difficult part.
I currently do them by hand, but I want to enlist a hand-crank at some point. Keeping the tension on the bobbin is the most important and yet difficult part.
I just use Minwax gloss polyurethane. I do not use the wipe-on as it's too thin. The satin has elements for less shine, and therefore is also less viscous.
Warning, the bigger coils do not fit through the lid in the smaller cans.
Warning, the bigger coils do not fit through the lid in the smaller cans.
I've had this problem with some of my Iron Core Inductors from PE.
My solution was more hot glue. 😊
My solution was more hot glue. 😊
Thanks fkr the replies. I've come to realize this type of vibration - distortion mechanism has more affect on what we hear a speaker reproduce having passive crossover components in the critical midrange. I now see the benefit of having the xover in a separate environment from the vibration rich location in the speaker enclosure itself. That's probably the worst place to put a crossover in, especially larger inductors that have to deal with alot of midrange energy. Even if the structural integrity of the coil was perfect, vibration born energy will still cause some amount of distortion and having a huge magnet in proximity of the inductor will likely make that effect more pronounced. The question is how much does which component contribute to the issue? Being i build alot of speakers that can do higher than average SPLs, this issue is of greater concern to me than someone else would have with small 2-way bookshelf speakers.
I guess the hardest part in winding your own coils is keeping constant tension on the wire and getting the layers perfect in the coil. It would be more consistent to put friction tension on the wire itself than the supply spool, as the supply spool diameter of wire plus its sloppy winding will cause the tension to vary quite a bit. Putting drag on the wire itself would fix that, but the friction itself can potentially compromise the enamel surface of the wire. Hmmmm....
I guess the hardest part in winding your own coils is keeping constant tension on the wire and getting the layers perfect in the coil. It would be more consistent to put friction tension on the wire itself than the supply spool, as the supply spool diameter of wire plus its sloppy winding will cause the tension to vary quite a bit. Putting drag on the wire itself would fix that, but the friction itself can potentially compromise the enamel surface of the wire. Hmmmm....
good finding by the way I am wondering what's the AC current going to both woofers at the time when this happening?
So do you think that the strange distortion was caused by the changing inductance when the coils moved? Or could you possibly have been heaingr the coil itself?
It was clearly audible through the LF driver, more than inudctor rattling mechanically itself. The loudest distortion component sounded like a rubbing VC, so it was more or less a non-linear type of distortion.
Judging by the frequency range that excited the distortion most, it was in the minimum impedance range of the LF drivers, mainly between 100 and 800 Hz. It sometimes sounded like crossover distortion of an under biased amp and a rubbing VC.
To isolate xover vibration from any driver effect, why not connect the xover to a dummy load and crank it up.
I use electric water heater elements for the load when testing amps....pretty easy to make 4, 8, or 16 Ohm loads on the cheap.
Just put the elements in a bucket of water if amp is high powered.
Prosound guys have laughed about having sing-a-longs with passive xovers 😀
I use electric water heater elements for the load when testing amps....pretty easy to make 4, 8, or 16 Ohm loads on the cheap.
Just put the elements in a bucket of water if amp is high powered.
Prosound guys have laughed about having sing-a-longs with passive xovers 😀
The tension in winding needs to remain at both ends of the process. The wire on the receiving hub being most important. When you start the process after doing a pause and scrape-check of value, the line becomes taut again as the winding starts. The binder clip (as in my jig) is removed and you continue the process. There has to be a clip or clamp at the receiver end in case you need a break. The supply spool can sit on something to provide friction like a fixed rubber rail, or thick wire jacket over a dowel.
I have an idea of how to take my jig to the next step, I just have not implemented it.
I have an idea of how to take my jig to the next step, I just have not implemented it.
Thanks. Did you happen to measure the inductance change before fixing it? It would be cool to know how much - or how little - a change became audible.
@mark100 Good idea with the water heaters elements. I used small oil bath radiators. They are close to 8 ohms at full power and of course have a built in heatsink.
Actually I've had an inductor output some intelligible sound, with no speaker connected. However, that was a Class-D ferrite core inductor (Ferroxcube 4C65 material), and I guess it was due to magnetostriction. I had also posted the occurrence here on the forum over 10 years ago, received no help, so changed to Micrometals powdered iron core -2 grade.
https://www.diyaudio.com/community/threads/ucd-for-s-woofer-appln.175304/page-8
However, I am not very aware of strong magnetostriction (or any other non-linearity) in air-core inductors, as is the case in this thread.
I like your idea better ! Heatsink included !!!!
I had no idea those kind of heaters had an audio range line resistance. Just gave away two nice ones....
My water bucket, heater element rig, is kinda nice because the 4000w elements are very close to 16 ohms and it's of course easy to makeup standard loads with that.
Well. 2 ohms is still a big mess ... Lol.
Biggest prob i have is emptying the bucket, and taking it all down before it goes rancid... haha
Since the coil wires were moving around in a DC magnetic field (from the speaker), they would generate their own back-EMF, just like the voice coil in a speaker does.
The back-emf from the vibrating inductor coils is in series or parallel with the voltage going to the loudspeaker (depending on filter configuration), so it would directly drive the loudspeaker with an additional (noise) voltage.
This unwanted noise voltage (back emf) from the inductor would be largest at the vibrational frequencies of the inductor coils, i.e. at the mechanical resonance frequencies of the loose winding.
My hypothesis is that this is what the OP (profiguy) was hearing.
-Gnobuddy
The bad inductor effects were very noticeable at roughly 28V rms drive into 4 ohms in the lower mids. Thats a lot of juice - about 200 w. I did manage to get some microphonics from the coil when tapping it firmly with a screwdriver handle while the input was shorted. That was definitely caused by the LF magnet some inches away from the inductor.
In hind sight it would have been better to mount the face of the inductor at a 90 deg angle to the LF driver magnetic field to lessen the effects, even if it wasn't the loose coil windings causing it. The inductor value was roughly the same as the other channel within maybe 1 to 2 percent and no distinct change before and after according to my LCR meter. Now imagine if you were dealing with a poorly built, thin walled cab and crossover mounted dead center on an un-supported or un-stiffened back panel right behind a huge LF driver with a monster sized motor. That would have to do something bad aside from hearing the acoustical effects of severe panel resonances.
In hind sight it would have been better to mount the face of the inductor at a 90 deg angle to the LF driver magnetic field to lessen the effects, even if it wasn't the loose coil windings causing it. The inductor value was roughly the same as the other channel within maybe 1 to 2 percent and no distinct change before and after according to my LCR meter. Now imagine if you were dealing with a poorly built, thin walled cab and crossover mounted dead center on an un-supported or un-stiffened back panel right behind a huge LF driver with a monster sized motor. That would have to do something bad aside from hearing the acoustical effects of severe panel resonances.