Competency check on diy crossover please

[Joe Hart]

Ferrite core is a way to increase inductance using fewer turns than air core; it is a cost- and weight-savings measure which introduces non-linear behavior.

Coupling between magnetic fields is the issue. The two coils act like a transformer, with flux inducing currents in the opposite coil. This why inductors must be separated and, better yet, in orthogonal planes. Since field strength falls off as r^2, distance is the crucial factor.

May need to look into iteration 2 for the build layout... OR Can I put a material between the inductors to block the magnetic fields (Lead, Aluminium etc etc)

 

[YNWOAN]

You can't put material between the magnetic fields as it will interact with said magnetic fields and alter the value of the coil (this can easily be measured). If it doesn't effect the coil then it isn't doing any shielding either.

 

[martinab2]

I must say that your soldering needs work, you should make a mechanical connection by hooking wires together or twisting them then solder.

Pete

From the pictures, it looks like your soldering iron may have been struggling to get the joints hot enough to allow the solder to flow properly. You will normally need a fairly powerful iron for big cables and teminals. What you are aiming for is to have the whole joint liquid at the same time so that you get a smooth finish and a strong joint. Solder should flow like melted chocolate. If it behaves more like jam then you need a bigger iron. I'm off to have my breakfast now

 

[Fatmarley]

How to turn a schematic into a crossover.

 

[walriley]

The Lpad won't interfere with the zobel. Lpads exist to reduce driver output without altering the driver resistance that the crossover 'sees' so to speak.

 

[Joe Hart]

I wonder how an L-Pad works v a series resister

 

[Joe Hart]

The Lpad won't interfere with the zobel. Lpads exist to reduce driver output without altering the driver resistance that the crossover 'sees' so to speak.

How do I design an L-Pad?

 

[walriley]

http://www.sengpielaudio.com/calculator-Lpad.htm

 

[Fatmarley]

I wonder how an L-Pad works v a series resister

There's no need for a series resistor, you will have enough resistance In the series capacitor. Try a 47r In parallel with the tweeter. If you need more attenuation try a smaller value.

 

[demotivated]

Is this the tweeter, cos I can't read the datasheeet well?

https://cdn.shopify.com/s/files/1/0809/2387/files/XT19TD00-04.pdf?7425721213548150932

They certainly make a lot of tweeter types.

You already have way too many people advising you, but this is what I would do. The peak is not too bad as I guessed as I wasn't expecting it to be a 3.5 ohm tweeter. This plan uses stuff that is already soldered in place and will lose you about 7dB of treble. I have assumed the components are as typed, cos the pen says different on the zobel drawing. If it's the written values you will lose about 8dB so try it. This is probably more cut than you wanted, so give it a long time to adjust because it will be drastic.

Disconnect the input to the 10uF C4 leave it unconnected.

Disconnect the negative [return] connection to the bottom of the 2 zobel components R1 and C5. It does't matter which order you soldered them as long as the driver end is still connected. Then connect the input feed that used to go to C4 to the unconnected end of the zobel [this is now the input]. This gives you a smaller input cap so the tweeter doesn't come in until about 6kHz instead of 4kHz before. It also reduces the gain as above. All you have lost is a zobel which you don't really need with that very low inductance tweeter. The very high treble will lift a few dB without it.

If you like it, you have a pair of 10uF caps for future use

 

[Fatmarley]

If the speaker has a wonky frequency response my suggestion wont change that, It will only drop the tweeter level. My suggestion Isn't anything new, It was used In the Royd Minstrel, and to be honest It had me confused for a while until I simulated It.