4
\$\begingroup\$

I'm trying to come up with a basic two-way crossover for a speaker I'm building out of scrap. (Working with what I've got - no new parts.) My amp handles 8-16 ohms.

Here's what I've come up with for a speaker design so far:

schematic

simulate this circuit – Schematic created using CircuitLab

I've just got the one inductor, but I've got a ton of different capacitors. I'm planning to build the capacitor out of a combination of smaller ones in parallel. (10, 5, 1, and 0.5.)

My primary question is: How would I calculate the nominal impedance?

Secondary questions: - I believe this crossover is set to 12kHz; did I do my math right for this? - Since the tweeter is 6 ohms, will be attenuated enough so that it won't overwhelm the woofers?

Additional info: - I'm not looking for this to be a fancy system. Just something that will sound ok and not blow up my amp. - This is really just a re-working of the existing components in the original speaker. My intention was just to see if I could adjust a few things and increase the impedance to >8 ohms. (So it won't blow up my amp...) - I based my design on the series circuit in figure 1.1 here.

enter image description here


A bit of background: I pulled apart a pair of Samsung surround speakers (3 ohm) and I'm trying to make it into something I can use with my RCA STAV-4090 receiver. The exact driver models are a mystery - no spec sheets available, unfortunately. Two of the four speakers say "F30C-D998-1 75108" and the other two say "F30C-D998-1 74208". (The difference being 51 vs 42 - any ideas what they could mean?)

Samsung speaker front

Samsung speaker back

From what I can see, here's what the circuit looked like in the original:

schematic

simulate this circuit

I believe this circuit was a cheap attempt at a three-way crossover? The tweeter gets a high-pass, the center driver effectively gets a low-pass from the leftover, and the bottom driver has its own low-pass filter. (Is this right?)

\$\endgroup\$
2
  • \$\begingroup\$ I'm a bit late on this, but what I see in the manufacturer's crossover is Tweeter filtered with 1st order cap, woofer 2 is a slight sub with inductor filter to trim high Hz and woofer 1 is unfiltered. However 12 kHz is very high for first order. Depending on your age, may not hear that range much. Try 6 kHz. \$\endgroup\$ – Ben Aug 28 '20 at 16:01
  • \$\begingroup\$ Learn by experimentation. In a first order crossover there is a fair bit of overlap. I don't see that the unfiltered speaker will get less power - in fact I'd put a resistor inline with the mid-range (full range in here). \$\endgroup\$ – Michael Nolan Jan 6 at 13:48
1
\$\begingroup\$

You can't just throw any random speakers together and expect a good result. Each speaker has its own impedance curve which can vary dramatically with frequency and has to be taken into account when designing the crossover.

Also their sensitivities are probably not matched so you could need padding resistors to balance them, and the tweeter usually can't handle as much power so it needs to be protected against getting too much at lower frequencies.

I haven't done any calculations on you design, but it looks iffy to me because even below 12KHz the capacitor partially shorts out the drivers and puts high power into the tweeter, and the amp might not like the low impedance load.

The two larger speakers should be fine in series provided that they are identical. If the original circuit is correct it suggests there may be some difference between them - or perhaps L1 is wired like that to modify the impedance and/or amplitude curves.

The main difference between that design and yours is that it uses the natural inductance of the drivers as part of the high pass filter. Since driver impedance increases at high frequency this transfers power to the tweeter without increasing load on the amp. However without a full analysis with all speaker characteristics taken into account it's impossible make an accurate calculation.

\$\endgroup\$
7
  • \$\begingroup\$ Thanks for the answer! Yes, I'm completely aware that this won't be an amazing speaker - I'm just looking for usable. (I edited the question to clarify my intentions.) \$\endgroup\$ – James Nov 9 '17 at 20:49
  • \$\begingroup\$ Since I don't know the specs for these speakers, is there any way to reasonably guess at what they may be? \$\endgroup\$ – James Nov 9 '17 at 20:50
  • \$\begingroup\$ Actually - you're quite right about there being a difference between the speakers! I initially thought all four were identical, but a second look revealed a small difference in the number on the back. Is there any way to know what they mean? I can't find data sheets based on anything in these numbers. \$\endgroup\$ – James Nov 9 '17 at 21:13
  • \$\begingroup\$ It's probably an in-house number, and may not have any significance. I would measure the DC resistance of each speaker and compare their construction (cone shape, suspension etc.) - if they look identical then they probably are. To get the impedance curve I would wire a resistor in series, then measure AC voltage across the speaker while putting a tone through it that sweeps across the audio band. For frequency response I would use a good microphone with flat response and known sensitivity (or just do a listening test, but my ears have 30dB loss above 2kHz so....). \$\endgroup\$ – Bruce Abbott Nov 9 '17 at 22:16
  • \$\begingroup\$ Thanks - they do look identical, and the dc resistance is the same. 8.2~8.4 ohm \$\endgroup\$ – James Nov 9 '17 at 22:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.