Hi Henry,
I hope this doesn't come across rude, but between this and your 3 way crossover one, it seems like you're spending a lot of time using those kind of basic Web calculations and then needing to go over what their results mean here in the forum. It might be more productive if you download something like VituixCAD & learned to use it, that way you can see the actual results of each component change in real time for yourself. I find that kind of learning much more satisfying than having to ask heaps of questions, though of course, everyone's minds work differently.
HTH,
David.
I hope this doesn't come across rude, but between this and your 3 way crossover one, it seems like you're spending a lot of time using those kind of basic Web calculations and then needing to go over what their results mean here in the forum. It might be more productive if you download something like VituixCAD & learned to use it, that way you can see the actual results of each component change in real time for yourself. I find that kind of learning much more satisfying than having to ask heaps of questions, though of course, everyone's minds work differently.
HTH,
David.
You can measure the DC resistance of a driver (Re) with a LCR meter, simply connect the terminals to the meter. I didn't bother to investigate about how to measure inductance (Le) for many reasons (simply put, it doesn't offer real value in crossover simulation/design), but I highly suspect that you can't measure it with a simple meter, as it will vary in frequency.
Having said this, what is the purpose of a Zobel network? It is a way to create a constant impedance, thus allowing to calculate easily a HP/LP filter. However this approach doesn't take into account natural roll-off of a driver, thus works as intended only if the FR of that driver is flat in the crossover region, if not the resulting HP/LP filter will have a higher acoustical slope than is the electrical slope of the HP/LP calculated filter. I find this approach pretty useless.
With the use of a crossover simulator (XSim or VituixCAD), more often than not a Zobel is simply unnecessary to design a HP/LP filter with the correct acoustical slope.
I second what David said here before me, but I also have to add that for a design with a driver for which no information is given, you need measurement equipment, i.e. calibrated mic and a least a gig to measure impedance.
Ralf
Having said this, what is the purpose of a Zobel network? It is a way to create a constant impedance, thus allowing to calculate easily a HP/LP filter. However this approach doesn't take into account natural roll-off of a driver, thus works as intended only if the FR of that driver is flat in the crossover region, if not the resulting HP/LP filter will have a higher acoustical slope than is the electrical slope of the HP/LP calculated filter. I find this approach pretty useless.
With the use of a crossover simulator (XSim or VituixCAD), more often than not a Zobel is simply unnecessary to design a HP/LP filter with the correct acoustical slope.
I second what David said here before me, but I also have to add that for a design with a driver for which no information is given, you need measurement equipment, i.e. calibrated mic and a least a gig to measure impedance.
Ralf
Very nice write-up.
So all you need is a resistor (80-200 Ohms should work fine), a bunch of alligator test leads, and REW, as per the link to the 17-foot labyrinth above. But I would definitely also "test" some known-value resistors, say about the value of your drivers and 30 Ohms, to provide benchmark curves for comparison.
Since the measured impedance curve is likely to be pretty flat in any range of freq that matters, the change in impedance (and hence the value of fussing with a Zobel circuit) is minor. (But the distortion that might be introduced by sticking a large electrolytic capacitor in the Zobel circuit may not be minor.)
Also to remember, the XO is either "transparent" or "opaque" everywhere except near the XO frequency. So even if there were a substantial inductance changing with frequency, it wouldn't change the power filtered by the XO, except near the XO frequency.
Like with tapped "horns", phase adjustment, and Karlson enclosures, Zobel "correction" is a topic loved at this forum but not many other places.
Industry standard for inductance is measured @ 1K, with the VC immobilized. Glued in place.
You did not say what you are trying to do. Is this just for curiosity to see what the designer was thinking? Or is there some issue yo want to address? Crossover design had many tradeoffs, most are not obvious until you have made the mistakes several times.
To design a crossover , you need both impedance with impedance phase over frequency ( .zma file) and the amplitude with phase over frequency ( .frd file)
With these, as well as the measured offsets and baffle dimensions, you can use network simulators to design the crossover.
A DATS is a handy way to get impedance files but you can do it with a sound card and some resistors. You need a calibrated mic for the frequency response file. With any of the free software, you can be in business measuring a driver for under $250. When I started, it would have been $250,000! So we did things by ear.
Zobels have their use. One of many tools. There is no "right way" and all of the on-line calculators make assumptions, many different. That is not how to design a crossover.
Also be aware, the .zma files and .frd files are made with the drivers on an IB. Great to select them for a project, but useless for design. You need to measure in a prototype box.
You did not say what you are trying to do. Is this just for curiosity to see what the designer was thinking? Or is there some issue yo want to address? Crossover design had many tradeoffs, most are not obvious until you have made the mistakes several times.
To design a crossover , you need both impedance with impedance phase over frequency ( .zma file) and the amplitude with phase over frequency ( .frd file)
With these, as well as the measured offsets and baffle dimensions, you can use network simulators to design the crossover.
A DATS is a handy way to get impedance files but you can do it with a sound card and some resistors. You need a calibrated mic for the frequency response file. With any of the free software, you can be in business measuring a driver for under $250. When I started, it would have been $250,000! So we did things by ear.
Zobels have their use. One of many tools. There is no "right way" and all of the on-line calculators make assumptions, many different. That is not how to design a crossover.
Also be aware, the .zma files and .frd files are made with the drivers on an IB. Great to select them for a project, but useless for design. You need to measure in a prototype box.
Ebay has lots of different ypes of signal generators. What type would I need to measure impedance and phase response variation with frequency?
Now I wonder if there is free software, that when used with a computer, elimnates the need for a signal generator?
https://midwestaudio.club/resources/easyzma/
The quality of the soundcard is what differentiates between a good and a so-so measure.
Ralf
The quality of the soundcard is what differentiates between a good and a so-so measure.
Ralf
Room EQ Wizard (usually referred to as REW) can do impedance measurements using a simple jig of a few wires and one accurately measured resistor, assuming you already have a remotely decent soundcard. The online help articles have a pretty good explanation of the steps.
Speaker measurements are like playing horse shoes. Close counts. Sound cards are fine.
Depends on what you want to do. Some cheap sound cards aren't suitable for T/S calculation. OTOH they can be pretty useful in obtaining an impedance graph that can be used for crossover simulation. See here for an example of a sound card which doesn't measure well under 200 Hz, but can be used (something) above that frequency:
https://www.diyaudio.com/community/threads/problems-with-impedance-measurement.319740/#post-5365046
Ralf
https://www.diyaudio.com/community/threads/problems-with-impedance-measurement.319740/#post-5365046
Ralf