driver impedance and measurement rig

[AnalogBeforeDigital]

I was hoping to better understand the ways a cabinet affects impedance measurements and how to quickly (reliably) measure drivers.

so a couple of questions:

1. can I measure a (8 ohm) mid-bass driver and (4 ohm) tweeter (2-way or 2 and a half way) at the same time (wired together) to gain a sense of how the impedance of the tweeter affects the mid bass and vice versa or do the need to be measured separately.


2. assume two cabinets - with differing dimensions (A and B) - yet same internal volume (1 cu ft) - will the measurement for the same driver in box A be the same as box B and if not what is affected.

3. in terms of the cabinets in 2 above - do you have any ideas about making a measurement rig where it is easy and cheap to quickly change a box volume and measure the drivers incrementally... if so could you share ideas on such a rig?

4. in three above - if if I used magnetic strips to hold the box together - would the presence of magnets interfere with the measurement enough to rule it out as a construction method for testing?

5. similarly - would the relative positioning of the drivers (tweeter to mid-bass) also affect the measurements in terms of destructive interference across frequency range or could that for all intents and purposes be ignored

got a few other questions - but this should get things moving (hopefully)

Thanks!

 

[Matthew J Poes]

1. can I measure a (8 ohm) mid-bass driver and (4 ohm) tweeter (2-way or 2 and a half way) at the same time (wired together) to gain a sense of how the impedance of the tweeter affects the mid bass and vice versa or do the need to be measured separately.

MJP: sure of course. This is exactly how a finished speakers impedance is tested and calculated. Nothing wrong with this.


2. assume two cabinets - with differing dimensions (A and B) - yet same internal volume (1 cu ft) - will the measurement for the same driver in box A be the same as box B and if not what is affected.

MJP: it can make a difference but not much. Resonances, internal reflections, and the space between the driver and back wall can have small impacts. For the most part this is uninportant and likely would only show up if the back wall was very close to the driver.


3. in terms of the cabinets in 2 above - do you have any ideas about making a measurement rig where it is easy and cheap to quickly change a box volume and measure the drivers incrementally... if so could you share ideas on such a rig?

Not really. I suppose you could make a box that could be partitioned and then make the front baffle removable. You need to be sure that the panels are held in place in a rigid fashion and air tight. Small air leaks can cause some measurement anomalies in an impedance sweep.

4. in three above - if if I used magnetic strips to hold the box together - would the presence of magnets interfere with the measurement enough to rule it out as a construction method for testing?

I can’t imagine how this would work. The wood panels won’t likely be true or flat and the magenta couldn’t compensate so it wouldn’t be airtight. It also probably wouldn’t be strong enough to support a speaker. IF the magents were that strong then yeah they could throw off the results. A bucking magnet changes the T/S of a speaker. It raises BL typically. I think small magents won’t make a big difference but powerful magents could cause problems.

5. similarly - would the relative positioning of the drivers (tweeter to mid-bass) also affect the measurements in terms of destructive interference across frequency range or could that for all intents and purposes be ignored

MJP: are we still talking about impedance? If so, then no, this won’t be an issue. The phase and resistance is measured electrically, not acoustically. Acoustical interferences don’t have an impact. The reason I suggested the acoustic effect of a very close boundary might impact it is that it could cause a reflection that is significant enough to modify the cones movement.

 

[Kerry Armes]

I was hoping to better understand the ways a cabinet affects impedance measurements and how to quickly (reliably) measure drivers.

so a couple of questions:

1. can I measure a (8 ohm) mid-bass driver and (4 ohm) tweeter (2-way or 2 and a half way) at the same time (wired together) to gain a sense of how the impedance of the tweeter affects the mid bass and vice versa or do the need to be measured separately.

2. assume two cabinets - with differing dimensions (A and B) - yet same internal volume (1 cu ft) - will the measurement for the same driver in box A be the same as box B and if not what is affected.

3. in terms of the cabinets in 2 above - do you have any ideas about making a measurement rig where it is easy and cheap to quickly change a box volume and measure the drivers incrementally... if so could you share ideas on such a rig?

4. in three above - if if I used magnetic strips to hold the box together - would the presence of magnets interfere with the measurement enough to rule it out as a construction method for testing?

5. similarly - would the relative positioning of the drivers (tweeter to mid-bass) also affect the measurements in terms of destructive interference across frequency range or could that for all intents and purposes be ignored

got a few other questions - but this should get things moving (hopefully)
Thanks!

1. You can measure them together but it will only give you an idea of what the two drivers operating with no crossover will look like. There really isn't a reason to measure them like you are proposing unless you have the crossover on each.

2. It usually takes significant changes in dimensions, like going from a cube to a long and skinny tower, to see much difference. This will also normally show up as a blip in the impedance somewhere by changing the cabinet resonance or internal standing waves. It could potentiall change the port tuning by making your cabinet go from a traditional BR to a TL type design, intentionally or unintentionally.

3. You build to the biggest size and use blocks to take up volume internally that you could add through the driver cutout or something similar. Although I'm not sure why you would really want to do this. Modeling programs are good enough these days that if you measure the free air T/S parameters properly and don't screw up the internal volume on your box, you should be right on with tuning. This can be verified with an impedance measurement as well.

4. If the magnets were close to your measurement equipment or the driver magnet then it could.

5. Relative position between the two drivers definitely affects frequency response but does not change impedance. Significant changes in the woofer location could have similar effects as changes in the box dimensions depending on the shape of the enclosure.

 

[AnalogBeforeDigital]

okay - so I am hearing I need to step back a bit and not get in the weeds too much on the measurements - except for where they may be positioned (tweeter and woofer) in relation to each other. My assumption on this is due to the destructive interference patterns which may be directly affected by the distance between acoustic centers.

which brings me to a separate question regarding positioning the woofer and tweeter in such a a way that their acoustic centers align vertically.

How much does this (vertially aligned centers) really matter in terms of delay to a laymans ears?

As to the magnetic strips - If I use paper backed foam core in place of wood and use thin magnetic strips instead of glue would I have a 'good enough for government work' rig - or do I need something better for assembly (duct tape?) to ensure its relatively air tight

 

[Kerry Armes]

okay - so I am hearing I need to step back a bit and not get in the weeds too much on the measurements - except for where they may be positioned (tweeter and woofer) in relation to each other. My assumption on this is due to the destructive interference patterns which may be directly affected by the distance between acoustic centers.

which brings me to a separate question regarding positioning the woofer and tweeter in such a a way that their acoustic centers align vertically.

How much does this (vertially aligned centers) really matter in terms of delay to a laymans ears?

As to the magnetic strips - If I use paper backed foam core in place of wood and use thin magnetic strips instead of glue would I have a 'good enough for government work' rig - or do I need something better for assembly (duct tape?) to ensure its relatively air tight

Relative positioning of the woofer and tweeter is only relevant for acoustic measurements (frequency based) , not electrical (impedance based) measurements. Yes, it is constructive and destructive interference that plays a roll in this but it dictated based on overall 3D location on an x/y/z coordinate system, not just aligned acoustic centers. For example, you could have two speakers that have the acoustic centers aligned between their woofer and tweeter, but one could have a very far offset tweeter on a large baffle. The acoustic centers are still aligned but the summation is completely different.


Good crossover design software (and designers who know how to use it) takes into consideration the relative acoustic centers of the driver (which may or may not be aligned) and uses the phase information of the two drivers to calculate the summation. Using minimum phase allows the software to calculate this over multiple points in space and not just at the measurement point where the original measurements were taken. The proper delays are added during crossover design to get the drivers to sum flat. In passive crossovers, this is done by adding inductors and capacitors, which change the phase behavior of the speaker and create the desired frequency response. In active, a delay can be set in the software to align the acoustic output and then slopes are added to create the desired response. If you are doing active, there is essentially no reason to try to align the acoustic centers. If you are doing passive, it can be required to use symmetrical slopes on your woofer and tweeter. Most designers use different slopes to create the proper summation.


Testing with foam can work but it is not ideal and the measurements likely won’t be as accurate as if you have a proper enclosure. You might see this as peaks and dips in the frequency response caused by cabinet measurements or in the impedance as a little wrinkle in the curve. Leaks in your box will also cause this.