When measuring compression drivers with waveguides/horns.....

[dkalsi]

Is there is a need to include a series capacitor to protect the CD? I'm not measuring at loud volumes on anything.

If the answer to the above question is "Yes" - would it not impact the measured phase? I'm currently measure using ARTA's semi-dual method. If I use a 50uf or 80uf capacitor to protect the CD - it seems to impact the response (albeit slightly). How does one typically account for the impact of the protection capacitor on the measured response and phase when one gets to the crossover design stage.

[Robh3606]

The reason that you would do that is to protect the driver from DC especially turn on thumps from amps. It really depends how well you know your measurement system. I have never used a cap in series while taking measurements, which is a risk, but I have confidence in my set-up, system and methodology. I have a dedicated amp that has a very well controlled start up. I also won't hook up a driver unless the amp is essentially turned off with the attenuators at max after turn on is complete and the amp stable. Feed in a 1 K sine slowly bringing it up while monitoring the SPL. Once at the correct level I will start measuring.

Bi-amping for sure.

As far as phase you will get a shift as far as FR well use a big cap or several at least an octave lower than horn cut off.

If you have any doubts just use the caps and work around them. Just remember you have an extra pole in there if you are designing a passive crossover. May not be significant but don't forget it's there.

Curious to see if I am alone in this as many use measurement systems on the site.

Rob

[pos]

If you need a protection cap in the final biamped system then it make sense to measure and adjust it with the (same) cap in place.
The high-pass magnitude and phase effect have to be taken into account in the final crossover design, just like the natural rolloff of the compression driver and horn combo: you simply add electrical filters and EQ in order to obtain a given acoustical target.

[dkalsi]

If you need a protection cap in the final biamped system then it make sense to measure and adjust it with the (same) cap in place.

Sorry - to be clear, this won't be for active/bi-amp setup.

I'm just delving into speaker design and learning how to take measurements and then using simulation software (e.g., Xsim, VituixCAD, etc.) to design a crossover.

I have to following parts on hand:

Waveguide B-52 PHRN-1014 - which I believe is said to be similar (if not identical) to the QSC HRP152i horn that was popularized in the Econowave builds.

Tymphany DFM-2535R00-08

Ultimately, I would like to design a simple 2-way speaker system using the above described CD/WG combo. Still deciding what to use for mid-bass (affordable recommendations welcomed :-)).

To have accurate files for crossover simulation, I understand I would need good measurements. If I use a protection cap in series, and I know that cap would NOT be included in my final crossover (e.g., 80uF cap for protection), then is it possible to remove the impact of the included cap in my measured frequency response/phase?

Things would be a lot simpler if one can safely measure without the cap :-)

[pos]

Oups, sorry for the confusion.

I have no experience with passive design software, but I guess they will let you compensate for the cap and calculate the raw magnitude/phase response for you.

If not then you need to measure without the cap.
If that is your first experience measuring drivers you should probably keep the cap in place for most of the measurements, and remove it for the final set of measurements when you are confident enough in your process and hardware, as per Rob recommendations above.

Most measurement software will let you either high-pass the measurement signal with a linear phase filter so that phase is not impacted, and a sharp slope so that magnitude is also left pretty much left untouched a few fractions of octave above cutoff.
The other solution is to use a sweep and make it start at a given frequency, but you will loose precision in the lower part of the range that way, making the stop band harder to analyse.

[dkalsi]

If that is your first experience measuring drivers you should probably keep the cap in place for most of the measurements, and remove it for the final set of measurements when you are confident enough in your process and hardware, as per Rob recommendations above.

.....The other solution is to use a sweep and make it start at a given frequency, but you will loose precision in the lower part of the range that way, making the stop band harder to analyse.

This is great idea. I will take all measurements with cap in place, and then remove it for my final measurements. I am using sweeps to take measurements - so I'll adjust the bottom end to start the sweep at a little higher frequency.

Thank you again.

[1audiohack]

I almost never use a cap while taking measurements. I usually limit my sweeps to ≈300Hz for comp drivers once I am above 1V drive.

I use a D45 for a test amp for most things as it's a pretty low powered well behaved supply.

I do sweep through zero on woofers as my measurement platform comes back through zero in quadrature giving more data points at low frequency where resolution is typically poor.

All the best.
Barry.