ARTA impedance measurement anomalies

[Paul Ebert]

I got my new soundcard in today. Time to do some ARTA work!

So, I get everything ready to do an impedance measurement on a RS180 I have and this is what I get:



I'm not thrilled about the little blip at ~65 hz, but that's nothing compared to what happens below resonance and above ~13K.

Any ideas as to what's going on?

[Jed]

Did you try a higher sampling rate? Check the settings and see where the cutoff is set.

[Paul Ebert]

The sampling rate is 48k which is the highest my soundcard will go. I discovered that I can eliminate the problem above 12K by using pink noise (and FFT) for the measurement instead of the stepped sine approach, but the high impedance below 20 remains.

I wonder if I need to use an amp between the card and the speaker. Perhaps I'm not able to provide enough current. I'll give that a try when I next get a chance.

[jbateman]

If the sound card has capacitor coupled output it will give less accurate measurement the closer you get to DC (0 hz)

A small power amp should fix things. The little glitch at 60hz is just leakage/noise from the 120volt house power.

[Paul Ebert]

The little glitch at 60hz is just leakage/noise from the 120volt house power.

It would be nice if that were true, but I don't think so because I'm using a laptop and it doesn't appear on the other two drivers I've measured (RS225 and RS52). I hope the driver isn't defective. I've got three other RS180s, I'll see what I get with those.

Time to dig out my old NAD integrated amp.

[mazurek]

I get pretty good measurements using the headphone output on my soundcard, though I can sometimes get some hash if the output level is very low (I faintly here the drivers during testing when I have good levels). Did you make your jig as suggested in the manual? Make sure you use their jig with added protection if you start using an amp.

[Paul Ebert]

The soundcard I'm using is the Behringer UCA202. It's a supercheap ($30) USB card. I'm pretty sure the jig is correct, but I'll check it again. I think I'll also pick up some 1/8" to RCA cables and try the soundcard in the laptop.

I probably should have just picked up a M-Audio transit since ARTA is known to work with that. I'm really trying to watch the budget, but perhaps I went too far with it this time.

[Paul Ebert]

So, if I insert an amplifier, do I connect the reference resistor to the speaker output of the amp? The diagram in the LIMP manual seems to indicate so. I am using a 10 watt resistor, so it should be OK if that is, in fact, what I want to do.

[Paul Ebert]

OK, so I got my M-Audio Transit soundcard and it seems to be working fine. But I got the following impedance measurement from one of my RS225-8s:



What might cause a woofer to not have a resonance?

[Dennis H]

Something's still wrong. No clue what.

[Paul Ebert]

Well, I figured it out. The voice coil seems completely frozen . The cone doesn't move at all.

[Dennis H]

Well, that would do it. Looks like your measurements are working now so the good news is you can see if something's wrong with a driver.

[Paul Ebert]

Yes, I am getting good impedance measurements. But, the woofer is trashed. I guess I'll need to buy another one.

Tomorrow, I'm hoping to get my frequency measurements.

[Dennis H]

PE is very good about replacing bad drivers for free. Call them.

[JonP]

Sounds like you've figured out the problem, at least with the RS225... It's gotta move to have a resonant peak!!

Does Arta have some kind of loopback test? It would be good to verify that the test setup is OK. You could just test an 8 ohm (or other low value known) resistor, to see if you get a flat and accurate response. If not, something about the measurement setup (card, wiring, probes) is not right. Some cards need a bit of delay to get things flat. Some cards are hopeless... hopefully not yours!

Check if the soundcard input monitoring (will pipe input signal out the output, very bad) is on, and disable it. That causes all kinds of weird response errors, etc... sometimes it mysteriously gets turned back on.

[JonP]

So, if I insert an amplifier, do I connect the reference resistor to the speaker output of the amp? The diagram in the LIMP manual seems to indicate so. I am using a 10 watt resistor, so it should be OK if that is, in fact, what I want to do.

Yep, you're on the right track. Measuring resistor goes in series with the amp and driver, and your input probes will go on either side of the resistor, just like in the soundcard output setup.

Assuming I'm getting the right picture of what the "reference" resistor is. I realized you could have meant "reference probe resistor" as in the reference soundcard input with a protection resistor, but that's still OK, in that the reference input goes on the amp side, and the measuring input goes on the driver side of the measuring resistor.

Clear as mud? :B

[Jed]

Well, I figured it out. The voice coil seems completely frozen . The cone doesn't move at all.

Same thing happened to me. Only it was on a pair of Accuton C79s!!! 8O

Luckily I was able to return them without questions. Anyway, back to the regular programmed schedule.

[Paul Ebert]

A pair! OMG!

Unfortunately, I bought these used and I recall that the box had a dent in one of the corners, so I suspect that the damage was done in transit from the previous owner. I wouldn't feel right returning it or even going back to the previous owner about it. It's also been quite a while since I got them.