Hi,
I've made myself one of those little impedance jigs that uses a reference resistor and the in/out on a sound card, but have very mixed results, often extremely inaccurate or close but not quite. I'm fed up and ready to by a WT2 or DATS.
The issue I have with buying one of these systems is I'd like to be able to take impedance measurements with different drive levels and get accurate results. I'm not sure these systems are all that accurate.
Are there any retail systems that are driven with an amplifier rather than USB? Is there any way to make a system like I want that is simple and safe for my computer? I am terrible with soldering and making circuits. I could buy one from someone, but making something complex might be tough.
Thanks for any help.
Ryan
I've made myself one of those little impedance jigs that uses a reference resistor and the in/out on a sound card, but have very mixed results, often extremely inaccurate or close but not quite. I'm fed up and ready to by a WT2 or DATS.
The issue I have with buying one of these systems is I'd like to be able to take impedance measurements with different drive levels and get accurate results. I'm not sure these systems are all that accurate.
Are there any retail systems that are driven with an amplifier rather than USB? Is there any way to make a system like I want that is simple and safe for my computer? I am terrible with soldering and making circuits. I could buy one from someone, but making something complex might be tough.
Thanks for any help.
Ryan
Even a simple impedance jig driven by the headphone output of a sound card can be very accurate if the software you are using allows you to calibrate out the effect of the input impedance of the sound card. (LIMP from the ARTA test suite allows you to do this).
It is easy to measure the speaker impedance at higher levels if you are prepared to add a simple chip-amp to drive the impedance bridge. You will of course have to add well-matched attenuators to the reference and the measuring inputs of the sound card to drop the signal level. The facility to calibrate out the effect of the input impedance also means that you can get away with normal 1% resistors in the attenuators, as they then don't have to be perfectly matched.
It is easy to measure the speaker impedance at higher levels if you are prepared to add a simple chip-amp to drive the impedance bridge. You will of course have to add well-matched attenuators to the reference and the measuring inputs of the sound card to drop the signal level. The facility to calibrate out the effect of the input impedance also means that you can get away with normal 1% resistors in the attenuators, as they then don't have to be perfectly matched.
Hmm, I'll have to look into this. I'm using LIMP, but it hasn't been accurate for me. Perhaps I've still not set it up right. I don't remember entering the input impedance of my soundcard. I don't even know what the input impedance is.
Ok, see that's a little over my head. I'd like to do this, but don't know what it is. I'd like to measure impedance at different drive levels. Is there a place I can find how to do this? Or buy a device that does this?
Thanks.
I'll capture the settings for the LIMP calibration later on today if I get the time. One thing you must remember if using the output from the sound card on its own is that it has to be able to drive an undistorted signal into the load that the jig and speaker presents. Eg, if you are using a 10 Ohm reference resitor in the jig, and the speaker impedance might drop to 4 Ohms somewhere in the range, then the sound card has to be able to drive 14 Ohms. Not all headphone outputs will be able to do this.
Always perform a "sanity check" by connecting (say) a 100 Ohm resistor in place of the speaker initially and check that the measured result is an impedance of 100 Ohms flat across the whole measuring frequency range. Also make sure that you are using the correct connections to the L and R input channels of the sound card from the impedance bridge. If they are swopped over you will get meaningless results.
Always perform a "sanity check" by connecting (say) a 100 Ohm resistor in place of the speaker initially and check that the measured result is an impedance of 100 Ohms flat across the whole measuring frequency range. Also make sure that you are using the correct connections to the L and R input channels of the sound card from the impedance bridge. If they are swopped over you will get meaningless results.
I guess you have downloaded the LIMP user manual from the ARTA site? Figures 3.1 and 3.2 in there (in the V1.7 user manual anyway) show the simple connection and the connection using a power amplifier to test the impedance at higher signal levels. I see that a 100 Ohm ref resistor is shown for the simple case, which all headphone outputs will be able to drive.
Section 4.9 in the manual talks about calibrating the jig to the sound card. I perform this slightly differently to the way described in the manual. Rather than connecting the L and R inputs to the sound card together, I leave them attached to the jig, but with no load connected to the terminals on the jig where you would normally attach the speaker to be measured, then the calibration run will automatically compensate for the slight attenuation caused by the ref resistor feeding into the input impedance of the sound card. To be honest, when measuring normal 4 or 8 Ohm speakers the error will be small (most soundcards have an input impedance of around 10k Ohms), but with 70/100V line PA speakers which I work with, it can give a significant impedance error at high frequencies if not calibrated out.
Section 4.9 in the manual talks about calibrating the jig to the sound card. I perform this slightly differently to the way described in the manual. Rather than connecting the L and R inputs to the sound card together, I leave them attached to the jig, but with no load connected to the terminals on the jig where you would normally attach the speaker to be measured, then the calibration run will automatically compensate for the slight attenuation caused by the ref resistor feeding into the input impedance of the sound card. To be honest, when measuring normal 4 or 8 Ohm speakers the error will be small (most soundcards have an input impedance of around 10k Ohms), but with 70/100V line PA speakers which I work with, it can give a significant impedance error at high frequencies if not calibrated out.
Thank you for this information Ouroboros. I'll take another look at those sections of the manual. Been a while since I've looked. Reading how you described calibration, I know I didn't do that. So I'll try again. Thanks.
Long ago I had the exact same problem you are having, except that I was using SpeakerWorkshop and a Wallin Jig. After much trial and error and many posts, someone told me that most soundards have some kind of feedback inhibition which affects the recorded signal if you're recording and playing at the same time. Long story short - I gave up on impedance measurements through my soundard and spent $100 on Dayton's WT3 and haven't regretted it.
I was able to find this old thread, but not the one where someone explained the soundcard feedback loop problem.
http://www.diyaudio.com/forums/equipment-tools/185515-about-give-up-wallin-jig-speaker-workshop.html
I don't mean to discourage you. I would certainly try Ouroboros' recommendations before giving up.
I was able to find this old thread, but not the one where someone explained the soundcard feedback loop problem.
http://www.diyaudio.com/forums/equipment-tools/185515-about-give-up-wallin-jig-speaker-workshop.html
I don't mean to discourage you. I would certainly try Ouroboros' recommendations before giving up.
I use an E-MU 0404 usb external sound card both at work and at home with no problems. I have used an M-audio Transit usb sound card and that also worked well. You need to ensure that there is no internal digital path between signal out and in, so check the Windows mixer settings! As I said earlier, a quick check with a known resistor as the load will give you confidence that everything is working (or not...)
I also had this problem with a Wallin jig I made (accurate R measurement one time inaccurate the next). The problem was the switches. I changed them to more expensive low resistance ones and the problem went away completely.
Last edited:
Thanks everyone. I'm still going to try and get this sorted out, but a WT3 came up in the swap meet section, so I jumped on it. Looking forward to finally having reliable results.
Thanks again for all the suggestions.
Thanks again for all the suggestions.
- Status
- Not open for further replies.