Hello guys.
I recently decided to get the Arta software (primarily to measure T/S parameters in LIMP) and to use it i had to construct the Arta box. So i did. There are some issues with the measurements, tho, and i am not entirely sure what is going on. I will try to describe what i am doing and the results i am getting. I have the M-Audio Audiophile 2496 sound card, that has been (apparently) used successfully before with ARTA/LIMP.
1. I connect everything up- the card input (via 2 rcas) goes to the box, the output (2 more rcas) goes to the amplifier. I have speaker cable going from the amp to the box and another one going from the box to the speaker being tested (when i calibrate i take out the speaker load and the switch is at calibration).
2. I set up the software- first the sound card drivers- i tried both with the WMD and Asio drivers and the results are pretty close, also the wave format is at float; then i set up the reference resistor value (27) and ref channel (left). Also disabled the direct monitoring from the asio control panel; didn't mute any channels for now, as the m-audio drivers are a bit confusing to me and there isn't a mic channel.
3. I calibrate- there is a nice 0.03db difference between both channels, which i read is what it is supposed to be; levels are not clipped, hence green.
4. I run the record, using stepped sine at full 20Hz-20kHz and then go to Analyze - Loudspeaker parameters - to calculate parameters. Also tried the added mass method with 10g (2x20p coins, 5g each, taped on the membrane), just to see what happens.
Here are my issues with the measurements: the parameters i get look a bit suspicious.. Z1 looks too high (not sure here) and the program does not give me anything for Qes, hence no Qt and the Le number looks outrageously high to me.
Here are some images: Sound card drivers, Calibration, Measurement setup, Parameters, Added mass overlay, Impedance data (Z1,Fs)
As i am completely new to this, i understand there is a learning curve to be expected. I really appreciate the help. Thank you in advance.
I recently decided to get the Arta software (primarily to measure T/S parameters in LIMP) and to use it i had to construct the Arta box. So i did. There are some issues with the measurements, tho, and i am not entirely sure what is going on. I will try to describe what i am doing and the results i am getting. I have the M-Audio Audiophile 2496 sound card, that has been (apparently) used successfully before with ARTA/LIMP.
1. I connect everything up- the card input (via 2 rcas) goes to the box, the output (2 more rcas) goes to the amplifier. I have speaker cable going from the amp to the box and another one going from the box to the speaker being tested (when i calibrate i take out the speaker load and the switch is at calibration).
2. I set up the software- first the sound card drivers- i tried both with the WMD and Asio drivers and the results are pretty close, also the wave format is at float; then i set up the reference resistor value (27) and ref channel (left). Also disabled the direct monitoring from the asio control panel; didn't mute any channels for now, as the m-audio drivers are a bit confusing to me and there isn't a mic channel.
3. I calibrate- there is a nice 0.03db difference between both channels, which i read is what it is supposed to be; levels are not clipped, hence green.
4. I run the record, using stepped sine at full 20Hz-20kHz and then go to Analyze - Loudspeaker parameters - to calculate parameters. Also tried the added mass method with 10g (2x20p coins, 5g each, taped on the membrane), just to see what happens.
Here are my issues with the measurements: the parameters i get look a bit suspicious.. Z1 looks too high (not sure here) and the program does not give me anything for Qes, hence no Qt and the Le number looks outrageously high to me.
Here are some images: Sound card drivers, Calibration, Measurement setup, Parameters, Added mass overlay, Impedance data (Z1,Fs)
As i am completely new to this, i understand there is a learning curve to be expected. I really appreciate the help. Thank you in advance.
Last edited:
I did not read everything (I merely skimmed through) however, your resistance value is too low. You should be using a minimum of a 100 ohm resistor. Also, you do not use an amplifier. The soundcard offers enough voltage to create the measurements.
Follow the steps below listed in the hyperlink
Measuring the loudspeaker parameters
Follow the steps below listed in the hyperlink
Measuring the loudspeaker parameters
This is the official jig offered by Arta's website. It states the reference resistor at 27 ohms. I am aware of the other jigs out there, also i know there is a way to do it without an amp. Using either method should work, however. I am gonna try the simple method without an amp, tho, just to see what the difference would be.
I recall always getting dodgy results when the impedance rise of the speaker exceeded the value of the resistor. Possibly ARTA has modified things with their newer versions. I do recall literally purchasing a 100 ohm Resistor because that was the requirement for LIMP years ago. These days I use anywhere from 100 ohm to 330 ohm resistors based on the impedance rise of the loudspeaker.
I am uncertain how you came to the conclusion that you have a measured resistance of 5 ohms in the Re (I am assuming you measured the driver using a multimeter) when a separate amplifier is going to add its own resistance in the equation once you have a separate amplifier within measurement chain. That may be the reason why, you are hovering nearly 2000 ohms.
I am uncertain how you came to the conclusion that you have a measured resistance of 5 ohms in the Re (I am assuming you measured the driver using a multimeter) when a separate amplifier is going to add its own resistance in the equation once you have a separate amplifier within measurement chain. That may be the reason why, you are hovering nearly 2000 ohms.
I see what you mean and it does make sense. But, as i mentioned before, i totally built the box their own website suggests for all measurements, so i don't know whether or not that's the problem. Gonna get myself a 100r 1% resistor and try the simpler jig without amplification and report the results.
I did measure the speaker's resistance with a DMM. It is a kef driver off a subwoofer and is supposed to be 4 ohms, but it measures 5.
However, as i am still learning to use the software, i am more concerned about not getting any data for the Qes. I am fairly certain LIMP is supposed to give you something that's not zero and i have no idea why it doesn't.
I did measure the speaker's resistance with a DMM. It is a kef driver off a subwoofer and is supposed to be 4 ohms, but it measures 5.
However, as i am still learning to use the software, i am more concerned about not getting any data for the Qes. I am fairly certain LIMP is supposed to give you something that's not zero and i have no idea why it doesn't.
Just splice some RCA patch cords, wire them together accordingly and stick a resistor on the end. It is simple and effective. You do not even need a switch suggested in the photo below. You can just connect the resistor directly to the loudspeaker terminals and, disconnect the resistor from the loudspeaker terminal for calibrating the sound card.
I recall the first time using LIMP I used the on-board soundcard on my Desktop and configured everything using 1/8" plugs. To this day, I follow this method and have not created a fancy box.
You are getting 0 on the QTS in addition to 0 on the QES because your wiring scheme is configured incorrectly. It may be the amplifier you added in the equation, it may be the box. Why not just start over fresh instead of dwelling why the fancy box is not working properly?
At a resistance of 5 ohms, there is no reason you are supposed to be hovering around 2000 ohms at all times unless something is adding more resistance in the measurement chain.
I recall the first time using LIMP I used the on-board soundcard on my Desktop and configured everything using 1/8" plugs. To this day, I follow this method and have not created a fancy box.
You are getting 0 on the QTS in addition to 0 on the QES because your wiring scheme is configured incorrectly. It may be the amplifier you added in the equation, it may be the box. Why not just start over fresh instead of dwelling why the fancy box is not working properly?
At a resistance of 5 ohms, there is no reason you are supposed to be hovering around 2000 ohms at all times unless something is adding more resistance in the measurement chain.
Last edited:
That's exactly the scheme i am gonna do. Will add the switch, tho, as i have one laying around. I plan to connect 3 of the 910 ohm 1% resistors i have in parallel to get the value of 303.33 and use that as reference. Should have went with the simple jig first and then build the fancy one, i knew that's a mistake 🙂
cheers
cheers
Okay. So i made this simple jig
but instead of soldering a permanent resistor, i added alligator clips so i can change the resistance or remove it for calibration. Connected it to the sound card and now it won't calibrate. Says there's 2.83db difference between both channels. Tried to calibrate with the fancy box and amp and it works fine, it calibrates and everything. Nevertheless, i tried measuring with the simple jig (303 ohms of resistance) and i still got the outrageous number for Le and no Qes / Qts. I am starting to think the box isn't the problem at all, maybe something is wrong with my cables or idk :O
EDIT: Turns out i can calibrate; using no amplifier, it needs some load (loudspeaker connected) to be able to calibrate. Measurements are still off, tho, at least in my opinion. At least this time the magnitude seems normal and i got a normal number for Le, but Qt=0.01 is really strange 🙂
Images:
An externally hosted image should be here but it was not working when we last tested it.
but instead of soldering a permanent resistor, i added alligator clips so i can change the resistance or remove it for calibration. Connected it to the sound card and now it won't calibrate. Says there's 2.83db difference between both channels. Tried to calibrate with the fancy box and amp and it works fine, it calibrates and everything. Nevertheless, i tried measuring with the simple jig (303 ohms of resistance) and i still got the outrageous number for Le and no Qes / Qts. I am starting to think the box isn't the problem at all, maybe something is wrong with my cables or idk :O
EDIT: Turns out i can calibrate; using no amplifier, it needs some load (loudspeaker connected) to be able to calibrate. Measurements are still off, tho, at least in my opinion. At least this time the magnitude seems normal and i got a normal number for Le, but Qt=0.01 is really strange 🙂
Images:
Last edited:
I can't seem to get it working properly. I don't think i am supposed to be adding load in the calibration process, but if i don't the software won't calibrate. Someone else has any ideas? I thought this is supposed to be pretty simple. I am doing everything by the book and getting totally weird results. I don't believe i made any mistakes this time, it is pretty simple and straight forward and i was pretty careful. Out of frustration, i might slice up some rca cables and take the number of components even lower.
Last edited:
You do not remove the resistor for calibration. The resistor value is taken into account once you add the value in the resistance value in LIMP.
Please re-read what I wrote:
Sound Card Calibration
An externally hosted image should be here but it was not working when we last tested it.
Measuring The Loudspeaker
An externally hosted image should be here but it was not working when we last tested it.
Do it exactly that way and do not add any modifications. No alligator clips, no additional amplifiers, no switches. Just how it is shown.
Last edited:
Looks like you are not able to generate the impedance curve so T/S param measurement is not possible
1) Ensure that "Setup->Audio devices" has the right devices selected
2) Switch to Pink PN, its quicker than stepped sine
2) Ensure that the drivers are correct/latest for the card
3) Ensure that your connections are good, especially dry solder
4) Ensure that your reference resistor value is entered at "Setup->Measurement"
5) With the simple jig connected, go to "Setup->Generator->Test". Post the snapshot of the screen here. In this condition, you should hear a faint noise thru the driver under under test. Disconnecting and reconnect the driver you should see the impact in this dialog, mostly on one channel
1) Ensure that "Setup->Audio devices" has the right devices selected
2) Switch to Pink PN, its quicker than stepped sine
2) Ensure that the drivers are correct/latest for the card
3) Ensure that your connections are good, especially dry solder
4) Ensure that your reference resistor value is entered at "Setup->Measurement"
5) With the simple jig connected, go to "Setup->Generator->Test". Post the snapshot of the screen here. In this condition, you should hear a faint noise thru the driver under under test. Disconnecting and reconnect the driver you should see the impact in this dialog, mostly on one channel
Something to note is the screen shots you have posted are showing the phase response of the measurement. The magnitude trace is off the top of the screen. Note that the cursor measurement at the bottom of the screen says that the cursor is at 20Khz and 1.1K ohms. Click on the "Fit" button on the right side of the screen to set the range of the display so you can see the impedance trace.
Something is not setup correctly. After you run calibrate try measuring a resistor with LIMP. Do not proceed with testing speakers until you can measure resistors in the 1 to 100 ohm range with repeatable results.
The calibrate function only matches the levels of the 2 channels of the sound card.
To verify that things are set up correctly use a high accuracy resistor in the 5 to 20 ohm range to check that LIMP is measuring correctly. The value of the sense resistor that you enter into LIMP can be varied up or down to fine tune the measurements. The actual value of the sense resistor is not critical, but should be in the 20 to 100 ohm range. I've been using 35 ohm for years...
If you build a measuring jig include switching so that you can verify that LIMP is working correctly each time you set up.
My jig (loosely based on the "Wallin" jig from speaker workshop 12 years ago) includes 4 ohm and 20 ohm calibration resistors so correct functioning can be verified quickly before making a measurement.
Something is not setup correctly. After you run calibrate try measuring a resistor with LIMP. Do not proceed with testing speakers until you can measure resistors in the 1 to 100 ohm range with repeatable results.
The calibrate function only matches the levels of the 2 channels of the sound card.
To verify that things are set up correctly use a high accuracy resistor in the 5 to 20 ohm range to check that LIMP is measuring correctly. The value of the sense resistor that you enter into LIMP can be varied up or down to fine tune the measurements. The actual value of the sense resistor is not critical, but should be in the 20 to 100 ohm range. I've been using 35 ohm for years...
If you build a measuring jig include switching so that you can verify that LIMP is working correctly each time you set up.
My jig (loosely based on the "Wallin" jig from speaker workshop 12 years ago) includes 4 ohm and 20 ohm calibration resistors so correct functioning can be verified quickly before making a measurement.
Okay, my bad, i know what i am supposed to do, just messed up my post.
This is exactly what i am doing. Anyway, i removed all modifications and made it cleaner. The result is the same. It won't calibrate. Gonna post screenshots later, cuz i am late for my basketball game now.
Hey.
1. I am pretty sure i have the right devices selected, cuz otherwise there is no signal to calibrate from.
2. I have tried Pink PN, with the same results.
3. There is only one set of drivers for Windows 7 x64 SP1 on the M-audio website and i have them installed and the ASIO control panel of the sound card is working.
4. I have both the left channel and the reference resistor selected in the measurements menu.
5. The generator is also used in the calibration menu and i have a screenshot of that from earlier. There's nothing significant to see here, the levels are both green (not clipped) and when i am using the simple jig i get more than 2db difference, hence i can't calibrate. Tomorrow when i connect it again i am gonna take a screen shot juts for the sake of it.
Thanks
As of right now, i can't calibrate using the simple jig. I can only calibrate using the fancy box (that is, with an amplifier in the circuit). I might have done something wrong with the wiring (although i was pretty careful), so i am gonna post some photos of the jig itself tomorrow.
I really like the idea of measuring resistors first, to confirm that the measurements are correct, before actually measuring speakers. Going to definitely do that once i get it working.
cheers
Another thing to note is LIMP can measure capacitors and inductors. Run an impedance curve of the part you want to measure, click to put the cursor at 1Khz or what ever frequency you want to know the impedance for, then click Analyze and choose "RLC impedance values at cursor position". LIMP will give you the resistance, inductance, or capacitance value plus the impedance and phase numbers at the cursor location.
The measurements match up nicely with my ESI RLC bridge.
Last edited:
Here are some pics of my simple jig and wiring.
Connections:
Soundcard:
Connections:
Soundcard:
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Some more pictures:
Audio devices and asio driver control panel:
Measurement setup menu:
Calibration / generator menu:
More than 2db error with current setup:
Audio devices and asio driver control panel:
Measurement setup menu:
Calibration / generator menu:
More than 2db error with current setup:
- Status
- Not open for further replies.