The data you are showing here looks really noisy. I can't determine the cause because I do not know what ARTA data normally looks good.ravslanka said:
There are many ways to do things. I selected using a jig because I wanted consistent, controlled, test conditions without the risk of damaging my computer. Doing this give a better insight to the equipment I use for testing so that when I see some funny things, I know what to consider and what to ignore.
john k... said:
ravs,
I don't think soongsc means you need a multi test level jig like his, just that a jig of some sort is highly recommended (I agree).
What John and soongsc have both recommended is that you adjust the pulse delay setting in the MLS window until the phase response in your sound card loop-back test is flat. Like John said, start with 0.015 and increase or decrease the value until the phase response is flat.
Attachments
Some engineering humor:
A tech that had be assembling driver for many years finally wanted to get some sort of measurement system setup. Not having much experience with computers, he bought a second hand notebook, and called me up for some help because he wanted to get SoundEasy. I agreed to give him an old USB sound card that I was no longer using, and agreed to meet with him and help him get started with the software.
The day came, and we decided to meet at a MOS burger. He brought his notebook along and we sat down to turn the computer on. I pressed the power button, nor response. I tried a few times and finally asked him, "are you sure the computer is good?"; his reply was, "sure, my friend was using it before I bought it from him." Then I asked, "did you charge the batteries before you came?" His response, "What! You need to charge these things?"
End of part one.
A tech that had be assembling driver for many years finally wanted to get some sort of measurement system setup. Not having much experience with computers, he bought a second hand notebook, and called me up for some help because he wanted to get SoundEasy. I agreed to give him an old USB sound card that I was no longer using, and agreed to meet with him and help him get started with the software.
The day came, and we decided to meet at a MOS burger. He brought his notebook along and we sat down to turn the computer on. I pressed the power button, nor response. I tried a few times and finally asked him, "are you sure the computer is good?"; his reply was, "sure, my friend was using it before I bought it from him." Then I asked, "did you charge the batteries before you came?" His response, "What! You need to charge these things?"
End of part one.
I am using a M-audio firewire 410 and the trouble I am having is measureing imp.
When I measure imp anything below 20hz has lots of peaks
I have notice that it is only below 20hz everything above that is fine.
I am also using the power adapter .
Any suggestion of what it could be ?
I am starting to think that it could be my pci firewire card.
When I measure imp anything below 20hz has lots of peaks
I have notice that it is only below 20hz everything above that is fine.
I am also using the power adapter .
Any suggestion of what it could be ?
I am starting to think that it could be my pci firewire card.
hi
soongsc,johnk,lhwidget and ultrachrome
sorry i was out of my net connction broken fibre.
ok i played with pulse deay for inter channel latency with loopback with various ampling rates.most flat was with 44100 and spl smoothing of 1/24 db octave.but not as much smooth.
below are he screenshots.
44100 with pulse dlay of 0.0144
http://i44.tinypic.com/2nsz909.jpg
with 48000
http://i41.tinypic.com/2urov8g.jpg
and with 96000
http://i39.tinypic.com/2j2u7nm.jpg
and with driver conneted with amp
44100
http://i43.tinypic.com/24pb1i8.jpg
48000
http://i41.tinypic.com/ei5il1.jpg
is there any way to furter modify the reponse or smoothness.
icreasing spl smothing dips or increase the phase response.
thanks in advance for all you guys for all your efforts
ravs
soongsc,johnk,lhwidget and ultrachrome
sorry i was out of my net connction broken fibre.
ok i played with pulse deay for inter channel latency with loopback with various ampling rates.most flat was with 44100 and spl smoothing of 1/24 db octave.but not as much smooth.
below are he screenshots.
44100 with pulse dlay of 0.0144
http://i44.tinypic.com/2nsz909.jpg
with 48000
http://i41.tinypic.com/2urov8g.jpg
and with 96000
http://i39.tinypic.com/2j2u7nm.jpg
and with driver conneted with amp
44100
http://i43.tinypic.com/24pb1i8.jpg
48000
http://i41.tinypic.com/ei5il1.jpg
is there any way to furter modify the reponse or smoothness.
icreasing spl smothing dips or increase the phase response.
thanks in advance for all you guys for all your efforts
ravs
john k... said:
hi john k
this is impedance curve with 15 ohm test resistor
http://i42.tinypic.com/sc54bc.jpg
ravs
That pretty much says you are good to go. Nit picking, it looks like your sound card has some chanel to channel high frequency variation, but I wouldn't be too worried about it. It doesn't look like this would be the cause of the wiggles in the driver Z at high frequency you see. I would say at this point they are something associated with the driver.
Have you looked at othjer driver's Z measurements? Only way to tell for sure is to measure a number of drivers. Also, look at Z made with different signal levels.
Have you looked at othjer driver's Z measurements? Only way to tell for sure is to measure a number of drivers. Also, look at Z made with different signal levels.
Hey ravs,
Glad you're up & running. Hope you have fun with it
John,
I've always wondered, for impedance testing, what signal level do you use? I've seen one reference in Bohdan's manuals referring to use of a 2.83V signal level. The thing that drives me nuts is that I can move Fs up using lower levels and down by using higher test levels.
I'm just wondering if you have found a specific test signal level that will yield the best prediction of a driver's in-box response?
Soongsc, ravs, ultrachrome, oldbar, and John, I enjoyed the thread.
Glad you're up & running. Hope you have fun with it
John,
I've always wondered, for impedance testing, what signal level do you use? I've seen one reference in Bohdan's manuals referring to use of a 2.83V signal level. The thing that drives me nuts is that I can move Fs up using lower levels and down by using higher test levels.
I'm just wondering if you have found a specific test signal level that will yield the best prediction of a driver's in-box response?
Soongsc, ravs, ultrachrome, oldbar, and John, I enjoyed the thread.
I think I mentioned this before. The variation of Fs and Qts, etc with signal level is a function of the quality of the driver motor and suspension. Higher quality drivers don't show significant variation in T/S parameters with signal level. Cheaper drivers can show quite a bit.
I don't test at any particular level. Some manufactures state the level they test at but it also matters if it is MLS, impulse, swept sine, etc.
What you will often find is that the higher Fs also means higher Qts with the ratios such that it has little effect on box design. This isn't always the case as it is a function of how suspension compliance and Le vary with excursion. Best you can do for a box design is test at several levels and see what the T/S parameters are and how any variation effects box design.
I don't test at any particular level. Some manufactures state the level they test at but it also matters if it is MLS, impulse, swept sine, etc.
What you will often find is that the higher Fs also means higher Qts with the ratios such that it has little effect on box design. This isn't always the case as it is a function of how suspension compliance and Le vary with excursion. Best you can do for a box design is test at several levels and see what the T/S parameters are and how any variation effects box design.
I'm not sure how you set the delay values, nor do I know what other values you used. But they don't look like the way I recommended in the section that I mentioned.ravslanka said:
Additionally, to repeat myself again:
soongsc said:
If I don't see the results of the setup I recommended, there is no way I can help diagnose the problem.
ravs,
I think soongsc is right. After looking at your sound card loopback test results, the undulation (wave) at the high end of the phase response plot looks similar to the undulation in the driver's phase response.
The 96 kHz sampling rate loopback test shows obvious problems.
soongsc,
Are you thinking about trying the loopback tests now with the pc optimized?
I think soongsc is right. After looking at your sound card loopback test results, the undulation (wave) at the high end of the phase response plot looks similar to the undulation in the driver's phase response.
The 96 kHz sampling rate loopback test shows obvious problems.
soongsc,
Are you thinking about trying the loopback tests now with the pc optimized?
lhwidget,
Optimizing the pc does not effect the interchannel latency as far as I have experienced. But if data are not consistent and the impulse location is not at 0, the optimizing the pc for audio is the first thing I look into.
"Setting the interchannel latency" section is where I put an example on determining the delay increment. If this works, then good, if not, then I will investigate to see what the problem is.
Optimizing the pc does not effect the interchannel latency as far as I have experienced. But if data are not consistent and the impulse location is not at 0, the optimizing the pc for audio is the first thing I look into.
"Setting the interchannel latency" section is where I put an example on determining the delay increment. If this works, then good, if not, then I will investigate to see what the problem is.
soongsc said:
hi soongsc,
sorry for delay facing network problem no link for 2 days.
here are the screenshots with loopback ,rectangular, no spl smoothing
w/o pulsedelay
http://i42.tinypic.com/10ne1l2.jpg
with pulse delay
http://i41.tinypic.com/x0u1ck.jpg
it is flattest one .starting increasing from o.o120 to 0.0155
ravs
soongsc said:
hi soongsc,
here it is
http://i43.tinypic.com/2i88a6a.jpg
iam using creative pci128 card based on ct5888 chipset
tested wth yamaha pci card too which is ymf 724v-e series with tea2025 amplified line out.on board sound cards line in left channel is dead.
ravs
hi soongsc,
here are the screenshots with yamaha sound card
after calibration of sound card
withoud pulse delay
http://i44.tinypic.com/vo3ihh.jpg
with pulse delay
http://i43.tinypic.com/2ly0i6a.jpg
calibration scopes are these ok
http://i44.tinypic.com/34g0w8z.jpg
and with sampling rate of 96000
http://i44.tinypic.com/htzgw0.jpg
ravs
here are the screenshots with yamaha sound card
after calibration of sound card
withoud pulse delay
http://i44.tinypic.com/vo3ihh.jpg
with pulse delay
http://i43.tinypic.com/2ly0i6a.jpg
calibration scopes are these ok
http://i44.tinypic.com/34g0w8z.jpg
and with sampling rate of 96000
http://i44.tinypic.com/htzgw0.jpg
ravs
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.