help measurin imp in soundeasy

[ravslanka]

Please do that following and post the results of loop back.
1. check "Show Phase"
2. set window to "rectangular"
3. set smoothing to "none"

hi soongsc

here is the screenshot of loopback with amp hooked

http://i44.tinypic.com/1zewdhw.jpg

 

[ravslanka]

This is not an amp or wire wound resistor problem, unless the amp has a flating ground? Ahigh inductance resistor whoud show Z rising with frequency not decreasing.

The SPL loop back looks ok.

The next question is about cables and leads and probes. First try reversing the probes. That is, change move the probe on the In cable to the Ref cable and vise versa. If one of the probes is not correctly made it could interact with the cable capacitance and give the high frequency roll off. Also, look for poor grounds and make sure the interconnects for the measurement are exactly the same.

hi johnk

here is the screenshot with reversed probes

http://i42.tinypic.com/fmjrrr.jpg

do i need to change the cables

thanks

ravs

 

[soongsc]

hi soongsc

here is the screenshot of loopback with amp hooked

http://i44.tinypic.com/1zewdhw.jpg

Notice how the phase droops? If you follow the steps to
set the "pulse delay" that I described in the "My SoundEasy User Manual" the section "Setting the interchannel latency". You will get the phase to be totally zero (flat). Then you have this part calibrated. The manual file is in the Yahoo SoundEasy User Group.

I'm getting to sound a bit like tech support...

 

[soongsc]

You know lhwidget, I had run into all this before. That's why I just sorted some standard procedures to people getting in can use as a jump start calibration.

Usage of a Jig, which is a modified wallin jig, is very valuable to reduce possible connection errors. Got this from runing SpeakerWorkshop.

Once the system is set up. There are really lot's of things one can do using the tools that SoundEasy provides. I've actually taught a course in speaker design using this software.

 

[ravslanka]

You know lhwidget, I had run into all this before. That's why I just sorted some standard procedures to people getting in can use as a jump start calibration.

Usage of a Jig, which is a modified wallin jig, is very valuable to reduce possible connection errors. Got this from runing SpeakerWorkshop.

Once the system is set up. There are really lot's of things one can do using the tools that SoundEasy provides. I've actually taught a course in speaker design using this software.

hi lhwidget,

i will re read the your manual and come up with the results

thanks

ravs

 

[soongsc]

Section 5 and section 6 cover all the initial setup part prior to using SoundEasy.

 

[ravslanka]

soongsc,
I think you may have hit the cause in your earlier post. Latency issues/problems. I tried this freeware program:

http://www.thesycon.de/eng/free_download.shtml

It's called DPC latency checker (you may know a better one). I found something on my laptop that grabbed the system every minute or so (I suspect the infra red port).

I created a new hardware profile with the following turned off:
1) on-board sound chip
2) ethernet port
3) wireless port
4) I-R port
5) parallel port
6) serial port

I think this may help ravslanka, it cleared up two things on my laptop:
1) Occasional measurements that didn't make sense (obviously incorrect imp & SPL measurements taken when the offending device grabbed the system)
2) Before turning everything off with the second hardware profile, even when I got seemingly good results, the impedance would either drop on the high end, or not rise as high as it should have. Not as much as your impedance plot rav, but in the same way.

The attached pic is the obvious latency problem showing on my laptop. It took 1 to 2 minutes for it to show.

Just another thought (follow soongsc's & John's recommendations first)

Jay

hi lhwidget,

i have tried the latency checker. it is running normally while running
soundeasy and running mls.showing 146 ms as absolute maximum

after that i disabled everything that is not needed at the time of measurement ie onboard lan/audip/parral port/ serial port/fdc/
but that didnt help

ravs

 

[john k...]

hi johnk

here is the screenshot with reversed probes

http://i42.tinypic.com/fmjrrr.jpg

do i need to change the cables

thanks

ravs

I didn't mean to reverse the connections, just switching the probes on the end of the cables. In this picture you obviously have the IN and REF channels reversed. Your original connection was correct.

Try this. Do you have a 1k ohm resistor? If so, use the 1k resistor as the Ref resistor and if you have a 100 ohm resistor use it for the resistor to be tested. Then make a measurement but do not use any preamp, amp or probes. See the figure below for the set up.

An externally hosted image should be here but it was not working when we last tested it.

Connect the resistors in series. Then connect the hot side of the output of one channel of the sound card to the free end of the 1 k resistor. Connect the ground side of the output to the free end of the 100 ohm resistor. The connect the left channel sound card input to the hot side of the 1 k resistor. This will be ref input. Connect the right channel sound card input between the 1k and 100 ohm resistors. Change the reference resistor value to 1 k ohm in SE. Then make a measurement. You may not get exactly 100 ohms, but the result should be constant.

This test will remove all external components (amp, preamp, probes) from the result.

If you don't have exactly 1k ohm use something close. Same for the 100 ohm resistor.

I do not think you have a latency problem, or at lease the impedance roll off is not related to such. If you did , it would only effect phase, not the roll off of Z as F increases. Your SPL loop back was ok for amplitude. To check latency redo the SPL loop back and look at the phase too.

 

[ravslanka]

I didn't mean to reverse the connections, just switching the probes on the end of the cables. In this picture you obviously have the IN and REF channels reversed. Your original connection was correct.

Try this. Do you have a 1k ohm resistor? If so, use the 1k resistor as the Ref resistor and if you have a 100 ohm resistor use it for the resistor to be tested. Then make a measurement but do not use any preamp, amp or probes. See the figure below for the set up.

An externally hosted image should be here but it was not working when we last tested it.

Connect the resistors in series. Then connect the hot side of the output of one channel of the sound card to the free end of the 1 k resistor. Connect the ground side of the output to the free end of the 100 ohm resistor. The connect the left channel sound card input to the hot side of the 1 k resistor. This will be ref input. Connect the right channel sound card input between the 1k and 100 ohm resistors. Change the reference resistor value to 1 k ohm in SE. Then make a measurement. You may not get exactly 100 ohms, but the result should be constant.

This test will remove all external components (amp, preamp, probes) from the result.

If you don't have exactly 1k ohm use something close. Same for the 100 ohm resistor.

I do not think you have a latency problem, or at lease the impedance roll off is not related to such. If you did , it would only effect phase, not the roll off of Z as F increases. Your SPL loop back was ok for amplitude. To check latency redo the SPL loop back and look at the phase too.

hi johnk

here is screenshot with 1 k as ref and 100.2 ohm test resistor both are metal film types.

impedance
http://i40.tinypic.com/6ooc44.jpg

and for spl measurement

http://i39.tinypic.com/2vlti6q.jpg
ravs

 

[soongsc]

Still have not compensated for interchannel latency. Don't expect to get the right results.

 

[ravslanka]

Still have not compensated for interchannel latency. Don't expect to get the right results.

hi soongsc

i dont want to make that jig.right now i justwanted sound easy to measure the drivers.not planning to measure frequency response or spl measurements yet.read alot of manuals of speakeshop,lsplab,lspcad,sample champion,and arta iam totally confused.i want to test drivers with fixed power level 15 to 20 watt not various power levels .

 

[soongsc]

You don't need to use a jig. But when you do the loop back as john k mentioned, you need to do the interchannel latency compensation.

 

[john k...]

hi johnk

here is screenshot with 1 k as ref and 100.2 ohm test resistor both are metal film types.

impedance
http://i40.tinypic.com/6ooc44.jpg

and for spl measurement

http://i39.tinypic.com/2vlti6q.jpg
ravs

The SPL result does show a latency problem. I would suggest that you try a 0.015 msec pulse delay. Fine tune it so that the phase in the SPL loop test is flat. I did some tests and latency does effect both magnitude and phase of a Z measurement. However, it effects only phase in the SPL measurement. Soongc was correct on this.

 

[john k...]

hi soongsc

i dont want to make that jig.right now i justwanted sound easy to measure the drivers.not planning to measure frequency response or spl measurements yet.read alot of manuals of speakeshop,lsplab,lspcad,sample champion,and arta iam totally confused.i want to test drivers with fixed power level 15 to 20 watt not various power levels .

Don't read too much. I don't think you really want to test at such high power levels. 20 watts will yields an SPL of around 100dB for a typical driver. You would stand a real change of damaging some driver at that level with an MLS signal.

 

[ultrachrome]

When I was having problems getting consistent data out of SoundEasy, I did try ARTA for impedance measurements. It's actually pretty simple to use and you can use the same probes.

It uses a stepped sine wave (I think that is what it's called) which takes a fair bit longer than MLS. Just check the preferences and tell it which channel is probe and references. I didn't take more than a minute or for me to get it working.

I recommend trying it to see if you get a similar result.

 

[soongsc]

You can do the same thing in SoundEasy using analog sweep. Just less flexibility in data interpretation and usage. I use this sometimes if I am interested in finding out how sine wave tests will differ from MLS type signals. It helps to understand what's going on, specially in SPL testing.

 

[ultrachrome]

You can do the same thing in SoundEasy using analog sweep. Just less flexibility in data interpretation and usage. I use this sometimes if I am interested in finding out how sine wave tests will differ from MLS type signals. It helps to understand what's going on, specially in SPL testing.

ARTA was my sanity check. I was getting quite frustrated with SoundEasy. Turns out most of my problems where user error. I still get an odd result 10% of the time.

Downloading your guide now.

 

[ravslanka]

ARTA was my sanity check. I was getting quite frustrated with SoundEasy. Turns out most of my problems where user error. I still get an odd result 10% of the time.

Downloading your guide now.

hi ultrachrome

j have already tried arta very straight forward working with both probes but i was thinking to go with sound easy as it is the complete
sol for audio testing.screen shot of the same driver used in sound easy

http://i39.tinypic.com/hvqotj.jpg

it did show difference when using sin wave and pink noise with pink noise fs was 89 and that for sine wave 84 hz q parameters were different too very huge difference
above is the one measured with pink noise.

hi soongsc,
is there no easy way out avoiding the jig for different power levels

ravs

 

[john k...]

Different test signals will yield different results regardless of the software. DIfferent signals will also yiled different results. This is largely a function of the driver motor structure/suspension and linearity there of. Well made drivers with high quality linear motor and suspension show less drift with signal level/type.

 

[ultrachrome]

My suggestion was to see if other methods yielded an expected result.

Wasn't suggesting relying on another method or software to accomplish your goals.