Home Theater Forum and Systems banner

Impedance Measurement with REW

18K views 26 replies 3 participants last post by  jtalden 
#1 ·
#2 ·
Text Diagram Line Parallel


Does this help? I just added the ground wires. It shows that the RSense needs to break into the positive wire output to the speaker at some point.

I just did this with the help of a barrier strip like the ones linked below to connect up all the wires. I just cut the ends of standard wires to connect to the barrier strips as needed. Some people put the resistor in a box with connectors installed to so standard wires can be plugged in.

https://www.radioshack.com/search?q=barrier+strip
 
#13 ·
Hi everyone,

I have just complete my jig to do some impedance measurement of a Great Plains Audio 416-8B speaker in a classic 250L bass-reflex box. My jig, as far as I know, is wire as the one p.36 of REW user manual. The only thing is that I don't have a 100R resistor with me. I tried today a 150R and a 75R. Could I do it with these values via the headphone output?

This said, I was not able to do a proper impedance measurement. Result were not realistic. Headphone output is able to create a signal which goes to the speaker, but the signal is quite low. Plus, I don't get any "In" and "Ref in" signal. I tried to exchange different connection but it still doesn't work.

A small difference compared to Jtalden picture below is that my headphone only has one ground wire, not two. I connected this one with speaker negative terminal.

Any help is welcome.

Thanks,

Vieux Knowl


View attachment 112969

Does this help? I just added the ground wires. It shows that the RSense needs to break into the positive wire output to the speaker at some point.

I just did this with the help of a barrier strip like the ones linked below to connect up all the wires. I just cut the ends of standard wires to connect to the barrier strips as needed. Some people put the resistor in a box with connectors installed to so standard wires can be plugged in.

https://www.radioshack.com/search?q=barrier+strip
 
#8 ·
Assuming you mean; using an inductive resistor:
The error amount using a wire-wound resistor depends greatly on its value and the values of all the other elements in the measurement circuit. A circuit analysis would be necessary to determine the impact.

It is only a big deal if it effects your standards for accuracy. So, I don't have any way to answer your question.

Typical carbon or metal film resistors are common and very inexpensive.

That said, if this is a first time casual project for the fun of learning the process, then don't worry about it.
 
#9 ·
Thanks for all the help guys - I'm making some progress. :)

Now I'm trying to take into account the resistance of the test leads.

The tutorial says:

"When measuring very low impedances the resistance of the test leads may become significant. To measure this, first calibrate the rig as above, then make a measurement with the leads shorted together at the point the load is attached (set the RLEADS value to zero before making the measurement). The measurement should be fairly uniform, perhaps showing variation at very low frequencies depending on the low frequency limitations of the headphone drive stage. If the result is more than one or two tenths of an ohm, check that the connections are tight and the leads are not too flimsy. Enter the resistance from the flat part of the measurement into the RLEADS box on the measurement panel."

Where exactly must I "short" the leads? Before the resistor? After the resistor? And how exactly do I short them out i.e. what do I connect to what? (a diagram would help if possible) :)

Thanks,



Ryan

:wave:
 

Attachments

#10 ·
Okay,

I'm actually getting a measurement now.

I only have a 1,000 own resistor so maybe that's throwing everything out of whack!!!

This is the reading I get without any smoothing. Seems something is very wrong.

Could it be that the vibration of the speaker lying on my table is causing problems?

Any suggestions?
 

Attachments

#12 ·
High Compared to what? Have you compared it to the manufacturer's spec sheet? There is a wide range of normal depending on the design of the driver. If the TS values are in the same general range as the spec, that is all that can be expected. There is often a significant difference between published values and measured values.

I have not reviewed or used this feature in several years and my recall is often not reliable, but my thoughts are...
> The shape of the last measurement is normal. The ones with lots of noise are not correct unless the driver is defective.
> If the calibration process of the measurement setup was followed and worked as expected, then the chart values are probably correct.
> The values are normally taken using a small signal and the process is best done that way. Cone motion should be very small for the test. I mention this noting that your label indicates 'High Output'. Hopefully you followed the suggested test levels.

Just a side note:
Normally we would not twist the leads to the speaker. The tighter the twist and greater the number of turns, the greater the inductance of the cable will be. This is not tightly twisted and is very short so it is probably not a factor here. I just mention this FYI.

Normal zip cord (untwisted) works fine for speaker connections.
 
#14 ·
Yes, there is normally a shared ground in a headphone jack so my picture is wrong in that regard. A single ground wire is used for both channels. Since the left channel is not used in this case that positive line is just left open and the common ground is connected as shown. It is not only connected to the speaker, but also to the left and right channel input ground wires.

In practice, I did something very similar to that shown in Post 10. I used cables that matched the input and output plugs on my soundcard and then cut them as shown in that post to make the needed connections at the junction where the resistor is. I just used a barrier strip instead of soldering and taping the connections.
 
#15 ·
I'm not sure about the corresponding fonctionnality of each wire of my headphone cable. I've got red, black and white. I put the red to positive, white to negative and black to ground. Good?

There is still the question of the resistor. Can I use a 75R or 150R instead of 100R with my headphone ouput?

I have done some more test and still no realistic impedance curve... I still have no "In" response or "Ref in". Normal?

Pretty discouraging....

VK
 
#16 ·
I think either size resistor is probably okay to use.

I don't think there is any standard for the wire color scheme. The plug 'tip' wire is the left positive, the ring wire is right positive and the sleeve wire is the ground for both channels. If the plug is a 1/4" TRS type it can be opened and inspected. If the plug is molded as most smaller plugs are then you would need an ohm meter or use another method to confirm the color scheme.
 
#18 ·
There are hints that you may be setup correctly for the measurement, but you are correct that curve is not as defined and smooth as expected for bass reflex design. Something is going on. Did the calibration curve of the impedance measurement setup look normal? Do your near field audio measurements with a mic look normal? If so then it is maybe a setup or mechanical issue with the speaker itself rather than a measurement issue.
 
#22 ·
The level is way too low. Confirm that the gain of the 2 inputs and the headphone output are acceptable using 'check levels' before the measurement. If the level does not adjust with changes to input and output gains on the M-Audio then the wiring is still not correct.

I have not had to troubleshoot these types of problems with the impedance measurement process so hopefully JohnM or someone else here with more experience has better suggestions.
 
#24 ·
Some precisions:

1) with my 1/8 jack connected to headphone plug in my soundcard, I've got the "In" in the red zone without any possibility to decrease it. And I've got "In" at -59 dBFS and "Ref In" at -63 dBFS;

2) with my 1/8 jack connected to "Insert" (channel 1) in my soundcard, "Output" level's the same, but I get an easy adjustment at -26 dBFS in the "In". "Ref in" still around -70 dBFS. Impedance measure with this setting is like the one below.
It seems that it recognized my 74R4 resistor. Good.

3) with my 1/8 jack connected to "Insert" (channel 2) in my soundcard, "Output" level's the same, but I get an adjustment at -26 dBFS in the "Ref In", with nothing more in the "In".

Still working.

VK
 

Attachments

#25 ·
I only have experience with using the 2 line inputs, not insert jacks.

I found the attached sketch in my setup notes:
> Headphone output used a TRS/RCA adaptor and then an RCA splitter.
> Line inputs used RCA/TS adaptors.

This method avoided confusion of which wires to use as the adaptors control that issue. Also, cheap RCA cables were cut up for connections at the barrier strip. I can post a photo of my wiring harness if necessary.

View attachment Impedance Jig connections for measuring T-S Parameters.pdf
 
#26 ·
Hi there,

Well, one week later, I was able to get my impedance jig working. Thanks to Jimbee on the French forum Mélaudia who helped me a lot. Many problems were discovered. The first is regarding one of the three cables in my headphone cable which was not conducting! Then, I realized that I was not using my input on my M-Audio/M-Track. I was using RCA output. Finally, I used the two "Inserts", one by channel, with 1/4 TRS jack on each using only their R=ring and S=sleeve.

I provide a picture of the set-up and a measure to what it gives. Please note that the impedance jig is in the center of the picture. On the left side's a filter for my speakers.

Thanks,

VK Subban
 

Attachments

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top