New speaker impedance tool, SimpeS

Being the cheapskate I am, I was looking for something like LIMP, a very nice tool, totally free to add to my cheapskate toolkit I am writing up to go with PSD-Lite. For myself, I have Woofer Tester II, so I don't need one.

I found SimpeS. Seems to export a three column file just fine.
AcustikA - Home

It does not plot until it is done, so when you run it, you are not sure it is working. Stepped sine, so like WT-II, not blindingly fast. Still in Beta.
 
WoW!!!!!

Thank You so much for posting the link to this software!!!! :)

I have just been looking for exactly this type of program in the last few days to plot Impedance's.
I have been using Visual Analyzer the find reactive values and such but I have no idea how to make it plot curves (yet).

This has made every thing that I am working on so much easier!!!!

I did a simple test of a transformer that I am testing for step-up use on ESL's and it seems to be right in line with the values I am getting from VA and one other program that I use for for finding values and impedance's with single frequency's.

I will check in more detail just to be sure though.

Here is the plot that I just did using a modified Antek AS-1206 with the 120V winding's open and with the magnetic and electrostatic shields removed.

What took less than 5 minutes saved me hours and hours and hours of having to plot each individual frequency manually by hand on some graph paper!!!!

Thanks again!!!!

:cheers:

jer :)
 

Attachments

  • antek1206 test.jpg
    antek1206 test.jpg
    88.3 KB · Views: 845
Last edited:
Glad to help. When I started, we did everything by single measurement and graph paper, so I know your pain.

Look at the values above 10K carefully. I used a 10 Ohm load resistor and I think the impedance droops. I am about to retest with 200 Ohms.

LIMP from ARTA is very nice too, but you need to send them a small shareware fee if you want to save the plot values. Easy and cheap if you happen to have a European electronic fee account. I use ARTA sometimes too, but going to figure out how to send a postal order to them is not on this weeks schedule. We really should as that's what keeps this stuff coming. SimpeS would like a donation.
 
I just did a sweep of a 67pf capacitor using a 1 megohm reference rsistor and it was pretty darn close.

It came up with about 125k at 20khz instead of the predicted 118k.
Even though this is a 5.6% error it is still only a matter of a few picofarads in this range at 63.5pf.
My cheapy digital meter registers it at 66pf.

The first time I tried measuring the capacitor I used a 100k reference resistor and the reading was way off by about 2 to 3 times.
If I can find my one and last 10megohm resistor I will try it and see if it is any closer.

Also, I just remembered I forgot to measure the resistance of the 1megohm resistor and this much error is typical of resistors in this range as well.

I have an old 1Ghz P3 that I am going to setup just for doing such measurements.
I would love to back into programming and have been doing so very slowly.
It seems that none of the programs that I have used so far takes advantage of the 24bit sound cards and that would be really something if they did.

Visual analyzer is supposed too, But I have not had good luck with it and it crashes unless I keep it in 16bit mode.

jer :)
 
Last edited:
I hear ya jer,
If Zelscope would do 196K! That would be handy. 24 bit is not that big a deal, I want sample rate.

I an emu 1616m and M-Profire that were 192, 24 bit, but both had issues that made them unsuitable for speaker building. I dropped to a much cheaper Focusrite 2i2 USB 96K as at least it works. For speakers it is fine, to look at an amp, I guess the old Tektronix is still my only bet.

I can crash my pretty snappy AMD running an RTA in fast mode for a period of time. All those FFT's start to mount up. Dropped in another 2G to see if that helps. Funny I went from 1.8 to 3.2 available. Seems it decided to take more for the video since it was there.
 
Yes,at least Va does at least do 192Khz and that is a big plus.
Lately I have been thinking about just building a faster AD/DA acquisition system.
I have been learning how to interface to the PC and this is not as hard as it seems.

The trick will be associating the hardware as sound card drivers so that our common software's can access it.
This sounds easy to me but then again I'm not a programmer at least not since the days of BASIC and the 8bit machines.

My next project is to add DC measurement using an old modified SB16 card and if one were to be able to add all of these really cool functions into a program that doesn't crash because of windows or what ever other reason would be so cool!!!

Machine language anyone?!!!!

Today our Cpu's run at +3Ghz and it seems to me that a very simple 8bit or even 16bit A/D interface could be made to run fairly fast for just a few bucks!
I used to do this on the older 8bit machines running at only 2Mhz !!!

I know this is done using the USB interface.
But I want to stay away from the latency issues of the USB way of doing things and use the straight forward parallel buss method way of doing things.
Even on a PCI buss of 32 to 100 Mhz would be fine with me.
I even have a few ISA machines sitting around doing nothing as well and these would be a few prime candidates for something like this as well.

Sorry to go of topic.
Just a few thoughts. :)

FWIW

jer :)
 
Last edited:
USB-2 is a lot better than it was. I was warned by Bodzio to keep with firewire as it was a streaming protocol, but alas, it is all but dead. Sweetwater support told me the USB-2 drivers have gotten much better in the last year. So far, so good on my test set, where I am playing with SimpeS. Some of it is also thread management in the kernel. We lost the best real time OS in True64 when HP destroyed what was left of DEC.

We are all waiting for the new HD compatible protocol, but I bet it is packet based too. The answer used in studios is a shared external clock.

So, ya want to play geezer?
I'll match your Basic and raise you 2911 bit slice machine code. :)
 
I was not getting a nice graph with a 10 Ohm resistor as I used in all my other tools. The values were quite a bit off and the phase was drooping at the high end. I switched to 200 Ohms and it is very close to the values from WT-II.

Tip, go to view and turn on the status bar. You will see it is actually doing something before it starts making noise.
 
I use Holmimpulse, driving the speaker through a 1k resistor and connecting the speaker voltage as line input. Calibration is by measuring a 10R resistor. You get a very nice graph.

They make it hard to donate and get the "save" function as they only take an European money pay thingie. If they did Pay-Pal, they would have my money now. HOLMImpulse is a very good tool.
 
Been here too long. LIMP is where it hard to send a donation. Holm software is free. Their hardware is astronomical.

I am not sure how you do impedance measurements with HOLM, or how you could calculate the T/S parameters, which is why we are fiddling with LIMP and SimpeS.

Be aware, a "feature" of the USB protocol is some timing issues. It is packet based and was never intended for streaming audio. When you are running the test, yo hear ticks when the signal changes. Yo don't on a Firewire and probably don't on a buss sound card. SimpeS does this on my other DACs, so it is inherent to USB-2. It works fine, so don't be alarmed.
 

Neil Davis

Member
Paid Member
2004-12-07 3:23 am
Reston, Virginia
I was not getting a nice graph with a 10 Ohm resistor as I used in all my other tools. The values were quite a bit off and the phase was drooping at the high end. I switched to 200 Ohms and it is very close to the values from WT-II.

I've been tempted to build up a circuit with the AD5933. It wouldn't have any problems with the source resistance, as it uses a constant current source. However, if you are able to get good enough measurements using a sound card and a simple jig, there probably wouldn't be much demand for the AD5933 board. The chip itself isn't cheap--$27 in single quantities.

http://www.analog.com/static/imported-files/application_notes/236037846AN_843.pdf
 
I have used both Holms and REW to get impedance curves easily.
But, That is all you get is a curve.
How do you convert the db scale to to impedance?

I think a really cool method could be to use the data of a stereo sweep using Audacity in 24bit/192Khz mode and plot the curves from that using another program of some sort as there doesn't seem to be and import function using SimpleS.
Unless I am missing something.

jer :)