Thanks Erik.
Great to see such an active community!
I know of the existence of miniDSP, but for what I need, I think it's a bit overkill. This should be easy to fix with analog means.
Yeah, I have been trying to minimize the parts, you can see that from the remainders that I left in the schematic
Great to see such an active community!
I know of the existence of miniDSP, but for what I need, I think it's a bit overkill. This should be easy to fix with analog means.
Yeah, I have been trying to minimize the parts, you can see that from the remainders that I left in the schematic
Bill,
When you update Xsim, please include the inductor's ESR value on the schematic below the inductance value. Currently, if you need to share a schematic picture you must add ESR resistors in series with the inductor on the schematic. Although less important, the capacitor's ESR value and resistor's wattage could also be printed on the schematic. Schematic picture would "tell all".
When you update Xsim, please include the inductor's ESR value on the schematic below the inductance value. Currently, if you need to share a schematic picture you must add ESR resistors in series with the inductor on the schematic. Although less important, the capacitor's ESR value and resistor's wattage could also be printed on the schematic. Schematic picture would "tell all".
OK, when I get to programming mode again (maybe this winter). But you can add text anywhere on the schematic as it is.
If I put EVERYTHING about every component on the schematic, then schematics won't be able to be very many parts!
If I put EVERYTHING about every component on the schematic, then schematics won't be able to be very many parts!
Click on the Curves menu on the graph, you'll see various options you can look at.
The AddGraph menu on the main window gives more graph types you can add.
Play around, you'll find what you need.
-Peter
Hi there,
I'm looking to work with XSim on a Mac OS X Yosemite version 10.10.5. Is that right that XSim doesn't work with Mac? If so, what would be the easiest way to get a program to make it works on Mac?
If not, is there another similar program which I can use on Mac? PCD?
Thank you,
Sébastien
I'm looking to work with XSim on a Mac OS X Yosemite version 10.10.5. Is that right that XSim doesn't work with Mac? If so, what would be the easiest way to get a program to make it works on Mac?
If not, is there another similar program which I can use on Mac? PCD?
Thank you,
Sébastien
WINE or Crossover Mac (commercial version of WINE)
Virtual Box, Parallels, Fusion all need a copy of Windows.
Or you can use Boot Camp to install Windows natively.
dave
I have discovered another cool hack for XSim: Blending port and woofer together. I tried doing this in OmniMic and for the life of me I cannot understand ANY of the math functions. I know what they do, I just can't get them to work, except for cutting and splicing but that's more of a curve action than adding. Anyway...
Do your normal close and far field woofer measurements. Measure port strictly in near-field. Cut off the upper limit of the port data off as necessary. I usually use -20 dB as my upper cut off, but make yourselves happy.
In XSim create two speakers, one with the woofer FRD and one with the port FRD. Make sure to "mod delay" the port if needed.
Make sure you have all the phase data. Import your complete far-field data. This will be your amplitude guide. Adjust port amplitude until your simulation and far-field charts agree. Export the system FRD. Voila.
Biggest advantage over mathematical models... there's no math, and no chance to make a mistake because of the math.
Best,
Erik
I tried doing this in OmniMic and for the life of me I cannot understand ANY of the math functions. I know what they do, I just can't get them to work, except for cutting and splicing but that's more of a curve action than adding. Anyway...Do your normal close and far field woofer measurements. Measure port strictly in near-field. Cut off the upper limit of the port data off as necessary. I usually use -20 dB as my upper cut off, but make yourselves happy.
In XSim create two speakers, one with the woofer FRD and one with the port FRD. Make sure to "mod delay" the port if needed.
Make sure you have all the phase data. Import your complete far-field data. This will be your amplitude guide. Adjust port amplitude until your simulation and far-field charts agree. Export the system FRD. Voila.
Biggest advantage over mathematical models... there's no math, and no chance to make a mistake because of the math.
Best,
Erik
Last edited:
Lazzer, if you just want to do it for "fun" put an 8 Ohm resistor in parallel with the driver.
You an also create a ZMA file with 2 data points. I'm not acutally sure how many it needs, but 20 Hz and 20,000Hz should do it. The format is:
F, Z and Angle, so:
20 4 0
20000 4 0
Should do it.
If you are trying to simulate a real speaker, it's best to import the actual impedance curves.
Best,
Erik
You an also create a ZMA file with 2 data points. I'm not acutally sure how many it needs, but 20 Hz and 20,000Hz should do it. The format is:
F, Z and Angle, so:
20 4 0
20000 4 0
Should do it.
If you are trying to simulate a real speaker, it's best to import the actual impedance curves.
Best,
Erik
IBlending port and woofer together.
Well, it goes like this:
For combining drivers playing at the same time as if into one microphone, you would SUM the drivers' outputs. The responses don't reallt modify each other, but the graph shows what you'd see from playing them both toward the mic simultaneously.
For playing something out of one driver, then picking it up with a microphone and then playing that again through another driver (or EQ box, or room response effect, or....), called a "cascaded" response -- you would use FILTER. The one response acts as a filter to the other, in other words.
For taking a frequency response and removing the effect of another frequency response that was included in it (such as from FILTER, or from natural cascaded responses like playing a driver in a room or through an EQ) you would NORMALIZE the curve with the one you want taken out from the cascaded total.
Or, if you have a group of curves and want to see how they all differ from one of those curves, you could NORMALIZE them to the one curve. Normalizing a curve by itself gives a flat horizontal line. That is done a lot with off-axis sets of curves, when you aren't too concerned with the actual on-axis shape (which can always be changed with an equalizer) but want to see what the polar off-axis changes are.
(the above is equivalent to an OmniMic "SUM", btw....). But to get it right, you should modify the SPL levels relative to the cone and port cross-sectional areas. When you look near-field, your pressure mic only sees the pressure over its small capsule area, but the total output from cones(s) or port(s) depends on the pressure over their total areas, most of which won't register over the tiny mic capsule. In other words, having a given pressure over only 1 square cm would be a lot less sound output than having that same pressure appearing over a square meter! So you have to adjust the port output according to the ratio of port area to effective cone area.
Or if you're only looking to combine a woofer's cone output with it's port output, here's a trick I learned from Laurie Fincham: Keep making repeated measurements while moving the mic within the space between the (very close) cone and the port, find the position where the lowest frequencies drop way down. At that point the very low frequencies are cancelling (as they must, ported speakers being unable to do steady DC pressure) and the respective cone and port outputs are then in the correct proportions. That trick does require pretty good SNR, though, so you may have to turn it up, or wait till LF noise in your testing area is low enough!
The least tricky one is the math with "added curves". Load an added curve (or as many as you want) then choose what kind of math you want to do with it. You'll be prompted for the frd file you operate on it with (to sum with, filter, etc). Then the result curve(s) will be shown in place of the prev displayed "added" curves. The files of the added curves DONT get changed, just the display. But you can save the modified curve (just get to the added curve's properties window, theres a button to "save as modified".
There are similar menus for math on the "live" curve too (the math is applied with each new acquisition/sweep/refresh before they display). So you could measure fr as if it had an eq applied, for example, or summed with another response curve).
There are similar menus for math on the "live" curve too (the math is applied with each new acquisition/sweep/refresh before they display). So you could measure fr as if it had an eq applied, for example, or summed with another response curve).