Many GRS woofers/subwoofers parameters aren't working well in Winisd

[Weber]

So I'm looking for replacement 12" woofers for cabinets I already have and once I input the parameters into winisd, it says failed integrity check. After I ignore that warning and proceed to box volume I'm getting things like 0.005 ft3 for a sealed box and other crazy volumes for vented. However according to the project page it says 2.72ft³ for recommended sealed volume which is what my box is exactly. So does anyone know why? Are the Thiel smalls not correct?

[wogg]

Careful of units, the PE pages have a mix of metric and imperial units typically. Sometimes I see published specs cause a problem, but not that much of a problem.

[a4eaudio]

Too ensure success in the integrity check...
1) Turn on the auto calculate option 2) Only enter the minimum number of parameters needed. (you can find this in the WinISD help file, in threads on this forum and the tutorial at midwest audio club)

[Weber]

Careful of units, the PE pages have a mix of metric and imperial units typically. Sometimes I see published specs cause a problem, but not that much of a problem.

I've doubled checked them and even entered them in again a second time with the same results.

[Chris Roemer]

Clear ALL driver data (or start a new entry) and enter ONLY:
Qes, Qms (Qts will autocalculate), Fs, Vas (cf OR Liters - if that's your bag), Re, Le, Sd (almost always given in sq.cm. - cm^2), Xmax, and Pe.
That should do it.

[LOUT]

Looking at some GRS listed specs VS measurements from 3rd parties I get the feeling that the manufacturer specs might not be accurate.

[Weber]

Clear ALL driver data (or start a new entry) and enter ONLY:
Qes, Qms (Qts will autocalculate), Fs, Vas (cf OR Liters - if that's your bag), Re, Le, Sd (almost always given in sq.cm. - cm^2), Xmax, and Pe.
That should do it.

I'll try it but it has errors with several preloaded speakers as well. Almost all of them if not all. And it does this even with a brand new project, but I will try what you suggested Chris. Thanks

[billfitzmaurice]

I'll try it but it has errors with several preloaded speakers as well.

That's because whoever entered them into the data base did it the same way as you. WinISD calculates to a higher degree of accuracy than most data sheets list. If, for instance, the data sheet says Qts is 0.3 but the program calculates it as 0.31 it will read as an error. If you follow the instructions Chris gave you will never get an error message.

[Weber]

That's because whoever entered them into the data base did it the same way as you. WinISD calculates to a higher degree of accuracy than most data sheets list. If, for instance, the data sheet says Qts is 0.3 but the program calculates it as 0.31 it will read as an error. If you follow the instructions Chris gave you will never get an error message.

So the program is broken to some extent then? I don't like having to use workarounds to get something to work correctly. Are they even trying to fix these problems?

[Chris Roemer]

Nope, the program works fine. You wouldn't say that your car was broken if you put some bad gas (with water in it) in the tank, would you?? There have been MANY releases of WinISD. I'm using a (fairly old - now) vers. 0.50a7, a.k.a. the "Pro" version.

There might be some drivers entered using data that was entered in a "weird" way. I know the vers. I use has double entries for quite a few Eminence drivers like the "Eminence Alpha 10" and the "Eminence Alpha 10" (with TWO spaces ahead of "Alpha"). Don't think Juha entered all the data though.

[tvrgeek]

Several issues.
Frist, publishes specs give you a very good idea on what to buy, but you really need to measure the drivers yourself for actual design.
Second, especially with large magnet woofers and high inductance, you really need to use the semi-inductance model. HornRes and Vituix can use it. Not easy deriving the parameters. I was not aware of how dramatic a difference this made in the model until recently. For one sub, it shifted the peak by 40 Hz!
Nothing beats building a prototype and measuring it. No model is going to really tell your about your room.
I use .50a7 version. I found the newer one to be cantankerous.
Many of the supplied TS parameter sets do not pass basic back-calculation as the newer versions of WinISD will tell you.
If you failed integrity check, your data is bad. Be sure the units are correct. Don't enter parameters it will calculate. In other words, don't enter Qts and then Qes and Qms. I do not know the precision of the integrity check, but it is pretty tight.

All said and done, even published specs were good enough to make a purchase decision. After that, it is all measure, prototype and tune anyway. It feels great when the model is really close, but in reality, does not really tell you a lot.

[Weber]

Nope, the program works fine. You wouldn't say that your car was broken if you put some bad gas (with water in it) in the tank, would you?? There have been MANY releases of WinISD. I'm using a (fairly old - now) vers. 0.50a7, a.k.a. the "Pro" version.

There might be some drivers entered using data that was entered in a "weird" way. I know the vers. I use has double entries for quite a few Eminence drivers like the "Eminence Alpha 10" and the "Eminence Alpha 10" (with TWO spaces ahead of "Alpha"). Don't think Juha entered all the data though.

Well the latest version is somewhat broken so I switched to the .50a7 version and it works perfectly, no issues so far! Thanks

Several issues.
Frist, publishes specs give you a very good idea on what to buy, but you really need to measure the drivers yourself for actual design.
Second, especially with large magnet woofers and high inductance, you really need to use the semi-inductance model. HornRes and Vituix can use it. Not easy deriving the parameters. I was not aware of how dramatic a difference this made in the model until recently. For one sub, it shifted the peak by 40 Hz!
Nothing beats building a prototype and measuring it. No model is going to really tell your about your room.
I use .50a7 version. I found the newer one to be cantankerous.
Many of the supplied TS parameter sets do not pass basic back-calculation as the newer versions of WinISD will tell you.
If you failed integrity check, your data is bad. Be sure the units are correct. Don't enter parameters it will calculate. In other words, don't enter Qts and then Qes and Qms. I do not know the precision of the integrity check, but it is pretty tight.

All said and done, even published specs were good enough to make a purchase decision. After that, it is all measure, prototype and tune anyway. It feels great when the model is really close, but in reality, does not really tell you a lot.

The .50a7 version is much better, thanks for the recommendation.