Omni, thanks,
re Bagby, I'd like to try it, but unlikely as I don't have Excel; and noted re SW - I think I'll have to concentrate on it. The limitations I'm finding are probably mostly me not doing things right. I did find a good link to network optimizer help in it, so when I've worked it out, I'll let you know.
Very kind of you offering to trace my P25. Unfortunately I don't have a scanner and its been scribbled on over the years. So many thanks anyway. My 'eyeball' imports in SW look pretty close though.
I'd love to be able to try the FRD progs, but from memory most, if not all, require Excel. Cheers, grant
re Bagby, I'd like to try it, but unlikely as I don't have Excel; and noted re SW - I think I'll have to concentrate on it. The limitations I'm finding are probably mostly me not doing things right. I did find a good link to network optimizer help in it, so when I've worked it out, I'll let you know.
Very kind of you offering to trace my P25. Unfortunately I don't have a scanner and its been scribbled on over the years. So many thanks anyway. My 'eyeball' imports in SW look pretty close though.
I'd love to be able to try the FRD progs, but from memory most, if not all, require Excel. Cheers, grant
Sreten, Oops, I made another ommission, Sorry to be a nuisance!
Zaph's and Seas response/impedance do look essentially the same at 300Hz, in the different boxes and baffle step also included.
*I missed the point* that my 1.5 Litre mid box could drastically alter things. Sorry about that! And there is no way to get appropriate files without testing in this enclosure. Looks like a no-go to me, unless there is a web design somewhere with it used as a mid. Apologies, grant (sometimes I can't see the wood for the trees!, hehe)
Zaph's and Seas response/impedance do look essentially the same at 300Hz, in the different boxes and baffle step also included.
*I missed the point* that my 1.5 Litre mid box could drastically alter things. Sorry about that! And there is no way to get appropriate files without testing in this enclosure. Looks like a no-go to me, unless there is a web design somewhere with it used as a mid. Apologies, grant (sometimes I can't see the wood for the trees!, hehe)
Hi,
Use WinISDpro to estimate the zma of the CA15 in 1.5L,
and the response below 300Hz in the frd file.
Should be near enough for modelling simple c/o's.
If you are following omni's logic then you can use Zaph's infinite
baffle frd, use the above info and model your box baffle, then
use the combiner to give a predicted response.
Note that the CA15 on its own will run out at ~ 80dB/W with full
BSC, not surprisingly bass extension is very good. An 86dB/W box
needs to be ~ 4 x the size to have only the same extension.
But bass extension and maximum bass levels are very different things.
🙂/sreten.
Use WinISDpro to estimate the zma of the CA15 in 1.5L,
and the response below 300Hz in the frd file.
Should be near enough for modelling simple c/o's.
If you are following omni's logic then you can use Zaph's infinite
baffle frd, use the above info and model your box baffle, then
use the combiner to give a predicted response.
Note that the CA15 on its own will run out at ~ 80dB/W with full
BSC, not surprisingly bass extension is very good. An 86dB/W box
needs to be ~ 4 x the size to have only the same extension.
But bass extension and maximum bass levels are very different things.
🙂/sreten.
Sreten,
I think you just saved my project! Awesome, thank you.
I need to get Excel to follow Omni's path. Also, I did D/L WinISD Pro (Alpha) previously on your recommendation to check driver parameters with it. But had a problem. I will try again. Wonderful stuff! thanks, grant
I think you just saved my project! Awesome, thank you.
I need to get Excel to follow Omni's path. Also, I did D/L WinISD Pro (Alpha) previously on your recommendation to check driver parameters with it. But had a problem. I will try again. Wonderful stuff! thanks, grant
Omni,
yes I would like to keep in touch, but I didn't want to put up an email
address here due to spam, then remembered I have another one that
I rarely use: bonviveur@lycos.com
this would lead to my real one! grant
yes I would like to keep in touch, but I didn't want to put up an email
address here due to spam, then remembered I have another one that
I rarely use: bonviveur@lycos.com
this would lead to my real one! grant
Grant, cool.........I am sorry I forgot you mentioned you didn't have Excel, but now that I come to think of it, you may have mentioned it in an earlier post............At any rate these projects are full of contingency plans............I got a kick kick out of your statement: I need to get Excel to follow Omni's path..........Let's just HOPE this doesn't turn into a scenario of the blind leading the blind into hell........lol.............Any way, we will just have to keep on trudging.............Currently, I am trying to work out a few bugs in BDS, along with dealing with my baffle paint job...........I need to sand again, in order to attempt a final paint coat..........So hopefully, by the time I get that done I will have finished simulations................Grant: you mentioned a few problems in Speaker Workshop. Are these problems gonna make crossover modeling not doable or are they just minor glitches ? Thanks for your e mail link......If we get to the point of being too longwinded on this forum or want lots of specifics, we certainly can correspond that way..........................Respectfully...........Omni
Hi Omni,
I'm back again! Re: 'the blind leading the blind', at least you've got a walking stick! hehe.
You seem to be making better progress than me, although I have managed to 'play with' SW a little. SW being really my only Sim tool now. The unofficial tutorial, while very good, and a fine effort, does leave me with more questions. I'm still not sure if I'm doing things
correctly i.e. as intended by the prog. The other three? tutorials also help but some also raise even more questions. Its a matter of now trying to find the time.
Its good that you can do simulations as a break from enclosure finishing.
I'm sure xo modeling is very do-able in SW, it really is a great program. The limitations mostly seem to be due to my lack of comprehension at the moment. Someone more experienced with it would probably have no difficulty at all. I'll try again in the SW forum(s).
However:
It does have a more complex method of calculating driver 'equivalence', eg, R1 & L1. (Which is a bit of a mystery to me currently until I investigate the driver parameter estimation etc, as recommended by the very kind Sreten. Btw, I did manage to get
WinISDPro working, and have briefly scanned the help file. It also seems very good, this might also help with box/port Q values etc, also).
I did have a little time to try 2nd order (for my old drivers - just for fun!) which was a bit of a tedious exercise, and involved a little 1dB FR smoothing to make them look a tad more ideal. In order to better understand the process. Eventually, by trial and error, starting
with Cookbook LR2 values, I did get a response +-< 1.5dB! But of course, the values to get this response look nothing like the original. Which raises the question:
If LR2 Q=0.49 for each 'section', then how on earth are the 'proper' values derived if you can't believe cookbook formulas? I haven't checked yet, but I bet my new values are more like Butterworth/Chebychev, or maybe my own 'oddball' concoction! So this is a big mystery. If I could remember any filter theory from decades ago, I might have more luck, lol.
The mid in particular seems very difficult to get right. Assuming correct acoustic rolloff of woof/tweeter, and also tweeter attenuation. It makes for a laborious, yet absorbing puzzle.
Changing banpass HP/LP values alters not only slope but centre 'placement' and overall SPL as well. Then of course there are other factors like attenuation, bandpass gain and baffle step. How Sreten does all this without measurements is beyond me - he must be a genious!
How I got to the above mentioned response must have been a fluke, lol. I assume SW includes bandpass gain? Baffle step might be approximately included from the specs? .... It is good fun.
Then of course, there's the 'leap of faith'. Can I really assume that in reality, it will be feasible and more viable than cookbook? Guess, I just have to spend the cash and see! (slight problem: just got a huge quote for urgent roof repair etc, in rental property).
My longest 'rant' yet?, lol (yes maybe email is better!) .....best regards, grant
I'm back again! Re: 'the blind leading the blind', at least you've got a walking stick! hehe.
You seem to be making better progress than me, although I have managed to 'play with' SW a little. SW being really my only Sim tool now. The unofficial tutorial, while very good, and a fine effort, does leave me with more questions. I'm still not sure if I'm doing things
correctly i.e. as intended by the prog. The other three? tutorials also help but some also raise even more questions. Its a matter of now trying to find the time.
Its good that you can do simulations as a break from enclosure finishing.
I'm sure xo modeling is very do-able in SW, it really is a great program. The limitations mostly seem to be due to my lack of comprehension at the moment. Someone more experienced with it would probably have no difficulty at all. I'll try again in the SW forum(s).
However:
It does have a more complex method of calculating driver 'equivalence', eg, R1 & L1. (Which is a bit of a mystery to me currently until I investigate the driver parameter estimation etc, as recommended by the very kind Sreten. Btw, I did manage to get
WinISDPro working, and have briefly scanned the help file. It also seems very good, this might also help with box/port Q values etc, also).
I did have a little time to try 2nd order (for my old drivers - just for fun!) which was a bit of a tedious exercise, and involved a little 1dB FR smoothing to make them look a tad more ideal. In order to better understand the process. Eventually, by trial and error, starting
with Cookbook LR2 values, I did get a response +-< 1.5dB! But of course, the values to get this response look nothing like the original. Which raises the question:
If LR2 Q=0.49 for each 'section', then how on earth are the 'proper' values derived if you can't believe cookbook formulas? I haven't checked yet, but I bet my new values are more like Butterworth/Chebychev, or maybe my own 'oddball' concoction! So this is a big mystery. If I could remember any filter theory from decades ago, I might have more luck, lol.
The mid in particular seems very difficult to get right. Assuming correct acoustic rolloff of woof/tweeter, and also tweeter attenuation. It makes for a laborious, yet absorbing puzzle.
Changing banpass HP/LP values alters not only slope but centre 'placement' and overall SPL as well. Then of course there are other factors like attenuation, bandpass gain and baffle step. How Sreten does all this without measurements is beyond me - he must be a genious!
How I got to the above mentioned response must have been a fluke, lol. I assume SW includes bandpass gain? Baffle step might be approximately included from the specs? .... It is good fun.
Then of course, there's the 'leap of faith'. Can I really assume that in reality, it will be feasible and more viable than cookbook? Guess, I just have to spend the cash and see! (slight problem: just got a huge quote for urgent roof repair etc, in rental property).
My longest 'rant' yet?, lol (yes maybe email is better!) .....best regards, grant
Grant, I share many of the notions you pointed out in you last post, especially about reading tutorials, and becoming more informed, which then ultimately leads to MORE questions................Which then drives me to simply play with the programs, just for ***** and grins, in order to see SOME kind of results, HOPEFULLY which will be accurate.......... So, now, I am trying to simplify and follow some writings which are somewhat "step by step"...............In the FRD Consortium on the rjbaudio.com site there are 2 tutorials which I have found useful: 1- Using FRD Consortium Tools To Design A Speaker. This one basically provides some step by step instruction in the use of the Trace, BDS, Unibox, Frequency Combiner for preparing files for crossover simulation. Obviously they require Excel, and I know you don't have Excel, however, this Tutorial is written in laymans terms, and it may simply provide some understanding......... 2-Using Speaker Workshop to Simulate Crossovers: This tutorial is program specific for the Speaker Workshop program, and again, it provides a step by step approach in laymans terms..............I definately believe you will benefit from this reading. I will be using that procedure to model my crossover in Speaker Workshop, after I finish processing the FRD and ZMA files in the other programs. So check these out if you haven't yet seen them. They may prove useful..............This is LABORIUS, and with the multitude of files required, becomes OVERWHELMING, but if I do just a little bit at a time, eliminating the overwhelming feeling, I think it will be worth it and work out.................When confusion pervades with the use of these tools, and the associated operation manuals, I have E mailed Roman at rjbaudio.com to get clarification, and he has graciously returned my E mails with definative answers to my questions, and, in a very timely manner.........I will keep you posted................keep in touch..........Respectfully........Omni
Hi Omni, thanks so much!
Briefly, re: overwhelming, yes it is almost! but you don't 'eat an elephant at one sitting!' , lol I've done nothing since my "last post" - reason below....'***'
Re: SW tutorials, thanks. I had read all of them, mostly in depth.
Re: Roman - great idea! Its wonderful that he's so helpful for you. If you get any clarification, and have the time ( I know your'e very busy!) I'd like to know, briefly would be fine. Thanks.
If you have a minute or two to spare, you'd be very welcome to send a short hello via email.
Sorry, I have nothing new to offer right now. Yes Roman does have some nice designs. I once read the 'Galatea' I think, and others.
Very best wishes Omni!...from Grant
'***' mum is here, she fell yet again and broke another rib. I can't watch her .every. second unfortunately. Seven hours in emergency department, home at 5am. I'm a bit clueless right now.
Briefly, re: overwhelming, yes it is almost! but you don't 'eat an elephant at one sitting!' , lol I've done nothing since my "last post" - reason below....'***'
Re: SW tutorials, thanks. I had read all of them, mostly in depth.
Re: Roman - great idea! Its wonderful that he's so helpful for you. If you get any clarification, and have the time ( I know your'e very busy!) I'd like to know, briefly would be fine. Thanks.
If you have a minute or two to spare, you'd be very welcome to send a short hello via email.
Sorry, I have nothing new to offer right now. Yes Roman does have some nice designs. I once read the 'Galatea' I think, and others.
Very best wishes Omni!...from Grant
'***' mum is here, she fell yet again and broke another rib. I can't watch her .every. second unfortunately. Seven hours in emergency department, home at 5am. I'm a bit clueless right now.
hi Omni,
Back again! How are you progressing? I had a little time, so I had another short Sim in SW. The combined FR graph now looks very flat for 2nd order, *if I believe the component values*.
So, I had an idea:
When you get your combined (frd/zma) files, including baffle-step, impedance in box, etc, measured for your drivers (in your specific scenario), it would be interesting to see how we would both independently arrive at an xo? Maybe you could email your final files, and see what we come up with and cross-check our results. What do you think? It would be a good exercise for me anyway. Just an idea. ...grant
Back again! How are you progressing? I had a little time, so I had another short Sim in SW. The combined FR graph now looks very flat for 2nd order, *if I believe the component values*.
So, I had an idea:
When you get your combined (frd/zma) files, including baffle-step, impedance in box, etc, measured for your drivers (in your specific scenario), it would be interesting to see how we would both independently arrive at an xo? Maybe you could email your final files, and see what we come up with and cross-check our results. What do you think? It would be a good exercise for me anyway. Just an idea. ...grant
Sreten, are you still here?
I've been reading many of your posts in other threads.
I'm wondering if you could personally recommend your favourite budget design (2way or otherwise)? Just for interests sake. Would the woofer be ALU or paper, etc. This question is probably 'out-of-bounds' but I'm curious to know where your design instincts are at! And what would, in your opinion, give the 'best bang per buck'!...thanks grant
I've been reading many of your posts in other threads.
I'm wondering if you could personally recommend your favourite budget design (2way or otherwise)? Just for interests sake. Would the woofer be ALU or paper, etc. This question is probably 'out-of-bounds' but I'm curious to know where your design instincts are at! And what would, in your opinion, give the 'best bang per buck'!...thanks grant
Hi,
The best designs are those suitable for purpose, the worst
are those built because it is believed it is what is needed.
Other issues are the builders capabilities on the mechanical
side of things, many cabinet designs are very mundane and
not very effective at minimizing resonances.
Speakers need to be matched to systems, e.g. a power hungry
small metal coned system is not going to work with smallish
valve amplifers or ever crank out Nirvana at realistic levels.
But in other circumstances it could sound sublime for the investment.
But if pushed to suggest something rather good
I'd say a properly built floorstanding version of :
http://www.zaphaudio.com/audio-speaker17.html
Should be something special with a decent CP player and amplifier.
Bang for Buck - any version of
http://www.zaphaudio.com/audio-speaker13.html
is going to be tough to beat at the total build cost.
🙂/sreten.
The best designs are those suitable for purpose, the worst
are those built because it is believed it is what is needed.
Other issues are the builders capabilities on the mechanical
side of things, many cabinet designs are very mundane and
not very effective at minimizing resonances.
Speakers need to be matched to systems, e.g. a power hungry
small metal coned system is not going to work with smallish
valve amplifers or ever crank out Nirvana at realistic levels.
But in other circumstances it could sound sublime for the investment.
But if pushed to suggest something rather good
I'd say a properly built floorstanding version of :
http://www.zaphaudio.com/audio-speaker17.html
Should be something special with a decent CP player and amplifier.
Bang for Buck - any version of
http://www.zaphaudio.com/audio-speaker13.html
is going to be tough to beat at the total build cost.
🙂/sreten.
Grant, I agree with your idea about crosschecking our design profiles...............I have been quite busy with things in addition to this speaker project, so I have not been too communicative lately. I have been in touch with Roman at rjbaudio.com discussing the frd tools, and have been trying to hammer out some issues with the baffle diffraction tool, meanwhile, studying the frequency combiner tool for when I get there............Another snag on the baffle paint job, so I gotta re-sand, AGAIN, and give it another spray. The cost alone, on the baffle finish is beginning to wear on me, but not to worry, it will get there, hopefully sooner, rather than later.........It's delaying my final cabinet construction, so I can mount the drivers, but Oh Well............can't wait to get to the crossover..................I hope your mom is fairing well........................warmest regards........Omni
Sreten, thank you very much!
I am begining to appreciate your advice even more as my understanding improves a little.
Yes, Zaph's designs seem excellent but its good to hear your confirmation! I was at one stage very keen to build his L18 design - may still yet if funds available. I have 2 amps, both around 100W, so
there shouldn't be a problem driving them. Also, I don't think SilverFlute drivers are available in Au - negative google. thanks again, grant
I am begining to appreciate your advice even more as my understanding improves a little.
Yes, Zaph's designs seem excellent but its good to hear your confirmation! I was at one stage very keen to build his L18 design - may still yet if funds available. I have 2 amps, both around 100W, so
there shouldn't be a problem driving them. Also, I don't think SilverFlute drivers are available in Au - negative google. thanks again, grant
Hi Omni, glad your still here!
Sure, please send me the 'combined' files when you're ready. It would be fun to do and compare results. I'm now more confident of SW component value calculations and I read in one of the
tutorials that a 2nd order electrical filter MAY yield a LR4 acoustic, dependent on driver rolloff, of course.
Sorry I can't offer any input yet on the FRD tools.
Oh no! another re-sand? You are a perfectionist, lol. Are you after a mirror finish?
(Many thanks re: my mum, she's in 'respite care' now for a month or so....) Very best wishes, grant.
Sure, please send me the 'combined' files when you're ready. It would be fun to do and compare results. I'm now more confident of SW component value calculations and I read in one of the
tutorials that a 2nd order electrical filter MAY yield a LR4 acoustic, dependent on driver rolloff, of course.
Sorry I can't offer any input yet on the FRD tools.
Oh no! another re-sand? You are a perfectionist, lol. Are you after a mirror finish?
(Many thanks re: my mum, she's in 'respite care' now for a month or so....) Very best wishes, grant.
Alright, I am gonna try to attach my baffle simulation driver response chart here............Sreten, can you provide an interpretation of it ? Tinitus, if you're still here please chime in. This was done on Baffle Diffraction Simulator with my baffle design and driver placement, 8 degree slanted baffle. One curiosity I have though.............The operation manual states that when driver sizes are entered that the piston figure is calculated, however, it did not appear to happen for me. Do you know why, or am I gonna have to enter them in myself ? Since I am using standard inches, how do I convert square cm to square inches?.....................Driver 1=Peerless SLS 830669 Woofer.............. Driver 2=Seas 27TDFC Tweeter......... Driver 3=SeasCA15RLY Midbass...............Anyone else who reads this thread and can interpret this chart, please feel free to chime in. That would be greatly appreciated...........Respectfully..........Omni
Attachments
Hi Omni,
Square cm to square inches: Lalena site says to multiply metric unit by 0.155 to get imperial.
Re: the baffle graph, I really don't know, but is it the baffle step in dB for each driver from 100 to 20kHz? Or is there more to it than that?
warmest regards, grant
Square cm to square inches: Lalena site says to multiply metric unit by 0.155 to get imperial.
Re: the baffle graph, I really don't know, but is it the baffle step in dB for each driver from 100 to 20kHz? Or is there more to it than that?
warmest regards, grant
Hi,
Without details of the actual baffle layout its hard to interpret.
It implies that you have a trapezoidal tapered front baffle.
For simulation you enter the effective, not actual diameter.
So for a tweeter = 1", for a 5" bass/mid= ~ 4".
1 square inch = 6.45 square cm.
Having been there if the finish of a coat of spray paint is not good
enough another sand and coat will never fix it. Stopper, sand, then
primer (which is much bulkier), sand, very fine sand, then paint.
Use layers of primer to get a finish, not paint.
🙂/sreten.
Without details of the actual baffle layout its hard to interpret.
It implies that you have a trapezoidal tapered front baffle.
For simulation you enter the effective, not actual diameter.
So for a tweeter = 1", for a 5" bass/mid= ~ 4".
1 square inch = 6.45 square cm.
Having been there if the finish of a coat of spray paint is not good
enough another sand and coat will never fix it. Stopper, sand, then
primer (which is much bulkier), sand, very fine sand, then paint.
Use layers of primer to get a finish, not paint.
🙂/sreten.
Omni,
I've been reading Roman B's Audiofiles article 'Locating the tweeter to reduce diffraction effects'.
Its quite good and may give the explanation you're after. Apparently, if I've understood it correctly, the baffle step is a 'smoothish' rise/falloff to 6dB, modulated by diffraction giving a ripple effect in the worst case of +- 2.5 dB. I assume diffraction also applies to the mid. He explains the 2/4 PI stuff and the Golden Ratio quite well. Luckily, my tweet/mid drivers are about OK for placement
and the baffle edges are rounded.
I hope this is useful to you.
I read in 'the Edge' doco (a BSC compensation program by Tolvan, ie Basta) that baffle-step can be negated to an extent by a contour? network. Interestingly, ( I read from somewhere else) that
if the woofer in a 3way is crossed low enough (in our case 300Hz) with an appropriate order electrical filter, then BSC mostly applies. But, because the CA15 SPL is ~85dB and therefore not attenuated in relation to your Peerless, then you probably would have to compensate for it (and the tweeter). I'll try and find the reference for this if you're interested. I hope some of this makes sense, grant
I've been reading Roman B's Audiofiles article 'Locating the tweeter to reduce diffraction effects'.
Its quite good and may give the explanation you're after. Apparently, if I've understood it correctly, the baffle step is a 'smoothish' rise/falloff to 6dB, modulated by diffraction giving a ripple effect in the worst case of +- 2.5 dB. I assume diffraction also applies to the mid. He explains the 2/4 PI stuff and the Golden Ratio quite well. Luckily, my tweet/mid drivers are about OK for placement
and the baffle edges are rounded.
I hope this is useful to you.
I read in 'the Edge' doco (a BSC compensation program by Tolvan, ie Basta) that baffle-step can be negated to an extent by a contour? network. Interestingly, ( I read from somewhere else) that
if the woofer in a 3way is crossed low enough (in our case 300Hz) with an appropriate order electrical filter, then BSC mostly applies. But, because the CA15 SPL is ~85dB and therefore not attenuated in relation to your Peerless, then you probably would have to compensate for it (and the tweeter). I'll try and find the reference for this if you're interested. I hope some of this makes sense, grant
- Status
- Not open for further replies.