It's okay guy, I'm just going to try to see if I can get something to work here. You're not being a pain at all, simply presenting me a modeling challenge!
Can't argue with your method or results but in this case it's a HUGE unbraced area (almost 4 sq ft?) and a couple of very powerful drivers. I personally wouldn't do it.
I can't get either one to fit properly anyway, it appears that I have to go back to the drawing board.
If I could have gotten the second option to work, I would probably have suggested some sort of cut-out brace that makes the driver accessible.
If I could have gotten the second option to work, I would probably have suggested some sort of cut-out brace that makes the driver accessible.
An externally hosted image should be here but it was not working when we last tested it.
This is the simplest way to show why it won't work accurately. See the lines I've circled? They get thinner and fatter. Unless you can recreate this effect in plywood your sim won't match your finished product. This is a small degree of error but it does make a difference.
EDIT - I'm not sure what's going on with the circle on the right, it shouldn't have a problem but the one on the left is due to having a bend at S2 and S3 in the sim which can't be accurately built with flat plywood.
I'm not at all sure your software is handling the bends correctly either.
The errors I've circled probably aren't enough to make a whole lot of difference but if the bends are also not right this all adds up to inaccuracy.
Last edited:
guy that was just a rough, the final design would have properly spaced .75" material thickness. C4D will only get me half of the way there, I would still CAD it out in rhino and it would be exact dimensions
The pic you are citing is more just a proof of concept that cinema4d can interpolate a horn flare along a spline and read the current length of the spline. That image took me about 5 min to create
The pic you are citing is more just a proof of concept that cinema4d can interpolate a horn flare along a spline and read the current length of the spline. That image took me about 5 min to create
Do you have any other recommendations for fold patterns? I cannot get this one to fit
Ignore the places that overlap and make t's, that is just the pipe overlapping itself
Ignore the places that overlap and make t's, that is just the pipe overlapping itself
Last edited:
I added an edit but it was too late. Here it is. EDIT - I'm not sure what's going on with the circle on the right, it shouldn't have a problem but the one on the left is due to having a bend at S2 and S3 in the sim which can't be accurately built with flat plywood.
When we get further along and you have a fully drawn and completed plan I'll check it for accuracy.
Not tonight but I can draw up some more later. Basically it's just drawing it out in different ways until something works. Sometimes nothing will work and you have to change the sim.
Another way to do this is just draw a picture and simulate it and repeat over and over that's how the Othorn and Gjallerhorn were designed. Iterate until you get something you like.
I see what you mean with the S3-S4 flare. Though looking at it, it is so minimal I don't think it will make a difference. If I understand this process from reading through threads in the past, subtle errors often don't have that huge of an impact. I mean there is a massive cone sitting right in the throat and we do nothing to correct for that?
Now that I am really starting to get a grip with my space limitations, the only way I can see this working is with some kind of exponential flare or a higher compression. Just barely big enough S2 and just barely big enough S3/S4, with an exponential flare along L23 [and possibly L34 as well] to save space until we get to the end tap where the driver needs to fit.
This is a big topic, believe it or not.
First the cone in the throat. Yes we do something about that. That's what Vtc and Atc in Hornresp are for. There's even a feature in Hornresp to help determine how big the cone divot is so you can correctly specify Vtc and Atc. But if you play with these settings a bit you will find that it doesn't make much difference which is why I don't even bother to sim it unless I'm at the stage where I'm actually drawing up plans. Some people even use cone correction, which is placing blocks of wood in the in the horn path adjacent to the cone divot to "fill in" the hole that the cone makes. Personally I think that would do a lot more harm than good but I haven't tested it and probably won't bother.
On the other hand, the basket in the mouth can make a pretty big difference in response and velocity, especially in a horn as small as yours where the basket severely chokes the path. And yes, I do something about that too. I sim it in Akabak. Most people don't bother.
Moving on, "subtle errors" is subjective so let's qualify that. Different types of errors in different parts of a horn cause different effects. Here's the sim I provided on page one. The light grey line is as shown, the black line is the effect of making S3 50.8 cm larger (a 1 cm error in the plans based on a 20 inch internal height).
An externally hosted image should be here but it was not working when we last tested it.
It's subtle but it's definitely a difference.
There are dozens of things in the real world that sims don't consider at all so sims will never really match measurements of finished products. Lilmike is a member here (although he is much more active elsewhere) that is EXTREMELY accurate with folding, I've verified it myself. And he's always saying that a large percentage of his projects end up in the burn pile because they didn't turn out as expected.
So for example if real world factors contribute to an unexpected dip at a certain frequency and folding errors contribute to even more of a dip at that same frequency there is a good chance that you just wasted a bunch of time and money on something that has no use other than roasting hot dogs.
In the end it's pretty simple, is it really worth saving a few minutes on doing things right if the results are sloppy? Personally I don't have a lot of money so I'll take the extra time to do it accurately every time. I can't afford to burn a few sheets of baltic birch.
LOL I think I've mentioned this a few times.
You can't do exponential flares in a single segment unless you are prepared to bend the plywood. If you are making a flare that has parallel walls it has to be PAR.
The entire flare can be an exponential shape but it has to be made up of PAR segments.
Anyway the problem is not space so much as trying to fit the sim into the predetermined dimensions. You have plenty of space, you just can't make it fit into the desired shape. I'm pretty sure I could do it but I use Sketchup and draw it in panel by panel and reiterate with different layouts as necessary (or use other methods to make it fit) so I'm very slow. (But accurate.)
Thanks for sharing your process. I guess I was hoping my beginners luck would hook me up here, but alas it was a no! Ha
Is there any disadvantage to having several folds as opposed to just 5 or 6? I see many different variations on that
Is there any disadvantage to having several folds as opposed to just 5 or 6? I see many different variations on that
I'm working on a new sim in the vein of your first post here, exponential (parabolic sectioned) flare but with a higher compression ratio of 3.7:1 and properly sized S4 for a 20" internal height. First of all, how much is too much compression? Is this a no-go for this driver?
The response graph makes me happy- nice and flat
System volume is slightly higher, but I have a sneaking suspicion it may be easier to fit with my box dimensions. Mainly, I am wondering if I am wasting time with the higher 3.7:1 CR
The response graph makes me happy- nice and flat
System volume is slightly higher, but I have a sneaking suspicion it may be easier to fit with my box dimensions. Mainly, I am wondering if I am wasting time with the higher 3.7:1 CR
Last edited:
If y'all get tired of seeing my progress let me know, I don't want to be redundant or annoying with this.
More sim / model loop, getting much closer now (as long as 3.7:1 CR is okay)
I know this isn't advanced centerline, but I'm just kind of roughing stuff until I can get a good fit. Then I plan to go in for the kill
More sim / model loop, getting much closer now (as long as 3.7:1 CR is okay)
I know this isn't advanced centerline, but I'm just kind of roughing stuff until I can get a good fit. Then I plan to go in for the kill
I found another few inches of depth in the back of the truck! 😉
Everything matches sim, horn length includes advanced centerline
Everything matches sim, horn length includes advanced centerline
Not really. More wasted space in the bend corners, more fold complexity, but as you found out it can help make things fit. Your new drawing is close to what I was going to draw for you as a new layout to try. If front to back doesn't work try side to side, and variations of that.
Ask different people you will get different answers. Danley usually doesn't go too much higher than 2:1 but I've seen people use 8:1 with smaller drivers and get away with it. If the cone is strong enough and you don't get crazy with the volume (you are limited with power anyway) it will probably be fine.
Anyway, MMS of your driver give some kind of idea how strong it might be. A higher weight could indicate strength, but it doesn't necessarily mean a high value HAS to be strong.
Also you can check diaphragm pressure at full volume in Hornresp.
Click Calculate, go to Diaphragm Displacement, Tools, Diaphragm Pressure, Total.
Check that against your ported box model. If it isn't showing a whole lot more pressure on the diaphragm than the ported box it will probably be fine.
Nice and flat isn't usually best for horns.
Can you test t/s parameters? The parameters you are using to design the box are small signal taken when the driver is cold. As you beat on it with 1200 watts for a few hours the t/s will change and they will usually indicate that the driver would prefer to be in a box that is larger than the one you made for it. It will show an overdamped response. So it's usually better to make the box a bit underdamped to begin with (a bit bumpy response) so it can "grow into" the box as it heats up instead of "grow out of" it.
If you can test t/s test the driver when it's cold and then test it again when it's hot, plug both sets into your model and see what happens.
Is this your final answer? When you get it done I want to check the fold for accuracy.