If you can still find them, Peerless P830870 (HDS 106 PPB)
They were used A LOT on Philips surround sets.
That is a pretty good THD @ 90dB for just a 4 inch!
Dayton RS125-4 is also not bad (don't know the price atm)
SB12CACS25-4 is great, but slightly out of budget I think?
(all @ 90dB btw)
They were used A LOT on Philips surround sets.
That is a pretty good THD @ 90dB for just a 4 inch!
Dayton RS125-4 is also not bad (don't know the price atm)
SB12CACS25-4 is great, but slightly out of budget I think?
(all @ 90dB btw)
What aspect of the perfromance would you like to see improved? The directivity on the low end? or is it the high frequency behavior that you would like to improve?
I guess I was taken in by the manufacture's marketing, but SB calls it a mid range. But it is the only non Satori driver they list as a mid range. Here is the link that lists some of the points Jim made.
https://sbacoustics.com/product/4-sb12mnrx2-25-4-norex/
The dip on axis in the top octave. Looks just like SB's waveguide version, so not worse. But I think I could improve it with a different contour. I can print both and send them to you and you can decide what you like. I don't have time to do measurements myself at the moment.
That is a generous offer, thanks... but I might want to hold off on that line of action for now. For a few more $ the SB26ADC has better performance, so that may be the way I go...
When you pair a waveguide/tweeter with a low frequency driver, i.e. a midrange in my case, how do you decide how large of a waveguide you need? in other words, how do you select the frequency at which the waveguide starts to provide directivity control? Is it a function of the directivity of the mid driver? Or is it related to the crossover frequency?
The only other waveguide project I have done used a Purifi 6" driver and a Satori TW29BNWG, and I chose those drivers in advance before any simulation. I knew those drivers could work well together because others had done so. So I did not have to decide which waveguide to use, because that tweeter comes with its own waveguide.
I use VituixCad for my simulation work, and it is difficult to simulate the directivity of a waveguide tweeter without actually building a prototype and measuring.
So I am interested in your thoughts on this...
@fluid - I would also be interested in any thoughts you might have... guidelines on matching waveguide directivity to midrange directivity.
j.
Last edited:
I think Erin had good answer to this couple of weeks ago in one of his videos.
A part depends on your own taste. So figure out how much directivity you prefer.
The only very important thing I would like to add, is that it also depends on the room acoustics as well.
Which is an inevitable tango dance between keeping reflections at bay as well as having the sound stage you want.
As for answering the question about what size of waveguide is needed to make sure the directivity doesn't get a big jump; rule of thumb is at least the same diameter waveguide as the woofer.
A bigger waveguide is often also fine (or sometimes better), but it just looks totally awful I think.
Although with a 4 inch you probably don't need a waveguide if you like a very wide directivity.
(very traditional 3-way sound btw)
One other note;
I personally always advice people to set a total budget and not a per driver budget.
The reason why is because you can shift and cater the budget better.
Maybe you can save a few bucks on the woofer as well as tweeter, just to get a better mid-woofer/mid-range that just suits the whole project better.
But that also depends where your preferences are or the goal is of the project.
I totally believe you, it just doesn't feel right to spend 3 times more money compared to some other alternatives that already perform so much better on paper.
Or spend the same amount for drivers that are even better.
Based on your first post where you set your goal, here are a few suggestions:
- If the potential builders arent hi-fi nerds, looks will play a role in the decision to build a DIY speaker. I think similar looking drivers of one brand look better than 3 drivers with different colors of black, for example. If SB12MNRX2-25-4 is used for midrange, maybe SB23NRXS45-4 can be used as woofer. Similar price and similar specs to the proposed Dayton.
- You wrote: "This means it is easy to build with readily available parts". That should exclude a waveguide. I think with a small midrange driver, good dispersion should be possible without a waveguide. SB Acoustics SB29SDAC-C000-4 could fit. Quite wide dispersion, matching the midrange driver. Might be preferred in a living room context.
@b_force I bought mine a couple of years ago from Wilmslow audio in the uk. At the time it wasn’t especially expensive, certainly not 3 times the price of similar units! However a moot point, it seems cheaper and better are now available and I’m looking forward to seeing Jim’s final line up.
It depends on what you want the waveguide to do. Is it to push the crossover frequency down to avoid breakup in the mid driver, is it to achieve a specific radiation angle, is it to keep the directivity constant or to keep a smooth rising slope? The physical size will be mainly about how low in frequency you want any form of directivity control to happen. For crossovers 3K and up the size is more like a tweeter face plate and the difference between well done bevelled edges and a designed waveguide may well be not much. No waveguide performs as well as it can with any amount of flat baffle area between it and the baffle edge.
ABEC or AKABAK can give you very accurate directivity information to throw into a Vituix simulation to see how a particular waveguide and cabinet combination will work with a mid. I know you can't do this yourself but I have the SB26ADC profile already programmed and it isn't too much work to make a simulation of a waveguide in a cabinet if you want that done.
For some reason the tag didn't work but you have my thoughts anyway
* Unfortunately these tags don't work if you add them to a post during an edit. They need to be part of an original post.
This is (for the most part) my reason for using a waveguide. I want to control the directivity index through the crossover region so that it either stays constant or smoothly rises.
Intuitively it seems that if the area of the waveguide mouth is the same as the Sd area of the midrange driver, then their directivities would be well matched through the crossover region... but I know the situation can be complicated by other aspects of baffle geometry, and I also realize that no two waveguide/tweeters are alike... So I would be rather shocked if "intuition" was of any help here...
That would be quite helpful. I appreciate that offer.
I assume you will need the following information: a box 292mm wide x 648mm high x 222mm deep. Assume a 25 mm bevel/round-over on the sides and top. Assume the tweeter is centered horizontally, and I will let you decide the best vertical position.
This is very helpful... Thanks in advance...
j.
I have been thinking about the difficulty of using a waveguide... having it printed, routing a recess which may not be circular... yes it starts to get complicated.
The SB29SDAC does simulate quite well with 4" and 5" drivers in a well beveled cabinet.
@hifijim
With a cabinet 292mm wide, you can use a 10” woofer?
Any reason not to, Jim?
I implore you to consider this option.
IME a single 8” just isn’t enough bass to make it worthwhile. Twin 8” or single vented 10” is a minimum to make a 3 way worthwhile.
If I had to do it again I would not do a single 8” 3 way.
An 8” two way yea!
A single 8” way? Never again with a standard driver (Revelator and Purifi not withstanding)
The challenge with soft domes in waveguides is that since they are always operating in breakup mode (albeit small breakups) it is more challenging to simulate in a waveguide. A hard dome has more predictable behaviour, making the correlation between simulation and actual real life measurement much more likely to be accurate.
Addit: Having worked with SB29SDAC, it’s a fine tweeter. But in a waveguide? We’ll have to see (it has to be modelled, built and measured and several iterations).
I would recommend ADC or CDC, if budget allows.
Or a dimple dome or ring radiator (fixed centre of the dome reduces breakups)
With a cabinet 292mm wide, you can use a 10” woofer?
Any reason not to, Jim?
I implore you to consider this option.
IME a single 8” just isn’t enough bass to make it worthwhile. Twin 8” or single vented 10” is a minimum to make a 3 way worthwhile.
If I had to do it again I would not do a single 8” 3 way.
An 8” two way yea!
A single 8” way? Never again with a standard driver (Revelator and Purifi not withstanding)
The challenge with soft domes in waveguides is that since they are always operating in breakup mode (albeit small breakups) it is more challenging to simulate in a waveguide. A hard dome has more predictable behaviour, making the correlation between simulation and actual real life measurement much more likely to be accurate.
Addit: Having worked with SB29SDAC, it’s a fine tweeter. But in a waveguide? We’ll have to see (it has to be modelled, built and measured and several iterations).
I would recommend ADC or CDC, if budget allows.
Or a dimple dome or ring radiator (fixed centre of the dome reduces breakups)
Last edited:
You've gotten some good answers from others so I don't have a much to add. Ideally you want the directivity of the waveguide to match that of the woofer at the intended crossover frequency, and have the DI overall that you are looking for. Practically speaking, it so happens for my particular style of waveguide where the width is the same as the woofer, and the acoustic center depth about the same also, you have directivity that matches the woofer well. So you don't have to overthink my waveguides, use the 4" waveguide with 4" woofers, 5" waveguide with 5" woofers, etc.
Having the driver and waveguide of an equal size to each other does generally ensure that there will be a match of directivities at a sensible crossover point but this may or may not be what you want. A round or equal height and width guide will tend to allow a more constant directivity overall, where an elliptical or squashed vertical will cause the directivity to rise at high frequencies more steeply. The SB26 does rise at HF itself but not as much as the large surround drivers do.
https://www.diyaudio.com/community/threads/vituixcad.307910/post-6541438
What tends to happen is that the directivity is kept to a lower frequency but then all of a sudden lets go unless the directivities are blended with a phase mismatch. You have had a bit of experience with crossover phase matching or mismatching to change the directivity so you may have formed an opinion on whether this matters to you.
It is simplest to make a box of the right width and depth but not the full height to allow for 1/4 symmetry and more mesh resolution in the model. Moving to half symmetry for a full box is easier at mid and low frequencies. What crossover frequency are you targeting?
One of the design goals of this speaker is to be compact. The 8" woofer design is about 21 liter overall with 5 l for the mid enclosure, leaving 16 l for woofer volume. Another design goal is to allow for easy expansion, where a second 8" woofer cabinet can be used in conjunction. The current design allows stacking, so the second cabinet has the same profile as the first, and the two together become a floor standing speaker... I thought this would be useful in large rooms.
However, as I thought about the SPL capabilities of the single woofer version, in a sealed box, with +6 dB of boost to get an F3 of 40 Hz, I decided that the second woofer would almost always be needed, except in the smallest of rooms. if the second woofer is going to be needed all the time anyway, the speaker is no longer compact, and I can't fairly claim that the drivers cost $250 per side. This has caused me to evaluate the possibility of a larger woofer. I have found a 10" driver that gives me +3 dB more capability in the bass, extends the bass response from F3=40 to F3=37, and it requires the woofer volume to grow just slightly from 16 l to 21 l. I can probably stay within the budget. The cabinet is about 1/2 inch wider, 1 inch deeper, and 2" taller. It is still a very compact speaker. If this option is chosen, then there will be no need for an expansion capability. So yes, a 10" driver is under serious consideration. I will post more details tomorrow when I am at my other computer.
When I said "The SB29SDAC does simulate quite well with 4" and 5" drivers in a well beveled cabinet", I was speaking of a non-waveguide cabinet, with tapered bevels. This combination simulates quite well.
Thanks!
j.
As I mentioned in the previous post, I am reconsidering my woofer requirements.
I was very close to a final decision selecting the Dayton RS225-8 woofer. From a cost standpoint, at $90 it is correctly priced for this project. From a performance standpoint, particularly distortion, it is significantly better than the lower cost options in the $40 - $80 range. To achieve a significant step up in performance would have required spending almost double the price, so it is a good value. The preliminary cabinet design was a 21 liter box, 23” H x 11” W x 8.5” D, 16 liter woofer volume + 5 liter midrange enclosure.
One of the design goals was this: “A modular design which would allow me to design/fabricate a second bass driver/cabinet later. This would extend the bass response and expand the SPL capability.”
Simulation showed that this 8” woofer could generate 96 dB SPL at 1 M, at 40 Hz, at Xmax. Please note that this is Xmax calculated by voice coil geometry, not based on Klippel type performance metrics. I started to think about how this speaker would be used, and I realized that nearly every end user would prefer to have the second woofer cabinet. The combination of sealed box alignment with Linkwitz Transform EQ provides a very high quality bass reproduction, but it requires a lot of volume displacement (Sd and Xmax). I had a clever design for a second cabinet which would serve as a stand and bring the height up to 38”. However, at that point it is no longer a compact speaker, and it now has a driver cost of $350/side.
The Dayton RS270 series has the same performance features as the RS225, but with 40% more volume displacement as Xmax. The cost is about 45% higher than the RS225 at $131. Once again, to get a 10” woofer that is significantly better would require spending about double the cost.
So I looked into an alternate cabinet that would use either the RS270-8 or RS270-4. Since I would no longer need to allow for an expandable second woofer cabinet, I was now free to consider 4 Ohm woofers.
Both drivers would work well. The RS270-8 has a slightly higher SPL capability, but it requires a 25 liter woofer volume, which deviates too far from my original goal.
The RS270-4 could work in the original 21 liter cabinet, but there was not enough height in the baffle to accommodate the drivers, the bevels, and the necessary spacing. Slightly increasing the cabinet volume allows the RS270-4 to perform even better, and it provides the required baffle space.
The net result is a 3 dB increase in bass SPL capability, and and extension of F3 from 39.4 Hz to 36.7 Hz, and F6 from 31.3 Hz to 28.3 Hz.
The new cabinet is 25.5” H x 11.5” W x 8.75” D (648mm x 292mm x 222 mm), with 25.5 liter total internal volume, 20.5 liter for the woofer and 5 liter for the mid. It is just slightly larger than the original, and yet still very compact for a speaker with a projected F6 of 28 Hz.
If I decide to go with the RS270-4, there is no need for an expansion capability, so this requirement goes away.
j.
I was very close to a final decision selecting the Dayton RS225-8 woofer. From a cost standpoint, at $90 it is correctly priced for this project. From a performance standpoint, particularly distortion, it is significantly better than the lower cost options in the $40 - $80 range. To achieve a significant step up in performance would have required spending almost double the price, so it is a good value. The preliminary cabinet design was a 21 liter box, 23” H x 11” W x 8.5” D, 16 liter woofer volume + 5 liter midrange enclosure.
One of the design goals was this: “A modular design which would allow me to design/fabricate a second bass driver/cabinet later. This would extend the bass response and expand the SPL capability.”
Simulation showed that this 8” woofer could generate 96 dB SPL at 1 M, at 40 Hz, at Xmax. Please note that this is Xmax calculated by voice coil geometry, not based on Klippel type performance metrics. I started to think about how this speaker would be used, and I realized that nearly every end user would prefer to have the second woofer cabinet. The combination of sealed box alignment with Linkwitz Transform EQ provides a very high quality bass reproduction, but it requires a lot of volume displacement (Sd and Xmax). I had a clever design for a second cabinet which would serve as a stand and bring the height up to 38”. However, at that point it is no longer a compact speaker, and it now has a driver cost of $350/side.
The Dayton RS270 series has the same performance features as the RS225, but with 40% more volume displacement as Xmax. The cost is about 45% higher than the RS225 at $131. Once again, to get a 10” woofer that is significantly better would require spending about double the cost.
So I looked into an alternate cabinet that would use either the RS270-8 or RS270-4. Since I would no longer need to allow for an expandable second woofer cabinet, I was now free to consider 4 Ohm woofers.
Both drivers would work well. The RS270-8 has a slightly higher SPL capability, but it requires a 25 liter woofer volume, which deviates too far from my original goal.
The RS270-4 could work in the original 21 liter cabinet, but there was not enough height in the baffle to accommodate the drivers, the bevels, and the necessary spacing. Slightly increasing the cabinet volume allows the RS270-4 to perform even better, and it provides the required baffle space.
The net result is a 3 dB increase in bass SPL capability, and and extension of F3 from 39.4 Hz to 36.7 Hz, and F6 from 31.3 Hz to 28.3 Hz.
The new cabinet is 25.5” H x 11.5” W x 8.75” D (648mm x 292mm x 222 mm), with 25.5 liter total internal volume, 20.5 liter for the woofer and 5 liter for the mid. It is just slightly larger than the original, and yet still very compact for a speaker with a projected F6 of 28 Hz.
If I decide to go with the RS270-4, there is no need for an expansion capability, so this requirement goes away.
j.