Finishing the design of my ZRT cabinets.

[TheD]

Hi,

After many years of having the drivers and crossovers and a number of failed attempts I am finally putting in the work to finish my ZaphAudio ZRTs.

Due to the cabinets being both no longer available and the huge shipping costs that I would have to bare to get a pair to me, I have started designing my own cabinets in AutoCAD for them (based off the drawings on Zaph's site).

The problem I am having is that I am not sure what I should have the box volume at and if I need to rebate slots for bracing or if it is fine to just glue the bracing flush to the walls (the parts are getting cut with a CnC machine, so that will push up costs).

With regards to the box volume I am at 37.499 liters for the box sans bracing (braces will be 3 bits of 25 mm MDF), but based on my calculations the Madisound MD38 (which Zaph's drawings seem to be based) is 40.57 liters with the 3 braces taking up less than the 2.57 liters needed to push it back to 38 liters.
Should I stick to the Madisound cab volume or make my cabinets a real 38 liters or is it OK to have a cab with only 36ish liters (after bracing)?

Do I also need to take into the size of the crossovers and acoustic cabinet lining?

Any ideas would be welcome.

P.S: I might post up the dxf or dwg for the design after I have finished, in case someone else wants to build a pair but does not have cabinets.

[JonW]

Hi TheD,

I don't have any additional info on the ZRT's, but I did make a speaker using that same driver pair. And they really do sound amazing, in my opinion. So, regardless of what you decide, you should end up with a very nice sounding pair of speakers.

When I was trying to decide on what volume and bass tuning to pick, I experimented. I built test cabinets and tried various tunings and listened a lot. What I found was that getting more bass extension came at a little bit of clarity in the mids. Just a little. We're splitting hairs here. Anyways, in the end, I went with a 34Hz tuning and around 39.3 liters volume. I seem to recall Zaph going for a lower tuning, maybe it was 30 Hz? It's a small difference and will be good either way, but I went a little higher than Zaph did.

Here is the thread from my build, with all the gory details. The bass tuning experimenting is probably in there somewhere, buried deep. I am half way around the world right now, on a work trip in Singapore. My head is spinning with jet lag and being tired. So I thought that I'd check out htguide when I can't sleep. But I'm not feeling like reading through the whole thing again at the moment. I can find it later if you want, in some days after my travel calms down and I am sleeping again.

https://www.htguide.com/forum/showth...0-amp-18W-8531)

Have fun with the build!

[TheD]

Yeah, I remember seeing your thread.

Unfortunately I will not be able to make a bunch of test cabs (due to not having the room for a router I am having the cabinets CNCed and even then I have had to change my design to use butt joints for the baffle to save on costs).

As it stands my design is about 38 liters when taking into account the bracing (3 25mm bits of MDF with a depth of 40mm from each side of the cabinet).

What I am now worried about is if I should take into account the space taken up by the glue in each joint and if butt joints could even hold the baffle.

Thanks for the help.

[JonW]

Hi again,

No need to make any test cabinets. Just pick which tuning you want and make the cabinet volume accordingly. Alternatively, you might be able to keep with a given volume and change the port length to reach a desired tuning.

Regarding tuning and volume, like I mentioned, Zaph and I came to different choices on which tuning is best. You don't need to test anything, just pick one and go with it. I found that the difference between 30 and 34 Hz tuning was not large. So you're OK either way or in between. You could even go higher or lower.

Volume is just actual volume "seen" by the woofer. So you take the internal volume of the cabinet and subtract out the braces, the crossovers, even the woofer cone, depending on how fancy and precise you want to get. The volume of the glue is negligible.

If you want to prove to yourself what matters and what does not, get a tuning program. WinISD online is easiest for this type of quick thing. Pick a volume and tuning. Say 38 liters and 30 Hz or something along those lines. Then change the volume by, say, 0.5 liters. See how much (or little) difference it makes. Then try to factor in your glue volume or the crossover volume. Then you'll have a feel for what matters. Here you go:

http://www.linearteam.dk/?pageid=isdonline

Regarding butt joints or rabbets for the braces, both can work. I've gone both ways with success. I guess it all depends on your woodworking skills. In theory, the rabbets are better. But if you can't make them well, it could be worse than a butt joint. A good butt joint, glued well, can be excellent. If you can get everything to stay straight. Sometimes that can be a challenge, depending on the project.

Good luck!

[TheD]

Hi,
I think I will keep with the original tuning, already got a far amount of things to think about with this design and I do not really want more.

With regards to the glue volume, I was thinking less about the woofer and more in regards to if the volume of the glue would make the cabinet rebates not line up/fit (don't want a cabinet bursting at the seams).

I have no wood working skills (reason for CNC), the lack of rebates on the baffle means I am looking into glue and screwing it on (I think ferrous screws should not be a problem). Only other way I can think of that does not require rebating the back of the baffle is rebating the sides and top panels and having a smaller baffle that kind of slots into the front, but I have no idea if I could get it smooth.

I also can make the sides thicker (and oddly cut the cost) by changing the MDF from 18 mm to 25 mm, but I am not sure of the effects having the side walls closer would have on the rear reflections from the woofer.

Late edit: What rebate depth should the drivers have? I have 6 mm on the tweeter and 7 mm on the woofer. Sadly the driver specs are not really clear on how much the foam compresses.

[JonW]

Hi TheD,

The original tuning will work very well. You are good there.

Ahh, OK. Now I see what you are trying to do with the cabinet. Rebating or recessing the front baffle to fit into the cabinet will work. Also, making the cabinet with the recess and dropping the front baffle in there will be fine. A third option is to just glue the baffle onto the front. Option #4 is to screw the baffle on flush. (I’d try to find non-ferrous screws but it probably does not matter.)

It all just depends what you are comfortable with. I have had recesses into which I put cabinet braces and I have also just glued them in with butt joints. They each have pros and cons and I do not prefer one way over the other by too much. Maybe, in the end, sliding shelf braces into recesses/rabbets is a better way to go, if I had to choose. But it depends in the situation. For a really small cabinet, butt joints are fine. The grooves are more helpful with larger cabinets. In the end, it’s one of those things that you will have to develop your own feel for things, from experience of trying different things on different projects.

If pieces do not match up perfectly, you can always sand down some parts. And add putty/bondo to fill gaps.

One thing that I will recommend is to use high quality birch plywood instead of MDF. I find that it is more stable and things shrink or expand less once the wet glue is applied. I do not think that glue will influence your volume at all. Yes, things may slide around a bit once the glue is applied. That’s why the rebates/recesses might help.

You can change the wall thickness without any problem, I will guess. As long as the 25 mm does not get close to the woofer. See how much clearance there is with the Zaph cabinets versus the 25 mm sides. 25 mm should make for a more stable, less resonant cabinet, generally speaking. But definitely do not change the width of the front baffle or relation of the drivers to the edges.

For driver rebates, you want the flange of the driver to be flush with the cabinet face. You can get the specs from the driver sheet but, for this one, I usually wait to have the drivers in hand. Then I router down the rebate until I have a flush fit. In your case, to make the CNC specs, I would recommend getting the drivers in hand, measuring those, and pick the dimensions accordingly. You can add in a mm or so for the foam but I probably would not. The foam will compress to almost no width once the drivers are screwed in tightly.

[JonW]

For what it's worth... I thought about it a little more. And I think that I will recommend having the front baffle as one piece. Then glue on a second piece to the back of the baffle. This second piece will be smaller that the main baffle, so that it forms the rebate/recess and then the double thickness baffle drops into the rest of the cabinet. That way, you have a clean, one piece front baffle. I think that is what I might do if I were looking to make a CNC package out of it all. Subject to change and all that but it might be the way that I would go.

[TheD]

The problem I see with gluing an other bit onto the back of the baffle is lining them up correctly.

I am not sure plywood is going to be an option with the place that is doing the CNC (and I am not sure how it paints vs MDF).

25 mm MDF sides nearly line up with the edge of the rebate for the woofers, so I am not really sure if that is a problem.

I do have the drivers, but I am not sure how far the foam compresses (woofer shows 7 mm to half width of the foam on it's spec sheet).

[JonW]

At some point, you are going to have to line up all six sides. I guess it's up to you to decide at which of those points will yield the best results. It's one of those things that will be easier to predict once you have personal experience. Different people like different clamping methods, for example.

I usually veneer my ply so I do not know how it paints. Probably fine with good sanding. Ply also allows the option of a clear or stained finish. MDF is generally smoother and might paint better if that's the finish you are after. But MDF also dents more when dropped, swells when wet, and strips more easily with screws. Your call. Either is fine. Especially if you do not have to deal with the MDF dust because it will be CNC'd.

If you go 25 mm, be sure to keep the front baffle of the same dimensions for width and height. You might need to extend the depth of the cabinet to account for the added wall thickness, so that you maintain the desired cabinet volume/tuning.

I would not worry about the foam compression much. You can add in 0.5 or 1 mm if you want and that would be fine. I probably would not. A good way to know for sure is to test it out yourself. Route out a hole, add foam, screw in the driver, and see how much depth the foam added, if any.

[TheD]

At some point, you are going to have to line up all six sides. I guess it's up to you to decide at which of those points will yield the best results. It's one of those things that will be easier to predict once you have personal experience. Different people like different clamping methods, for example.

Gluing an other panel to the baffle would require something to hold it in place and what ever it was could get stuck on as well.

I usually veneer my ply so I do not know how it paints. Probably fine with good sanding. Ply also allows the option of a clear or stained finish. MDF is generally smoother and might paint better if that's the finish you are after. But MDF also dents more when dropped, swells when wet, and strips more easily with screws. Your call. Either is fine. Especially if you do not have to deal with the MDF dust because it will be CNC'd.

I plan on getting a painting done on them by an artist I know, so finishing them an other way is not relevant for me.

If you go 25 mm, be sure to keep the front baffle of the same dimensions for width and height. You might need to extend the depth of the cabinet to account for the added wall thickness, so that you maintain the desired cabinet volume/tuning.

Yeah, I have changed the volume a few times and left the baffle alone.

I would not worry about the foam compression much. You can add in 0.5 or 1 mm if you want and that would be fine. I probably would not. A good way to know for sure is to test it out yourself. Route out a hole, add foam, screw in the driver, and see how much depth the foam added, if any.

Well, I can not cut test holes, but I think what I have is fine.

[JonW]

The "something to hold it in place" would be a clamp. As they say in woodworking: "A man can never have too many clamps."

Yes, I think that your dimensions for the drivers should be fine without worrying about the foam too much.

Painted by an artist sounds cool. You have to post photos when they are done. Good luck!

[JonW]

Oh, one other tip that you can feel free to ignore: Maybe make some sides of the cabinet a little bit large, in order to leave yourself some extra room for lining things up, slipping during glue up, etc. And then use a router with a flush trim bit to get everything nice and even. Good luck!

[knowledgebass]

Gluing an other panel to the baffle would require something to hold it in place and what ever it was could get stuck on as well.

It would be helpful to use some 'pins'. Some people use toothpicks, small dowel would work, but finishing nails work too. Line everything up, drill two to four holes from the back side about the size of the pin being used - just don't drill through the front of the baffle - apply glue, slip in your pins and clamp. The pins would help prevent the pieces from slipping until the cauls and clamps are in place. Have the pin holes included as part of the CNC run to help with precision. The problem in this application is you would have glue at the seam which you'd want to clean out before it hardened as it could mess with the fit when mating the baffle with the box.

If going CNC, why not just have them cut the rim/ridge around the baffle for you? If using a 25mm baffle, just have them cut the lip all the way around the baffle for you. No need to glue pieces together. Even 1/4" inset to your box would add plenty of security for clamping the baffle to the box.

[TheD]

The "something to hold it in place" would be a clamp. As they say in woodworking: "A man can never have too many clamps."

I meant something to make sure it could not shift when I glued it and put the clamps on, I figure it could be a bit hard and messy to move it by hand once the glue was on.

Yes, I think that your dimensions for the drivers should be fine without worrying about the foam too much.

I hope so.

Painted by an artist sounds cool. You have to post photos when they are done. Good luck!

I will post pics when it is done (might take some time), not sure what it will turn out like when it is finished though (not sure what to do to protect the painting, maybe a clear coat of something).

It would be helpful to use some 'pins'. Some people use toothpicks, small dowel would work, but finishing nails work too. Line everything up, drill two to four holes from the back side about the size of the pin being used - just don't drill through the front of the baffle - apply glue, slip in your pins and clamp. The pins would help prevent the pieces from slipping until the cauls and clamps are in place. Have the pin holes included as part of the CNC run to help with precision. The problem in this application is you would have glue at the seam which you'd want to clean out before it hardened as it could mess with the fit when mating the baffle with the box.

If going CNC, why not just have them cut the rim/ridge around the baffle for you? If using a 25mm baffle, just have them cut the lip all the way around the baffle for you. No need to glue pieces together. Even 1/4" inset to your box would add plenty of security for clamping the baffle to the box.

The reason that I do not want the rebates around the front baffle is the cost of CNCing two sides of the same panel is much higher than only doing one side due to the things they have to do to make sure it does not move when they are doing it (including doing it slowly).

Thanks for those ideas anyway.

[TheD]

Hmm,

Just looking at my crossover boards I can see a strange orange substance on the ends of most of the capacitors, at first I thought it was just some flux from the solder but I have seen it on caps that have not even been soldered yet!

Do I have a batch of bad caps?

[JonW]

I don’t know but those do not sound good. Given how much time and money are invested in a speaker build, I would not use caps that look to have gone bad. If you post a photo, someone around here might be able to help. Some caps are made from oil and paper. Maybe you have some of those and they are leaking. Good luck with it.

[TheD]

They are Jantzen audio CrossCaps

[JonW]

I don't know what that stuff is. But I know little about electronics. If it were me, I'd pitch them and start with fresh parts. It's just not worth bothering with dodgy things, given how much time and money go into a speaker build. But that's me and we all have different priorities or perspectives on value. If you want to try and use them, I'd put a meter on them and see if they are still within spec. If you do not have a meter, maybe a local electronics shop can measure them for you.

[TheD]

Yep, had a shot at measuring the caps resistance with my cheap DMM and one of them is dead (the other woofer cap that had not been used and did not show as bad signs as the others), the DMM shows it stuck at 50-60K ohms (and stays that way when the DMM is reversed) when the other ones quickly rise to 1999K ohms then open circuit (and then have a negative reading then start charging back the other way).

So I am thinking that I can not trust any of the caps to be correct if one that has not even been used is dead and the others have strange stuff on them and thus I am going to have to spend $150 on new caps

[TheD]

Yeah, I have confirmed with a DMM that can check capacitance that the cap is dead (too bad the new DMM turns out to only give correct readings some of the time).
Other ones still seem good though.