Hi all,
I picked up some GW-8004 coaxial full range drivers, they have a built in crossover. I picked them up because they have high Qts and seem to have quite a bit of excursion travel potential. This is the driver:
https://www.parts-express.com/Goldwood-GW-8004-8-8-Coaxial-Speaker-290-378?quantity=1
Here's the manufacturer graph response:
Here's the manufacturer T&S:
I went ahead and measured it myself on DATS V3. All 4 drivers measured very similar to this:
Here's a near field (3" non-anechoic) no baffle measurement (driver suspended in the air more than 3 feet away from surfaces, etc). I didn't gate or anything just to see a general idea of what it was doing. Mostly looking at that 3~6khz dip. The roll off is expected under 1khz I think here. And the rest was ok.
I used Basta to look at potential baffle sizes. After playing in that and doing some calculations for baffle width and all, I realized, it won't matter because I'll never make this large enough to support really low frequencies. I potentially would like to get down to 100~120hz with these after EQ with some boundary gain (they will mount near walls all the time). But I calculated a rather large 3.4m baffle to do that. Obviously impractical. And so the differences from 24 inches and smaller didn't seem to matter. So I figured, I'd just measure it and see what I could do.
I made a few baffles that were around the sizes I wanted to work within and made some temporary wings to extend the baffle while reducing the front baffle width here and there.
I started with 12" x 18" baffle with and without 3" wings on the sides.
Here's a few measurements. These are just near field but non-anechoic, I know they're not accurate, just looking at the overall shape of the response based on the baffle which should show up ok with fairly near field. The black line is no wings, so the baffle width for bass is 12" x 18". The other lines are two different position measurements with 3" wings ont he sides and they were more consistent but it was clear the extra baffle width helped lift the bass under 400hz.
Next, I wanted to shrink the front baffle some so I reduced to 10" x 15" and kept the 3" wings.
I later added some cut out notches around the tweeter to see if it would matter. I know this is a practice to explore with tweeter + woofer two-ways, but I didn't know if it would help with a coaxial driver like this. I don't think it matters at all, so I likely will leave the notch out in the final version. Overall I think the 10" x 15" with 3" wings (16" baffle width, by 15" height, for bass) still seems to keep enough bass for this purpose near a boundary (walls, ceiling, floor, etc). I know it will have different response from a vast distance. I will measure that soon. For now, just trying to get quick measurements just to see the baffle width changes and the bass response.
Blue line is no notch.
Black line is with a notch centered around the tweeter the size of the tweeter diameter.
It didn't seem to matter much. It maybe took down some output from the mids and treble by 1db to 2db or so here and there, but it also lost 1~2db in bass.
=================================================================
So, some comments and questions for anyone willing to help out:
1 - These are going to be used as surrounds so they will not get a lot of primary use. I mainly am looking at getting 100~120hz to 15~16khz bandwidth from them. They don't have to be flat and I'm going to EQ some. I plan on increasing the low frequency a bit with boosting and maybe some cutting on the 1~2khz range mids and a bit at 7~14khz treble ranges a little, but not a lot. That's just for context of use. It's going to be mostly ambient stuff, effects, rarely ever going to be voice or instruments and will not be the primary source when listening. Again, these will be surrounds, just for context.
2 - I'm trying to figure out what I can do about that 3~6khz null. It's the driver. You can see in the free air measurement and on different baffles, it doesn't change or move. I measured a 2nd driver to compare and it has a similar one but its shifted to different frequencies a bit. I imagine this is just a constant I will have to live with, with these drivers. Some sort of cone break up or interference between the tweeter and woofer? Any help on what this is really and if there's anything that can be done with it (I doubt it)?
3 - Any suggestions on baffle sizes and/or the wings or notches? I don't think the notches matter here since the tweeter is in a coaxial. But I'm open to any suggestions. Ultimately I'm going for a slim depth on these as they will be near walls, so figure they will pick up some boundary gain a bit. I'll measure that soon after getting the baffle width ironed out. Any suggestions on 10" and 3~4" wings? They probably will end up in a U shape, as I think doing angled wings will be a bit beyond my craftsmanship abilities on long panels. I can do that with a miter on studs and stuff, but I don't have anything to do that on long panels.
After some more tinkering, I'll do a rough baffle and wings build with plywood and do some on-wall measurements after figuring out, with any help here, any changes to the baffle.
Thanks 🙂
Very best,
I picked up some GW-8004 coaxial full range drivers, they have a built in crossover. I picked them up because they have high Qts and seem to have quite a bit of excursion travel potential. This is the driver:
https://www.parts-express.com/Goldwood-GW-8004-8-8-Coaxial-Speaker-290-378?quantity=1
Here's the manufacturer graph response:
Here's the manufacturer T&S:
I went ahead and measured it myself on DATS V3. All 4 drivers measured very similar to this:
Here's a near field (3" non-anechoic) no baffle measurement (driver suspended in the air more than 3 feet away from surfaces, etc). I didn't gate or anything just to see a general idea of what it was doing. Mostly looking at that 3~6khz dip. The roll off is expected under 1khz I think here. And the rest was ok.
I used Basta to look at potential baffle sizes. After playing in that and doing some calculations for baffle width and all, I realized, it won't matter because I'll never make this large enough to support really low frequencies. I potentially would like to get down to 100~120hz with these after EQ with some boundary gain (they will mount near walls all the time). But I calculated a rather large 3.4m baffle to do that. Obviously impractical. And so the differences from 24 inches and smaller didn't seem to matter. So I figured, I'd just measure it and see what I could do.
I made a few baffles that were around the sizes I wanted to work within and made some temporary wings to extend the baffle while reducing the front baffle width here and there.
I started with 12" x 18" baffle with and without 3" wings on the sides.
Here's a few measurements. These are just near field but non-anechoic, I know they're not accurate, just looking at the overall shape of the response based on the baffle which should show up ok with fairly near field. The black line is no wings, so the baffle width for bass is 12" x 18". The other lines are two different position measurements with 3" wings ont he sides and they were more consistent but it was clear the extra baffle width helped lift the bass under 400hz.
Next, I wanted to shrink the front baffle some so I reduced to 10" x 15" and kept the 3" wings.
I later added some cut out notches around the tweeter to see if it would matter. I know this is a practice to explore with tweeter + woofer two-ways, but I didn't know if it would help with a coaxial driver like this. I don't think it matters at all, so I likely will leave the notch out in the final version. Overall I think the 10" x 15" with 3" wings (16" baffle width, by 15" height, for bass) still seems to keep enough bass for this purpose near a boundary (walls, ceiling, floor, etc). I know it will have different response from a vast distance. I will measure that soon. For now, just trying to get quick measurements just to see the baffle width changes and the bass response.
Blue line is no notch.
Black line is with a notch centered around the tweeter the size of the tweeter diameter.
It didn't seem to matter much. It maybe took down some output from the mids and treble by 1db to 2db or so here and there, but it also lost 1~2db in bass.
=================================================================
So, some comments and questions for anyone willing to help out:
1 - These are going to be used as surrounds so they will not get a lot of primary use. I mainly am looking at getting 100~120hz to 15~16khz bandwidth from them. They don't have to be flat and I'm going to EQ some. I plan on increasing the low frequency a bit with boosting and maybe some cutting on the 1~2khz range mids and a bit at 7~14khz treble ranges a little, but not a lot. That's just for context of use. It's going to be mostly ambient stuff, effects, rarely ever going to be voice or instruments and will not be the primary source when listening. Again, these will be surrounds, just for context.
2 - I'm trying to figure out what I can do about that 3~6khz null. It's the driver. You can see in the free air measurement and on different baffles, it doesn't change or move. I measured a 2nd driver to compare and it has a similar one but its shifted to different frequencies a bit. I imagine this is just a constant I will have to live with, with these drivers. Some sort of cone break up or interference between the tweeter and woofer? Any help on what this is really and if there's anything that can be done with it (I doubt it)?
3 - Any suggestions on baffle sizes and/or the wings or notches? I don't think the notches matter here since the tweeter is in a coaxial. But I'm open to any suggestions. Ultimately I'm going for a slim depth on these as they will be near walls, so figure they will pick up some boundary gain a bit. I'll measure that soon after getting the baffle width ironed out. Any suggestions on 10" and 3~4" wings? They probably will end up in a U shape, as I think doing angled wings will be a bit beyond my craftsmanship abilities on long panels. I can do that with a miter on studs and stuff, but I don't have anything to do that on long panels.
After some more tinkering, I'll do a rough baffle and wings build with plywood and do some on-wall measurements after figuring out, with any help here, any changes to the baffle.
Thanks 🙂
Very best,
3" ( 7,62 cm) is not very good distance for speaker measurements.
Better is to make real near filed (as close to cone as possible) and 1 m distance (or more) measurements.
Better is to make real near filed (as close to cone as possible) and 1 m distance (or more) measurements.
Thanks, yes I realize, again was just trying to get quick examples of what the baffle was going to do overall to bass.
I cannot measure extremely close to the woofer due to the tweeter; its position is static on a post and the woofer moves around it.
I'll try a few other ways to get closer and measure it though. Or just do some gated 1 meter measurements. I just wanted to get an idea of the baffle size before getting deeper into more involved measurements.
Very best,
On near filed measurements you actually not see baffle size influence, for this 1 m or more distance is needed.
No worries; I appreciate any feedback. I picked these up on sale, cheap, just because they were full range coaxials with a built in crossover so I had less work to do. Definitely not up to the standards I'm sure of this forum's expectations of a full range driver by any means. This is a $30 per driver experiment and learning platform for me either way.
Very best,
Thanks; hrm, not sure if I want to tear out that tweeter and crossover, lots of effort for what is basically a throw away driver. If it's a re-build, then it's just junk and I can move on in that sense.
Something to fiddle with while waiting for either the Markaudio or Fostex drivers I'm eyeballing to be available again.
Very best,
Ok, got some more measurements.
This time I did everything at approximately 1 meter distance. I tried middle of the room so that there was greater than 48" in all directions so I could gate it, but that only allows me to look at mids & treble. I lose bass metrics when I gate for the room for obvious reasons. So I looked at it at a few steps. 3ms (300hz) and 10ms (100hz).
I did a small 10" x 15" baffle, no wings. A little inverted T on the back to help it stand in place.
For measuring, I took the center of the driver to 48" from my floor and it was 48" from the ceiling and more than 48" from any wall.
That 3kzh to 6khz null moved. I think it was however I had it mounted previously, maybe it was resonating real bad or loose. I re-oriented the driver for new screw holes so they're clean and that null moved away. Now, I only see the 7khz deep null that is shown also on the manufacture release graph with the same expected null. So this is promising. It's still bad but not as terribad as it was, lol.
Here's IR gated to 10ms, 100hz to 20khz
I tightened it up to gated to 3ms, limited to 300hz to 20khz and this is the result
Here's the distortion from the sweeps 100hz to 20khz. I have something nasty going on down near 100hz and you can see its harmonic spikes chasing it upstream. That may be the mounting or the baffle itself. I'm using 3/8th inch plywood here and it vibrates like a toy. So if I ignore those obvious harmonic spikes, maybe its less than 10% down at its worst. I can live with that, but I'll try to get a better idea of things on a better built baffle that is more rigid.
Since the data is there, I went ahead and looked at waterfall for the sweeps, all the way up to 105db peak, I took it out to 300ms, then limited it to 85db at the bottom simply because it was already fallen off by 120ms.
Finally, I loaded Foobar2000 and started playing the Donald Fagen with some heavy EQ boosting and some rolled off treble over 16khz. I listened at loud levels for a while just to see how it behaved and how it sounded. It sounded like a cheap speaker does typically to me. It wasn't particularly good at anything, at least in this configuration on this baffle. Maybe it would be better in an enclosure sealed up. It was ok though, tolerable. Not what I would want for a primary listening speaker. But, I didn't hear nasty stuff going on with the boosting. The excursion seemed to handle it ok without gross effects.
Here's the RTA down to 40hz, I had a substantial house curve on it in Foobar2000 to playback some music and it sounded balanced to my ear, so just took RTA data to confirm and it looks much more flat if you can see the invisible line going through all this mess. This was pretty loud, louder than my usual listening, I just wanted to stress it a bit to see how bad the EQ boosting would be.
Here's the driver well past its typical excursion, handling it fine, sounds ok. No mechanical noises or anything. I'm sure distortion was much higher. Just wanted to test it in case it was sent to 105db at any point with a boost.
Very best,
This time I did everything at approximately 1 meter distance. I tried middle of the room so that there was greater than 48" in all directions so I could gate it, but that only allows me to look at mids & treble. I lose bass metrics when I gate for the room for obvious reasons. So I looked at it at a few steps. 3ms (300hz) and 10ms (100hz).
I did a small 10" x 15" baffle, no wings. A little inverted T on the back to help it stand in place.
For measuring, I took the center of the driver to 48" from my floor and it was 48" from the ceiling and more than 48" from any wall.
That 3kzh to 6khz null moved. I think it was however I had it mounted previously, maybe it was resonating real bad or loose. I re-oriented the driver for new screw holes so they're clean and that null moved away. Now, I only see the 7khz deep null that is shown also on the manufacture release graph with the same expected null. So this is promising. It's still bad but not as terribad as it was, lol.
Here's IR gated to 10ms, 100hz to 20khz
I tightened it up to gated to 3ms, limited to 300hz to 20khz and this is the result
Here's the distortion from the sweeps 100hz to 20khz. I have something nasty going on down near 100hz and you can see its harmonic spikes chasing it upstream. That may be the mounting or the baffle itself. I'm using 3/8th inch plywood here and it vibrates like a toy. So if I ignore those obvious harmonic spikes, maybe its less than 10% down at its worst. I can live with that, but I'll try to get a better idea of things on a better built baffle that is more rigid.
Since the data is there, I went ahead and looked at waterfall for the sweeps, all the way up to 105db peak, I took it out to 300ms, then limited it to 85db at the bottom simply because it was already fallen off by 120ms.
Finally, I loaded Foobar2000 and started playing the Donald Fagen with some heavy EQ boosting and some rolled off treble over 16khz. I listened at loud levels for a while just to see how it behaved and how it sounded. It sounded like a cheap speaker does typically to me. It wasn't particularly good at anything, at least in this configuration on this baffle. Maybe it would be better in an enclosure sealed up. It was ok though, tolerable. Not what I would want for a primary listening speaker. But, I didn't hear nasty stuff going on with the boosting. The excursion seemed to handle it ok without gross effects.
Here's the RTA down to 40hz, I had a substantial house curve on it in Foobar2000 to playback some music and it sounded balanced to my ear, so just took RTA data to confirm and it looks much more flat if you can see the invisible line going through all this mess. This was pretty loud, louder than my usual listening, I just wanted to stress it a bit to see how bad the EQ boosting would be.
Here's the driver well past its typical excursion, handling it fine, sounds ok. No mechanical noises or anything. I'm sure distortion was much higher. Just wanted to test it in case it was sent to 105db at any point with a boost.
Very best,
It wouldn't let me edit...
Here's the EQ'd response from 1 Meter, psych smoothed, 60hz to 500hz. I put a heavy house curve on 100hz and dropped it towards 300hz. For this measurement, I dropped everything as I approached 500hz just to limit and keep the tweeter quiet for this.
Here's the distortion with this boost. That 100hz spike is around 2%, at 104db at 1 meter.
Time domain just because it's fun to see. Something near 150hz seems to want to linger past 300ms upwards into 80~85db range. Maybe that's a wall reinforcement standing wave?
Very best,
Here's the EQ'd response from 1 Meter, psych smoothed, 60hz to 500hz. I put a heavy house curve on 100hz and dropped it towards 300hz. For this measurement, I dropped everything as I approached 500hz just to limit and keep the tweeter quiet for this.
Here's the distortion with this boost. That 100hz spike is around 2%, at 104db at 1 meter.
Time domain just because it's fun to see. Something near 150hz seems to want to linger past 300ms upwards into 80~85db range. Maybe that's a wall reinforcement standing wave?
Very best,