I'll only mount magnets on four plates at this stage; I'll have to test the suspension more for the membrane before I go big.
Also, I'm not done with evaluating the membrane's material.
I'm not so happy with the distortion levels; I need at least 10 dB less THD (from -40 down to -50 dB).
I've used 5 mm wide 3D printed spacers, 20 mm long, almost as the oval cut outs in the plate.
The epoxy glue that I'm using cures fast, so I must remove the spacers fairly soon after the magnets are glued to the plate.
Also, I'm not done with evaluating the membrane's material.
I'm not so happy with the distortion levels; I need at least 10 dB less THD (from -40 down to -50 dB).
I've used 5 mm wide 3D printed spacers, 20 mm long, almost as the oval cut outs in the plate.
The epoxy glue that I'm using cures fast, so I must remove the spacers fairly soon after the magnets are glued to the plate.
Here's some PEN membrane distortion measurements with 3 mm and 6 mm wide aluminium traces.
Quite a difference I would say.
The mid membrane with 3 mm traces (red) THD is at -60dB; goal achieved!
At THD -50dB (blue), the tweeter is almost there. There's still some bumps to sort out though.
The jury is still out when it comes to having return traces in the membrane or a separate return lead.
I did the the weeding a little different for these membranes; instead of weeding before applying the 3M 77 spray lim, I weeded after.
By doing so it's a lot easier to remove the backing paper afterwards. On the downside the weeding is a lot messier.
Next up: 20 µm BOPP. This time 3 mm wide traces only.
Quite a difference I would say.
The mid membrane with 3 mm traces (red) THD is at -60dB; goal achieved!
At THD -50dB (blue), the tweeter is almost there. There's still some bumps to sort out though.
The jury is still out when it comes to having return traces in the membrane or a separate return lead.
I did the the weeding a little different for these membranes; instead of weeding before applying the 3M 77 spray lim, I weeded after.
By doing so it's a lot easier to remove the backing paper afterwards. On the downside the weeding is a lot messier.
Next up: 20 µm BOPP. This time 3 mm wide traces only.
Last edited:
Congratulations on reaching your target goal of -60 dB THD!
Have you measured the efficiency in dB @ 1m / 1W?
And interesting that you got a distortion reduction with the 3mm wide traces. I got a distortion increase but that might be because of how heavy my membrane is compared to yours.
Have you measured the efficiency in dB @ 1m / 1W?
And interesting that you got a distortion reduction with the 3mm wide traces. I got a distortion increase but that might be because of how heavy my membrane is compared to yours.
Thanks, OllBoll.
As I'm only measuring on a less than a fifth of the complete SMAPPP, it's hard to predict the sensitivity.
As an indication however, here's an SPL measurement on the mid membrane with 1 W at 1 m. The membrane had a 4 ohm resistor in series:
Yes, there could of course be other reasons other what you might expect.
I can make a 6 mm BOPP membrane as well, just to be sure.
As I'm only measuring on a less than a fifth of the complete SMAPPP, it's hard to predict the sensitivity.
As an indication however, here's an SPL measurement on the mid membrane with 1 W at 1 m. The membrane had a 4 ohm resistor in series:
Yes, there could of course be other reasons other what you might expect.
I can make a 6 mm BOPP membrane as well, just to be sure.
Nice!
I just remembered about the distortion increase in 3 vs 6 mm.
I think I had reduced distortion for the 3 mm at low levels but when I pushed it to higher levels then the distortion was higher on the 3 mm than the 6 mm wide traces.
In your -60 dB test, I assume you tested at your target SPL or close enough / 5?
I remembered another reason why you probably will have less distortion: yours will be taller and run full power along the whole length where mine is shorter and will be shaded, so I need to push mine harder. Hence I will need to test at higher power levels where you can optimize for lower and most likely end up with less distortion overall.
I just remembered about the distortion increase in 3 vs 6 mm.
I think I had reduced distortion for the 3 mm at low levels but when I pushed it to higher levels then the distortion was higher on the 3 mm than the 6 mm wide traces.
In your -60 dB test, I assume you tested at your target SPL or close enough / 5?
I remembered another reason why you probably will have less distortion: yours will be taller and run full power along the whole length where mine is shorter and will be shaded, so I need to push mine harder. Hence I will need to test at higher power levels where you can optimize for lower and most likely end up with less distortion overall.
Then perhaps there are different reasons for the distortion.
The 3 mm trace shows lower distortion as long you are within the linear part of the gap, exceeding that and you'll get high distortion.
The 6 mm trace has higher distortion at lower levels compared to the 3 mm,
but at higher levels the 6 mm trace doesn't have to travel outside the linear partion of the gap at any comparable output.
This as the 6 mm trace is more efficient.
I think it was around 4 watts. I haven't measured the distortion at any higher wattage but I did a compression test.
Using pink periodic noise and REW's RTA, I increased the input by 3dB and the output was 3dB higher all the way up to 0 dB.
Yes, we have different use cases.
The 3 mm trace shows lower distortion as long you are within the linear part of the gap, exceeding that and you'll get high distortion.
The 6 mm trace has higher distortion at lower levels compared to the 3 mm,
but at higher levels the 6 mm trace doesn't have to travel outside the linear partion of the gap at any comparable output.
This as the 6 mm trace is more efficient.
I think it was around 4 watts. I haven't measured the distortion at any higher wattage but I did a compression test.
Using pink periodic noise and REW's RTA, I increased the input by 3dB and the output was 3dB higher all the way up to 0 dB.
Yes, we have different use cases.
Some measurement with 20 µm BOPP, 3 mm traces. Three traces for the mid (about 2 ohm) and one for the tweeter (about 0.7 ohm).
This is how it looks before corrugation:
To keep the corrugation, the long side edges of respective membrane parts and between them has be enforced with 60 µm aluminium tape.
I have done some test with return leads in the aluminium foil, but they measured as well as having outside return leads
Both membranes has a 4 ohm resistor in series during measurements.
This is how it looks before corrugation:
To keep the corrugation, the long side edges of respective membrane parts and between them has be enforced with 60 µm aluminium tape.
I have done some test with return leads in the aluminium foil, but they measured as well as having outside return leads
Both membranes has a 4 ohm resistor in series during measurements.
First the mid membrane:
Perhaps 1dB over all lower distortion compared to PEN, but it's a lot calmer at the low end!
Perhaps 1dB over all lower distortion compared to PEN, but it's a lot calmer at the low end!
So what about compression? Here's the mid membrane again from 1W (-21dB input level) to around 50 W (-3dB input level) in 3dB steps.
At 50W, brown trace, the amplifier distorts.
As I see it, there's no compression.
Distortion rises with increased input level, until 50W that is!
Compression for the tweeter:
Light green is the 50W trace.
I'm not so happy with the higher distortion levels between 5 and 6 kHz.
At 50W, brown trace, the amplifier distorts.
As I see it, there's no compression.
Distortion rises with increased input level, until 50W that is!
Compression for the tweeter:
Light green is the 50W trace.
I'm not so happy with the higher distortion levels between 5 and 6 kHz.
My conclusion is that BOPP is a keeper for the mid membrane, but I need to work on the tweeter membrane a little bit more.
First I need to have two separate membranes, and the tweeter membrane needs to be a lot lighter to get the a better response level.
Having only one trace, it's 5dB lower than the mid membrane if they are of the same material.
I have some very thin PET that I can test...
First I need to have two separate membranes, and the tweeter membrane needs to be a lot lighter to get the a better response level.
Having only one trace, it's 5dB lower than the mid membrane if they are of the same material.
I have some very thin PET that I can test...
Nice Build !!!! that measures really good, and also looks really nice ! that plotter works wonders!
what foil do you use for the tweeter (thickness substrate and alu) by the way the distortion above at 5khz, looks a bit like it will when having breakup in the foil. (when conductor to conductor distance is to big the undriven area will resonanate) maybe you need a bit more driving surface then one trace ?
is it possible to make it a 2 traces wide and use the space that is now filled with undriven alu as returns ? you can put one return on either side of the 2 driven traces if it wont fit on one side (if the magnet motor alows it ofcourse)
is it possible to make it a 2 traces wide and use the space that is now filled with undriven alu as returns ? you can put one return on either side of the 2 driven traces if it wont fit on one side (if the magnet motor alows it ofcourse)
By the way whats the benefit of using BOPP vs Mylar ? i was reading backwards to find out 🙂 but i could not find it, so i might have missed it somehow. by the way maybe we should have a Skype call sometimes. might be fruitfull for both, as well as i think it would be fun ! love the effort you put in builds ! you got allot of patience !
side note.. nothing to do with the build, but since all the people in here have the same passion for making drivers like this. anyone interested in thin 13 micron alu tape ? since i am and are talking to supliers but i bet it will cost me allot more then i want as ussual. maybe i can sell some (for the original price) to keep costs down?
since we all use a plotter. i really like the method i use for bass panels. cutting alu tape, and then chose whatever substrate to put on there. no glue needed. and no glue at places it should not be. acrylic glue is thicker though then a spay i noticed. 🙁 but the damping is really nice
since we all use a plotter. i really like the method i use for bass panels. cutting alu tape, and then chose whatever substrate to put on there. no glue needed. and no glue at places it should not be. acrylic glue is thicker though then a spay i noticed. 🙁 but the damping is really nice
I use 7 µm aluminium foil with paper backing, you now the Danish sandwich paper.
I find it esay to cut and weed.
As I corrugate the tweeter membrane I thought that it would be impossible for any undriven part to resonate/breakup?
I don't want any more driving area, 3 mm aluminium trace in the 5 mm gap between two magnets, as it will increase the distortion levels but I can have return leads outside the two magnet rows.
In fact, I have already tried it, but I didn't find it better.
The benefits of BOPP is that it had a slightly better performance.
It was also easier to deploy onto the cut membrane.
It's thicker though (20 µm) than the others (PET and PEN) I have tested so far.
I have still to test these PETs:
1 micron clear
2 micron clear
5 micron clear
7.5 micron clear
5 micron aluminised
10 micron aluminised
12.5 micron aluminised
First step is to cut the combined mid/tweeter membrane into two separate membranes; they don't need to be of the same materials and the tweeter might suffer from from the mid membrane.
Also, it could be could to have a lighter membrane to increase the output.
Does the 13 µm aluminium foil have any backing?
How come you don't need any glue for your bass membranes? (I have seen your Youtube videos, but don't remember this).
I find it esay to cut and weed.
As I corrugate the tweeter membrane I thought that it would be impossible for any undriven part to resonate/breakup?
I don't want any more driving area, 3 mm aluminium trace in the 5 mm gap between two magnets, as it will increase the distortion levels but I can have return leads outside the two magnet rows.
In fact, I have already tried it, but I didn't find it better.
The benefits of BOPP is that it had a slightly better performance.
It was also easier to deploy onto the cut membrane.
It's thicker though (20 µm) than the others (PET and PEN) I have tested so far.
I have still to test these PETs:
1 micron clear
2 micron clear
5 micron clear
7.5 micron clear
5 micron aluminised
10 micron aluminised
12.5 micron aluminised
First step is to cut the combined mid/tweeter membrane into two separate membranes; they don't need to be of the same materials and the tweeter might suffer from from the mid membrane.
Also, it could be could to have a lighter membrane to increase the output.
Does the 13 µm aluminium foil have any backing?
How come you don't need any glue for your bass membranes? (I have seen your Youtube videos, but don't remember this).
Thanks for the explaination, i must say any foil thicker will have there pro's (in distortion) since you are using 20 micron i am not sure if its the BOPP or the fact it has better damping due to its weight. in my opinion 12 micron Mylar is the all rounder. for tweeter 6 might work to, going lower brings all kinds of troubles with resonances. so you could.. (but not going to) skip the 1 and 2 micron. they worked out fine for my in headphones. but tweeter duty or anything else became troublesome. if you gap is 5 mm and use a 3mm trace, you can easilly use 2 x 3 mm, i know efficiency wise this is not the best thing. but if you look at my older tweeters the traces extend almost to the middle of the magnet. and they look muuuuuch cleaner then lets say a Neo 3 in the top end. i think the reason is as i said undriven mylar is not nice. even when the field is much lower on top of the magnet it still drives it allong at a lower volume making it not break up but follow the excusion be it lower. (you can adjust the difference even with wider traces in stronger field at the cost of efficiency)
The tape 13mic foil will not have a backing only a linner (release paper) and acrylic adhesive. i am talking with a few supliers... one can make 13 mic alu with adhesive combined it would be 48micron thick... a bit to heavy for my taste. another one said they could do 15 mic alu and 15 mic adhesive. if thats true... witch always is a gamble. that one would be my go to.
The bass panels do not need glue because i use huge rolls of alu tape 30 mic 50 mic or 100 mic that already has adhesive on it. so there only will be adhesive at places where alu will be. the rest is clean mylar. something Magnepan recently found out... or copied in there LRS+ (or cleaned up the glue after, witch would be tricky)
The tape 13mic foil will not have a backing only a linner (release paper) and acrylic adhesive. i am talking with a few supliers... one can make 13 mic alu with adhesive combined it would be 48micron thick... a bit to heavy for my taste. another one said they could do 15 mic alu and 15 mic adhesive. if thats true... witch always is a gamble. that one would be my go to.
The bass panels do not need glue because i use huge rolls of alu tape 30 mic 50 mic or 100 mic that already has adhesive on it. so there only will be adhesive at places where alu will be. the rest is clean mylar. something Magnepan recently found out... or copied in there LRS+ (or cleaned up the glue after, witch would be tricky)
Ooh i noticed spraying the alu first , then cut and weed, makes cutting easier when using thin foils. i know this is not a problem with the paper backed stuff. but it will be with lets say 12 micron al without paper backing, it tends to rip. but it is less likely to rip when you first put on the spray glue. just like the paper makes it easier to cut