DML PA systems

I have never managed to get a plate to distort really, apart from if there are issues with mounting. I did manage to burn some exciters from overheating, but then distortion eventually happened as a consequence of mechanical damage to exciter or mounting. So I don't think the plates can distort really, at least it is not enough to use a cluster of the most powerful models.

I have not done any calibrated SPL tests yet, but have a basic SPL meter now to use as reference, as well as a scope, so will test sensitivity and max SPL soon!

I have done rough SPL calibration when doing my most recent tests with REW, and tried to play really loud then to pick up as much distortion as possible, but I'm limited by what measurement mic can handle, so end up measuring around 100dB at 30cm, and then have very minimal distortion.
My best guess based on the level I give them then, is that a single plate with 4 exciters at least can get close to 110dB.

I print small blocks using TPU, which is a flexible thermoplastic. I find that they need to be soft but most foam that is appropriate softness will degrade with time. TPU is much more stable, and when printing I can make it hollow to adjust how soft they are.
 
  • Like
Reactions: 1 user
Weather getting better here, gone from freezing and snow to +18c in a few days :)
So planning to take a sub and a couple of plates outdoors and do some measurements. I finally have a scope, and since I now can measure AC voltage properly I was planning to measure response with calibrated sensitivity, but turns out that is not that easy with a class-d amp :(
There is a sine wave at 240kHz from the switching in the amp, and while it is quite many dB down from the max output of the amp, it is constant regardless of amp level, and with an overpowered amp like this the signal is still 1.33Vrms. So turns out I need a RF filter to be able to check the output voltage with decent accuracy.

Not sure I will bother making a filter since absolute sensitivity figures are not that important to me, and at least after a little testing, trying to ignore the FR component from the scope readings, my meter read 105dB at 2.83V into 4 ohm at 1kHz. That would mean sensitivity of 102dB with 1W/1m, so I'm at least confident sensitivity is very good. I can have misjudged the RF component a little bit, but should still be over 100dB. When doing outdoor test I will also test with higher voltage to reduce the error from RF signal.
 
  • Like
Reactions: 1 user
Hi Leob, glad it's warming up for you there, we are having a lovely Autumn on the other side of the world.

That's great your plates are so sensitive, so in theory you won't need a lot of watts for your output needs even at party levels.

Glad you can do some playing with your system now you have the subs to try it all together.

I have found your suspension method is great as it lets you tune the plates by moving and adding more clamping blocks in different places, I got rid of a broad 8db peak around 500hz by adding a third clamping point on one long side. I also had a 8db peak at 9000hz that I removed by adding a little felt infront of the exciter.

I sourced some 12mm epdm tape which I am using as blocks to clamp them with now instead of the closed cell foam. It holds better and I think I was getting distortion from the plate and foam rubbing/buzzing at higher SPL levels.
 
  • Like
Reactions: 1 user
Weather getting better here, gone from freezing and snow to +18c in a few days :)
So planning to take a sub and a couple of plates outdoors and do some measurements. I finally have a scope, and since I now can measure AC voltage properly I was planning to measure response with calibrated sensitivity, but turns out that is not that easy with a class-d amp :(
There is a sine wave at 240kHz from the switching in the amp, and while it is quite many dB down from the max output of the amp, it is constant regardless of amp level, and with an overpowered amp like this the signal is still 1.33Vrms. So turns out I need a RF filter to be able to check the output voltage with decent accuracy.

Not sure I will bother making a filter since absolute sensitivity figures are not that important to me, and at least after a little testing, trying to ignore the FR component from the scope readings, my meter read 105dB at 2.83V into 4 ohm at 1kHz. That would mean sensitivity of 102dB with 1W/1m, so I'm at least confident sensitivity is very good. I can have misjudged the RF component a little bit, but should still be over 100dB. When doing outdoor test I will also test with higher voltage to reduce the error from RF signal.
I await your report including video with great interest
 
  • Like
Reactions: 1 user
Was asked for some more info on how I prepare the GPS plates. While I think the hide glue with shellac combination should work well on most styrofoam, the thickness you want might vary somewhat depending on the density of the material. And perhaps for example PVA works better on some materials.

I mix 1 part hide glue granules with 3 parts water and heat in a pot in a water bath until just above 60c.
Use a paint roller and apply, making sure to roll off excess glue before rolling on the plate to get a nice even thin layer.
I missed to measure how much alcohol I added to the shellac flakes, but you want a much thinner mixture than a usual solution, probably almost 10 times the weight in alcohol compared to flakes. Soak the flakes overnight in half the alcohol, then dilute with the rest and apply in the same way as the hide glue.
Often it is recommended to use high purity isopropanol alcohol, but I just use the kind of cheap mostly isopropanol alcohol you can buy in bottles in petrol stations in red colour with horrible smell (T-Röd for the Swedes). Seems to be fine for this application...might be a bit more fuzzy when you want to make a nice french polish on furniture.

Both the hide glue and shellac should dry very quickly, so you have to work fast. If you have thin plates, you also have to be careful to avoid bending when plate dries. Paint both sides quickly after each other before they dry to minimize the risk of that happening.

I found a building preservation shop in Sweden that sells both the glue and shellac:
https://www.byggnadsvardsbutiken.se/product/hudlim-3
https://www.byggnadsvardsbutiken.se/product/schellackflingor-tn-orange100g
It is often sold by specialist shops for instrument building and restoration as well.
Hey there, I was just wondering if you do the coating for durability reasons or sound reasons?! I didn't find any info about that... many thanks for all the interesting information!