Piezo Horn Tweeter EPIC FAIL in the MC-1800 Mini Mach One Project Part 3
HTML-код
- Опубликовано: 10 сен 2024
- Oh no! Just when this Mini Mach One Speaker Project was going so well...
The Piezo Horn Tweeter isn't working out quite as well as I expected.
There is some good news though. The Realistic MC-1800 DIY Loudspeaker build is making serious progress.
Thanks for joining me as I experience the pros and cons of DIY speakers.
Well... I figured it out. Did the filming and still need to edit and publish. I'll keep you in suspense until the video is published... Cliffhanger
lol, I look forward to watching it.
I've built with piezo tweeters before. You can't have inductors inline with piezos. They essentially act as a straight capacitor, in reality with some internal resistance. But adding an inductor will create a straight LC circuit dropping impedance at that resonant frequency to 0 ohms. And this can cause amp burnout or shutdown and current oscillations between the inductor and capacitance of the piezo element, risking burnout of the piezo, a $5 mistake but still nonetheless.
This is why you need to run a 1st order filter only. This is accomplished by adding a parallel resistor to the tweeter and using a single capacitor inline to filter out unwanted lows, for convenience sake use 8 ohms, as this is system impedance and will maintain flat impedance throughout frequency range. This forces the tweeter, to act as a voltage following device and ensures that system impedance at resonance never falls below the DC resistance of the tweeter and there's no resonant frequency from an LC circuit to drop impedance. Since we're running a theoretical 8 ohm system impedance the capacitor used to create a 3 khz 1st order crossover is simply 6.63 uF, but experience tells me the tweeter impedance is about .1 uF so use about a 5.53 uF inline capacitor. This should keep the crossover point pretty close to 3khz.
There's also another trick, to lower distortion keep the crossover above 3kz. I've tested these tweeters in and out of horn. Those tweeters you've selected are the low compression, wide dispersion Motorola CTS clones about 90-93 dB efficient but wide dispersion which is good for better SQ. In my opinion better than the high compression small diameter ones that do about 96 dB but tend to beam at high frequencies.
The low compression large horns typically have bad rolloff at higher frequencies but the piezo helps solve that by having better response at ultrahigh frequencies above 12khz. They're used in pest control situations because of their ultrasonic efficiency. So even though it's a 1st order filter with the natural rolloff at about 3 khz this creates a psuedo -12 db/octave rolloff which mates well with a 3khz -12db/octave standard crossover for the 8" woofer.
But another plus one is the 1st order crossover is the nearly perfect impulse response of 1st order crossovers. Less components in crossovers always means less stored energy in the system leading to better impulse response. There's so much mystery and legend regarding piezo tweeters but there is solid reasoning why they gained popularity, good enough sound and minimized crossover complexity, albeit to the point of everyone losing the original knowledge and just poo-poo'ing piezos blanket. When properly implemented they really do have a special airiness, immediacy and presence that's hard to replicate, especially when you consider the price and simplicity of simple 1st order crossovers.
Wow thank you for watching the video, commenting and all the information!
So... I've got some questions:
First, based on your experience with piezos, I'm wondering if I've got the resistor in the correct location. The piezos I tried have an 8 ohm resistor I wired in parallel, (across the positive and negative terminals). I read that placing the resistor in parallel, can help the piezo behave behind a crossover, more like a conventional transducer. It's actually similar to a Zobel minus a capacitor and it apparently serves a little different purpose. Still for impedance equalization, albeit to combat two different electrical Phenomenon. However, there's other information that is common and recommends the use of a resistor in series, except fairly large values, like 20 to 30 Ohms. I'm not sure which is correct, or if the recommendations are for two entirely separate circuits, e.g. resistor in parallel for use with crossover, and resistor in series when not using a crossover possibly?
My next question, is that I've got that same pair of piezos still kicking around and have them planned for an 8” or 6-1/2" 2 Way, that I'm planning on using either no crossover, or a vintage first order RadioShack adjustable crossover. Which would you recommend and what way would I use the resistors? I'm assuming that the piezos I have with the 8 ohm parallel resistors could be used behind the first order crossover as you mentioned using a cap in series... The crossover I'm planning on using has I believe 4 x-over points to choose from.
I've also got some CTS Piezos that arrived from parts Express this past week, slated for use in more upcoming projects. I'm wondering what you think about those. They're very similar in appearance to the ones in this video.
By the way, the problem I had with the 2nd order 3500 Hz crossover and the piezo with the 8 ohm resistor in parallel was that it was very quiet. It sounded decent, and I could hear it working, but it wasn't sufficient to keep up with the mid bass driver. Plus there was very poor high frequency extension. Was that because it was behind a 2nd order network? It seems odd that 2nd order or 1st order would matter, because the higher order is simply limiting the lower frequencies at a steeper rate.
I've built speakers with piezos way way back when and I remember them sounding nice... Very airy yet super quick.
As a PS, the pair of piezos that I whacked off the piezo and glued on a conventional dome tweeter actually sound very nice and have a surprisingly decent frequency response, which tells me even though the speaker was cheap, somebody did a decent job on the horn.
Anyway I'll stop blabbing... looking forward to your response.
You sir have the knowledge! Ive been experimenting with piezos for many years myself and I agree with everything you've mentioned. My final version uses a large rectangular horn 13" wide and 5.5" tall. Have a 8 ohm resistior in parallel and a 3.3uf Sonic cap in series. The sound very good! ( finally) This is paired with a vintage 10" alnico full range in open baffle with a small .4mH air core inductor to roll off their top end. Bass is handled by 15" open baffle subs. Lots of people can't believe the big horns are piezos.
their you go.that did it my friend..
After making some speakers incorporating piezo tweeters, I went back to coil tweeters. Like chalk&cheese. So much smoother!
Piezo tiny discs are stiff and brittle, prone to resonance. I stuck a bit of Blue Tack on the back of the piezo disc. It helped a lot to mitigate the resonance. I also made inductors from 20 winds of thin speaker wire around a pencil ✏. That helped a lot to mitigate the "scratchy" upper treble sound. I found that no cap sounded the best for me.
Pieto tweeter is a capacitive load. It is necessary to put 10-15 ohms in line with it.
Then a standard crossover can be applied, which is also loaded with a resistor of nominal size (4 - 8 ohms). Greetings from Serbia.
Siberia? Wow that's a ways away! Thanks for watching and the comment. My understanding of Piezo tweeters, (which is limited), is that a resistive load in series makes little to no difference and can adversely affect the high frequency output of the Piezo, because unlike a coil in a conventional transducer which has it's highest impedance at its lower frequencies, the Piezo's impedance rises potentially to 100 ohms or more at the highest frequencies. Some argue that the 20+ ohm series resistor is unnecessary and isn't much of a change when there's that much impedance and one adds even more. The thought process behind the 8 ohm resistor in parallel, is to correct the nominal impedance as the crossover sees it, so while the frequency increases, it will present only an 8 ohm resistance to the crossover and/or amplifier, which obviously is better than 100+ ohms.
Disclaimer, I'm not a Piezo expert, nor have I measured the varying resistance of one. I'm simply regurgitating the most reliable information I could find online.
@@wattspeakers It's not Siberia. Serbia, part of former Yugoslavia.
@@wattspeakers Piezo tweeters have a capacitive character of several nF. That capacitance can cause the amplifier to oscillate. That's why you should put 5-30 ohms regularly. Then the crossover is loaded with several hundred to those 5-30 ohms, as the frequency changes from 1000-20 000+Hz. The crossover requires, say, 8 ohms, so it should be loaded in parallel with those 8 ohms. after that goes 5-30 ohms and piezo.
@@wattspeakers It is not Siberia, Serbia, the former republic of Yugoslavia, the Balkans, Europe. More precisely, Belgrade was bombed in 1999.
I would put your ohm meter on the speaker inputs on the back (This may help look for problems with the crossover)and check the DC resistance. going into the speakers and also check the DC resistance on each driver. should be around 5.8 Ohms or so for the woofers Not sure about the tweeters but be good to see what they show. If you purchase the Dayton Audio Dats V3 it can test your speakers and show an impedance graph on your computer screen to show where the impedance may dip to cause problems with your amplifier, I have one of those too.
dang buddy they sound pretty awsome i like them..im impressed you outta get into speaker building and alot has to do with what reciever your using.crank that baby up...make me a pair🙂🙂🙂🙂
Very interested to see the mod with the new tweeter in those horns.
I like that look too.
Thanks. I think it looks much more macho
they sound clean..
i cant help commenting they do sound tight i like crisp clean and bassy sound...
👍 these are a unique design by accident. The pro mid-bass driver isn't top shelf but gets the job done and was perfectly suited to the box size. If you noticed, the port tuning was very close to perfect, bringing these down to the mid 50 Hz region before rolling off. That 40 Hz to 200 Hz range is where much of the snappy dynamics come into bloom. The efficiency is nuts as well, so these get loud very fast and positively pound without any strain. I got lucky mating a 93 Db tweeter to a horn, bringing the top end into the mix without over emphasizing it, as is the case with most PA speakers. In other words, through spare parts, cheap old speakers, some cheesy off the shelf crossovers and a little ingenuity I managed to make a hybrid high-fidelity home PA speaker, with a relatively balanced amplitude across most of the average person's hearing range.
Crisp and clean is a great way to describe them, thank you. The PA driver's relatively huge motor, combined with a stiff paper cone really gets the job done nicely. The aperiodic vent creates a little extra low bass, enough to run them without a subwoofer, or combined with a subwoofer to fill in that 100 Hz range creating a larger sound stage.
Thanks again for watching and commenting!
you put your load resistor in parallel to the tweeter, when I believe you wanted to put in series. Even after you removed the resistors from the circuit depending on the resistance of the tweeters and how the matching load in the crossover works those tweeters may be causing too low impedance for the outputs on your amp? Try wiring up the resistors in series on the positive terminal of the tweeters and check the impedance across the tweeter(should be >the load resistors value). Hope this helps?
cheers!
ron
Thanks for the comment Ron.. I meant to put the resistor in parallel with the Piezo, from what I've read, it presents an 8 ohm load for the crossover / amp, because the Piezo impedance is funky. A resistor or L Pad in series is apparently used for another purpose, although the purpose escapes me at the moment.... Maybe something to do with eliminating sky high impedance at high frequencies from newer amps and receivers? Anyway... The parallel resistor is needed with a crossover, so the X-over works as intended, while presenting a more stable load to vintage amps. Without a crossover, the Piezo can be used with nothing at all, or a small value capacitor in series that does something for audio quality.
I built a pair of coaxial speakers with a piezo disc inside the voice coil tube of a 4" full range driver (along with added woofer). So I didn't need the 8 ohm resistor. The FR driver was the resistor. 😅 (No cut off from the amp even when cranked).
It's a good idea to also put at least 2 ohm resistor in series with the piezo tweeter cos (apparently) they can go down to nearly zero ohms @ 20 khz. Although for lower frequencies the piezo tweeters can be like 1,000 ohms. Maybe your amp triggers cos of that. Your amp may be sensitive to the highest frequencies impedance. I put 4 ohms in series with my piezos. I got the information from a piezo nerd speaker builder on RUclips.
What may be happening is the amplifier is going into OSCILLATION due to the capacitive loading. Amps don't like caps across the outputs. I got a Marantz 250 that destroyed my Magnepans. I've seen in articles that you also need to put a series resistance in the Piezo circuit just for that reason. But thats not gonna help with the level issue.
Hi Richard thanks. Actually what had happened was I had mounted the crossover but one or two of the sharp solder points on the backside, had penetrated the butyl and we're making contact with the foil on the sound deadener I used. Mounted them better so they can't touch the foil and it worked fine. I did end up swapping the Piezo's for conventional tweeters however because they didn't have a very good upper treble.
I ran some test... Your crossover really does need to see a load for the tweeter otherwise it sends most of the signal straight to ground through the inductor on that portion of the second order crossover. this would cause your impedance to drop pretty low and potentially causing the amplifier to go into overload and shut down. You could try a 10 or 12 ohm resister... Or unsoldering one end of the inductor on the tweeter portion of the crossover. If you were using a first order crossover then this would not be a problem if you had no resistor in parallel with the tweeter. Second order or higher crossovers must have a load on all outputs. Since your tweeter looses to much output when using a resistor I would snip that inductor and then the crossover would be like a first order for the tweeter. and load would not matter to the amp.
That's good info and makes sense. If I try some other project with Piezo it might be worthwhile... Or probably cheaper to buy or make a low pass, then just add the capacitor in series with the Piezo. I think that's why Piezo builds 30 years ago worked no matter what because 1st order crossovers were common.
I had to go back and watch the video again looks like your amp kicked out Before and After you removed that resister. The Dats V3 would really help in troubleshooting. Since your crossover is the new untested part. You could start with that. (after resting Re on all drivers.) and use your meter to verify 8 ohms on those resistors too.
A piezo nerd told me that piezos can go to nearly zero ohms @ 20 khz. At lower frequencies they go to like 1,000 ohms. The low resistance @ the highest frequencies could maybe trigger some amps.
Those 8" PA type are really shouty and have a serious medrange level! I deal with them cos I build coaxial drivers out of them since some have a 2" voice coil. I use a 5.6 uh bass inductor and a 1.8 uh med inductor. And a 10 mf cap resisted. I have to pull the speakers apart to tell you the resistance (I forgot). And in a big box for bass intensity. It's a type 1 crossover and you could use a smaller med inductor with a type 2 circuit.
Thanks for the comment! Budget build $100 per speaker. The Dayton MB-820 actually sounds less PA style than one would imagine. It's actually pretty musical and blends well with horn loaded tweeter and a subwoofer. I agree that some PA speakers a.k.a. pro drivers, can be "shouty" or just plain annoying. This one I can easily chill to all day, or crank it up and it seems to be very natural, neutral and quick.
send me the specs you used..ill push them with my TUNER! BUT REALLY MY FRIEND THEY SOUND AWSOME...
Piezoelectric need a cap 3.3uf and 10 ohm to give amp stablity,ie a load the network you hold is for magnet / voice coil 8 ohm drivers.
Thanks Larry! So you've had success with a simple 1st order cap in series and 10 ohm resistor in parallel?
What ohms are they running at?
Hi Daniel thanks for the question! It's built to be an 8 ohm nominal.
It's hard to hear you
Hi Glenn, I'm sorry, yes. Trying to work that out, shopping for microphones which are what most RUclipsrs utilize, so they're much louder during playback. I appreciate the constructive criticism and I agree it's hard to hear me.
Those tweeters are a waste of money, even if they only cost 6.99, a pair, they ALL sound low rent.
I love that analogy, "low rent" seems like a really good hook for a video title. I'll be doing a smaller low buck RestoMod project with Piezo's so your comment is very valuable thank you!
Hi as an electronics engineer, I think after research and testing. Putting your 8 ohm resistant across the piezoelectric unit will only. Serve to give the useful high frequency another path to ground, very wasteful!! I would strongly suggest you wire it in series ". So it's still loaded BUT" no energy goes to ground ". The none reactive load of a piezo , is much better direct to the amp anyway". Cheers
Thank you for watching and the suggestion! As someone who recognizes the value in what engineers have to say, I will take that advice and try it. I've got more piezo tweeters that I'll be using in upcoming projects. They're two ways and the plan was to not use a crossover at all with them. The ones from this video are still hooked up with the resistor in parallel, so the plan was to see what they do without a crossover the way they are now, and if they still sound crummy, I'll try resistors in series. Thanks again for watching and commenting!
@@wattspeakers hi. Cheers for reply , iv always professionally never used anything. Straight to the amp. This way there are no losses and in 10 k setups with multiple piezo tweeters never had a single failure.