Op-amp spec to maximize speaker power (RMS?)

wayne42

Member
2020-01-04 8:34 am
This is my first audio project so I don't know very much ...

My goal is to power a 30W, 8 Ohm horn speaker so as to generate maximum loudness. Sound quality is not important.
I will be playing short 1-2 second audio clips about once per minute for 3 hours, so most of the time the speaker will be silent.
Ideally, the volume would be greater than 120dB @ 1m, but if not, the louder the better.

The audio input will be coming from a DFPlayer MP3 arduino module. The power source is a single 12V battery,
so I will need a voltage booster as well as an op-amp. I'm trying to work backwards from the speaker to determine
the minimum specifications for the op-amp (and voltage booster) in order to drive the speaker at its maximum power.

The first confusion I have is how the speaker power is specified. I've heard that they are commonly measured based
on RMS rather than peak to peak signal? I initially computed an (ideal) input voltage of sqrt(8 x 30) = 15.5V @ 1.93 amps.
But if that is based on RMS, then I need 2.828 x 15.5 = 43.8V peak to peak?

Next confusion relates to single vs dual voltage supply. For example TDA7292 has a maximum supply voltage of +/-33V,
so with dual voltage supply of +/-33V that would easily give me 43.8V peak to peak, but with a single voltage supply
(from single battery) it would only give 33V peak to peak?

Ideally I would like to purchase a prefabricated op-amp board (with heat sink?) and voltage booster board
that I could simply wire together as my soldering skills are not great. Are class D op-amps such as TPA3116D2 better
for my needs because they require only a single voltage source?

I'm happy to pay up to $30 each for the op-amp and voltage booster.

I'm looking to both increase my technical understanding as well as for specific board suggestions.

If I did purchase a variable voltage booster, how should I determine how high I can increase the voltage without
damaging the speaker or other components? I have a multimeter, but not an oscilloscope.
If I'm only playing for 1 or 2 seconds at a time, can the speaker handle peak power above 30W and will I still need a heat sink?

Thanks!

https://au.element14.com/yuasa/y12-12/battery-lead-acid-12v-12ah/dp/2083827
C2022 - Redback 30W 8 Ohm Plastic PA Horn Speaker - Altronics
 

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
30 watts rms into 8 ohm requires a voltage swing of...

W= voltage squared/R therefore V = squareroot of watts multiplied by R which is 15.5 volts rms.

The peak voltage is 15.5 multiplied by root 2 which is 22 volts. The peak to peak is 44 volts.

A chip with a -/+33 volt spec will operate on a 66 volt single rail. There is no loss of output doing this but the design then needs to be AC coupled and biased to suit.

A speaker will handle more power than its rating if the duty cycle is lower but it is uncharted territory. A bass speaker would certainly be limited by such things as mechanical cone excursion and so on.

Class D is more efficient and almost always runs in a bridged mode thus increasing voltage swing (and so power into the load) available from the single rail supply.

A very common mistake we see all the time is running chips at their maximum ratings... simply answer these are absolute maximum values under 100% ideal test conditions and it is much safer using lower voltages (say no higher than -/+28 for a -/+33 volt part.

Ask yourself if you really need that power into a horn speaker. This will give you a good idea:

A Test. How much Voltage (power) do your speakers need?
 

wayne42

Member
2020-01-04 8:34 am
What is it for?

To start athletics races outdoors, with a possibly 50m distance from start speaker to the furthest athlete, with background crowd noise.

The start gun that we currently use generates greater than 120dB at 1m.
The new horn speaker that we are planning to use is rated SPL 1W @ 1m: 106dB so at 30W should be close to 120dB @ 1m.

I don't care about operating the op-amp at is maximum, I just want the speaker running at it's maximum power. Ideally the op-amp will have the headroom, which was why I was asking about gradually increasing the supply voltage until the speaker was at it's maximum power.
 
Opamps are not power amps. I think you are meaning a chip-amp?


Definitely consider something like a cheap and cheerful class-D board, for instance something like a TPA3118 or TDA7498 board which will have lots of power to spare and run cool and efficient (most of the time the thing's silent anyhow). Most class D chips are bridged, so that 24V single supply can give 30+W into 8 ohms.
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
www.diyaudio.com
What is it for?

To start athletics races outdoors, with a possibly 50m distance from start speaker to the furthest athlete, with background crowd noise.

The start gun that we currently use generates greater than 120dB at 1m.
The new horn speaker that we are planning to use is rated SPL 1W @ 1m: 106dB so at 30W should be close to 120dB @ 1m.

Short answer: forget it.

maybe you "measured" 120dB with some cheap general purpose meter; or even worse, with some kind of "App".

Wrong measurement, maybe because it was too short to be registered, buy way worse because it certainly surpassed microphone and preamp range.

truth is you are ORDERS OF MAGNITUDE WRONG.
From those who know:

[Peak sound pressure levels of gunshots from starter's pistols]. - PubMed - NCBI

I cut and paste:

Starter's pistols are often bought for self-defense, but can also be used for criminal activities (e.g. assaults, etc.). When a starter's pistol is loaded with blank cartridges and is fired, a powerful shooting noise results. The level of the noise produced is high enough to cause acoustic trauma. For legal examinations and giving an expert opinion further information is needed about the power of such noise. We examined how high peak sound pressure levels were of the gunshots of blank cartridges and whether there existed any directional characteristics from the noise emissions. In all, 15 different models of starter's pistols of 8 different calibres were examined. In addition to blank cartridges, 8 mm tear gas cartridges were also examined. Four transducers were situated in the horizontal plane around the muzzle: 0 degree (shooting direction), 45 degrees, 90 degrees, and 180 degrees (towards the firer). The distances between the transducers and the muzzle were 25 cm, 50 cm, 100 cm, and 200 cm. At a distance of 1 m and in the 0 degree shooting direction the peak sound pressure levels of nearly all weapons tested exceeded 160 dB. At a shooting distance of 25 cm the peak sound pressure levels reached 181 dB. In addition, we observed a directional characteristic concerning the emission of noise: pistols produced higher peak sound pressure levels to the front than backwards towards the firer.

Forget about reaching such levels with any electromagnetic transducer driven by any sensible amplifier.
 
Last edited:

wayne42

Member
2020-01-04 8:34 am
JMFahey: I think you are referring to the start "pistols" used in north america - which are more like real guns, but with black cartridges. In Australia we use

Double Hammer Cap Starting Gun | Starting Devices | HART Sport

which has paper caps which a similar to children's toy cap guns. Yes, we may not be able to reproduce even that volume, but my point is that we need the sound to be as loud as possible as the cap guns we currently use will not be able to be used in the future.
 

wayne42

Member
2020-01-04 8:34 am
BTW, I don't want to use off the shelf commercial start systems as I want to be able to save the start times to SD card and RF transmit start to finish line display board and other components used in photo finish system.

Thanks to all that replied with suggestions. Based on your suggestions I'm currently considering the following voltage booster and amp combination:

150W DC-DC Boost Converter 10-32V to 12-35V 6A
Digital Amplifier Audio Board TDA7498 Audio Amp 2.0 Class D Amplifiers Stereo

Obviously more powerful that what I need, but allows speaker to be upgraded in future.

Do you see any problem with this plan?
 

pcan

Member
Paid Member
2015-12-31 4:57 pm
The amplifier you selected does have two BTL channels, so you may connect a second horn speaker to increase the volume. The horn speaker choice is critical. According to my experience, the JMFahey post is correct. I have installed a outdoor PA system on a industrial environement (4 big horn speakers, 100W) and the volume is barely enough for the job. The chime function of the amplifier has come handy, because it announces the message with a tone frequency so people can hear that something is coming. You may eventually add a klaxson horn, to be briefly operated by the arduino board before the audio play.
 
JMFahey: I think you are referring to the start "pistols" used in north america - which are more like real guns, but with black cartridges. In Australia we use

Double Hammer Cap Starting Gun | Starting Devices | HART Sport

which has paper caps which a similar to children's toy cap guns. Yes, we may not be able to reproduce even that volume, but my point is that we need the sound to be as loud as possible as the cap guns we currently use will not be able to be used in the future.
Ok, I see Australia is also hell bent on suppressing "guns", even toy cap type ones :rolleyes:

Are pyrotecnics fully forbidden also?
Any kind of firecracker will be able to provide way more peak SPL than any amplifier+transducer system and you might even devise an electrical system to fire it.
Search "hobby rocket starters" .

Are nail guns also forbidden?
Used without any kind of nails, just the cartridge, (which can be bought at construction site suppliers) can provide a hefty Bang!!, definitely louder than the toy cap version.

The link you provided says they are out of stock and can´t resupply :(

My point being you can not "play a gunshot recording" at realistic levels.

In any case, if you insist, I would suggest you forget the amplifier and playback route entirely, instead experiment with charging a capacitor (say, 1uF for starters) and discharging it into a PA type horn speaker.

Experiment rising voltage (and capacitance) until you destroy voice coil and then back down to half that power (or 70% voltage) for safety.
You will get a very loud CLICK! which may be enough,once you found proper level you build a small dedicated supply.

Wire things so positive voltage pushes diaphragm against phase plug, at those extreme levels it´s better to get positive pressure than negative.

Plan B: forget explosions of any type and modify rules so a foghorn or similar can be used.
Or a large gong or bell, not kidding.
 
Last edited:

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
www.diyaudio.com
> you can not "play a gunshot recording" at realistic levels.

He does not need true gunshot level. He needs an impulse louder than crowd noise which will cover the entire start-line.

The 22 Pistol was a common varmint (rats, possums) tool so (with blanks for safety) became the standard start-signal. Over-kill but cheap and convenient.

Cap-guns make sense to me, for smaller venues and polite crowds.

I think the problem with caps is that a BOX of caps, ignited in a storage-hold, makes a nasty panic. So they are getting harder to get.

The delay on a fused firecracker is a small problem. Worse, because the correction for a mis-start is a second shot, no delay. Yes, an electric squib in a firecracker could give instant bang. But the total market for start-bangs may not support such a thing. And electric squibs have their own hazard. (But have other uses, including gun-shot on film, so maybe there is a suitable product ready to expand its market?)

As we become more gun-adverse, maybe we need a different sound. A pitched "PINgg!" will still startle but is clearly not gun-fire.

He's right that 106dB SPL on axis is a reasonable spec for a loudhailer horn, and 30Watts above that is 120dB SPL. *At 1 meter!!* So 100dB SPL at a runner 31 feet from the starter. And in all but the loudest crowds, 100dB SPL will bite the ear.

The speaker's power rating is probably for an hour, with initial time constant of several seconds. The bulk of the impulse should be less than 3 mS; indeed races are timed to milliSeconds. So first approximation the speaker can be hit with 30,000 Watts without melting. In fact it would rip apart. When Gauss developed their horn driver, rated 30W long-term, they eventually used a 3,000 Watt amp to pulse-test for weakness. Few makers go this far. But 100-300 Watt amp is reasonable for this duty cycle.

IAC, I would start with any handy amplifier and runners. I suspect the exact wave-form sent to the speaker will be more critical than power.
 
Last edited: