Circuitry to limit voltage going to driver.

probably best to try something before the signal is amplified.

there are several amps that have such variable limiters built in.

several equalizers allow for the same type of limiting.

A lot of plate amps already have this as part of their limiter circuitry. Unfortunately for the DIY crowd it's not exposed as an end user tweak or adjustment. It's just like the high pass / boost filter in the amp. It's all a big secret. However, If you're an OEM and want to use the amp you can get the data / instructions on how to modify the limiter, boost, and high pass "subsonic" filter on the amp.

The most tweakable conventional plate amps out there are the two BASH amps PE sells and they only have a documented adjustable low pass / boost filter.

Adjusting the voltage limiter should just be a matter of changing the resistor divider that feeds the output voltage back into a comparator circuit in the amp.

Eminence has such a product in the works, the D-Fend. The concept seems simple, the realization is anything but.

Then buy a commercial product like the Submersive where someone spent the time to dial in the amp to make the final product virtually foolproof.

If you want that DIY we're all going to have to get out torches and pitchforks and demand the plate amp makers give us the information they're currently only giving their OEM customers.

as far as my understanding goes,
voltage limit = excursion control.
current limit = coil power handling control.

both of which are available to the diy community in the form of limiter function in dsp processing like in minidsp and behringer inuke limiter. there's also a clipping detection and limiter circuit available but that doesn't do frequency specific limiting.
current limiting function are widely available in a lot of diy amplifier design.

So, you just need some sort of compression circuitry? Look at many different PA options. For example
http://www.behringer.com/EN/Products/MDX2600.aspx
http://www.americanmusical.com/Item--i-BEH-DEQ2496-LIST?src=Y0802G00SRCHCAPN&gclid=CLn_6_j21LICFQTOnAod4nMAWg

You can set how the compression works. For example, you may have it set so when you crank it to 7 and the output is 7, crank it to 8 and the output is 7.75, 9 is 8.5, and 12 is 10.
*edit* let me try this again.
To my understanding, the current is set, regardless of voltage, since it's a function of the speaker's impedance. Voltage remains variable and is what increases or decreases since amplifiers are voltage gain devices. So that means current limit doesn't exist unless you're able to dynamically change the speaker's resistance. Ohm's law.

Definitely not.

Looney's suggestions are best. Set it up, place it out of sight and forget about it.

Outboard compressor and limiters are an crude solution. A driver might have enough excursion to take 2kw at 60Hz, but not at 20Hz. A PA style compressor / limiter doesn't know that so you're just going to tamp down output across the whole frequency range to to the "safe" power level at 20Hz unless they support progressive limiting curves that are frequency dependent.

When properly setup a good plate amp can do a better job keeping the driver "safe" without reducing output across the entire frequency range. Unfortunately, it seems the half of the DIY crowd using plate amps has decided they don't want to use the limiter(s) already on the plate amps (that OEMs do) so no plate amp makers are releasing the information on the limiter circuit(s) to DIYers. The other half of the DIY crowd is enamored with mega power pro amps which don't offer any limiting features, but are relative cheap for the power output.

You're giving sellers of DIY plate amps too much credit. Limiting and compression features are readily available in some products like Digmoda's amps and a few others with DSP front end. The BASH & HPSA amps from PE both have their own limiting which makes for a softer clipping, but no supported user adjustment, as that's beyond what PE even knows about the amps. The reality is that unless you are purchasing from the company who has the engineer who designed the amp, which no DIYer will be for any amp sold at DIY distributors like Parts Express, at best you will get a schematic and you are left to sort it out on your own. Getting inside and tearing it apart and re-soldering components brings in all sorts of warranty issues, and a very small percentage of users are comfortable changing the resistors for a high pass filter.

Simple Voltage limiting on the amplifier output (after any front end EQ) often is all that is needed and done on many products. Sometimes a high pass placed properly will make it that much more bulletproof. For anyone interested, a ~$300 DCX2496 affords you quite a few such options to play with both limiting and level dependent EQ.

There isn't. In theory something using zener diodes could work, but none exist capable of handling the power required. As a hard limiter can be had for as low as $100 no one's going to develop 100 watt plus zeners for the purpose, especially as a limiter is easily adjustable, whereas a passive component based solution would not be.

Of course they are, but show me anything else that a plug and play solution - it's better for most users to have something in situ like this than to damage drivers. I can design or modify and build anything I need to do the required job, but as Mark mentions, few people on these boards are even capable or comfortable changing a couple of passives on a PCB so suggesting a more technically challenging approach will fall on deaf ears.

You are correct in one regard re plate amps offering better protection than large pro amps, but mainly because few will offer enough power to do damage in the first place.

There are very high power plate amps from SpeakerPower.net, miniDSP, and Digmoda. They're just not cheap.

True, but they are considerably less common especially around here and the most common plate amps are generally lower power (say <500W). Most look for value and pro amps with a MD for processing works out cheaper almost all the time especially with more than one sub.

The problem is that excursion limited output is frequency dependent.

You need to run the signal through a filter with output proportional to excursion and use that to trigger a compressor. For sealed/IB/OB implementations it's a low-pass; although for ported enclosures you'd probably add a notch too.

I really liked Eric Weitzman's implementation for his Orions where the dipole 18dB/octave excursion increase for a constant SPL output really calls for such a solution.

His web page documents the wedge micrometer measurements he used to arrive at the required transfer function.

http://oriongateway.org/woofer-protection-limiter/

That's not a problem, you limit to the voltage that doesn't exceed xmax anywhere within the speaker passband. You need to know what that voltage is of course, but modeling software will tell you that.

Some of us don't want to either

1) Give up 12dB of dynamic range in the second octave (which we use for life-like playback of recordings with decent dynamic range) to protect our drivers from last octave material that shouldn't be there in a musical power spectrum or film sound track screen channel but is.

or

2) Loose our last octave of extension which still has some content

or

3) Quadruple our bass driver count ($2700 for me if my wife didn't divorce me for the space increase and a hundred times that if she did)

or

4) Stuff our drivers in boxes small enough that the amplifier can't bottom the driver, with equalization providing whatever bass extension is desired ( This precludes dipole bass and may introduce thermal compression issues).

A low-pass filter triggered compressor sucks less than all those options and is surprisingly inoffensive when it does trigger.
Edited by Drew Eckhardt - 9/28/12 at 9:40am

That's a good well thought out and written article.

Because my new drivers are not replaceable (FTW21), I have been thinking about something similar as a 'just in case', using switchable HPFs to ensure that I cannot overdrive the cones beyond Xmax with excessive LF content and the power I have available.

I have been playing around with the Dfend circuit for about a year. It works very well-but is not in production yet. It is amazing how hard I can drive a loudspeaker with all sorts of nasty material-and it still be just fine.

The hardest thing is doing a good bit of loudspeaker measurements that allow you to determine the real parameters of the loudspeaker.

A loudspeaker is NOT a simple circuit-so a simple 1 number solution is not the answer.

Such as-what about a sine wave at 200Hz? How much can it take? What about a sine wave at 8Khz? It cannot take anywhere near as much. So a different type of limiter is needed.

What about over excursion? A voltage that is fine at 200hz may produce overexcurion at 30Hz, even if the power limit is not reached. So a different tyep of limiter is needed there.

The DFend does all of this-BUT ONLY if you know the exact limitation of your loudspeaker. You will not see these on a spec sheet.

Hence the reason to measure them.. ANd it is not exactly an easy task to simply "measure it".

Any circuit is ONLY as good as the data that is put into it.

Another issue is the D-Fend instructions, which are barely comprehensible even if you know what you're doing, impossible if you don't. I hope they fix them when they release the product.

Whilst interesting in an abstract way, the current DFend discussion is relevant to a non technical, non OEM DIYer, how exactly?

I have seen any instructions. What I started out doing was to measure how the various parameters "reacted" to various settings. Then I started measuring the loudspeaker in ways to work with those settings. Then appy-retest and adjust as needed.

It is NOT a simple process-or a 1 size fits all type of approach.

But when you get it dialed in-it can be darn near blowup proof-no matter what you do.

For example My first test was with one of our 300 watt rated loudspeakers that goes down to around 80Hz. I used an amp that could produce 4400 watts into the rated impedance of the loudspeaker. I can do all sorts of VERY abusive stuff (or would be).

Such as running regular music so that the amp is into FULL clip-ie the clip lights never go OFF-they are on constantly. Then apply HF sinewaves (like say 8Khz) that can drive the amp into full clip. And then let's say I want a bit more bass-so I add a 16dB boost at 40Hz. No problem.
Yes it sounds pretty bad. But as soon as the levels are dropped down to normal limits it sounds like there is nothing there (protection wise).

The LED indicators can be a bit deceving-and it takes some getting use to what they are REALLY telling you. I rely on my various measurement tools to give a good idea of what is going on-or what is really going to the loudspeaker.

I still stand by the fact that there is no way a typical user would be able to make the proper adjustments-at least to get close to pulling the maximum (before damage) from a typical loudspeaker.

But it could be used for some "general" settings that would offer a certain level of protection. But that ASSUMES that the specs they get for their loudspeaker are actually CORRECT-which is another can of worms. And the specs you generally get are very limited in terms of maximum input capabilities. Hence the reason to MEASURE the product itself to be sure. But that require the proper tools-which are not exactly cheap to the typical user.

I am a firm believer in the product-(they are still doing some upgrades-as far as control flexibility) and can't wait to start using it all the time.