Limiters - Finally understood

[escapemcp]

But in my sized rig, I like being able to make quick changes to the whole system from the driverack without having to touch each amp....and since my amps are on the bottom of the rack - I don't like standing on my head to do it.

Fair enough... each to his own.

And I like cheaper amps that I can keep adding to my system without having to buy redundant dsp each time.

That is a good point. I bought the DSP amps to get me up and running on the cheap (to avoid having to buy amp AND a driverack or Berry DCX). As such my system has ended up with DSP amps instead of a 2 box arrangement such as yours. I will be buying a DCX in the next week, but TBH, all my processing will still be done on the amps. The main reason why I am getting the DCX is so that I can split my signal to the amps (there is NO passthrough on the iNukes ). The DCX will be doing very little - it will produce a couple of mono channels (for a monitor) from a stereo input, but that's about it - all the other channels will just be passed straight through.

It will be nice to have the redundancy in case an amp does pop - for the first time since I built the rig, I would be able to borrow an amp if the need arises and get it to work as it should... one of the negatives of having a DSP amp is that if it does die, you lose BOTH the amp and the crossover/limiting functions - not good. Having a DCX in the rack is my insurance policy!

It's a choice everyone makes and I think there are some advantages each way - it's just important that everyone understand the differences and not think it's doing one thing when it's doing something else.

Knowledge makes us all better at what we do.

Yup, agreed.

[byacey]

byacey,
would you explain this a bit more please (i'm not sure if I'm understanding).

After all if you use a non DSP amp with an outboard limiter, then the limiter is most certainly before the amp attenuator.

So why would it not make sense to build a DSP amp the same way?

The way any simple limiter works is by some form of a voltage controlled attenuator (VCAT) or a (VCA) Voltage controlled amplifier that adjusts the signal passing through. In order to know when to start limiting, it has to sample the signal level at some point in the signal chain, and compare it to a chosen threshold reference. If the sampled signal is below the threshold reference, the signal is allowed to pass freely without any attenuation. However, if the signal level meets or exceeds the threshold reference, it triggers the VCAT or VCA to lower the level of the signal until the over threshold condition is removed.


With the driverack line level type setup, there isn't any feedback (voltage sensing) to tell the limiter when the threshold has been reached except at the line level output of the driverack; For this to work we need an operator with a voltmeter measuring the amplifier output and making the proper threshold adjustments in the feedback loop, and this is only a valid setting for that particular amplifier. It cannot compensate for any gain changes downstream from the driverack, -ie. someone adjusting the amplifier level controls.

However, in the old system I described, the signal level sample is taken from the speaker terminals, so in effect, the whole amplifier is included in the control feedback loop to provide the threshold sensing signal. If it's designed to limit at 30 volts, when the amplifier output produces 30 volts it will limit the signal feeding the power amplifier input. If the case of the amplifier level control being turned down, it will simply allow the line level signal to be increased until the amp output increases and reaches the 30V threshold, and then the limiter again clamps down on the line level signal feeding the amp.

In the case of onboard DSP, just stuff this circuit into the amplifier chassis, have the VCA control the signal right after the line level input jack, and sense the speaker terminals for threshold voltage, and you have a limiter that will always limit at the same output voltage regardless of how the input level control is set.

For further reading: http://www.rane.com/note127.html

[Grant Bunter]

awesome byacey,
thanks

[escapemcp]

For further reading: http://www.rane.com/note127.html

Had a look at that Rane site and came across http://www.rane.com/pronames.html - a quite interesting list of audio companies and how they got their names. Thought for a brief moment that FSR was one of Yoda'a creations (check it and you'll see what I mean!)

[byacey]

There's a lot more to it than what I tried to explain, but in simple terms it's an attenuator controlled by voltage feedback at some point in the signal chain.

What I was trying to describe in my previous, long winded explanation was if the voltage feedback sampling is taken from the speaker terminals, the limit point is unaffected by anything within the signal path.

[Bruce Weldy]

if the voltage feedback sampling is taken from the speaker terminals, the limit point is unaffected by anything within the signal path.

If they truly work like that - that would be great. I've just got this funny feeling that they shove a mini-dsp in front of the input knobs and call it a day.

Because if they did use the speaker output to control voltage, then why not just let you dial in a voltage instead of doing it by db?

Some high-dollar, high-end amps may very well do it this way.....but I'm doubting that the stuff we are talking about here are that sophisticated.

Short of tearing one apart (and knowing what the hell you're looking at), I doubt we will solve the "how does it work" question.

I don't have the chops for that..........but byacey?

[byacey]

It doesn't matter if the limiting is done within the digital domain or analog, they still work the same and both require some sort of voltage feedback from some point. If the DSP is integral with the amplifier, it would make the most sense to measure the speaker output for the limiter feedback sampling. If I was designing it, I would have the VCAT after the input attenuator, and have the limiter feedback taken from the speaker terminals.

Most audio engineer type people have (or certainly should have) a better understanding of db increments compared to volts; most likely this is why compression ratio and threshold is expressed in db.

If anyone has a full schematic of a dsp type power amp, I wouldn't mind looking at it.

[escapemcp]

Because if they did use the speaker output to control voltage, then why not just let you dial in a voltage instead of doing it by db?

I can...

It IS Voltage peak, so you have to divide by root 2 for the 'real' BFM voltage (so it indicates 70.7V for a 50V limit). If you fill in the impedance (marked 'load'), you can limit by power also

Am I winning you over yet Bruce??

[Bruce Weldy]

Am I winning you over yet Bruce??

Nah......I still like having the driverack. But, I can certainly understand why some would want it for their rig.

But, why do you have to do math to come up with the right voltage? I'd still put a meter on the outputs....you know what Reagan said....."Trust, but verify."

[escapemcp]

But, why do you have to do math to come up with the right voltage? I'd still put a meter on the outputs....you know what Reagan said....."Trust, but verify."

You have to do the maths (yes, it has a s! ) because we all know that Behringer only know of PEAK voltage/wattage... it's the same reason why the iNuke 3000 only puts out 2000W!

I'm with you on the double checking, as I wasn't initially prepared to trust my shiny new Lab12's to Behringer's supposed voltage... to be fair though, the value is pretty bang on

[escapemcp]

Am I winning you over yet Bruce??

Nah......I still like having the driverack. But, I can certainly understand why some would want it for their rig.

Fair enough... TBH if I had enough money when I bought my electronics, I would probably have had one also. Getting the iNuke meant that I had crossover and limiting sorted, but I then needed a DEQ for RTA purposes. Now it seems like I need some sort of signal splitter as I haven't got any through lines on my amps... so that's 1 DCX coming up (especially as they are on offer at the moment - £135 delivered )... so yeah, false economy - or more like it is a false economy buying a DCX to split the signal - I could just use Y cables to feed each amp with a full range signal, but where's the fun (and flashy lights ) in that! The flashy lights are actually to make my rack look more impressive, for those who hear with their eyes and not with their ears - there's a lot of them about you know!

I wasn't planning on having quite as big a rig as it has now ended up, hence me not seeing this problem earlier! So now I will end up with a DCX/DEQ combo - if I had got the driverack I would have saved space in my rack and I may have saved money (DCX/DEQ combo was £250, and of course the DSP amps are an extra £60 over the normal ones). At least I now have plenty of options for crossing over and limiting - looking on the bright side, the redundancy I have in that area is good

[Bruce Weldy]

the redundancy I have in that area is good

I would love to have a limiter in the amp, just to have a second backstop for safety.

But, I really don't need all the other stuff.....so, I'll just stick with what I've got. It's been nice to actually have the rig where I want it and not having to keep spending money on it.

[byacey]

But, I really don't need all the other stuff.....so, I'll just stick with what I've got. It's been nice to actually have the rig where I want it and not having to keep spending money on it.

That's a good feeling, knowing you have what's needed rather than trying to cobble something together in a make-do situation.

Escapemcp, How many inputs do you need to drive? You can drive quite a few amps before you start loading down the line. It's only when driving large multiples that I've seen distribution amps utilized.

[escapemcp]

Escapemcp, How many inputs do you need to drive? You can drive quite a few amps before you start loading down the line. It's only when driving large multiples that I've seen distribution amps utilized.

I could split the output... At most i'd have to drive 2 channels, so easily do-able with just a simple y-cable. Reasons for going with the dcx though:

I can easily create a mono signal from the dcx to drive subs and monitors... At the moment, the subs are being driven from the left channel only... I KNOW this is a bit of a moot point, as bass is usually (always?) panned to the centre anyway. For the monitor it will be handy though, although I am buying an active monitor to get me through the next few gigs. The active monitor has a left and right input that sums, but once I upgrade this to an Otop however, I will need the mono sum function of the dcx.

I also get extra outputs to run my dj facade from... The sound controlled LED controller needs a line in to function properly (if using the mic you only get about half the modes). This is yet another 'dead-end' for the audio in my system.

As I mentioned briefly above, I know that some of the people who run local club nights will be impressed by all those flashy lights!!... As such it may well net me an opening into said club nights... Stupid I know, but I'm playing to the audience on this one!!!

Future-proofing - I don't know what the future holds for my rig, but with a dcx in my rack, I should be covered for most eventualities

And as I can get them cheap at the moment, I'm taking that as a sign to get one!

[byacey]

I can easily create a mono signal from the dcx to drive subs and monitors... At the moment, the subs are being driven from the left channel only... I KNOW this is a bit of a moot point, as bass is usually (always?) panned to the centre anyway.

You lose 3db by driving the sub from one channel. Not a big deal, as you can make up the loss easily. For my purposes, I don't run stereo FOH mains, so it's not a concern.