Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Tim McCulloch]

Does Digam use line current limiting on 120V?  Can you power this at 240V and see if the amp drops in output?

[Mark Wilkinson]

2. Peak limiter. This is a true limiter and dynamically reduces the gain of the amplifier so as to prevent the (positive or negative) peaks from exceeding some threshold. The peak detection requires a specified attack and release time. The attack time may be set to zero. This is typically used in an attempt to prevent over-excursion. To be effective it needs to take frequency into account or, better, uses an accurate dynamic model of the speaker.

3. RMS voltage limiter. This dynamically reduces the gain of the amplifier to prevent the root-mean-square (RMS) voltage from exceeding some threshold in an attempt to prevent overheating of the voice coil. The RMS voltage serves as a proxy for power which is accurate for a constant and non-reactive load resistance. This may be preferable to a true power limiter in situations where multiple drivers are powered in parallel from a single channel. The averager has adjustable dynamics that are set with regard to the thermal time constant of the voice coil.

Thanks Frank, all 4 cases made sense.

One question....
I've been peering at peak and RMS limiters on a scope, and as far as I can tell they have the exact same effect....to lower gain.
They seem to differ only in threshold, attack, and release....but essentially both rms and peak limiters merely lower gain according to those parameters.
Is this correct? 

[Bradford "BJ" James]

Thanks all. I may have miswrote what I meant to ask in my original post, or perhaps it was interpretted incorrectly....
That particular amp is designed to behave the way I described, right or wrong, that’s what it’s designed to do.
I am simply asking if there is a way to defeat that behaviour so I can get a proper output voltage measurement so that I can properly set the limiters in the dsp.
From what I can remember working with these amps before, it clamps down after about a second of continuous input, so maybe I can just repeat the proceedure a few times to confirm I’m getting a reading that makes sense.
I’ll muddle through. For sure not a big fan of these amps.

[Frank Koenig]

Thanks Frank, all 4 cases made sense.

One question....
I've been peering at peak and RMS limiters on a scope, and as far as I can tell they have the exact same effect....to lower gain.
They seem to differ only in threshold, attack, and release....but essentially both rms and peak limiters merely lower gain according to those parameters.
Is this correct?

In short, no. While they both reduce gain -- with different attack and release times -- the fundamental  difference is that the peak limiter responds to the peaks of the waveform while the RMS limiter responds to the average of the squared instantaneous values -- all of them (and takes the square root of that). For a sine wave these will indeed track with the peak 1.414 times the RMS, but for a complex waveform they will be very different. Consider a signal with a high crest factor, such as a very low duty-cycle stream of pulses. The peak value could be high while the RMS value -- corresponding roughly to heating in the voice coil -- would be low.

Post one drink and off to dinner now. Busy tomorrow but can respond further Sunday or Monday. -F

[Peter Morris]

FWIW most modern switch mode amps will not produce their maximum output continuously.  They will however produce maximum power for long enough to satisfy the demands of most music – the old powersoft Digam amps when driven hard on some music would start to pull back a little.  Generally I consider this a safety feature as it protected my speakers.

It would be easy to build a switch mode so it would sustain full power continuously, but they would be much bigger, heaver and cost more for very little gain.

This is what happens with the newer K10 below - Output Vs Time

[Helge A Bentsen]

FWIW most modern switch mode amps will not produce their maximum output continuously.  They will however produce maximum power for long enough to satisfy the demands of most music – the old powersoft Digam amps when driven hard on some music would start to pull back a little.  Generally I consider this a safety feature as it protected my speakers.

It would be easy to build a switch mode so it would sustain full power continuously, but they would be much bigger, heaver and cost more for very little gain.

This is what happens with the newer K10 - Output Vs Time

Peter: Do your have data on the K20 as well?
K10 is one of the great low/sub amps on the market, and the K20 even more so.
I´m powering four dual 18" off a K20, its very nice.

[John Schalk]

Peter beat me to it.  If we assume that the Diagam 5000 shares the same basic design with the output of the K series then it should provide its 4 ohm stereo output for roughly 10 seconds.  If you are using it at 2 ohms, or bridged at 4 ohms, then you are only going to get that level of output for 1 second.

There is a long thread on this topic here and, after digging into, I decided against running my Psoft K2s at 2 ohms or bridged at 4 ohms.  For the musical styles that I work with, 10 seconds == continuous, but I was unwilling to define continuous as 1 second.

[Ivan Beaver]

FWIW most modern switch mode amps will not produce their maximum output continuously.

But with no load, they should produce the maximum voltage available long term.

You SHOULD NOT be setting limiters with sine waves with a loudspeaker connected.

Unless you want to buy new loudspeakers.

[Mark Wilkinson]

In short, no. While they both reduce gain -- with different attack and release times -- the fundamental  difference is that the peak limiter responds to the peaks of the waveform while the RMS limiter responds to the average of the squared instantaneous values -- all of them (and takes the square root of that). For a sine wave these will indeed track with the peak 1.414 times the RMS, but for a complex waveform they will be very different. Consider a signal with a high crest factor, such as a very low duty-cycle stream of pulses. The peak value could be high while the RMS value -- corresponding roughly to heating in the voice coil -- would be low.

Post one drink and off to dinner now. Busy tomorrow but can respond further Sunday or Monday. -F

Thanks Frank, yep, I'm good with all you say there.
My question is not so much about how they work in theory; I'm wondering how they really work in practice.

Because I know how voltmeters have a hard time with good RMS measurements for simple sine waves, so it makes me wonder how a processor or amp can be making any kind of even half-assed RMS measurement on complex waveforms.

Also, I've started peering at peaky waveform limiting on a scope (mainly using the SC-48 processor mentioned earlier).
Using the SC-48's tmax (RMS) limiting on really peaky waveforms like a Sinc function, the limiter kicks in well before you would expect based on the waveform's RMS output.
It makes me think the processor is using peak voltage as the RMS limiting trigger, and just scaling peak down by the numbers you previously explained.
Gonna take a look at other processors and amps, to see if the same pattern shows.

[Keith Broughton]

I have worked on quite a number of Digam 5000 amps and what you describe is exactly what they are designed to do.
Usually I can get a pretty good voltage measurement when applying the test signal a number of times, trying to get the measurement before the output drops.
That said, what Ivan suggested is an excellent solution.
BTW..I have them connected to a 208V supply.

[ProSoundWeb Community]