Print

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

[Ivan Beaver]

. It is only in auto mode you will have to worry about the RMS and Peak Limiter relationship.
You can actually see what voltage auto sets your peak limiter too if you scroll through the front panel of the amp.

But at least in the advanced mode-the RMS and the Thermal limiters are tied together.

[Ivan Beaver]

OK it gets even more interesting.

I went back and checked and recalculated my voltages and with the proper voltages entered, I still cannot get the output I am looking for.

Now either I have a bad amplifier-but everything appears to be working (except the settings in the limiter), so I will assume it is all good.

So I decided to take a different approach.  Figuring that somehow I am doing my “figuring” wrong.

And since the DSP does not actually measure the output voltage-or look at the output- (It assume a block with a certain amount of gain), there is room for a dB or so error-but what I am seeing is a good bit more than that.

So I recalled a JBL preset to see how the numbers matched up with the spec sheet numbers.

MODS-PLEASE don’t think this is some sort of  “discussion” about another product-(not my intention at all-but those are the only presets in the unit to use), but rather as a trying to figure out where all the numbers come from type of thing.

So I used the 4886- a passive 12 ohm full range cabinet.

It is rated at 900 watt continuous or 104V by my calculations.
The peak is 3600 watts or 207 volts.  And then multiply this by 1.414 and you get the peak voltage to be entered of 293 volts.

If you use Crowns suggestion of 80%  of the continuous if the exact thermal voltage is unknown you get 83 volts.

So opened up the limiter in the preset and here is what I found.  The Peak was at 180 (not the 293 I calculated), the RMS was at 90 (not the 104 I calculated) and the Thermal was at 79 (not that far off from the guessed at 83 volts).

So how much can I actually get out of the amp with these settings?  I used the internal noise as a source (trying to get rid of as many variables as I could).

When I cranked it up into limiting, I was getting an output of noise of 57-58Volts.

If you do the math-58 volts gives you 280 watts.  Nowhere near the 900-3600 watt rating of the loudspeaker.

So what is the problem(s)?  The numbers that are calculated from the spec sheet?  The numbers that are entered into the preset?  Or something else?

I tried another JBL product with a completely different set of parameters (both in response range-wattage and impedance) a 4883.  The results were kinda similar off-I won’t bore you with the exact numbers.

So, just to double check myself (I am having some doubts at this point), I recheck the main meter I was using, the HP 400e analog movement-against my other meters-a HP 3456 digital meter-my Wavetek VOM, A portable hand held O scope Chad had, and used a Sencore variac with analog meter as another source to put my measured AC in the same ballpark as I was measuring.

They are all right on the money.  So there should not be a problem there.

Something else is “wacky”-and I am open to any suggestions.

How come the JBL numbers are so different than the limiter settings in the preset?  How come my measurements don’t line up with those numbers? But when I put the same numbers into another DSP, I get exactly what I was expecting?

I will ask again, has anybody actually measured (with a voltmeter on the output terminals of the amplifier), the output with either music or noise as the source and gotten anywhere near the voltages that are input on the screen?

The sine wave numbers appear to be pretty close to the numbers on the screen, but it “appears” that something is happening in the process of trying to limit real signals as to give a different result.

And to go back to the original thing that started all of this.  The product in question was putting out what was expected SPL wise (with no limiting).  But when the numbers for limiting (based on the voltage calculations) were entered into the DSP, the output dropped about 4-6dB. 

It was only when the voltages were raised up quite a bit, that the output came back to where it should be.

Anyway-you have a confused Georgia boy right now.


 

[Justin Myers]

Dont know Ivan,
Seems odd. Even using music as the signal, the rms and peak voltages monitored by the osc. would reach set voltages.  I did not however have a thermal limiter set.
If there is a testing process you would like me to try on my amp let me know.
Justin

[John Roberts {JR}]

I would be more comfortable if Crown was participating in this exploration.

I already gather that it isn't acting like you expect.

I am curious too, but I don't perceive much forward motion.

Can you dial it in to work properly for one or more of your boxes? Confirm that it is set and working properly. Then share those "Crown" vs. actual threshold settings.

The truth is out there, and you should be able to triangulate to get the truth for your box.

Then we can ask Crown to explain their settings vs. what you see and measure they are actually doing.   

Either this is much ado about nothing or one of ya'll is wrong.


JR

[Ivan Beaver]

I would be more comfortable if Crown was participating in this exploration.

I already gather that it isn't acting like you expect.

I am curious too, but I don't perceive much forward motion.

Can you dial it in to work properly for one or more of your boxes? Confirm that it is set and working properly. Then share those "Crown" vs. actual threshold settings.

The truth is out there, and you should be able to triangulate to get the truth for your box.

Then we can ask Crown to explain their settings vs. what you see and measure they are actually doing.   

Either this is much ado about nothing or one of ya'll is wrong.


JR

When I had my conference call last week, they said their paper would help to clear things up.

Well I read it, and followed it, and still have the same results.

Yes I can get the results I want, but not in a "conventional" manner-ie using the voltages as entered into the fields.

To get what I want, I just raise the voltages real high and lower them until I get the "action" on the reduction meters and the voltage as read on the output that I want.

Then give those numbers "the settings".

It would be a lot easier if I could just calculate the voltages and then enter them and be done with it.

I will send these latest findings to Crown and see if they have any insight.

My new "concerns" are that the JBL wattages don't agree with the voltages as set on the limiter, and the settings on the limiter don't agree with the output voltage as read on the output terminals.

So out of the 3 things, nothing agrees or even comes close.  In this case I would consider 1dB to be close enough not to worry about it.

In the particular case in question, the head audio guy for the venue says they will never be allowed to operate the system as loud as the current limiters allow, the designer is happy with the system, but the installer wants more level-to be able to "push it to the max".

In my case, I just want some answers that I can rely on-you know-a standard approach for adjustment. 

Anyway-when I get more time-----------------------------

[Justin Myers]

Are you sure your measuring devices aren't all RMS measuring devices like Marty mentioned?

[John Roberts {JR}]

If I understand, you empirically measure actual threshold behavior that disagrees with the numerical entries. That suggests a scale error that should allow you to return to your math targets, just scaled to correct for Crown's internal math offsets, once you determine the scale factor(s).

Remaining questions are, does it scale reliably, or are there less than obvious interactions that also complicate the settings?

If you set up one box you should be able to determine scale factors. If you apply these scale factors to settings for another box, does it track or not?

This does seem a little too important to tolerate large errors that could be the difference between life or death for drivers. Another scary thought is what happens if they change it to agree with Ivan's expectations and all the custom tweaked settings are now wrong.

JR

PS: I apologize to Ivan I understand you are not trying to make Crown look bad, they are doing fine all by themselves by not engaging in public dialog.

[Irving A. Hammond Jr.]

It looks like their limiter settings are 2.5 db down from their rated power and the limiters limit peaks not rms like multimeters measure.

900w - 2.5db = 675w = 90 volts

80% of 900w = 720w - 2.5db = 540w = 80 volts

3600 - 2.5db = 2700w = 180 volts

Your 57 volt reading is 3db down from about 80 volts so it about a 3db reduction from thermal.

Sounds like JBL power specs are inflated or they decided to be conservative on their limiter settings.

[Ivan Beaver]

So here is the update on the voltage in the limiter section of the Itech amps.

One of our guys went to the Crown factory and spent a good deal of time with both harware and software Engineers.

I don't have the whole story yet, but what it appears is that there is a bunch of "special sauce" in the limiter section.

It was apparently not designed to drive a full range passive loudspeaker.

Although the JBL example I used is such a loudspeaker and the presets don't come close to the spec sheet numbers, but that is a different issue.

It appears that the only way we are going to the voltage numbers to actually matchup to what we want the limiter to do is to go into the "totally manual" mode.  This is not selectable in the software (as I understand it).  This is NOT the advanced mode.  It is one step further.

Apparently you have to "build" a block and then you have full access to what you want it to do.  I will learn more about this tomorrow when we do it.

So my origional conclusions appear to be correct.  If you take an XYZ loudspeaker with a specified power and impedance rating-then convert those to voltages, the end result is that the loudspeaker will not get as loud as you want it to.  Unless you know how to go into the "running with sissors mode"-as Crown calls it.

Apparently it is not that popular-at least to the people who have responded,as it has not been mentioned.

Yes you can actually measure the output signal (with noise or music-sine waves won't cut it due to the "special sauce") and then adjust the voltages to get what you want.  You just need to pretend that those are not voltages, but just rather "numbers on a scale".

But actually measuring and verifying appears to be a thing of the past.  I see this all the time in which people just blindly take the numbers that they are given as "gospel" and never check to be sure that they are correct.

But yet they use those same "numbers" to compare products of which they haven't verified either.  "So and so say so-so it MUST be right"

Seems odd to me-but that is just the old  bench tech in me-I expect a product to deliver what it says it will.

Oh well, as I find out more, I will update.

MEASURE PEOPLE MEASURE!!!!!!!!!!!!!!!VERIFY what you "think" is happening!

[Justin Myers]

Well I was about to post "I dont know what your caught up on Ivan, my measurments are fine."
I was using the Tektronic Osc at work today and realised what I thought was my Peak readings were actually Peak to Peak readings. The RMS should have been right still but I will need to check everything again.

If what you have discovered is true though, I'm not real happy about it. You spend tens of thousands on amps that have DSP in them that isn't true. I dont want to have to adjust my settings after the amount of money I've spent. 100 Volt limit setting should equate to 100 Volt limit.
I hope there is a silver lining to all this.

EDIT:
No, My measurements were right the first time. I did realise at the time it was Peak to Peak. I tried higher voltage thresholds this time and I'm getting the same result, not as tight, but still very close. With music as source the Rms and peak voltages were not far from set values. They were not falling short, sometimes exceeding thresholds momentarily.

I'm not sure on the product any more though after the depth Ivan has gone and found differences. Very interested in what the Crown factory had to say.

[ProSoundWeb Community]