ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 2 [All]   Go Down

Author Topic: DSP amplifier limiter calculation  (Read 6596 times)

Roberto Orozco

  • Newbie
  • *
  • Offline Offline
  • Posts: 10
DSP amplifier limiter calculation
« on: February 19, 2018, 08:17:39 PM »

hi, I hope somebody can help me with this issue. I recently bought an amplifier module for a pair of speaker cabinets I'm  building. the amplifier comes with DSP features, but is not programmed yet and I'm not sure how to do this. I contacted the manufacturer for help but they just sent me a link to download the software to adjust the different settings of the module. they told me is really easy!  it may be!  for somebody with the knowledge in the field.

I'm not a sound engineer, I been doing the sound with my band (small town band) for some years, we use an analogue mixer to connect all the signals, mixer to crossover, crossover to amps with no DSP, amps to speakers.
I been doing some research about this and looks like the DSP settings of an amplifier are base on the components that I'm going to be using in my cabinets and their specifications.

the amplifier has 3-channels
1500watts-low    500watts-mid    500watts-highs------at 8ohms
1800                  900                  900                 ------at 4ohms
RMS power.

the amp is to power 2 cabinets, each cabinet includes:

1 X 12'' subwoofer  550watts RMS  8ohms
1 X 6''   driver     100watts RMS   8ohms
1 X 1.75''  high frequency driver  100watts RMS  8ohms

if somebody could tell me how to setup the DSP features on the amplifier to use it with this speakers please!

I understand that to calculate the different settings, gain, limiters, crossover points, of the amplifier I need to know the specs of the speakers and the amp. I'll put the pictures attached with the information that the manufacturer gave me for the amp and the information I have from the speakers, I hope this helps.

I been trying to build my own speaker cabinets since long time ago, I hope I can find the help to complete my project  in this forum.

thank you so much for looking.     
Logged

Chris Grimshaw

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1822
  • Sheffield, UK
    • Grimshaw Audio
Re: DSP amplifier limiter calculation
« Reply #1 on: February 20, 2018, 03:54:34 AM »

Here's how I'd do it:

Limiters:
Sub - set a fairly long attack time (couple of seconds) and an even longer release time at -3dB.
Mid - 0.5s attack, 2s release, -6dB
High - fast attack and release, -6dB

Those will kick in if you hit the drivers with too much power and cut the signal down to something the drivers can manage thermally.

Dial in whatever crossover points the manufacturer states. I'd use 4th order Linkwitz-Riley on all of them*. Make sure you set a highpass filter on the 12" to cut off the very low bass. A poof of air into a microphone will launch cones across the room without that filter.

When you've done all that, play a test tone at the crossover frequency. Set it so the midrange driver has reversed polarity. Now, play with the delay on the midrange driver until you hear the minimum amount of sound. If you increase delay, you'll find that you can make that happen many times. You want to use the minimum delay required to still get that cancellation. Set the midrange driver for correct polarity and you should have lots of sound again.
Repeat this for the high-frequency driver. Reversed polarity for HF, play with delays, etc.

That should get your system somewhere near time-aligned. It's not the best method, but if you don't have a measurement microphone etc it's a pretty good way of doing it. It can be quite difficult with the high-frequency section - I'd recommend doing it outside so there aren't any reflections messing things up.

Play with the low/mid/high levels until you're happy with the system balance, and you're done.

*I like LR4 because it offers a decent amount of driver protection, and the two bands are electrically in-phase at the crossover point, which makes finding the delays much easier.

Chris
Logged
Sheffield-based sound engineering.
www.grimshawaudio.com

Roberto Orozco

  • Newbie
  • *
  • Offline Offline
  • Posts: 10
Re: DSP amplifier limiter calculation
« Reply #2 on: February 20, 2018, 12:23:11 PM »

thank you so much Chris for taking time to give me this information. I appreciate that!  if you don't mind, can you tell me how you calculate those numbers for the limiter?
I was trying to do it myself using amplifier limiter calculators on the internet but I couldn't do it because I was missing the gain voltage of the amp in dB. or some other number.

I think I can do what you say with the delay, the part that i'm not completely sure is how to reversed the polarity  to test the drivers.
excuse me if I can't understand completely what you are explaining to me, maybe is not really difficult but honestly I never had to dig inside of the DSP settings of an amplifier before.
the pictures that I attached are from the software itself, those are the screens that I'm going to be working on and there are numbers and boxes and numbers that I don't understand very well... but I would like to.  I'm not asking you to teach me all about DSP programming I just hope you be patient and if you have the time, help me to understand a little more what I'm trying to do.

I'll really appreciate your help and your time to replay. thank you.     
Logged

Mike Caldwell

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3091
  • Covington, Ohio
    • Mike Caldwell Audio Productions
Re: DSP amplifier limiter calculation
« Reply #3 on: February 20, 2018, 04:49:42 PM »

Here's how I'd do it:

Limiters:
Sub - set a fairly long attack time (couple of seconds) and an even longer release time at -3dB.
Mid - 0.5s attack, 2s release, -6dB
High - fast attack and release, -6dB

Those will kick in if you hit the drivers with too much power and cut the signal down to something the drivers can manage thermally.

Dial in whatever crossover points the manufacturer states. I'd use 4th order Linkwitz-Riley on all of them*. Make sure you set a highpass filter on the 12" to cut off the very low bass. A poof of air into a microphone will launch cones across the room without that filter.

When you've done all that, play a test tone at the crossover frequency. Set it so the midrange driver has reversed polarity. Now, play with the delay on the midrange driver until you hear the minimum amount of sound. If you increase delay, you'll find that you can make that happen many times. You want to use the minimum delay required to still get that cancellation. Set the midrange driver for correct polarity and you should have lots of sound again.
Repeat this for the high-frequency driver. Reversed polarity for HF, play with delays, etc.

That should get your system somewhere near time-aligned. It's not the best method, but if you don't have a measurement microphone etc it's a pretty good way of doing it. It can be quite difficult with the high-frequency section - I'd recommend doing it outside so there aren't any reflections messing things up.

Play with the low/mid/high levels until you're happy with the system balance, and you're done.

*I like LR4 because it offers a decent amount of driver protection, and the two bands are electrically in-phase at the crossover point, which makes finding the delays much easier.

Chris

I did that same thing and went with what I felt was the deepest null point and then later checked it with some measurement software that would display the magnitude and phase response and I found that most of my by "ear" delay settings were fairly close.

Paul G. OBrien

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1393
Re: DSP amplifier limiter calculation
« Reply #4 on: February 20, 2018, 11:35:57 PM »

the part that i'm not completely sure is how to reversed the polarity  to test the drivers.
Roberto... in the third pic down in your first post there appear to be buttons in the flow diagram that are labelled POL, I believe that is short for Polarity. I bet if you click on that you will see options for normal(+) or inverted(-).
Logged

Roberto Orozco

  • Newbie
  • *
  • Offline Offline
  • Posts: 10
Re: DSP amplifier limiter calculation
« Reply #5 on: February 21, 2018, 12:35:44 PM »

Roberto... in the third pic down in your first post there appear to be buttons in the flow diagram that are labelled POL, I believe that is short for Polarity. I bet if you click on that you will see options for normal(+) or inverted(-).

thank you so much Paul, I'll get to work on it. 
Logged

Chris Grimshaw

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1822
  • Sheffield, UK
    • Grimshaw Audio
Re: DSP amplifier limiter calculation
« Reply #6 on: February 21, 2018, 01:23:50 PM »

thank you so much Chris for taking time to give me this information. I appreciate that!  if you don't mind, can you tell me how you calculate those numbers for the limiter?   

I made the assumption (rightly or wrongly) that 0dB would be maximum power, and therefore -3dB would be half-power.
The time constants are just educated estimations of the thermal masses of the voicecoils in use for each band. You could try asking the manufacturer if they've got any recommended limiter settings for each driver, but I wouldn't bet on an answer.

Chris
Logged
Sheffield-based sound engineering.
www.grimshawaudio.com

Roberto Orozco

  • Newbie
  • *
  • Offline Offline
  • Posts: 10
Re: DSP amplifier limiter calculation
« Reply #7 on: February 21, 2018, 03:35:52 PM »

I made the assumption (rightly or wrongly) that 0dB would be maximum power, and therefore -3dB would be half-power.
The time constants are just educated estimations of the thermal masses of the voicecoils in use for each band. You could try asking the manufacturer if they've got any recommended limiter settings for each driver, but I wouldn't bet on an answer.

Chris
 
Hi Chris thanks again for your help.  well, like I mention before, I was trying to calculate the limiter for the amp using a limiter calculator, I don't remember the website, but they ask me for the power/voltage gain of the amplifier in decibels in order to do the calculation. I contacted the manufacturer letting them know what I was trying to do and this was the answer I get:

0dB for LF;  -10dB  for MF;  -12dB for HF.

exactly like that.
Logged

Chris Grimshaw

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1822
  • Sheffield, UK
    • Grimshaw Audio
Re: DSP amplifier limiter calculation
« Reply #8 on: February 22, 2018, 03:41:29 AM »


Hi Chris thanks again for your help.  well, like I mention before, I was trying to calculate the limiter for the amp using a limiter calculator, I don't remember the website, but they ask me for the power/voltage gain of the amplifier in decibels in order to do the calculation. I contacted the manufacturer letting them know what I was trying to do and this was the answer I get:

0dB for LF;  -10dB  for MF;  -12dB for HF.

exactly like that.

The calculator you were using needs to consider how much gain the amplifier has in order to work out your limiters - if you limit both to -3dB, but one has 36dB gain and the other has 46dB, then one will be putting out 10dB more signal.

With your amplifier, we're looking at the limiting compared to the maximum power it can give out.

In my opinion, setting for 0dB on the LF would be risky, as the amplifier does have enough power to burn out the woofers if you hit them with a high-power signal with a long duration. Like low-frequency feedback, for instance. The limiting I suggested (with a long attack/release time) won't activate unless there's been a high-power situation for a few seconds, in which case the power levels need to be decreased in order to save the woofers.
Their MF/HF settings are more conservative than mine. I can see why they've done that. It's going a bit safer, but losing a bit of peak output. It's all tradeoffs.

Some amplifiers have RMS limiting, which is really useful - you can tell them the peak/continuous power handling of your drivers, and the amplifier will look after the rest.

Chris
Logged
Sheffield-based sound engineering.
www.grimshawaudio.com

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9534
  • Atlanta GA
Re: DSP amplifier limiter calculation
« Reply #9 on: February 22, 2018, 07:31:34 AM »

The BEST way to set limiter voltages is to MEASURE them, using a sine wave a voltmeter.

Calculators can be useful-IF everything is correct that is entered into them.

I have know some amp gains to be different (as much as 4dB) than what the spec sheet says.

I know one major amp manufacturer who considers the gain to be normal if it is +/- 1 dB over published specs.

Simply put a sine wave in, adjust the gain so that the measured voltage on the output is higher (I like at least 10%) than the voltage you will limit to.

BE SURE the speakers are NOT hooked up at this time!!!!!!!

Then simply turn down the threshold until you get to the desired voltage.

If you are using multistage limiters, be SURE to do the peak or highest limiter first, and the continuous or heating limiter last.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

Roberto Orozco

  • Newbie
  • *
  • Offline Offline
  • Posts: 10
Re: DSP amplifier limiter calculation
« Reply #10 on: February 24, 2018, 11:03:09 AM »

The BEST way to set limiter voltages is to MEASURE them, using a sine wave a voltmeter.

Calculators can be useful-IF everything is correct that is entered into them.

I have know some amp gains to be different (as much as 4dB) than what the spec sheet says.

I know one major amp manufacturer who considers the gain to be normal if it is +/- 1 dB over published specs.

Simply put a sine wave in, adjust the gain so that the measured voltage on the output is higher (I like at least 10%) than the voltage you will limit to.

BE SURE the speakers are NOT hooked up at this time!!!!!!!

Then simply turn down the threshold until you get to the desired voltage.

If you are using multistage limiters, be SURE to do the peak or highest limiter first, and the continuous or heating limiter last.

Hi Ivan, I appreciate the information. like I mention before, my knowledge on the field limited (ha ha, talking about limiters! sorry)

I'm always willing to learn!  is there any video tutorial or link that could show me how to do that the you mention with my amplifier?

probably takes more than a video to understand and do it the proper way! maybe I need equipment that I don't have in my basement right now I can tell you that, I do have a voltmeter!   I just would like to finis my project with the best possible results.

thank you so much for the advise.
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9534
  • Atlanta GA
Re: DSP amplifier limiter calculation
« Reply #11 on: February 24, 2018, 05:54:13 PM »



I'm always willing to learn!  is there any video tutorial or link that could show me how to do that the you mention with my amplifier?

probably takes more than a video to understand and do it the proper way! maybe I need equipment that I don't have in my basement right now I can tell you that, I do have a voltmeter!   I just would like to finis my project with the best possible results.

thank you so much for the advise.
All you need is a voltmeter and a signal generator.

There are free apps, or you can download various sine waves.

Just follow these steps


LIMITER VOLTAGE-How to measure preset settings.
When determining what voltage limiters are set at, without a voltage number, you must measure the actual voltage output of the amplifier.
It is VERY IMPORTANT to use a voltmeter that has a flat response in the audio freq range.  MANY do not.

If you are using a meter that does not have a flat response, then you should ONLY use a test freq of around 60Hz.  This may mean bypassing various HP filters during the test, but be SURE to turn them back on after the test.

If you are using a flat response volt meter, then use a test freq around the middle of the response range of interest.  It does not have to be exact, just somewhere in the range before rolloffs

1: DISCONNECT the loudspeakers!!!!!!!!! And hook the voltmeter to the output of the amplifier in question.

2: We are going to “assume” a 2 stage limiter.  If there are more (3 is ideal), then simply add the additional steps, which are the same basic procedure.

3: Apply a sine wave to the input of the device/system and turn the level up until the voltmeter does not increase with increasing signal level.  MAKE SURE that no stages prior to the amplifier or the amplifier itself are clipping.  If they are, then readjust the gain structure of the system.
This voltage should be the or “heating”, or “RMS” limit of the system.  Note the voltage and write down the threshold setting.

4: Now either bypass the limiter in step 3, or raise the threshold to the maximum.  Now increase the drive level until the meter no longer increases in voltage. MAKE SURE there is no clipping anywhere in the system path-including the amplifier.  Readjust gain structure if needed.
This voltage should be the peak voltage.  Note this voltage

5: Re-engage the limiter in step 4 that was bypassed.

6: If there is a 3rd limiter, it would be the program limiter, and the step for it would be between 3 and 4 above.

7: Reconnect loudspeakers

ROUGH GUIDELINES FOR LIMITER SETTINGS
Voltage.  Determine the continuous voltage rating.  The formula is the square root of (continuous power rating x impedance).
Use this as a basis, the thermal or RMS limiter should be ½ of this voltage (1/4 power). 
Program limiter should be the continuous voltage rating
The peak voltage should be 1.414x the continuous voltage (2x power).  YES, that is 3dB down from the actual peak rating of the loudspeaker, but it has been determined that this is a safer setting to use than the “near death” rating.
ATTACK TIMES
Thermal/RMS   large subs 3seconds, general woofers 1 second, midrange 0.5 seconds, HF 0.1seconds 
     Release time should be 1.5 to 2x attack time.
Program    The attack time should be equal to the freq of the high pass filter for the xover.  Ms=1130/HP filter freq.    The release time should be between 10-20x attack time-15-16x is considered “ideal”.                                 
Peak should be as fast as possible.

Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

Roberto Orozco

  • Newbie
  • *
  • Offline Offline
  • Posts: 10
Re: DSP amplifier limiter calculation
« Reply #12 on: April 05, 2018, 09:59:28 AM »

All you need is a voltmeter and a signal generator.

There are free apps, or you can download various sine waves.

Just follow these steps


LIMITER VOLTAGE-How to measure preset settings.
When determining what voltage limiters are set at, without a voltage number, you must measure the actual voltage output of the amplifier.
It is VERY IMPORTANT to use a voltmeter that has a flat response in the audio freq range.  MANY do not.

If you are using a meter that does not have a flat response, then you should ONLY use a test freq of around 60Hz.  This may mean bypassing various HP filters during the test, but be SURE to turn them back on after the test.

If you are using a flat response volt meter, then use a test freq around the middle of the response range of interest.  It does not have to be exact, just somewhere in the range before rolloffs

1: DISCONNECT the loudspeakers!!!!!!!!! And hook the voltmeter to the output of the amplifier in question.

2: We are going to “assume” a 2 stage limiter.  If there are more (3 is ideal), then simply add the additional steps, which are the same basic procedure.

3: Apply a sine wave to the input of the device/system and turn the level up until the voltmeter does not increase with increasing signal level.  MAKE SURE that no stages prior to the amplifier or the amplifier itself are clipping.  If they are, then readjust the gain structure of the system.
This voltage should be the or “heating”, or “RMS” limit of the system.  Note the voltage and write down the threshold setting.

4: Now either bypass the limiter in step 3, or raise the threshold to the maximum.  Now increase the drive level until the meter no longer increases in voltage. MAKE SURE there is no clipping anywhere in the system path-including the amplifier.  Readjust gain structure if needed.
This voltage should be the peak voltage.  Note this voltage

5: Re-engage the limiter in step 4 that was bypassed.

6: If there is a 3rd limiter, it would be the program limiter, and the step for it would be between 3 and 4 above.

7: Reconnect loudspeakers

ROUGH GUIDELINES FOR LIMITER SETTINGS
Voltage.  Determine the continuous voltage rating.  The formula is the square root of (continuous power rating x impedance).
Use this as a basis, the thermal or RMS limiter should be ½ of this voltage (1/4 power). 
Program limiter should be the continuous voltage rating
The peak voltage should be 1.414x the continuous voltage (2x power).  YES, that is 3dB down from the actual peak rating of the loudspeaker, but it has been determined that this is a safer setting to use than the “near death” rating.
ATTACK TIMES
Thermal/RMS   large subs 3seconds, general woofers 1 second, midrange 0.5 seconds, HF 0.1seconds 
     Release time should be 1.5 to 2x attack time.
Program    The attack time should be equal to the freq of the high pass filter for the xover.  Ms=1130/HP filter freq.    The release time should be between 10-20x attack time-15-16x is considered “ideal”.                                 
Peak should be as fast as possible.


hi Ivan, first of all I want to say than you for your help and advise.
I tried to do what you told me, I'm not sure if I did the right way but, this is the result:

I connected the voltmeter low frequency  out put of the amplifier, I used I sine wave app at 120Hz, the readings on the voltmeter were 0.00 at this time. I started to turn the volume up but the voltmeter never stop moving until I reached the max volume on the amp, the numbers I get were 30.7  I was measuring AC volts.
after I reached the max volume the voltmeter stop.
the frequency cuts from the manufacturer for this speaker are 700Hz, 3000Hz, so for the 6" driver I used a sine wave at 1200Hz  and for the high frequency driver I used a 5000Hz sine wave. after repeating the process on each driver this are the numbers I get:
L-30.7
M-15.3
H-10.5
I did the test second time  using a really cheap voltmeter that I found in my garage but the readings were even smaller:
L-12.8
M-8.3
H-6.7
I'm not sure how good the battery was on this one but the number moved a lot.

my amplifier only have one box for each output  to setup the limiter, I thing that's what is called I stage limiter...not sure.

-the threshold only shows numbers from -1dB to -7dB
 
-the attack only shows number from 0.5ms to 50ms

the release from 100ms to 500ms

and the ratio:

I want to use the amplifier to drive 2 X 12'' woofers 550W RMS each 8ohms
2 X 6" drivers 100W RMS 8ohms
and
2 X 1.75 HF drivers 100W RMS 8ohms
not sure if that's actually possible!
I'll be using this cabinet as tops for my system, with a powered 18" subwoofer. the subwoofer has a HP filter at 120Hz (optional)

most of the time we will be using just 1 cabinet  per side with the 18" sub since we played at small venues or backyards.

I hope this information can help and if you don't mind give a most accurate number for my limiters so I can protect my speakers, thank you.

Best regards:
Roberto Orozco.
     
« Last Edit: April 05, 2018, 05:08:48 PM by Mac Kerr »
Logged

Ray Aberle

  • Classic LAB
  • Hero Member
  • *
  • Offline Offline
  • Posts: 3455
  • Located in Vancouver, WA (and serves OR-WA-ID-BC)
    • Kelcema Audio
Re: DSP amplifier limiter calculation
« Reply #13 on: April 05, 2018, 04:56:39 PM »

Move the [/quote] tag to just BEFORE your response, and it'll keep your reply out of the quoted text. :)
Logged
Kelcema Audio
Regional - Serving Pacific Northwest (OR, WA, ID, BC)

ProSoundWeb Community

Re: DSP amplifier limiter calculation
« Reply #13 on: April 05, 2018, 04:56:39 PM »


Pages: 1 2 [All]   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.027 seconds with 21 queries.