Setting Amps Gains by Ear and DMM

[UltraNutZ]

First off for the critics out there; I want to point out that to perform a 100% accurate test, an oscilloscope, the knowledge to use it, and multiple test tones/tone generator are required but this procedure will get you pretty darn close with a single test tone and a DMM.

*IMPORTANT* Disconnect speaker output wires from amplifier to speakers during this procedure otherwise you will stress the speakers and could possibly damage them when sending this 1000hz tone through the amplifier. Also disconnect any speakers that are connected directly to the HU.

This is the down and dirty way to test and set your gain controls on your amplifier with a DMM to properly drive your speakers without the amplifier clipping. To test a subwoofer, the process is the same; you simply need a lower frequency test tone for subwoofer frequencies such as 50hz. I don't suggest you use an iPod to do this mainly because you'll have to worry about the equalizer settings on that device too. I recommend on 5.25", 6.5", or 7.25" coaxial or 3 way speakers, as most are using on their bikes, that you use a 1000Hz 0db test tone to conduct this procedure. This procedure is the same regardless if using RCA inputs to the amp or line-level inputs (speaker wire inputs)



Abbreviations
HU – Head Unit or Radio
DMM – Digital Multi Meter
AP – All Pass – allows all frequencies to pass to speaker
HP – High Pass – allows all signals over a certain frequency to pass to speaker.  The frequency is adjustable on the amplifier.
LP – Low Pass – allows all signals below a certain frequency to pass to speaker.  The frequency is adjustable on the amplifier and used for sub woofers installations
F – Full – same as All Pass


Information and items needed prior to test procedure:

1- V = √PxR (see below)
2- Test tone CD or thumb drive. This tone can be obtained here and simply burn to a CD (or whatever device you are using) that will play in your HU. Leave this test tone at 320kbps otherwise this will be a futile effort.
3- You will need a decent DMM set to AC volts with auto ranging enabled or if you don't have an auto ranging meter, set it to measure at least 50vac.
4- Amplifier, HU, speakers already installed and fully tested for functionality.
5- Specifications for amplifier(s) and speakers.
6- For my example I will be referring to the Soundstream PN4.520d amplifier and Polk Audio MM651 speakers. Find your specifications in your owner's manuals for your specific product. DO NOT USE MAX POWER RATINGS OR BIRTH SHEETS!
a. Amp specs
RMS power 4 ohm load per channel (4 channels) – 100wx4=400W
RMS power 2 ohm load per channel (4 channels) – 130wx4=520W
b. Speaker Specs - You simply need to know what ohms your speakers are (2.7 ohms) and what ohm load the amp will see (2.7 ohms).
7- Set your HU to repeat play mode so it will continue to play the tone. The above tone is 2:00 in length so you have that much time before you will see the AC voltage drop on your DMM. If it does, no biggie, simply wait for it start playing again.
8-If you have an amp with 2 gain controls (usually a 4 channel amp), it is to be treated as 2 separate amplifiers and divide total watt output by 2.




Procedure:

1- What you now need to figure out is the V in the above equation. We know the P and R from the specs above. So, P = RMS output of amp at required ohms and R = ohms the amplifier will see from speaker(s) on that specific channel. In this example P=520/2 (260) (I'm running a 2.7 ohm load to each channel) and R=2.7. So P(260) x R(2.7) = 702. Now you need the square root of 702 which equals 26.49 so V=26.49. This is the AC voltage that you will achieve by adjusting your gain(s) on your amp.
2- Turn all equalizer settings, bass boost, etc. on HU to their minimum settings
3- Turn volume control all the way down on HU
4- Turn all gain controls, bass boost, etc. on amplifier to their minimum settings
5- Turn HP/LP filters to full range. On RF amps it's called AP and the Soundstream is called F
6- Connect your DMM + (red) lead to the front right speaker output + on the amp. Connect your DMM – (black) lead to the front right speaker output - on the amp.
7- Put test tone CD you burned earlier (or thumbdrive) in HU.
8-Turn your radio up to ¾ volume. On a HU that displays numbers turn your volume to max then divide that number by 4 to get your ¾ setting. On HD HUs there are 17 bars to max volume so set your volume at 13 bars.
9- Now look at your DMM. Whatever your AC voltage reading is coming out of the amp, adjust it up to 26.49 (or as close as you can get it)
10- On most 2 channel amps such as the RF PBR 300x2 has only 1 gain control. Most 4 channel amps will have a gain control for the front and one for the rear. The Soundstream PN4.520d has 2 gain controls, 1 for front, the other for rear so I'll need to repeat this adjustment for the rear channel output also.
11- After you have adjusted your rear gain control to match the front, turn your HU volume down to minimum and take out the test tone CD. Again, you do not want this test tone running through your speakers.
12- Turn everything off and reconnect your speaker wires.
13- Now you can adjust your equalizer or bass and treble back to your liking.




Setting your amp gains by ear
there are a couple of steps that MUST be followed prior to adjusting gain controls or the whole process is at best a futile effort as your gains will NOT be adjusted properly.

On stock H/K- Harley head units set your bass and treble to flat settings. Flat is the middle smaller bar on the display
On aftermarket head units, the easiest thing to do is press the reset button on the head unit to reset everything on it back to factory defaults. Otherwise make sure bass, treble, and ANY other sound enhancements are turned to their 0 settings.
Obtain a CD or iPOD (make sure EQ settings on iPod are off as well) with some dynamic music on it. Van Halen or Peter Gabriel is pretty dynamic and will suffice for this exercise.
Turn your gain control(s) on the amp all the way down. This is usually the counterclockwise direction.
With CD/iPOD playing, turn head unit volume up until you hear distortion then turn it back down a little until distortion is no longer present. If you can turn your head unit's volume all the way up without distortion then back it down to 3/4 and proceed
if you have multiple gain controls, adjust each seperately. Turn your gain(s) up until your hear distortion, then back the gain control down until the distortion is gone.
Proceed to other gain controls if you have them
UltraNutZ

[pp6000v2]

reading through the how-to, it still looks like you're measuring the AC voltage across a single channel, but are arriving at the desired voltage based on a 2-channel number.  Basically, it looks like you're setting the gain at twice the level you should be.  Is that correct?

Also, something that's been bugging me lately- for the stock HU with AVC turned on at whatever level, what impact does that have on gain settings if I'm setting the gain sitting still in the driveway?

[UltraNutZ]

not sure I'm following you on this.  Think of an amp with 4 channels (2 channels front and 2 channels rear) as each 2 channels as being 1 amp.  This is how the calcs are obtained and those calcs are applied to both outputs of that channel you're working with.  IF there were gain controls for each individual channel it would be a different ballgame altogether.

As for AVC, well if you can figure out a way to factor that into the equation I'm all ears but IMO its impossible because its a moving target.  As much as I like and use AVC, I think it induces a whole different set of problems to deal with when it comes to aftermarket audio.  I mean lets face it, no amp manufacturer, not even those slated as "motorcycle audio" manufacturers, can or will factor AVC into the mix.  Common sense has to factor into this somewhere.

[pp6000v2]

If it's putting 130w @ 2-ohm x 4, with two independent gain settings that's 260w across 2 channels. In the sticky example, you're coming up with 26.49v as the number you're looking for.  But then you're measuring across just the +/- of the Right Front channel in the example, rather than L+ and R-.  Wouldn't you want 18.73v (√130x2.7), not 26.49 (√260x2.7), if you're just measuring at one channel?


For AVC, at the end of the day, if it's distorting and clipping, it needs to be turned down; I just didn't know if there was a direct relationship between say AVC+4 and 4 bars on the volume

[UltraNutZ]

only if you would be bridging that channel (2 outputs).  if it's a 4 channel amp 130wx4@2ohms then that should be treated as 2 separate amps (1 amp per gain control).  The reason you're measuring across the left channel and then repeating for the right is "theoretically" they should be the same, at least most decent amps will be the same or really close anyways.  I personally haven't measured a 2 channel amp where output voltage in AC wasn't within 5% at a given volume with a given test tone.  Now if you're measuring AC voltage from the outputs with something other than a known test tone such as a music track then obviously your readings are going to fluctuate on that 1 output not to mention the other and then the two will normally be no where near the same.

[pp6000v2]

I may be totally talking in circles and thick.  But this is where I'm getting hung up:

6- Connect your DMM + (red) lead to the front right speaker output + on the amp. Connect your DMM – (black) lead to the front right speaker output - on the amp.

You came up with V = √[(520/2)x2.7] = 26.49.  To me, that math is saying that across both channels the gain is controlling, the voltage total should be 26.49.  But if I stick my probes on just one channel (the FR) and dial up 26.49, then stick them across FL+ and FR-, I get double that voltage.  If you're going to measure at just one channel, shouldn't the math be V = √[(520/4)x2.7] = 18.73v?  Or alternately, be measuring across the L+ and R- to arrive at that 26.49v?

[UltraNutZ]

the channels are independent of each other so what you calculate for one should be and has to be the setting for the other unless you've got a separate gain for the other ouput.  If you have an amp that's bridgeable you will see the numbers you're referring to.  If the amp is not bridgeable you will not get these high readings and to be honest I'm not sure what readings you would get.  The thing is you're not connecting L+ and R- to a speaker unless in bridged mode.  In bridged mode the calcs would be the same so you would then set V across L+ and R- (assuming those are the 2 leads to use for connectivity) to 26.49

[UltraNutZ]

here's something else to keep in mind.  For years with Oscilloscopes, signal generators, and DMMs, this has been the tried and true method.  Now you have geniuses out there such as Steve Meade who have taken it to a completely different level by using what's referred to as "reactive loads" which is constantly changing vs "passive loads" which is nothing more than an "assumed load" in testing for clipping and distortion points.  In other words when a speaker (tweeter, mid, sub, etc. doesn't matter) is operating at different frequencies it will have a highly fluctuating impedance load at the amp and that WILL ultimately affect the way your amp needs to be set.  And to further that, it's no different than tuning a bike in getting it dialed in exactly where you want it only to give that dialed in map to your buddy who has the exact same bike with the exact same mods as you and the map doesn't run worth a chit.  Speakers are the same way, no two will be the same.

[pp6000v2]

I'm looking at it like an electrical outlet.  If 120v is being supplied to each of the two outlets on the receptacle, plugging a device in to each outlet doesn't give each one to 60v, they're both going to get 120v.

So if your meter shows 26.49v at just the one measured channel (FL, FR, RL, or RR), you'd be pushing 260w at a speaker designed to run ~125w RMS/200w Peak, on an amp designed to push 130w per channel. That doesn't seem right.

I get why you want to treat a 4-channel amp as 2 logical amps when you have 2 gain controls, and I know 1kHz isn't a fluctuating, full-range signal that music would show, but why are you figuring the power of two combined channels (520/2) when you're measuring output on only one channel (which should be 260/2)? Or am I entirely missing something?

[UltraNutZ]

i think that's where you're confusing yourself.

AC has 3 wires, hot, neutral, ground.

we're talking 2 wires, hot and ground only.  There is also no common ground between outputs.

[pp6000v2]

So with load, the speakers are being fed only a share of the measured voltage, not all of it?

[Coyote]

Todd's procedure is measuring two amps at a time in bridge mode. That's why he calculated based on 260 watts. If you were to test just one amp, you would calculate based on 130 watts. I would expect the similar results either way.

[Garry in AZ]

Since this is a pinned subject, I thought it would be a good idea to provide a couple of corrections, and to provide a bit more insight into setting gains. I have no dog in the fight, and I absolutely mean no disrespect to any of the members who have already contributed to this thread. However, there are a couple of problems.
First of all, the procedure listed does not measure the amps bridged output, it specifically says to:
"6- Connect your DMM + (red) lead to the front right speaker output + on the amp. Connect your DMM – (black) lead to the front right speaker output - on the amp."
That is the correct connection for a single channel.
With that in mind, the target voltage stated is far too high. The amp will never reach that voltage on a single channel.
Without an oscilloscope, you are only guessing at the amps actual power, regardless of what it says in the published manual. I have measured lots of amplifiers with wildly optimistic published specs.
But if we assume the rated power is correct, then the target voltage using Todd's method would be 18.73 volts, just as pp6000v2 suggested.

Now for the "other" stuff... the test tone suggested indicates it's recorded at 0dB. In recording engineer speak, that means "maximum unclipped output" or sometimes we call it "all bits high".
This is fine if you want to set up a system for maximum dynamic range, and the best possible signal to noise performance. It does not however, work well if you have limited amplifier power.

Music is dynamic, and has higher and lower volumes throughout any given piece of recorded music. If you were to take the AVERAGE of the levels on a recording, and compare them to the maximum level possible, you would find that the actual average level is significantly lower than the 0dB tone. It has to be, otherwise every sound in the music would be exactly the same volume.
This difference between the AVERAGE level and the Max level is called Crest Factor.
The amount of crest factor in any specific piece of music is decided by the recording engineer. Sometimes, the engineer will want it louder, to sound "hot" when played on the radio, or if the music is usually used for dancing or clubs. A lot of Rap, Hip-Hop, and Grunge was recorded with about 5dB of Crest Factor. More mainstream music, Pop, Rock, and Country for example, is typically recorded with about 10dB of Crest Factor. Classical music, Jazz, and audiophile recordings can have as much as 15-20dB of Crest Factor.

Have you ever wondered why some recordings are a lot louder or quieter than others, even though you didn't touch the volume control? It's because of the amount of Crest Factor. The less Crest Factor the recording has, the louder the average level will be. If you are still with me, now I'll tell you what this has to do with setting the gains on your amplifiers.

If we were to measure the amps power in dB instead of watts, or simply provide a correlation between the two, you would see that for an amp to seem twice as loud, you need 6dB more power. But, 6dB means you would have to quadruple the watts from your amp to make it twice as loud. For example, if you had a 25 watt per channel amp, for the perception of twice as loud you need 100W per channel. For a 10dB increase, you would need 250 watts per channel.

So, back to Crest Factor and gains. If you listen to popular music with about 10dB of Crest Factor, the only time your amplifier will produce full unclipped power is when the source material has a brief segment at all bits high, AND, you have your volume control wide open. The rest of the time, you will not even approach the amps full power potential. This is absolutely fine if you have a huge amplifier and all the volume you need. But for most of us, we have to make do with small amps and need to maximize the performance we get for the dollars spent. Because of this, we use a trick called Gain Overlap.

Gain Overlap allows you to make use of much more of your amps power potential, but it does also allow the amp to briefly go into clipping on the all high bits portions of the musical signal. These portions are typically drum beats or other very brief notes that will not cause any damage to your system because they happen only for very short periods of time. But, a system set up with 10dB of gain overlap will sound 10 times louder at the same volume setting as a system with 0 gain overlap!
This is not something that was invented in car/bike audio either... Gain overlap originated in the early 1960's when amplified rock and roll music first became popular. The concert venues soon got way too big for the available amplifier power, and the mains power for that matter (back then concert systems used to be set up with 0 gain overlap) so a few clever sound engineers decided to simply increase the gain. Hence, gain overlap was born, and the amount of amplifiers and associated power supply were significantly reduced.

To summarize - there is nothing wrong with setting your system up with 0dB of gain overlap, except it simply won't play very loud.
If you do use gain overlap, you have to be careful to not apply too much, or speaker damage can occur. (usually after a case of beer and the volume knob at 11 for a few hours)
To perform the adjustment with differing amounts of gain overlap, you simply reduce the level of the test tone by a specific amount. There are commercially available CD's with test tones at 0dB, -5, -10, and -15dB. Google "Autosound 2000 CD-104".

I hope this helps the audiophiles in the group. I will now return to my usual lurking. :-)

For those of you who are wondering if I know what the hell I'm talking about, I have been working in the audio industry since 1976. I am an electrical and acoustical engineer, and I designed some of the products you guys are installing. I have also done some time behind a recording console, and I spent 7 years doing live sound for a variety of rock 'n roll bands. I'm also the technical editor for Performance Auto and Sound magazine.

Garry

[UltraNutZ]

Since this is a pinned subject, I thought it would be a good idea to provide a couple of corrections, and to provide a bit more insight into setting gains. I have no dog in the fight, and I absolutely mean no disrespect to any of the members who have already contributed to this thread. However, there are a couple of problems.
First of all, the procedure listed does not measure the amps bridged output, it specifically says to:
"6- Connect your DMM + (red) lead to the front right speaker output + on the amp. Connect your DMM – (black) lead to the front right speaker output - on the amp."
That is the correct connection for a single channel.
With that in mind, the target voltage stated is far too high. The amp will never reach that voltage on a single channel.
Without an oscilloscope, you are only guessing at the amps actual power, regardless of what it says in the published manual. I have measured lots of amplifiers with wildly optimistic published specs.
But if we assume the rated power is correct, then the target voltage using Todd's method would be 18.73 volts, just as pp6000v2 suggested.

absolutely no disrespect taken or assumed.  You're a hell of a lot smarter with this stuff than I am.  What folks need to understand too is simply this;  These are not "Todd's" statements or calculations; simply what we've used for many a years, created and tested by others.  I simply know what it is, how to correlate this info to the typical end user who doesn't want to bother himself with all the technical mumbo jumbo.

Now for the "other" stuff... the test tone suggested indicates it's recorded at 0dB. In recording engineer speak, that means "maximum unclipped output" or sometimes we call it "all bits high".
This is fine if you want to set up a system for maximum dynamic range, and the best possible signal to noise performance. It does not however, work well if you have limited amplifier power.

Music is dynamic, and has higher and lower volumes throughout any given piece of recorded music. If you were to take the AVERAGE of the levels on a recording, and compare them to the maximum level possible, you would find that the actual average level is significantly lower than the 0dB tone. It has to be, otherwise every sound in the music would be exactly the same volume.
This difference between the AVERAGE level and the Max level is called Crest Factor.
The amount of crest factor in any specific piece of music is decided by the recording engineer. Sometimes, the engineer will want it louder, to sound "hot" when played on the radio, or if the music is usually used for dancing or clubs. A lot of Rap, Hip-Hop, and Grunge was recorded with about 5dB of Crest Factor. More mainstream music, Pop, Rock, and Country for example, is typically recorded with about 10dB of Crest Factor. Classical music, Jazz, and audiophile recordings can have as much as 15-20dB of Crest Factor.

Have you ever wondered why some recordings are a lot louder or quieter than others, even though you didn't touch the volume control? It's because of the amount of Crest Factor. The less Crest Factor the recording has, the louder the average level will be. If you are still with me, now I'll tell you what this has to do with setting the gains on your amplifiers.

100% accurate statements, yes sir but what you must keep in mind is this is generalization and we're talking about motorcycles here.  If I'm putting together a car or home audio piece where audiophile quality sound is requested/required (and proper equipment purchased) there is a lot of testing going on during the install with many different variations of -10db to +20db test tones of varying frequencies utilizing an audio generator as well as oscope, SPL meter, and mics.

If we were to measure the amps power in dB instead of watts, or simply provide a correlation between the two, you would see that for an amp to seem twice as loud, you need 6dB more power. But, 6dB means you would have to quadruple the watts from your amp to make it twice as loud. For example, if you had a 25 watt per channel amp, for the perception of twice as loud you need 100W per channel. For a 10dB increase, you would need 250 watts per channel.

another very accurate statement.

So, back to Crest Factor and gains. If you listen to popular music with about 10dB of Crest Factor, the only time your amplifier will produce full unclipped power is when the source material has a brief segment at all bits high, AND, you have your volume control wide open. The rest of the time, you will not even approach the amps full power potential. This is absolutely fine if you have a huge amplifier and all the volume you need. But for most of us, we have to make do with small amps and need to maximize the performance we get for the dollars spent. Because of this, we use a trick called Gain Overlap.

Gain Overlap allows you to make use of much more of your amps power potential, but it does also allow the amp to briefly go into clipping on the all high bits portions of the musical signal. These portions are typically drum beats or other very brief notes that will not cause any damage to your system because they happen only for very short periods of time. But, a system set up with 10dB of gain overlap will sound 10 times louder at the same volume setting as a system with 0 gain overlap!
This is not something that was invented in car/bike audio either... Gain overlap originated in the early 1960's when amplified rock and roll music first became popular. The concert venues soon got way too big for the available amplifier power, and the mains power for that matter (back then concert systems used to be set up with 0 gain overlap) so a few clever sound engineers decided to simply increase the gain. Hence, gain overlap was born, and the amount of amplifiers and associated power supply were significantly reduced.

I 100% agree but will also state, that small amps have come a LONG way since way back when I was doing this professionally.  Back then to get 100w of pure unadulterated power from an amp you were forced to the Class A or B amps with heat sinks the size of a small computer and 20db noise level fans to keep them cool.  Now we're talking Class D, G, H, Boosted Rail, and many others that are FAR more efficient than some of their predecessors but also one must keep in mind, you get what you pay for.  I think we all get the fact that paying $99 for a Power Acoustik or Soundstream amplifier rated at 400x4 is a little far fetched at best but nonetheless, IMO and IME those types of amplifiers produce or allow enough head room to make for excellent choices within the motorsports industry.

To summarize - there is nothing wrong with setting your system up with 0dB of gain overlap, except it simply won't play very loud.
If you do use gain overlap, you have to be careful to not apply too much, or speaker damage can occur. (usually after a case of beer and the volume knob at 11 for a few hours)
To perform the adjustment with differing amounts of gain overlap, you simply reduce the level of the test tone by a specific amount. There are commercially available CD's with test tones at 0dB, -5, -10, and -15dB. Google "Autosound 2000 CD-104".

I hope this helps the audiophiles in the group. I will now return to my usual lurking. :-)

For those of you who are wondering if I know what the hell I'm talking about, I have been working in the audio industry since 1976. I am an electrical and acoustical engineer, and I designed some of the products you guys are installing. I have also done some time behind a recording console, and I spent 7 years doing live sound for a variety of rock 'n roll bands. I'm also the technical editor for Performance Auto and Sound magazine.

Garry

Garry your input on subjects like this is invaluable sir.  Thank you for taking the time to do so because even old hard-headed farts like me learn something from people like you every day.