Gain setting with DMM

All interesting or frequently asked topics will be moved here for reference. If you have a general question, please check here before posting. You may reply to threads here, but not post new ones. Chatter and off-topic remarks will be purged. Please contact a moderator to have a thread copied here.
VW337
Meatgician
Posts: 2780
Joined: Wed Dec 06, 2006 9:53 am
Location: PORK-Land OR

Gain setting with DMM

Post by VW337 »

Find the rated power of your amplifier per channel.

Do the math to calculate the target voltage you need in order to adjust properly with your DMM.

root(wattage * speaker load) = voltage


Now, play a 60 Hz or 50 Hz 0dB sine wave, and turn your HU up to just below clipping, which should be about 3/4 - 7/8 of the way up. Set your DMM to Volts AC, and measure across the amplifier's speaker output terminals while the speakers are disconnected. Turn to gain up until you get to your calculated Volts AC.
I think we've established that "Ka Ka" and "Tukki Tukki" don't work.
User avatar
stipud
Voltage Ohms
Posts: 14719
Joined: Sat Nov 26, 1983 4:00 am
Location: Burnaby, BC
Contact:

Post by stipud »

You can get your test tones here: http://www.soundbuggy.com/Eric/Test%20Tones/index.html
Ensure that you burn them to a standard CD, not MP3 disc.


One more item to note is that with a load matching amp (e.g. Xenon), you take the highest load that the amp is rated to play at, and set your gains there. Since Xenons make full power from 1-4 ohms, make sure you calculate with your load as 4 ohms, rather than the actual impedance of your setup.
User avatar
fuzzysnuggleduck
Soy Milquetoast
Posts: 4423
Joined: Wed Dec 06, 2006 1:08 pm
Location: The best place on earth
Contact:

Post by fuzzysnuggleduck »

Question: If I'm clipping the amp, should I be able to measure DC volts at the spearker (disconnected) or not?

I read something that insinuated that some amps have ways of not letting any DC signal created by clipping actually reach the voice coil... and thus I wouldn't be able to measure any DC volts this way?

I'm a bit confused... :(

Also, would it be wrong of me to let my M44 put more power out on a bridged channel than it's rated for? I know that's exactly what your instructions say but I always hear people talking about how many of the old PG lines being far more capable than what they are actually rated for...

More confusion :(
SOLD: '91 PG 4Runner
1moreamp
NOT justonemoreamp
Posts: 2504
Joined: Sun Dec 10, 2006 8:53 pm
Location: No where special

Post by 1moreamp »

fuzzysnuggleduck wrote:Question: If I'm clipping the amp, should I be able to measure DC volts at the spearker (disconnected) or not?

I read something that insinuated that some amps have ways of not letting any DC signal created by clipping actually reach the voice coil... and thus I wouldn't be able to measure any DC volts this way?

I'm a bit confused... :(

Also, would it be wrong of me to let my M44 put more power out on a bridged channel than it's rated for? I know that's exactly what your instructions say but I always hear people talking about how many of the old PG lines being far more capable than what they are actually rated for...

More confusion :(

Some amps have DC sense protection that will shut the amp down. I have seen this sort of protection on PG amps. I will check my notes about your M-44.
Amps are not supposed to pass DC to speakers, Speakers go Poof it that happens < Only Rockford Fosgate does this by my memory.

Passing a Square wave will pass Odd and even Harmonics along with the fundamental note through to your speaker IF your amp does not have a output Zobel filter, PG does have this.
So no you not passing DC to your speaker, BUT you are passing heat generating clipped information that can and will blow speakers by overheating the voice coils < this is not a good thing>. I can't even begin to tell you how many output Zobels I have repaired over the years, But its been a lot, Whew!

A clipped square wave is not the same thing as DC. DC is DC which by its very description means Direct Current, so you would have to be passing a single polarity output, not a square wave which is a AC signal as even clipped it will still swing to either side of the zero voltage point up to the clipped rail voltage level , that makes it still a full swing from rail to rail at the fundamental frequency your testing with. No DC here, but almost as bad with the Voice coil heating it causes.

You should set you amp so it has ANY headroom left for musical dynamics IMO. IMO I would use the amp as intended, and anything extra is a quality gift from the PG God's. Your amp will live longer, and like you better for this thinking < repairmans advice :lol: >

All that Headroom PG gave you will be swallowed up by musical Dynamics, :D
gridracer
RIP
Posts: 1853
Joined: Wed Dec 20, 2006 7:19 pm

Post by gridracer »

here is a link for figuring out the AC voltage to match with your DMM
http://www.caraudio911.com/tutorials/Ga ... oltage.htm
eyesofra
Posts: 613
Joined: Wed Feb 28, 2007 11:41 pm
Location: from Cameron Highlands MY to ACT AU
Contact:

Post by eyesofra »

Came across this amazing site while googling for test tones.

This site has plenty of it for free download and they've even got some tones as wave files .


http://www.realmofexcursion.com/downloads.htm
1moreamp
NOT justonemoreamp
Posts: 2504
Joined: Sun Dec 10, 2006 8:53 pm
Location: No where special

Post by 1moreamp »

Nice sight, I use my computer on my bench with a Audigy 2 card for some of my computer based scope work. This sight will come in very handy. :wink:
User avatar
fordtough1
Twisted's Boyfriend
Posts: 1438
Joined: Fri Feb 23, 2007 7:07 pm
Location: Are we there yet?

Post by fordtough1 »

I was using this procedure to set the gains on my zx475. I have the front channel going to my comps and the rear bridged to a sub.

When setting the front channels to 17.3 volts per the calculator, I set the right side, then put it on the left side and the voltage there was 14.1 volts. Is this normal, or is something wrong with my amp? How do I go about correcting it?
VW337
Meatgician
Posts: 2780
Joined: Wed Dec 06, 2006 9:53 am
Location: PORK-Land OR

Post by VW337 »

fordtough1 wrote:I was using this procedure to set the gains on my zx475. I have the front channel going to my comps and the rear bridged to a sub.

When setting the front channels to 17.3 volts per the calculator, I set the right side, then put it on the left side and the voltage there was 14.1 volts. Is this normal, or is something wrong with my amp? How do I go about correcting it?
In this situation you would take a look at your RCA's and HU balance settings as a quick fix option. Check the voltage coming off the RCA's if it is balanced L-R then you may have a bad Gain Pot in your amp or something worse.
I think we've established that "Ka Ka" and "Tukki Tukki" don't work.
maka78
Posts: 172
Joined: Sun Sep 07, 2008 7:41 pm

Re: Gain setting with DMM

Post by maka78 »

VW337 wrote:Find the rated power of your amplifier per channel.

Do the math to calculate the target voltage you need in order to adjust properly with your DMM.

root(wattage * speaker load) = voltage


Now, play a 60 Hz or 50 Hz 0dB sine wave, and turn your HU up to just below clipping, which should be about 3/4 - 7/8 of the way up. Set your DMM to Volts AC, and measure across the amplifier's speaker output terminals while the speakers are disconnected. Turn to gain up until you get to your calculated Volts AC.
Don't you have to look at the frequency with the highest voltage, not 60 or 50 Hz? For instance, if you have an EQ on your HU peaking at 35 Hz, and have, for example, 30V @ 35 Hz and 20V @ 60 Hz, you really want to look at the 35 Hz sine wave, because if you set the 60 Hz @ 30V, you will probably have something like 40V @ 35 Hz, which would be clipping.

This makes sense to me, but please let me know if I'm wrong here.
User avatar
stipud
Voltage Ohms
Posts: 14719
Joined: Sat Nov 26, 1983 4:00 am
Location: Burnaby, BC
Contact:

Re: Gain setting with DMM

Post by stipud »

maka78 wrote:Don't you have to look at the frequency with the highest voltage, not 60 or 50 Hz? For instance, if you have an EQ on your HU peaking at 35 Hz, and have, for example, 30V @ 35 Hz and 20V @ 60 Hz, you really want to look at the 35 Hz sine wave, because if you set the 60 Hz @ 30V, you will probably have something like 40V @ 35 Hz, which would be clipping.

This makes sense to me, but please let me know if I'm wrong here.
You should only ever use an EQ to CUT frequencies, not boost them. This is for the same reasons as avoiding bass boost -- the synthetic volume addition is noisy. Also, every 3 decibels of EQ gain causes the amplifier to try to put out DOUBLE the power at that frequency. That means 6dB = four times the power! So unless your gains are set really low, you don't want to be boosting if at all possible! The point of an EQ is instead to REDUCE the volume of LOUDER frequencies, not to boost quieter frequencies. So let's say your EQ has a boost of +6dB at 1kHz, and nothing else. Instead of that, you would set 1kHz to 0dB, and everything else to -6dB! So your quietest frequency should be at 0dB, and everything louder should be beneath it. In extreme cases, you can use a bit of boost, but do so sparingly!

With this in mind, you always want to defeat your processing, crossovers and EQ before you set your amp gains with this method. The choice of 50Hz and 60Hz is simply because this is the frequency at which household electrical outlets operate. 50Hz in Europe, 60Hz in North America. Because of this, all cheaper multimeters are calibrated for this frequency, and will only give accurate voltage readings at or near it!

The exclusion to this are the much more expensive True-RMS voltmeters, which can tell you the A/C voltage, regardless of frequency. If you have one of these, you can use whatever 0dB A/C test tone you like :)
maka78
Posts: 172
Joined: Sun Sep 07, 2008 7:41 pm

Post by maka78 »

stipud, so you're saying that my DMM is not reading a 6kHz reading as accurately as a 60Hz reading? I was always under the assumption the DMM can read all frequencies, as long as it's AC.

Let's take your case, where I decrease the 60Hz frequency by 6dB, and leave the 6kHz frequency a 0dB. I then measure the 60Hz frequency and set that to the appropriate power output for the amplifier (let's say 100w). Wouldn't this mean that the 6kHz frequency is now clipping because the amplifier is trying to push it past 100w at 6kHz?
maka78
Posts: 172
Joined: Sun Sep 07, 2008 7:41 pm

Post by maka78 »

Another question, most amplifiers state their rating at 14.4V, but when you're tuning your amps, you are most likely around 13.8V at idle - wouldn't you be clipping your amp if you set a 100w @ 14.4V amp at 100w output with 13.8V input?

I forget the calculations for this right now and don't have much time to look them up, but can look them up later if interested.
User avatar
fuzzysnuggleduck
Soy Milquetoast
Posts: 4423
Joined: Wed Dec 06, 2006 1:08 pm
Location: The best place on earth
Contact:

Post by fuzzysnuggleduck »

maka78 wrote:Another question, most amplifiers state their rating at 14.4V, but when you're tuning your amps, you are most likely around 13.8V at idle - wouldn't you be clipping your amp if you set a 100w @ 14.4V amp at 100w output with 13.8V input?

I forget the calculations for this right now and don't have much time to look them up, but can look them up later if interested.
If your amps are only have ratings at 14.4V, then don't set your gains to produce the 14.4V rated wattage at a lower voltage. However, I'm not so certain that if your electrical system is putting out 13.8V at idle that it will put out any more when rev'ing... wouldn't your voltage regulator be, well, regulating that?

As stipud said, most consumer level "digital multimeters" are calibrated to measure A/C for the frequency that the nation they were sold in runs their A/C circuits at. North American 120V AC circuits run at 60Hz thus most cheaper DMMs are calibrated for 60Hz. Your multimeter is going to need to say "True RMS" or it calibrated for a specific frequency.

Stipud also mentioned that you should take these measurement with your processing defeated (off!) so if you've got your 60Hz at -6db, you won't be measuring power output with the EQ cutting that frequency anyways since the EQ should be defeated during your measurement.
SOLD: '91 PG 4Runner
maka78
Posts: 172
Joined: Sun Sep 07, 2008 7:41 pm

Post by maka78 »

to add on to stipud... I cannot enforce this enough... make sure you use your judgement as well when listening to music.

When people convert a CD track to an MP3, they might tend to digitally increase the dB, so the song sounds louder than it did before (you'll notice some songs sound louder than other songs when you have an mp3 cd or ipod on). If this is so, even being at 3/4 volume (your new max as stipud said), you still might be clipping with certain songs.
User avatar
stipud
Voltage Ohms
Posts: 14719
Joined: Sat Nov 26, 1983 4:00 am
Location: Burnaby, BC
Contact:

Post by stipud »

You can't increase volume above -0dB. Those test tones are encoded at full bandwidth, so there's no way to make anything louder. If you tried to do that, the file would become "clipped" as well. I can't think of any MP3 encoder that would let you do that however. :?
maka78
Posts: 172
Joined: Sun Sep 07, 2008 7:41 pm

Post by maka78 »

I've seen a couple of mp3 encoders that let you increase the dB to the point that it's clipping. They amplify the signal within the track. I've blown previous speaker like this because I did not know that.
User avatar
stipud
Voltage Ohms
Posts: 14719
Joined: Sat Nov 26, 1983 4:00 am
Location: Burnaby, BC
Contact:

Post by stipud »

If that's the case, even if your gains were set low you would still be playing a clipped signal straight from the MP3, so there's no way to avoid clipping in this case. Standard MP3 encoding is done at the CD reference level, so you have nothing to worry about unless you purposely try to artificially boost your MP3's. Why you would do that.... god only knows.
maka78
Posts: 172
Joined: Sun Sep 07, 2008 7:41 pm

Post by maka78 »

stipud wrote:If that's the case, even if your gains were set low you would still be playing a clipped signal straight from the MP3, so there's no way to avoid clipping in this case. Standard MP3 encoding is done at the CD reference level, so you have nothing to worry about unless you purposely try to artificially boost your MP3's. Why you would do that.... god only knows.
stipud - can you not digitally amplify a signal without clipping it right away? I am not an expert in this field (frankly, I don't even mess with any digital editing of mp3s), but from my understanding, you can amplify the mp3's signal where it would not clip at low volumes, but can start to clip at higher volumes.

Correct me if I'm wrong here, but I thought this was the reason why certain songs sound louder than others when playing an mp3 cd - yes, some are lowered when converted, but some have volume increases (such as a digital bass boost, which boosts the low frequency signal in the mp3). These volume increases might still be played without clipping, but at lower volumes on the HU. Let me know your thoughts on this.

EDIT - btw, sorry everyone for jumping off topic.
User avatar
stipud
Voltage Ohms
Posts: 14719
Joined: Sat Nov 26, 1983 4:00 am
Location: Burnaby, BC
Contact:

Post by stipud »

maka78 wrote:stipud - can you not digitally amplify a signal without clipping it right away? I am not an expert in this field (frankly, I don't even mess with any digital editing of mp3s), but from my understanding, you can amplify the mp3's signal where it would not clip at low volumes, but can start to clip at higher volumes.

Correct me if I'm wrong here, but I thought this was the reason why certain songs sound louder than others when playing an mp3 cd - yes, some are lowered when converted, but some have volume increases (such as a digital bass boost, which boosts the low frequency signal in the mp3). These volume increases might still be played without clipping, but at lower volumes on the HU. Let me know your thoughts on this.

EDIT - btw, sorry everyone for jumping off topic.
You can't amplify a digital signal beyond -0dB.

Digital recordings are done by taking samples of the A/C wave at intervals (let's say 48kHz, i.e. 48 thousand samples per second). It then records the position of the A/C signal digitally. That way it can be recomposed into music, by "connecting the dots" on the digitally recorded signal.

Digitally you record with a certain number of bits, for example CDs use 20 bits to record the data. That is 20 1's and 0's... which in binary lets you count from 0 to 1048575. This is split in half (since A/C has a positive and negative). So 0 to 524 287 represents negative A/C signal, and 524 288 to 1048675 represents the positive signal.

A -0dB signal (that is to say, 0 decibels off of maximum volume) will go from 0 to 1048575 at both peaks of the positive and negative A/C wave... so it represents the full bandwidth of the recording. If you tried to amplify the signal beyond this, you would need more bits. But you can't just increase the maximum number of bits, so it would "clip" the music.

Let's do a simple example with three bits. 0-4 would represent negative A/C, 4-7 would represent positive.

Code: Select all

             Sample
000 - 0 x              xx
001 - 1  x            x  x
010 - 2   x          x    x
011 - 3    x        x      x
-----------------------------
100 - 4     x      x
101 - 5      x    x
110 - 6       x  x
111 - 7        xx
So let's say our AC sample is very simple, 01234567. Now we want to amplify the signal, so we increase the volume of the samples. To do this, we can add 1 to positive, and subtract 1 from negative signal. This would give us 00125677. Since 7 can't increase above 7 (would need more bits) it is "clipped" and remains at 7. 0 can't drop below 0, so it is "clipped" and remains at 0. Now let's "amplify" it one step further. 00016777. And one step further... 00007777. Finally, we have a fully "clipped" signal data. A "square wave", with no roll off at either end, completely maxed out at positive and negative.

When you record a CD to your computer in it's raw data, you have a 20 bit sample every ~48kHz. That means you have 20*48000 = 960000 bits per second, or 960 kilobits per second. MP3 usually records at 128 kilobits per second, in order to shrink the file down. So as you can obviously figure out, there is less accuracy in the MP3 data than in the raw CD data. This is why MP3s sound worse than CDs!

So, even in the digital realm, we have the same sort of limitations that you have in an amplifier. Digitally you can only record so much signal, and beyond that it gets "clipped", just like an amp can only produce so much voltage before it clips.
Last edited by stipud on Fri Sep 26, 2008 10:19 am, edited 3 times in total.
User avatar
HoseHead
Tim Horton Himself
Posts: 2262
Joined: Wed Dec 06, 2006 4:49 pm
Location: Ottawa, Ontario, Canada
Contact:

Post by HoseHead »

Smarty pants. :shock:

Well explained.

HH
The only stupid question is the one not asked .......
User avatar
stipud
Voltage Ohms
Posts: 14719
Joined: Sat Nov 26, 1983 4:00 am
Location: Burnaby, BC
Contact:

Post by stipud »

HoseHead wrote:Smarty pants. :shock:

Well explained.

HH
Thanks.

Here's a picture of a monkey riding a dog.

Image
maka78
Posts: 172
Joined: Sun Sep 07, 2008 7:41 pm

Post by maka78 »

lmao, what's up with the dog and monkey?
User avatar
stipud
Voltage Ohms
Posts: 14719
Joined: Sat Nov 26, 1983 4:00 am
Location: Burnaby, BC
Contact:

Post by stipud »

maka78 wrote:lmao, what's up with the dog and monkey?

Code: Select all

http://en.wikipedia.org/wiki/Non_sequitur_(humor)
A non sequitur (pronounced [ˌnɒnˈsɛkwɨtɚ]) is a conversational and literary device, often used for comical purposes (as opposed to its use in formal logic). It is a comment which, due to its lack of meaning relative to the comment it follows, is absurd to the point of being humorous or confusing. Its use can be deliberate or unintentional. Literally, it is Latin for "it does not follow." In other literature, a non sequitur can denote an abrupt, illogical, unexpected or absurd turn of plot or dialogue not normally associated with or appropriate to that preceding it.

Non sequiturs often appear to be disconnected or random comments, or random changes in subject, especially socially inappropriate ones. When non sequiturs are used frequently for comic effect this can be called "absurd humor".

The non sequitur can be understood as the converse of cliché. Traditional comedy and drama can depend on the ritualization and predictability of human emotional experiences, where the Theatre of the Absurd uses disjunction and unpredictability.
User avatar
Bfowler
Briaans..... BRIAAAAANNNNNNS
Posts: 10764
Joined: Wed Dec 06, 2006 11:06 am
Location: So easy, a cavewomen could do him

Post by Bfowler »

stipud wrote:Here's a picture of a monkey riding a dog.
yeah..i saw them live in mollala... :wink:

that was a good read btw. learned lots
my ex-girlfriend said "its car audio or me"
i've had tougher choices at a soda machine...
Post Reply