PG Amplifier Gain Testing Experiment

[Eric D]

In another thread there was a discussion related to amp gain and signal voltage. In an effort to clarify things, I did a bit of testing today and posted my results on a ZX500v2.

http://www.soundbuggy.com/Eric/Car%20Au ... index.html

Please let me know your thoughts, and I plan to perform a similar test on some other PG (and other brand) amps when I get a chance for comparison purposes.

Note: if you view that page and it is just two photos with minimal text, I have been having issues getting FrontPage to update my web space, and I have no idea what is going on. Occasionally the site is correct, then it changes randomly on me.

[kg1961]

Great info eric I have always loved to see this thanks

[ajaye]

Eric,

I'm not sure if this is in response to any posts I've made but here's my thoughts since you asked for them. As a forward: I don't dispute your results or methodology, I never have. In my OPINION, the test doesn't matter all that much TO ME. I don't have test equipment in my car when I'm listening to my stereo. I have a level I like for normal listening, and a level for when I want to hear it loud. For just about every change in source (the CD/album, or the one the digital file originally came from) there's usually at least a small adjustment of the volume knob needed to maintain the same level. There are MANY recordings out there that will never drive the preamp signal voltage to its maximum output. I don't care how high you have the volume knob, if the signal peaks 3db down on the CD and the head unit pushes "X" volts peak at 0db, and you don't have the ability to make up for it in another gain stage somewhere along your signal chain, you will never drive the amp to rated power with the input gain also set to "X" volts. Your options are set the amp more sensitive or make up the gain somewhere further up the signal chain. I think you will agree, in general, best practice is to apply as much gain as needed as close to the source as possible. This is what yields the best results for me. You need to be willing to assume a certain level of responsibility with the volume control with this method, yes. I will venture to assume that I have an above average understanding of recorded audio practices vs the average person so maybe I'm not explaining my point clearly enough. No disrespect intended, just my $.02.

I'm happy to delve further into specifics if someone wants to see examples or isn't understanding.

[ajaye]

Just to clarify, I appreciate the work you put into testing stuff and putting good info up for everyone and it is certainly helpful to people, myself included. In this particular case I just consider a few factors including the one you're demonstrating. Nice work nonetheless.

[Mr. Wild]

Good demonstration Eric! Gain setting is almost always too difficult to get just right without test gear.

ajaye wrote about gain overlap. Some decks have this "built in". I use a Kenwood KDC 6051 which reaches maximum signal level at volume setting 33/35 so there is a risk of sending a clipped signal to the amps. But on the other hand some recordings aren't at full level so you can just keep turning the volume knob to get closer to full power.

[Eric D]

Posting this information is not in any direct response to any one person, and the results actually debunk some of my own assumptions. I was pretty confident the ZX series would not handle more than 4V at minimum gain, but this was shown to be wrong, as they (or at least this amp) handles much more.

Even without test equipment everyone can gain some insight from what I did here. You can turn your gain on your amp all the way down, and then play the most dynamic audio track you think you have. You should very easily be able to determine the point where your head unit begins to clip. Once you find this point, back off slightly, and make note of this level as the maximum level you can go.

Now increase the gain on your amplifier with the head unit at this max point, and you can adjust for the maximum signal-to-noise prior to clipping (audible distortion) from the amplifier.

Mr. Wild, I highly suggest you give a Kenwood eXcelon head unit a try. I don't know if the modern ones still do this, but I have found all older models to not clip, and pretty much all non-eXcelon models do clip. I have a newer Kenwood in my daily driver which has a ton of on board processing. This needs to be disabled for the unit to have a clean output. With the signal processing on it clips easily. I would not doubt that by now though, all units do not allow you to disable the processing anymore.

[ttocs]

good work I too have been bothered that people think they have to have them or get more power with them as well. We need to figure out a way to introduce some noise to the signal to show what happens when you install it in the rear of the vehicle.

[Eric D]

Look at my last two screen shots off the scope. Channel 1 is the RCA input signal to the amp, and in the case of max gain (the last chart), you can easily see how much thicker (more noise on it) the signal is. Keep in mind music would be a much smaller signal than this 0db sine wave. So, in the last example I had only 250mV of signal, with music the noise would easily overtake the lower levels in the recording.

[ttocs]

Do you think if you put a relay next to the RCA signal wires and powered it up/down that it could induce a voltage in the signal that could be seen?

[Eric D]

I think I would start with my wife's hairdryer and see how it goes...

[ttocs]

that would definitely make some noise but I bet even a relay on/off could be picked up by the scope.

[groundpounder]

Yes nice demonstration Eric. It seams to me that you have Prooven the point of how the input signal voltage value effects the signal to noise ratio. And also how it effects the peak output of an amp, along with propper sequence for setting the gains when useing a line driver. I would also like to add that the amp will run with less effort and cooler with increased line signal voltage.

[Eric D]

Why would an amp run with less effort or cooler with more line signal voltage? I am not disagreeing with you, I just would like to know your reasoning on this.

If you are referring to the supply voltage to the amp (like 14.4V vs 12.6V), the amplifier would be more efficient at 14.4V than it would be at 12.6V.

[groundpounder]

Im sorry Eric. I ment to say with the amps gain turned down it will run cooler and with less effort. Whitch can help to be achieved by increasing the line signal voltage.

[kg1961]

Im not understanding if the is not clipping and making the same output power how would it run cooler? if the gain is turned up or down? its fan cooled

[ajaye]

Why would an amp run with less effort or cooler with more line signal voltage? I am not disagreeing with you, I just would like to know your reasoning on this.

If you are referring to the supply voltage to the amp (like 14.4V vs 12.6V), the amplifier would be more efficient at 14.4V than it would be at 12.6V.

This has been something I've been trying to figure out recently. I'm not as knowledgeable with the discrete functions of amplifiers' circuitry as a lot of you guys. In my basic understanding, if the power gain of an amp is a finite amount, then the output power it can achieve is determined by the strenght of the input signal and the strength of that signal once the gain the amp is capable of is applied. Starting with a stronger input signal means the amp will only need to apply some fraction of the maximum gain it is capable of to achieve maximum output. Less work (power gain) = less heat.

Also a guess, but I imagine the load on the amp would be as much a factor, if not more, depending on the amp's specific design as that will determine the voltage gain vs the current gain needed.

[Eric D]

It is true that an amplifier is a fixed source of gain. The gain control on the amplifier is to match its input to its preamp section.

An amplifier can only do so much output, before it starts to clip. This is regardless of the input gain or the input signal strength.

Say for example we have an amp which has a gain of 10. For every 1V of input we get 10V of output. To keep numbers simple, its maximum output is 10V prior to clipping.

The gain control on the amplifier generally cuts the input signal prior to the preamp, but it can boost it some depending on the design.

So, if we take our example amp, it would need 1V of signal at its preamp to achieve its 10V of output (10-1 ratio). It may be a design such that with the gain control at its minimum, 10V of input (from a head unit) will provide 1V to the preamp, and with gain at maximum, 1V will provide 1V of signal to the preamp section.

So in this example, 10V with the gain all the way down gets us 10V of output to the speaker, and 1V with the gain all the way up also gets us 10V to the speaker. In either case the gain of the amplifier output section is the same, and the output power of the amp is therefore the same.

The benefit to the 10V signal over the 1V signal is the improvement in signal-to-noise ratio, as with 10V a lot less noise will be heard from the speakers.

[ajaye]

It is true that an amplifier is a fixed source of gain. The gain control on the amplifier is to match its input to its preamp section.

An amplifier can only do so much output, before it starts to clip. This is regardless of the input gain or the input signal strength.

Say for example we have an amp which has a gain of 10. For every 1V of input we get 10V of output. To keep numbers simple, its maximum output is 10V prior to clipping.

The gain control on the amplifier generally cuts the input signal prior to the preamp, but it can boost it some depending on the design.

So, if we take our example amp, it would need 1V of signal at its preamp to achieve its 10V of output (10-1 ratio). It may be a design such that with the gain control at its minimum, 10V of input (from a head unit) will provide 1V to the preamp, and with gain at maximum, 1V will provide 1V of signal to the preamp section.

So in this example, 10V with the gain all the way down gets us 10V of output to the speaker, and 1V with the gain all the way up also gets us 10V to the speaker. In either case the gain of the amplifier output section is the same, and the output power of the amp is therefore the same.

The benefit to the 10V signal over the 1V signal is the improvement in signal-to-noise ratio, as with 10V a lot less noise will be heard from the speakers.

That makes sense thanks.

[zeropoint0.5]

i really like this test!!!! i'm very impressed about the results...........

thank you Eric for sharing it.......

in a simular test, would love to see the results of a zpa0.5 and original ms2250.....

when the pld1 reached full output, was the red peak led already flashing up or not......?

would love to see the same test on the output of a oldskool alpine 7939r........

at what volume level it would start to clip..........


but of course a 1Khz test tone is not the same as listening to music.....

and some cd's are recorded on a different volume, so what is the right test to set you're gains then.....

Mostly i set my gains based on listening to multiple cd's to make sure i never have a clipped signal.......

[Eric D]

On my PLD-1 the peak LED turns on at 4.85V RMS. It is on for all voltage above this level, so yes, it is on at max output of 9.4V RMS.

If you were listening to music on a PLD-1 and the light were flashing with the music, that would mean every flash is 4.85V or more.

Yes it is true that music is lower in output than sine waves are. However, I have set every one of my systems with 0db sine waves, and no clipping, all of which were louder than I could tolerate when set this way.

If someone were running 25W x 2 to their components, and 25W x 2 to their subs, then chances are a 0db gain overlap setup would not be for them. Their factory stereo might be louder. But, if someone has a large amp for their components and for their subs, 0db means it cannot clip, and it will still be very loud.

[zeropoint0.5]

i was always told, and thats for all line driver and also the bigger ms amps, the red peak led should never start to flash...

cause then you got a clipped signal.....so maybe this isn't true...............

[trickyricky]

A quick flash (where talking about milliseconds here) every once in a while wouldn't hurt it. Now if its flashing or staying lite for long periods of time, then that's another story.

[ajaye]

A quick flash (where talking about milliseconds here) every once in a while wouldn't hurt it. Now if its flashing or staying lite for long periods of time, then that's another story.

If you watch most people working a sound board either at a live show or in the studio, the meters will always just barely be blipping into the red on the loudest peaks. This is with analog of course, but yeah that is definitely common practice to wring every bit of loudness out of the speakers/recording without any actual audible distortion or damage to speakers. If the signal is being digitally processed/amplified, different story, you never want to exceed or even reach 0db as its impossible for a digital representation of any audio at or over 0db to exist in theory. An analog amp will still play a clipped signal it just will not be played linearly and if that's happening for fractions of a second you won't hear the difference unless you are a freak of nature or delusional.

[groundpounder]

Thanks for the lessons Eric.

[freshkryp69]

Excellent info Eric! As with running balanced, 1 advantage is the higher the V the better S+N.. I think my outputs are 15v or up to, havent measured it, Denon Z1. There seems to be a difference when I have run rca's vs the balanced. 4v vs 15v.. I had modded the nt's to accept rca's before I ran balanced.. I stuck with the balanced output's ran in direct mode, no bass & treble.. I think I got lucky, no eq.. Did have a AC dqxs but didnt really like it, no balanced outputs.. Been looking at the new zapco & Ground Zero dsp's, but really want something as high end as the HU as I dont want to loose my SQ to equipment down the line being not up to par..