Unclipped Amp Output Database

[Eric D]

I have been thinking it would be nice to have a database of the unclipped output for different amps, so then people could have some guidelines as far as gain setting is concerned, or the headroom of the amp.

We would probably need to specify the details of what is a valid test.

For example, attached is a photo of a test I ran for a MS250. I played a 0db 1kHz sine wave, and measured it for clipping with my o-scope. I then checked the true RMS voltage with my Fluke DMM, since my 10-1 scope probe is not calibrated, I had to use the DMM alone.

The MS250 is good for about 17.5V a channel unloaded.

[scottn29]

How does the 10-1 probe work? I got one of these scopes and dont exactly know how to use it or if it will even be good enough, I was told I need a 10-1 probe since its input was 50vpp

Here it is:

http://gizmodo.com/5252774/diy-pocket+s ... kit-for-33

[Bfowler]

thanks eric! thats a really good idea. i dont have the equipment to do that....but thats still awesome!

[PGjunkie]

Eric, this is interesting. What is the function of the little black box behind the DMM? The test tone is from the hu or your Dell laptop?

[andy600rr]

So, correct me if I'm wrong.....

with P = I x V

&

V = I x R

Power output into a 4 ohm load would work out to be 76.5 watts??

[Eric D]

So, correct me if I'm wrong.....

with P = I x V

&

V = I x R

Power output into a 4 ohm load would work out to be 76.5 watts??

Yep. But, that assumes the amp can sustain this level with a load on it. I tested it with no load, which is very easy for the amp to do. Some people test fully loaded, but I think that is crap. Music is not a constant full power tone, it jumps around. If you test an amp loaded, then you throw away all the headroom it has. Then if you set your gains to the loaded unclipped voltage, your stereo will be very quiet. I test no load, and assume the amp can keep up with music. Personally once installed in my system, I will then connect the o-scope and see if I can catch some clipping while music is playing. If I do, I back the gains off a bit. I have seldom had to back the gains off a bit on a PG amp, they handle dynamics well.

[Eric D]

Eric, this is interesting. What is the function of the little black box behind the DMM? The test tone is from the hu or your Dell laptop?

The little black box is my 10-1 voltage divider circuit, so my o-scope won't be overloaded. The test tone is a 1kHz 0db tone played off a CD on the Kenwood head unit.

[Eric D]

How does the 10-1 probe work? I got one of these scopes and dont exactly know how to use it or if it will even be good enough, I was told I need a 10-1 probe since its input was 50vpp

Here it is:

http://gizmodo.com/5252774/diy-pocket+s ... kit-for-33

Well, I can think of two ways to make a 10-1. One would be a voltage divider circuit, which is what I did. You set up a resistance ratio of 10-1 and then it will cut the input signal by a factor of 10. I have found some drawbacks to this though. To low of a resistance puts a load on the amp, and the resistors may cook. Too high a resistance and the input impedance of the o-scope will be in parallel with one leg of the voltage divider which will throw off the ratio.

The other way would be a 10-1 transformer, and I think some (maybe all, I really don't know) commercial o-scope probes are done this way. The beauty of this is you get full signal isolation, which will keep your scope from frying something, but the drawback would be no DC measurements.

[ttocs]

So, correct me if I'm wrong.....

with P = I x V

&

V = I x R

Power output into a 4 ohm load would work out to be 76.5 watts??

This is ac voltage we are talking about so are you talking about the peak voltage, peak-to-peak voltage or the rms voltage?

Am I really still the only on that thinks its a bad idea to tune a system with a meter and not a scope?

[fuzzysnuggleduck]

So, correct me if I'm wrong.....

with P = I x V

&

V = I x R

Power output into a 4 ohm load would work out to be 76.5 watts??

This is ac voltage we are talking about so are you talking about the peak voltage, peak-to-peak voltage or the rms voltage?

Am I really still the only on that thinks its a bad idea to tune a system with a meter and not a scope?

Since Eric is using a TrueRMS meter, the VAC is measured in RMS.

[lopezi]

So, correct me if I'm wrong.....

with P = I x V

&

V = I x R

Power output into a 4 ohm load would work out to be 76.5 watts??

This is ac voltage we are talking about so are you talking about the peak voltage, peak-to-peak voltage or the rms voltage?

Am I really still the only on that thinks its a bad idea to tune a system with a meter and not a scope?

He mentioned that his tests were a combination of scope (to make sure it's an unclipped signal) and then a meter because his scope probe wasn't fully calibrated....unless I missed something.

[Eric D]

I use the o-scope to find the unclipped voltage visually, then I measure it with my True RMS DMM to get an accurate RMS voltage level. Later if I am installing the amp in a car, so long as I know the signal to the amplifier is not clipped, I can set the gain on the amp using just the DMM, and adjusting until I get the voltage I measured before.

Now with anything there are glitches to this, like for instance my power supply is 13.8V, and my car might be 12.6V, or 14.4V if it is on. But, I personally think a good unregulated amp design will track clipping with input voltage.

[ttocs]

no you guys misunderstood me. An o-scope is the only tool that should be used as it shows what it actually putting out.

Your trying to measure the signal to keep its peak from clipping with a meter reading in rms voltage. Not only is the point at which it is clipped 30% higher they what the meter shows but once it goes into clipping the rms voltage will only make a very small rise.

On top of this have you seen how far off a meter can be when its battery starts to slip only a little? When did you replace yours last?

Now factor in that most people are not going to be using a very good quality meter(your stuff looks fine, I am not saying your meter is junk), that has probably not been calibrated and lord knows no one is doing the proper sequence to ensure the meter is reading correctly before testing.

I think all of these factors when taken into account is nothing more then setting a bunch of people up to fail.......

Whats wrong with yer ears when tuning equipment people? If it sounds bad, quit doin it.......

[Eric D]

My o-scope is not useful to give me a number, it is only useful to give me a visual of the signal. This is because my 10-1 divider is not actually 10-1. So, I have to use my DMM to get a number.

Since I got 17.64v RMS on my meter, if I installed this amp in a car and turned the gains up until I hit 17.5V, I would be before clipping, and the 0.14V would not be audible anyway.

I can't comment on other meters, but I know mine does not change when the battery voltage falls. It is consistent up until it actually is so dead it shuts off. I am confident any other good meter is going to give the same reading. If meters were not consistent, we would have some serious problems in this world designing and testing products.

Maybe we need to check our meters. We can stick them in the wall outlet and see if we get 120V AC. I can measure amps with 60Hz tones instead, and that should even allow cheap meters to be used as a tool.

I take it you like to use your ears, which is fine. I personally cannot hear the beginnings of clipping as well as I can see it on my scope. Additionally, by using a 0db tone, I know I am setting up my amp for the worst case scenario. I would hate to use music and then later play another song which is recorded higher and then clip my amp and further risk breaking something.

[ttocs]

why can't your scope give you a number? your using a bad lead 10-1 to measure it? How do you know its bad, how do you know your meters leads are good?

My last job was in industrial maint for a large semiconductor manf and one of the resposabilities I had for a time was calibration technition. If you have never had a tool taken to be calibrated then you will never have any idea how far off "calibrated" and high tech tools can drift from normal use, or from even just sitting around. Have you ever cleaned the leads of your meter? You would be suprised what a small amount of resistance can do in there, we threw out alot of leads that had accumulated resistance.......

I am really not sure why you guys can't hear when the signal starts to go/sound bad but are sure that when its at 17.5v that you are golden? I don't have dog ears but is it really that hard to hear when it sounds bad? I have never used a meter to tune an amp/speaker and also never blown a speaker....

I am really sorry to piss on your post and I will drop it. I just bite my lip everytime this gets brought up and its been bleedin a while. If you guys like a database then please feel free to delete my posts if you want.

[smgreen20]

My question to a lot of people, on all forums, who cares/what does it matter if some one uses their ears or a DMM? IMO a DMM is a way of "feeling" or knowing that the system is set OK w/o worry.

I for one, am for a database.

So now what Eric, send all our amps to you for output testing???

[fuzzysnuggleduck]

Well since we can't assure the exact same testing method is used by everyone when populating such a database, we must acknowledge that the numbers posted are a factor of how the person tested and may not be perfect.

However, I still think it's a great idea!

[Eric D]

I made my own 10-1 voltage divider circuit with two resistors. It is not actually 10-1, because I used cheap resistors. If I had bought expensive precision resistors, then I could build an exact 10-1 divider. Or I could buy a very high end potentiometer and tune in my signal to 10-1 ratio. If I use the meter function of my o-scope it will give me a number. That number is wrong, because I don't have an exact 10-1 ratio. So, I use the scope to get a visual of the signal. If I had a 9-1, or 11-1 ratio it will still show me when the clipping starts, it just won't let me measure the voltage in decimal form. So, I have to use a meter instead. My o-scope can only accept an input signal of around +-14V. If I exceed this, it clips my scope, when the amp might not be clipping. So I have to use something to cut the voltage to prevent clipping.

I don't see how others cannot add data to this database. If you own an oscilloscope, you can find the max unclipped voltage your amp will do. You then measure it with your oscilloscope (assuming you can, which I cannot), or use a quality DMM to measure it. If everyone posts how they got their data with what equipment, then we have some guidance as to how to duplicate the results.

If someone comes on here and says they have a ZX450 for example, and found it to do 24V a channel, then that gives others some idea what to shoot for. If my amp can do 24V, I can set it for 20V and know I am well before clipping. 20V a channel is still 100W into 4 ohms, which is far more than the rated output of 75W a channel.

[ttocs]

just curious, at what impedence is that output then if you are unsure of your voltage dividers circuit?

[Eric D]

I don't understand what you are asking.

In regard to the voltage divider, overall DC resistance is 13.9k ohms, and the leg leading to the DMM is 996 ohms. Both of these figures were measured with my Fluke 76 DMM shown above.

So the ratio is more like 14-1 than 10-1.

[ttocs]

by looking at the wiring you have the amp bridged right? What is the impedence of the load that you are measuring the voltage of? If you give the output of an amp, you need to tell what impedence it is running at just like a cea rating I would think.

[Eric D]

I am checking just one channel, so it is not bridged. There is no load connected.

I suppose you could call the 13.9k ohm resistance a load, but it hardly has any effect on the output voltage.

CEA ratings are for power, I am not measuring power, I am measuring voltage.

[ttocs]

so your saying the voltage output on the amp will be the same even if you change the resistive load that is being used to measure it?

[Eric D]

so your saying the voltage output on the amp will be the same even if you change the resistive load that is being used to measure it?

Exactly that indeed.

However, if you lower the resistance down near that of a speaker, it will draw significant current through the amp which will pull the rail voltages down and lower the output voltage.

[ttocs]

I didn't know that since I never do this.

So at that point you know that no matter if you slap a 8 ohm stereo or a 4 ohm stereo load that it is tuned? what if you bridge it?