IPR2 7500 noise floor question for users

john lutz

john lutz

Freshman
Jan 17, 2011
101
0
16
near Grand Rapids michigan
I am trying to find out what is actually normal or typical for a fully functioning unit. Anyone who has these amps care to comment on any self noise that may or may not be present at the outputs?

No inputs connected, no gain on front panel, and any typical monitor speaker plugged in.

Are both channels virtually silent? Are both the same? Or can you hear a small pitch shifting hum in one or both channels?

Thanks for any help on this.
 
Re: IPR2 7500 noise floor question for users

Seriously.. the noise output will depend on the sensitivity of the loudspeaker, perception will vary with ambient conditions.

You need to figure out how to put a meter on the output.

If careful you could feed it into a console input an look at those meters with some nominal gain. At least you could compare different amps to each other.

Of course do not play speaker level sound into a line or mic input.

JR
 
Re: IPR2 7500 noise floor question for users

Measuring an IPR2 7500 using Elenco LCM-1950 directly on outputs - I am reading .7mV to .8mV AC pulsating from channel A. Channel B, which has virtually no noise reads .3mV pretty steady.

For comparison measuring an IPR3000, I am reading .2 mV steady off both channels and 0mV to .1mV off both channels of a RMX2450. Both these amps are practically silent.

My test speaker is JBL MRX512m with published sensitivity of 97dB SPL 1W/1M. Anyone with normal hearing can hear noise at 1 meter......I have to get a little closer.
 
Re: IPR2 7500 noise floor question for users

john lutz said:
Measuring an IPR2 7500 using Elenco LCM-1950 directly on outputs - I am reading .7mV to .8mV AC pulsating from channel A. Channel B, which has virtually no noise reads .3mV pretty steady.

For comparison measuring an IPR3000, I am reading .2 mV steady off both channels and 0mV to .1mV off both channels of a RMX2450. Both these amps are practically silent.

My test speaker is JBL MRX512m with published sensitivity of 97dB SPL 1W/1M. Anyone with normal hearing can hear noise at 1 meter......I have to get a little closer.
Click to expand...

0.8mV is approx. 70dB down from 1W (assuming a 8 ohm load). This puts the noise floor at approx 27dB at 1m, which should be perfectly fine in applications where you need amplification.
 
Re: IPR2 7500 noise floor question for users

john lutz said:
So if I am using 2 of the same speaker then my noise floor is 30dB?
Click to expand...
I am not sure what you are referring to, but whatever it is, the answer is no.
OK EDIT
I was just thinking about my answer.

30dB (as stated in the question) is a meaningless number. dB is a RATIO-that is all. Without some sort of other designator who know what it means.

For example, 2 different uses of dB have been used in this thread. One is dB SPL (which I think the question is referring to) and the dB difference between 2 different voltages.

I was thinking in terms of the dB voltage difference, while I realized the question was about dB SPL.

Hence the reason for the "little things" that DO make a difference.

And to "reanswer" the question.

If the voltage stays the same, then the noise level will increase UP TO 6dB at some some freq. At other freq it will be lower because of the cancellation of multiple devices.
 
Last edited:
Re: IPR2 7500 noise floor question for users

Ivan Beaver said:
I am not sure what you are referring to, but whatever it is, the answer is no.
OK EDIT
I was just thinking about my answer.

30dB (as stated in the question) is a meaningless number. dB is a RATIO-that is all. Without some sort of other designator who know what it means.

For example, 2 different uses of dB have been used in this thread. One is dB SPL (which I think the question is referring to) and the dB difference between 2 different voltages.

I was thinking in terms of the dB voltage difference, while I realized the question was about dB SPL.

Hence the reason for the "little things" that DO make a difference.

And to "reanswer" the question.

If the voltage stays the same, then the noise level will increase UP TO 6dB at some some freq. At other freq it will be lower because of the cancellation of multiple devices.
Click to expand...

Read Rob Timmerman's response, above.
 
Re: IPR2 7500 noise floor question for users

Appreciate the help. Forgive me as I attempt to grasp what is probably basic for most here. I understood Rob's statement to mean that with noise at .8mV through the speaker given - assuming 8 ohms - my noise floor would be approx 27dB at one meter. Also that .8mV is 70dB down from the 2.83V 1W/1M.

What I then wondered (and do not yet know how to calculate for myself) what is my noise floor using two of same speakers.... so 4 ohms nominal.

If I understand what Ivan is saying.... 2.83V into 4 ohms the noise floor goes up depending on how speakers are positioned but at 1W/1M and 4 ohms - 2V - noise floor is approx the same at 27dB ?

Decades ago such a low noise floor would have been a miracle for me. Today though, I sometimes wonder if my system is turned on. At any rate, I'm not sure .8mV of gak is something I want to just accept given that it seems to be a random problem of variable degree with these amps.
 
Re: IPR2 7500 noise floor question for users

Rob Timmerman said:
0.8mV is approx. 70dB down from 1W (assuming a 8 ohm load). This puts the noise floor at approx 27dB at 1m, which should be perfectly fine in applications where you need amplification.
Click to expand...

Is the .7mV to .8mV hum I am measuring consistent with the published spec : Hum and Noise -106db, "A" weighted referenced to rated power @ 4 ohms. ?

Rated Watts 2ch x 4 OHMS: 2800 watts 20ms repetitive burst / 2450 watts 1% THD / 2020 watts 0.15% THD, both channels driven @ 1kHz
 
Last edited:
Re: IPR2 7500 noise floor question for users

Latest answer - Peavey is unable or unwilling to correct the problem and states that the noise is normal for this amp and amp is within spec. With assistance I have:

20 Log 98.99 / 0.0008 =

20 Log 123737.5

Log 123737.5 = 5.0925

20 x 5.0925 = (-)101.8 dB

The specs say:

Crosstalk > -85dB @ 1kHz @ 1000 watts power @ 8 ohms.
Hum and Noise > -106dB, “A” weighted referenced to rated power @ 4 ohms.

So technically this noise is +4.2 dB above the rated Hum and Noise specification................not up to the rated S/N spec...............not what I expected to find.
 
Re: IPR2 7500 noise floor question for users

john lutz said:
Latest answer - Peavey is unable or unwilling to correct the problem and states that the noise is normal for this amp and amp is within spec. With assistance I have:

20 Log 98.99 / 0.0008 =

20 Log 123737.5

Log 123737.5 = 5.0925

20 x 5.0925 = (-)101.8 dB

The specs say:

Crosstalk > -85dB @ 1kHz @ 1000 watts power @ 8 ohms.
Hum and Noise > -106dB, “A” weighted referenced to rated power @ 4 ohms.

So technically this noise is +4.2 dB above the rated Hum and Noise specification................not up to the rated S/N spec...............not what I expected to find.
Click to expand...

I haven't seen anywhere in this thread where you have explained how you are accounting for the A weighting of the noise curve, or any mention of the spectral content of the noise. When Peavey lists the noise spec at -106dB A weighted, they mean A weighted.

A straight voltage measurement without any frequency related specifications is pretty meaningless. What is the frequency response of your meter?

Mac
 
Re: IPR2 7500 noise floor question for users

Yup "A" weighting applies a Fletcher-Munson weighting curve to the noise to more accurately reflect human hearing. "A" weighted will discount HF and LF noise, so unweighted noise measurements will always be higher, especially if they include noise above 20 kHz (that we can not hear at all).

JR
 
Re: IPR2 7500 noise floor question for users

john lutz said:
Latest answer - Peavey is unable or unwilling to correct the problem and states that the noise is normal for this amp and amp is within spec. With assistance I have:

20 Log 98.99 / 0.0008 =

20 Log 123737.5

Log 123737.5 = 5.0925

20 x 5.0925 = (-)101.8 dB

The specs say:

Crosstalk > -85dB @ 1kHz @ 1000 watts power @ 8 ohms.
Hum and Noise > -106dB, “A” weighted referenced to rated power @ 4 ohms.

So technically this noise is +4.2 dB above the rated Hum and Noise specification................not up to the rated S/N spec...............not what I expected to find.
Click to expand...

I get slightly different results using the 8 ohm (not current-limited) values; putting the noise floor at around 103dB down from the maximum output. A SNR of >100dB is more than enough in practice (that's threshold of pain to the background noise in a concert hall), and even a 6dB difference is pretty easy to account for with the A-weighting curve, which is down more than that at 120Hz (without doing any math, it looks like the 6dB down points on the A-weighting curve is ~400Hz on the low end, and 15kHz on the high end).
 
Re: IPR2 7500 noise floor question for users

John Roberts said:
Yup "A" weighting applies a Fletcher-Munson weighting curve to the noise to more accurately reflect human hearing. "A" weighted will discount HF and LF noise, so unweighted noise measurements will always be higher, especially if they include noise above 20 kHz (that we can not hear at all).

JR
Click to expand...

So basically by using "A" weighting the amp can hum like crazy at 60hz and still be "ok"?
 
Re: IPR2 7500 noise floor question for users

mackerr said:
I haven't seen anywhere in this thread where you have explained how you are accounting for the A weighting of the noise curve, or any mention of the spectral content of the noise. When Peavey lists the noise spec at -106dB A weighted, they mean A weighted.

A straight voltage measurement without any frequency related specifications is pretty meaningless. What is the frequency response of your meter?

Mac
Click to expand...

The noise is around 60hz and harmonics. It oscillates a few Hz at a few Hz.

ELENCO LCM-1950
DIGITAL MULTIMETER
AC VOLTS (50Hz - 500Hz)
Ranges: 400mV, 4V, 40V, 400V, 750V
Resolution: 100mV
Accuracy: +(1.5% rdg + 4 dgts) on 400mV to 400V ranges
+(2.0% rdg + 4 dgts) on 750V ranges

Interestingly by disconecting the cooling fans the oscillation stops and the hum reduces to .5mV
 
Re: IPR2 7500 noise floor question for users

Rob Timmerman said:
I get slightly different results using the 8 ohm (not current-limited) values; putting the noise floor at around 103dB down from the maximum output. A SNR of >100dB is more than enough in practice (that's threshold of pain to the background noise in a concert hall), and even a 6dB difference is pretty easy to account for with the A-weighting curve, which is down more than that at 120Hz (without doing any math, it looks like the 6dB down points on the A-weighting curve is ~400Hz on the low end, and 15kHz on the high end).
Click to expand...

Thank you Mac, JR, Rob. I'm in over my head. I had help with the math - I don't fully understand all of it so can't argue or defend it.

Bottom line is the amp hums significantly and audibly more than any other of my "MI" grade amps, in fact the other amps barely show any noise at all. What I first thought must be a defect turns out to be a feature.
 
Re: IPR2 7500 noise floor question for users

john lutz said:
Thank you Mac, JR, Rob. I'm in over my head. I had help with the math - I don't fully understand all of it so can't argue or defend it.

Bottom line is the amp hums significantly and audibly more than any other of my "MI" grade amps, in fact the other amps barely show any noise at all. What I first thought must be a defect turns out to be a feature.
Click to expand...

Yes, "A" weighting will reduce significance of hum, just like human hearing does. I hope the amp is not singing along with you.

Does it hum audibly when input disconnected? Input shorted? Or is it related to the interface/interaction with other gear? OK I read that no input was connected, try a shorted input plug or perhaps a low impedance signal source (like mixer output) turned down. A floating high impedance input could pick up noise from the environment.

Are you comparing hum side by side with other amps, with same inputs and exact same gain structure? (I see you say no gain?? all amps have gain). An amp with higher gain (more output) will sound noisier until you play music through it and have to turn it down.

I have been outside the castle for long time, but don't expect overt design flaws from the amp design guys after decades of designing amps. It doesn't cost much to provide a decent input stage. Along with high power comes higher gain and elevated noise floors. 100 dB seems like a usable dynamic range.

JR

PS: and why am I doing customer service? :-(
 
Top Bottom