Post your noise!~

[Learner]

Hi guys,
I was wondering if we can accumulate a collection of noise figure on tube amps and tube dynamic processors?

Post your measurement on any of your tube gear(preferably all you have :green: ), may it be your DIY la2a or any commercial purchased unit. Post the output noise figure on the CRO or just the rms measurements, all data are welcome!!!

Any Fairchildriken taker!??? DIY La2A??? Pultec? TuB Tech? Monley??

 

[gyraf]

That wouldn't make much sense without specifying EXACTLY how to measure these figures. Remember, technical specs lie.

Jakob E.

 

[Learner]

[quote author="gyraf"]That wouldn't make much sense without specifying EXACTLY how to measure these figures. Remember, technical specs lie.

Jakob E.[/quote]

Hi Jakob,
What would you suggest as the most ideal method for testing noise? What I want to find out is how audible is the noise output when connected into a power amp. May be the measure the voltage on a 10 Meg probe??? Suggestions??

 

[gyraf]

The most sensible noise distance figure may be equivalent input noise with input shorted. That is, short input, measure output level, divide by unit gain.

But there's many other ways to specify this, everyone has their own favorite and different test equipment. This is why it really dosen't make sense comparing specs of different origin.

Jakob E.

 

[Learner]

[quote author="gyraf"]The most sensible noise distance figure may be equivalent input noise with input shorted. That is, short input, measure output level, divide by unit gain.

But there's many other ways to specify this, everyone has their own favorite and different test equipment. This is why it really dosen't make sense comparing specs of different origin.

Jakob E.[/quote]

Right, I see what you are saying.

It would make things a lot easier if there is a consistant method of approach to record the measurment, not to mention that the figure will be a lot more meaningful.

Since it is common for tube amps/dynamic processors to have an input gain control, I guess if that is turned all the way down it is pretty much shorting out the input.

So may be we can make that as part of the procedure of measure?

What I am interested to find out is the noise generated by the circuit inside the rack, and how audible it is when it is turned to half way mark on the output pot.

That's what you will hear when you connect it into the power amp, and that is ALL it matters imho.

If we can collect 10 voltage rms measurements of la2A output for example, we can find out the average noise figure of the la2a based on the half way mark of it's output capability. Needless to say, it doesn't make much sense to compare the noise figure of different circuit/amps.

All you need to do is turn down the input if there is one and turn up the output to the half way mark, stick your volt meter and take down the output measurment.

Thoughts?

 

[pstamler]

[quote author="Learner"]What I am interested to find out is the noise generated by the circuit inside the rack, and how audible it is when it is turned to half way mark on the output pot.

That's what you will hear when you connect it into the power amp, and that is ALL it matters imho.

If we can collect 10 voltage rms measurements of la2A output for example, we can find out the average noise figure of the la2a based on the half way mark of it's output capability. Needless to say, it doesn't make much sense to compare the noise figure of different circuit/amps.

All you need to do is turn down the input if there is one and turn up the output to the half way mark, stick your volt meter and take down the output measurment.[/quote]

What's "halfway"? 50% rotation? 6dB down from maximum output? And why that position? It seems arbitrary.

Peace,
Paul

 

[Learner]

[quote author="pstamler"]

What's "halfway"? 50% rotation? 6dB down from maximum output? And why that position? It seems arbitrary.

Peace,
Paul[/quote]

Yea, 50% rotation mid point of the dial.

I would imagine that it is the closest setting to a point, where each user would have experienced the most while using that amp. Also, when monitoring at a"low level" according to the knob position(anything less than half way) it would be a real worry if you can't turn the knob pass the half way position having to avoid appearent audible hiss to the user.

 

[Butterylicious]

Specs = specifications vs. speculations

 

[NewYorkDave]

I measure noise with the input terminated in the rated source impedance. This gives a "worse" figure than a shorted input, but is perhaps more reflective of real-world performance.

I generally specify noise as dB below maximum output or EIN, depending on my mood that day :wink: For instance, in the specs for the prototype of the One-Bottle Preamp, I could have just as easily quoted the noise as -141dB EIN (unweighted).

 

[Learner]

[quote author="Butterylicious"]Specs = specifications vs. speculations[/quote]

Speculation???

Forgive my ignorance, but I am not quite sure I understand what that means...

Perhaps you are refering to people purposely trying to fake a measurment figure in order to defend their expensive investment, to reassure the purchasing decision was worthwhile??

I don't understand why would people do that in exchange of finding out where their unit stand comparing to some other units, it would be benificial for all the owners to see the average and perhaps locate faults even by comparing the measurments. Well, that just my personal opinion anyway.....

C'mon guys, any takers???

Let it all hang out.

BTW, thanks Dave for sharing that with us!! :thumb:

:guinness:

 

[kvintus]

Don't forget to specify the bandwidth or express the noise voltage in nV/rtHz.

just as easily quoted the noise as -141dB EIN (unweighted).

For what source resistance?

/Anders

 

[Kev]

[quote author="gyraf"] This is why it really dosen't make sense comparing specs of different origin.[/quote]
I'm with Jakob here
... don't get tied up in specs
Is the unit useful in your working situation ?
if so ... move on

 

[NewYorkDave]

Anders,

As it says on the "spec sheet", all measurements were made with a 200-ohm source resistance.

 

[Learner]

No takers anyone??? :shock:

 

[adrianh]

all of the audio guys and microphone guys cannot
agree on how to do this. There are a few accepted
ways with said weighting filters and said test instrumention.

EIA and several others have a spec but who wants to pay
to get a spec.

I use the test instruments on either side of my head.

 

[WJS]

I've used a meter to hum balance tubes but the only other test i've used reliably is to listen to it my self.

its like when you are in the studio and you are balanceing the monitors so that music sounds about the same volume on each. there is no exact way to do it because they each reproduce sound slightly differently ns-10 and genelex sound different there is no right way, it is what ever sounds good that is right.

sorry learner, but i dont have any specs to post, besides:

my 83 bass pre: tons of hum(havn't fixed the psu) and constant distortion

my 6 channel g9 based pre, clean as a bell, some hum, needs a hum ballencer on the heaters

300b guitar amp sounds good, the driver stage and 300b stage are nice and clean for a guitar amp.

i doubt any of this is helpful... but um... if this is what you want

 

[Learner]

[quote author="adrianh"]
I use the test instruments on either side of my head.[/quote]

That's EXACTLY what I am looking for except in RMS value, since we can not all be there to participate.

I don't see the need of putting into dbs and calculations, since we are only trying to compare the SAME device except different units built by different people at different time.

No input or turn it down if there is one, then simply turn up the output/gain to half way and measure the RMS thats all.

What else is there to consider or agree with??? Please help me understand this anybody??

 

[Learner]

[quote author="WJS"]

my 6 channel g9 based pre, clean as a bell, some hum, needs a hum ballencer on the heaters

i doubt any of this is helpful... but um... if this is what you want[/quote]

If you could be kind enough to measure this hum, I would greatly appreciated!! :thumb:

All you need to do is turn the gain down to 0 and output to half way, use a voltmeter and measure the RMS.

Thanks!

 

[Kev]

sorry Learner
but I have to jump in here and say something

... turn the gain down to 0 and output to half way, use a voltmeter and measure the RMS.

but my pots may be different to your pots and so we have a different situation already.
Compressors and Mic-pres are all about gain structure and the battle between noise and clipping ... headroom

... I don't see the need of putting into dbs and calculations,

but you have to accept that what we are doing here is comparing and the whole point of dB's is to compare ... IN THE SAME STIUATION

... since we are only trying to compare the SAME device except different units built by different people at different time.

UNLESS
you can be sure these units are the same ( I don't think they will be )
a simple test like this is only a loose indication that you may have a major error somewhere

It might be cool to get a signal of known level running throught it first and then compare the RMS values.

The reason I say this is that unless you have a True RMS Meter it is likely we will measure the same signal and get different results.

I have more than a few DMM's and they all give different result for the same signal ... even the one of the same make and model

yeah yeah
I shouldn't buy the el-cheapo meters !!!

 

[Learner]

[quote author="Kev"]
but my pots may be different to your pots and so we have a different situation already. [/quote]

Oh definitely, even 2 pots from the same manufacturers will not be the same due to its tolerance range. Which is the reason why the method works on averaging, the accuracy will depend on the number of measurments collected not precision of 1 particular unit.

Compressors and Mic-pres are all about gain structure and the battle between noise and clipping ... headroom

Relevance??? lost me there.

but you have to accept that what we are doing here is comparing and the whole point of dB's is to compare ... IN THE SAME STIUATION

db is nothing but a ratio or in reference to a figure, still it needs a measurement of some sort to calculate db. The SAME STIUATION here is the circuit, there is nothing to be gained by converting the measurment into db since we are not trying to compare different circuits (eg. La2a vs Pultec)

UNLESS
you can be sure these units are the same ( I don't think they will be)

They won't be too much different except for the tolerance of components used, since THE CIRCUIT IS THE SAME unless the individual decides to mod the circuit which the unit's measurment will not be used for averaging.
It is actually harder trying to different than being the same, since the circuit is the same; the part list is the same. The only thing that might be different is probably the face panel, knobs, meter, panel marking and in many occasions even that is the same.

a simple test like this is only a loose indication that you may have a major error somewhere

Not only that, you get an idea of the noise level based on the averaging method and it serves as a reference point of that circuit. Also, for someone that has a low noise measurment it would be a lead for invetigation to see what the person has done to the circuit to obtain that. It could be real beneficial to all the builders of the same unit, if such an improvement has been discovered.

It might be cool to get a signal of known level running throught it first and then compare the RMS values.

The reason I say this is that unless you have a True RMS Meter it is likely we will measure the same signal and get different results.

I have more than a few DMM's and they all give different result for the same signal ... even the one of the same make and model

yeah yeah
I shouldn't buy the el-cheapo meter

Well, that sounds like the accumulative effect of components tolerance range. Just out of curiosity, how much of a difference do you get?? In decimals?? Not the that it really matters, since it will all be averaged out anyway.