Good dynamic range bench equipment please!

[TQ]

Hello, new forum user here
I have a few questions about gear that is worthy of designing low noise, low level analog with (besides your brain).  I like the Hameg HM2005-2, I read the manual and it's specs seem quality.  But $2200, it still doesn't match some older used models.  Yes, I'm willing to pay that much if it is worth it, just for the piece of mind that it's warrantied + accessories.  My goal for now is to improve analog stages on cheap modern audio designs.  My interest is pro-audio equipment mainly because I'm an audio engineer by trade, but recently earned a degree in electronics as well.  The Audio Precision 2722 is blatantly the best in this class, but lets be serious!  I'd like a lab where I can at least design something with specs worthy of being tested on one of those.  Thanks for any honest advice you can give.

-Sorry, I had to mod my initial post cause I was blabbering and I hate that.

[allanw]

Why not buy a good old Tektronix scope if you're willing to spend that much money?

You can get a 2465B for around $400-$600 and that was (and still is) the top-of-the-line analog scope.

The lower models are much cheaper, around $200.

Of course, they'll be used, and 20+ years old instead of brand new, but it costs so much less. They're also decently serviceable, as there's a big community of tekscopes people on yahoo.

[TQ]

Thanks allanw,

You are very right about the 2465B! I know a respected designer and he only uses analog Tek models on his bench, he has said the same thing.  Also think I might want a single sweep since I'm not going digital.

So besides the scope.
I'd like to design a mic-pre custom build for a specific mic that has 1.12mV output at 94dB SPL, and that's open circuit voltage, this means  only about 4uA at moderate SPL through a 300ohm 1/1 transformer (I don't want to bridge).  How in the hell do you characterize and learn from what's happening at your input, before amplification, when values are this low?  What type of resolution and accuracy do I really need to in order to trust and learn form measurements?  I like the Tektronix DMM4050 but am uncertain, your thoughts?  Any info would be appreciated.

[jahonen]

I once used a self-made AC-coupled 80 dB preamplifier (G=100 with SSM2019 followed by G=100 non-inverting amplifier with AD797) to measure the noise of some reference voltage output which needed to be low-noise. Scope used was TDS3034B. Input-referred noise level for that setup was some microvolts, enough to see the effect of PCB bending (clear change in output voltage). If you use time domain instrument like that, then you usually must add some bandlimiting filter after the preamplifier, to get meanful results, or calculate the noise density.

But, usually those audio things are usually characterized by AC-specifications (SNR, distortion etc.), so you don't actually get very far using just a scope. A good (external) soundcard with good preamplifier is probably better in that sense. For the software, use something like ARTA. Of course it's no AP but better than nothing.

Regards,
Janne

[Zad]

If you are looking for the ultimate linearity you shouldn't be using a transformer anyway.

[tinsmith]

You are very right about the 2465B! I know a respected designer and he only uses analog Tek models on his bench, he has said the same thing.

Scope brand often comes down to what you trained on. If you're going new and on a budget (i.e. if you have official funding) then Agilent tends to give slightly more meter per buck on the $1-2k end than Tektronix. Higher up, I'm becoming a bit partial to LeCroy...

But multimeter, HP/Agilent all the way. I <3 me some 34410A action.

[TQ]

Thanks, I have looked at the 34410/11A also.  That has a peak detect function that is supposed to be good to have for AC/noise, and the 11A has more readings than the rest and level triggering which seems like it could be useful.  Accuracy seems about the same, but I guess at the uV level I'll have to amplify and filter as Janne said anyway.  I've read a bunch about filtering noise for measurement accuracy or a few other purposes, anybody have any luck building their own test module for this?  I would be more than willing to learn since this is what I see myself having to do, especially with selectable input impedance.

As far as ultimate linearity, fat chance with the design I have!  Just looking for testing that can lead me to proper analytical thinking.

Also I looked at ARTA, thanks Janne.  I have a program called spectrafoo which does FFT and spectral analysis very well in the audio bandwidth much like ARTA.  I'm actually quite comfortable testing in this manner, but it hasn't gotten me far with my bench skills