Powering cheap field mixer - 48v dc needed?

[Doc Bernard]

Interesting about that 30-16k spec. It's only a neg -3dB. Not a +/-, I suspect there is an error somewhere. And since the manufacturer put the worse spec up, I would tend to believe that over the 20/20 spec. It's not unheard of for a retailer to change the plot to help sell a product. Use thicker lines, smaller graphs.

To decipher what I am talking about, you draw a horizontal line on a graph that represents 20 hz to 20,000 hz. That line represents a 20-20k @ 0dB. The length of the line you now draw is from 30 to 16k. At both 30 and 16k (actually it is slightly in from that) the drop-off in freq response is greater than -3dB. So, theoretically the frequency response could be under 0dB at all frequencies, up to -3dB. Which means a lower output.

OTOH, with a +/- of 1.5 dB That tells you it lives around the baseline of 0dB with peaks and valleys. -1.5 and 1.5 equal 3 db on a graph, but it is better than 0 to -3db. Think of a frequency plot as a snapshot of an EQ.

Basically, someone said "oh, the frequency plot that the manufacturer gave is not right" or they just plopped new numbers in. 20/20 @ +/-1.5 is a standard number 30/16@-3 isn't.

 

[j1clark@ucsd.edu]

Interesting about that 30-16k spec. It's only a neg -3dB. Not a +/-, I suspect there is an error somewhere. And since the manufacturer put the worse spec up, I would tend to believe that over the 20/20 spec. It's not unheard of for a retailer to change the plot to help sell a product. Use thicker lines, smaller graphs.

To decipher what I am talking about, you draw a horizontal line on a graph that represents 20 hz to 20,000 hz. That line represents a 20-20k @ 0dB. The length of the line you now draw is from 30 to 16k. At both 30 and 16k (actually it is slightly in from that) the drop-off in freq response is greater than -3dB. So, theoretically the frequency response could be under 0dB at all frequencies, up to -3dB. Which means a lower output.

OTOH, with a +/- of 1.5 dB That tells you it lives around the baseline of 0dB with peaks and valleys. -1.5 and 1.5 equal 3 db on a graph, but it is better than 0 to -3db. Think of a frequency plot as a snapshot of an EQ.

Basically, someone said "oh, the frequency plot that the manufacturer gave is not right" or they just plopped new numbers in. 20/20 @ +/-1.5 is a standard number 30/16@-3 isn't.

I guess I'd interpret that -3 dB at 30/16K Hz to be the 1/2 power points of the band pass of the device.

There's also 'flatness' of response in the band, which could be listed in +/- dB.

 

[Noiz2]

Doc. The 20-20k is from the MANUAL that is on the manufacturers site not some marketing sheet. Actually the marketing guys are probably the ones responsible for the web site.

If I had to guess I would guess that the response is flat from 30 to 16k and that the 20-20k is the point at which it has dropped by -3dB. But? They haven't gotten back to me or it got swallowed in a spam filter.

 

[Noiz2]

OK I got a response...

It's worse than either of the printed ones 30Hz - 16kHz +/- 3dB

That is NOT good. Truthfully for dialog it is probably fine but at almost 500 Euros that is no bargain.

If your going into a mini recorder those might be better than the "real" specs on the recorder, so if you can get this box at a VERY good price it's an OK deal.

BTW when I say "real" specs that is because a lot of specs printed for consumer gear take great liberties and use a lot of creative writing to imply they are much better than they actually are. Just about anything can be 20-20k IF you measure in the right place with un mentioned tolerances. Because these specs look so bad and this is obviously a company aimed at higher end users I suspect that these are brutally honest specs. A $200 micro recorder might advertise 20-20k but that is probably in and out of the converter. Since you have to go through the analog section to get to the converter those numbers meaningless. This is pretty obvious if you go to one of the comparison sites and listen to the "tests". The specs are all pretty identical but the feq. response is all over the map.