Deconstructing the Behringer NU3000

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am in tweaking mood again! My NU3000 (two different ones actually) have, for several months, running fanless. My next (and riskier) mod is to try running one without the output filters. I know there are reasons to not do this ... mainly to not turn into an AM broadcasting station. However, my speakers are bose 901s (no x-over, no tweeter to melt) and perhaps ideal for being directly driven. Lacking a schematic (or common sense perhaps) I am eyeing the board and ... lots of inductors and caps. So I will bypass and/or remove them a few at a time and see what happens. Not to worry, I have two fire extinguishers :)
 
I am approaching the dis-assembly of the output filtering with the assumption that the filters reduce the possible power output of what is (I assume) a full bridge per channel output pair. I am working without a schematic and even less electronic knowledge. However, the NU3000 is interesting in that, from inspection, there is a "big" output filter (inductors and caps) and then on a daughter board feeding the speaker outs, there is a little tiny inductor and two caps.
 
It would be interesting to monitor idle power input to the amp, load attached, with the original circuit and then with filters removed using a Kill A Watt or similar, if you don't have fancier test equipment. This would be an accurate enough baseline measurement to tell whether you're going to ruin your speakers even before you smell smoke. I don't think I'd call trying this without a schematic, "brave".
 
Last edited:
I am approaching the dis-assembly of the output filtering with the assumption that the filters reduce the possible power output of what is (I assume) a full bridge per channel output pair.

This being a classD amp its always running at full tilt - the filter prevents the ultrasonics getting out. Taking out the filter will dramatically increase the ultrasonic power output - and it does this irrespective of input signal level. With no filter I think you'd get better efficiency from a classAB, that is assuming the FETs in the output stage don't go into current limit first in which case even classA will beat it.
 
I removed the two output inductors and the amp goes into error mode. I think I will stop my experimentation now. Whether I have killed the amp or not, will have to wait until I reinstall the parts.

I am curious about the filtering. I would think the power output is the same...if the carrier is getting filtered by the LC filter or by the speaker voice coil. It goes to ground either way. More likely there is much I do not know about electronics :)

Also, I do have a Kill-a-watt but have read that they are unreliable when trying to read a switching power supply, which is what the Behringer NU3000 is, as far as I know.
 
Last edited:
Addend-dumb: I am now 99% sure that my earlier claim that the NU3000 is a full bridge (per channel) is incorrect. I found a reference that says a full bridged amp will have 4 MOSFETs per channel so the NU3000 is only full bridged when it is in [drum roll please] bridge mode!
 
Shorting output inductors should trigger protection. These inductors are there to control current slope, over-current will result without them (or kaboom if protection is slow or missing). Good news are that protection seems to work in this design hehe
 
Agreed. I am as clumsy a solderer as my nickname should imply. Soldering the two big inductors back on took 1-2 hours. Does anyone know how to fill a syringe full of molten solder ? I may try to put the little caps back in too. As you noted, and my experiments show, the Behringer protection kicks in which would probably make bypassing it very difficult. I give up, at least for now.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.