advertisement


Play time: NAP140 with SMPS

Another late thought - and before you re-wire - have you tried running the smps boxes you have at/close the specced 36v output,just in case that lifting the output 10-12% might have also moved their internal compensation into a corner it is not happy with the apparent loading ..?

Keep going! : )
 
That's a
martin clark said:
Another late thought - and before you re-wire - have you tried running the smps boxes you have at/close the specced 36v output,just in case that lifting the output 10-12% might have also moved their internal compensation into a corner it is not happy with the apparent loading ..?

Keep going! : )
Click to expand...
That's a good point, and easy to test. However, if that's the case, why did it not buzz with the passive resistors?
 
er. Um. Still thinking...

BTW - when you test this -is the amp signalinput open, or connected to a live preamp?
If the first - have you tried with the amp signal inputs shorted?
 
Please do; not that it ought to change anything- but if it does, we'll all learn something...well, perhaps one of these days

(I suggest shorting the inputs, to check possibility there is some sort of induced loop oscillation - smps radiated noise to input wiring, leading to HF unhappiness in the amplifier, reflected as some kind of interaction with the smps; or smthing like that.

And that's absolutely not a snarky comment on the very neat & compact loop-area layout and wiring you've shown...just another pondering ...)

ATB.
 
It turns out that I had already turned the voltage down to 36V, so no help there.

I shorted the input, but no change. The squeal remains.

Now that I have the NCC200 in place, I have a couple of options:
  • I've got a pair of LRS-50-48 (same power as the existing ones I'm testing, but 48V instead of 36V). It's the same footprint, and should be an interesting comparison.
  • I also have a pair of the LRS-200-48 (the same ones that Laverda is using). That should give the same results he's seeing.
I could do both of those with the same ground configuration as I have now. If the whine remains, then we know what to blame (or at least what to try next).
 
First test done: I replaced the LRS-50-36 pair with LRS-50-48 (from 36V to 48V, all promising 50W). The load is a single NCC200. There's still a whine, but perhaps it's a bit quieter (measured using my built-in irritation-ometer). I'm not sure what I've learned, though, given the output power of these is the same. Perhaps the 48V units are somehow more robust, and therefore more able to handle a "weird" load without complaint. (Recall that it the 36V handled basic resistance just fine.)

@laverda, you mentioned a potential plan to use some of these smaller units to power the input stage of the NCC300. Did you ever try that, or is your Mean-Well experience only with the heftier LRS-200-48? Perhaps these amp boards can only be supplied by the bigger units. If so, that would mean my only option is the power both channels from a single pair of supplies, rather than running them as dual mono.

My next step is to try the LRS-200-48 myself, to see if the whine disappears.
 
I just realized there are two additional differences between laverda's usage and mine, which I decided to test before trying the big supplies.
  • His units sit on top of each other. (That was how I eventually planned to position mine, but I wasn't there yet.) His prototype didn't do this, though.
  • His top unit is fed via a set of jumper wires from the bottom, whereas mine are fed with separate feeds from a connection block (much like you would wires all of the grounds back to a single point). His prototype also use this "jumper" configuration.
I made these two changes staying with the LRS-200-48, and the whine reduced further. I can definitely hear it if I'm a few inches away, but at a few feet it's almost imperceptible (albeit not silent, as when using a simple resistive load).

This suggests that the SMPS units are more sensitive to input supply than I would have expected.

I still plan to test the bigger supplies, just to see how that changes things, but I probably won't get to that until tomorrow.
 
Following this with interest ...as I'm sure are others!
Keep us updated....and good luck with sorting the issue
 
I think its probably the front end of the nap circuit which is sensitive to the ps noise, could you not seperate them from the back end and filter ala ncc200/220/300.
You could load the ps s with some resistors in parallel with the nap boards if there is a minimum recommended load for them, some high temp 2w 3k perhaps?
 
you mentioned a potential plan to use some of these smaller units to power the input stage of the NCC300. Did you ever try that, or is your Mean-Well experience only with the heftier LRS-200-48? Perhaps these amp boards can only be supplied by the bigger units. If so, that would mean my only option is the power both channels from a single pair of supplies, rather than running them as dual mono.
Click to expand...


No I didn't follow that up on a second input power supply, only ever used the LRS-200-48 in a mono block config.
 
hp1 said:
I think its probably the front end of the nap circuit which is sensitive to the ps noise
Click to expand...
It's not sound coming from the speakers. The output isn't even hooked up at this stage. The whine is coming directly from the SMPS units themselves, in response to the amp board(s) as loads. They don't whine for simple resistors with the same current load.
 
I substituted in the (much) bigger LRS-200-48 with the single NCC200, and there is absolutely no whine. For whatever reason, the smaller power supplies don't seem to like driving the amp modules, even though the current drawn is exceedingly small. I wonder which of the Mean Well SMPS units work versus not. Would the LRS-100 be enough? Or LRS-75?

That's good news and bad:
  • Good, as I can finally get past this conundrum. I'll use the in-hand LRS-200-48 pair to power an NCC200 stereo amp, and see how that sounds.
  • Bad, as I bought a few pairs of the smaller supplies to experiment, hoping perhaps to power the input stage of NCC200's in Voyager configuration. Perhaps they'll behave better with the input stage alone, better than they did with the entire board.
Fortunately, they're quite cheap, so I'm not really out that much, and I might be able to reuse them for something else.

My original inclination was to rig up a little stereo amp using the old NAP140 boards and two pairs of cheap MW SMPS modules to be "dual mono", but it turns out that none of the SMPS units I have here will do that. I'll wait to see what I think of the NCC200+SMPS amp, then do one of these:
  • If that sounds good, then I may order a pair of LRS-200-36 units for the NAP140 boards.
  • Another alternative is to buy a suitable transformer, as I have a bunch of CAP6 modules laying around.
  • Yet another option is to swap the NAP140 boards back into the upgraded NAP140, and salvage the NCC200 boards to put into another amp with a higher voltage.
  • Or I could just throw in the towel and sell them on eBay. :D
 
Last edited:
I finished putting the amp together today, and I'm actually quite pleased by the results. I was a bit concerned, given laverda's comments about the lack of bass with a QUDOS. At least with the NCC200 modules, it seems perfectly adequate. I can't discern an appreciable difference between this and a toroidal CT transformer with CAP6 or equivalent.

Granted, the two SMPS units together are rated to provide 400W, which is far in excess of the requirements of the two amp modules, which will likely draw at most half of that, and usually much, much less.
 
Mike Hanson said:
I finished putting the amp together today, and I'm actually quite pleased by the results. I was a bit concerned, given laverda's comments about the lack of bass with a QUDOS. At least with the NCC200 modules, it seems perfectly adequate. I can't discern an appreciable difference between this and a toroidal CT transformer with CAP6 or equivalent.

Granted, the two SMPS units together are rated to provide 400W, which is far in excess of the requirements of the two amp modules, which will likely draw at most half of that, and usually much, much less.
Click to expand...


Glad you got there in the end Mike.
 
BTW, here are a few pictures of the build, starting with the overview:



Next up is the power input, including the stupidly huge transformer for the UPC1237 speaker protection module. Yes, I'm Dr. Frankenstein, and this is my monster.



I'll admit that I'm a bit concerned about the switch. It's rated at a maximum of 200V, and I'm running 120VAC, which peaks pretty close to that. I have enough space to add a relay in there, or I might change it to higher rated switch. Any suggestions?

The switch has an LED, but I haven't hooked that up to the SMPS yet. It will take 12V straight-up, so I need to experiment with some inline resistors to figure out a suitable value to use when driving it with the higher voltage.



Here's how I've wired the SMPS to get 53V-0-53V. (The supplies are adjustable around 48V.) You'll see the FULARR connector used for the 0V reference. With all those grounds, this seemed easier than soldering them to a linking solid core wire, like laverda did. I suppose I could change it later, now that I know it works.



Here's a close-up of the NCC200 boards, with their beautiful copper heat spreaders.

Those with sharp eyes may notice that the power transistors on one board are the treasured BUV20 from ST Micro, while the other has the more common MJ15003G from ON Semi. I originally had these boards in another build with inadequate heat sinking, and one board went into thermal overrun and had to be repaired. I guess that means one channel sounds different/better than the other.

That's why I've replaced the Modushop Pesante's usual steel panel with a piece of aluminum 3mm thick. That should provide suitable heat dissipation.



Given the short run, the input uses unshielded wire, kinda-sorta twisted a couple of times, but I doubt that makes any difference in this case.



Finally the output with the UPC1237 board. Note that I've run the two output ground posts directly from ground central, and the protection board has its own ground reference wire from that same source. Only the positive signals pass through the protection board.



For my "serious" builds, I've generally moved to using 14-12 AWG for internal wiring, but this originally started as a silly project with my old NAP140 NAPA 5/5 boards. Consequently, most of the wire is 18AWG, the short input wires are 22AWG, and the output is 14AWG. (The stiffer 14 gauge stuff works well for starting the wiring harness.)

Any thoughts or suggestions for improvement?
 
You must log in or register to reply here.


advertisement


Back
Top