NC252MP High Pitched Whine from Speakers

[AcousticTheory]

Esteemed colleagues,

While researching my NCx500 purchase, I also became curious about the NC252MP. I found a unit built in the Ghent half-depth case used by Archimago and others for their NC252MP builds that was being offered for sale used at a decent price (but not a fire-sale price), and I bought it. I've been listening to it a little bit here and there while also acclimating myself to the NCx500 monoblocks, and during a very quiet moment when all other appliances and air conditioning were off, and there was no noise from outside, I decided to put my ear up to one speaker to see what the noise floor of the amps sounded like. That's when I discovered a very faint, high-pitched tone around 5.6kHz that was audible above the noise floor, which I could still discern at distance of 12 inches from the speaker in my room. Allowing the amplifier to warm up and stabilize for about an hour did not result in the tone dissipating, and the tone was audible even when the source device was disconnected. The internal cabling in the amp case is as short as possible, and positive/negative or hot/neutral pairs are twisted between the amp module and the rear panel connections to avoid noise infiltration. I verified all non-soldered connections were tight and I squeezed the spades to the IEC connector so they were tight when sliding on/off but still removable. The IEC socket is grounded to the screw of one of the chassis feet, so grounding is via the aluminum enclosure and the heatsink plate of the NC252MP module. I think the module is a V1 with red candy-drop output capacitors; I have seen photos of another module with blue output caps on the Hypex site but I haven't ever seen one of these V2 boards in the field.

Is this a known issue with the NC252MP modules, or could this be an issue with my module specifically? How has this quiescent tone not shown up in prior evaluations of this module I have seen? I would think this tone, if audible, would measurably contribute to SNR and would be visible in the spectrum of quiescent noise. It seems to be power supply noise; I noticed that when powering off the amp for a while then powering it on again, this tone would start at a lower pitch, then ramp up in pitch, then ramp down a little bit before stabilizing around 5.6kHz. I haven't noticed anything similar with my NCx500s; they use separate SMPS1200A700 power supply modules, so there is physical distance in the chassis. I'm interested to know what people think, or whether there's a known resolution for this.

Thanks,
RB

 

[ThatSoundsGood]

I have several NC252 and other model Hypex amps and none of them have this problem. However, I have exerpienced the problem and it was coming from a power supply into a mini DSP. I recommend moving your cabling around to see if it goes away. Also, unplug anything else that is drawing power from the same source as the amplifier to see if it might be a different device causing noise. Lights, bluetooth receivers, preamplifier, etc. can create noise that the amplifier then amplifies and you hear it. This can happen by just having them plugged into the same power strip/outlet as the amplifier. It can happen even if there is a device on another outlet (but the same circuit) in the house.

 

[DVDdoug]

It seems to be power supply noise;

I'd agree. It's probably noise from a switching power supply. I'm sure the switching frequency is above the audible range but there can be other electrical noise.

All analog electronics have SOME noise. If you're not normally listening 12-inches from the speaker there's no point in setting-up artificial conditions and TRYING to hear it. (Unless you are the designer/manufacturer.)

very faint, high-pitched tone around 5.6kHz that was audible above the noise floor,

That IS part of your 'noise floor".

 

[AcousticTheory]

An update: I tried using Studio Six AudioTools and my iPhone mic to try to 'see' the tone, and I was able to see a clearly-defined spike at around 4700 Hz at a level of 29 dB, at least 10dB above the adjacent noise; the amp has been idle for a few days now and this appears to be where this tone has settled (see my comments above on the tendency of this frequency to 'move' while the amp is warming up after power-on.) This was at a distance of one inch from the tweeter.

Normally, the 'noise floor' of an amplifier consists of a broad-spectrum, atonal rushing or hissing sound, and maybe a 60/120 hz component from the line, not a perceptible whistling at a certain audible frequency; this seems like an unfiltered power supply component that the amp has insufficient feedback to properly reject. If I had speakers with even higher efficiency, this sort of tone in the noise floor would be thoroughly unacceptable; it's a good thing I'm not trying to manufacture self-powered speakers or PA gear using this module. If Hypex is trying to sell an amp module on the promise of low noise and high SNR, they should make sure the idle noise of the amplifier itself is truly low, or at least benign.

I have several raw tweeter speakers around, so I will make a short harness (a few inches) for one, then go to an outlet in the house far from other gear, connect the tweeter to one amp channel only, and see if the amp continues to make this sound through that tweeter.

My workplace has an APx525 analyzer, so I will see about getting the amp to do this behavior on the test bench, then once I have that data, I will contact Hypex and see if they have a response, or if they can assist with a replacement.

The ICEPower 50ASX2 board I use at my desk doesn't produce any tones in this range, using the same Studio Six AudioTools app and same iPhone.

Also...now I need to see if there's anything similar from my NCx500 monos that I haven't previously noticed.

 

[ThatSoundsGood]

An update: I tried using Studio Six AudioTools and my iPhone mic to try to 'see' the tone, and I was able to see a clearly-defined spike at around 4700 Hz at a level of 29 dB, at least 10dB above the adjacent noise; the amp has been idle for a few days now and this appears to be where this tone has settled (see my comments above on the tendency of this frequency to 'move' while the amp is warming up after power-on.) This was at a distance of one inch from the tweeter.

Normally, the 'noise floor' of an amplifier consists of a broad-spectrum, atonal rushing or hissing sound, and maybe a 60/120 hz component from the line, not a perceptible whistling at a certain audible frequency; this seems like an unfiltered power supply component that the amp has insufficient feedback to properly reject. If I had speakers with even higher efficiency, this sort of tone in the noise floor would be thoroughly unacceptable; it's a good thing I'm not trying to manufacture self-powered speakers or PA gear using this module. If Hypex is trying to sell an amp module on the promise of low noise and high SNR, they should make sure the idle noise of the amplifier itself is truly low, or at least benign.

I have several raw tweeter speakers around, so I will make a short harness (a few inches) for one, then go to an outlet in the house far from other gear, connect the tweeter to one amp channel only, and see if the amp continues to make this sound through that tweeter.

My workplace has an APx525 analyzer, so I will see about getting the amp to do this behavior on the test bench, then once I have that data, I will contact Hypex and see if they have a response, or if they can assist with a replacement.

The ICEPower 50ASX2 board I use at my desk doesn't produce any tones in this range, using the same Studio Six AudioTools app and same iPhone.

Also...now I need to see if there's anything similar from my NCx500 monos that I haven't previously noticed.

But have you tried my suggestions of unplugging everything else from the same power source? Have you tried the amplifier on a different circuit in the house with nothing else on it? I have 2 Buckeye Hypex amplifiers and 2 Hypex FA253 amplifiers and none of them have this problem that you're talking about. Are you certain it's not coming from somewhere else?

 

[AcousticTheory]

But have you tried my suggestions of unplugging everything else from the same power source? Have you tried the amplifier on a different circuit in the house with nothing else on it? I have 2 Buckeye Hypex amplifiers and 2 Hypex FA253 amplifiers and none of them have this problem that you're talking about. Are you certain it's not coming from somewhere else?

I unplugged the amp, powered off the power strip with other gear connected (for good measure), took the amp into my bathroom and closed the door, shut off HVAC, plugged the amp into the outlet near the sink, and connected a tweeter (a 22mm ferrite Wavecor dome tweeter of modest sensitivity) to one channel (with no inputs) and upon power up, I heard a distinct single-frequency tone that ramped up in frequency, then ramped down, then ramped back up, and this time it went up to 11.2 kHz - almost too high to hear but not quite. In that environment I measured the tone at 20dB above the adjacent noise floor, instead of 10-15 dB as I had seen previously. So the issue appears to be 'real'. The speaker cable I used was a pair of 12" red/black test leads that I normally use with a multimeter, so exotic speaker cables or long/capacitive lengths are not in play, and the only other cable was an ordinary IEC power cord.

Next I will see if the APx525+AUX0025 at my office is able to capture this. Maybe I do have a bad unit, or one that has not aged gracefully. At least it should be easy to show this issue on a FFT plot using the AP analyzer.

The good news (since I've got more money tied up in them) is that my NCx500 monos from Deer Creek are still dead, dead quiet with the same speakers and speaker cables. I think they may be more pure-sounding overall; my subjective impression is that the NCx500 are just a straight wire with no discernible character except the signal, fully in-control of the speaker (and maybe a bit overdamped in the bass which could also be due to my speakers), while my dubious NC252MP module has sounded a little strained or harsh on some material. More to come, I'm sure.

 

[AcousticTheory]

Here is what the AP analyzer was able to show. I connected one channel of the amplifier to the AUX-0025, then from the AUX-0025 to the AP input, and 200 ohm termination was used. I connected an XLR cable from the generator output of the AP to the corresponding amplifier channel but I left the generator turned off. The amp output was connected to a 4" nominal-4-ohm woofer that was in the lab using a 2-foot length of speaker wire, and the output from that channel was connected to the AP using another 2' string of banana leads. I powered on the amplifier, and saw perfect visual confirmation of what I was hearing through the speaker. The strong 30dB peak at ~9.5kHz started low, around 4kHz-5kHz right after amp output turn-on, then wanders up to 11kHz, then back down to about 8kHz, then continues to very slowly wander up and down in that frequency range, and its harmonics move together with the peak. A couple of times I also saw the peak start and stabilize at a lower frequency. I also tried setting up a 5-watt sine tone at 1kHz run and only got a SINAD reading around 79dB. This behavior seems to show conclusively that this amp module is unwell.

I also took a phone video showing the procession of harmonics sliding up and down the frequency scale. See below:

I'm interested in anyone's thoughts about what could be happening. By this point I don't think there's anything in the amp build, setup, or external to the amp that is contributing to this, it's entirely caused within the amp module itself.

Note: I made a mistake in the image above, the peak is clearly shown to be outside the range from 6kHz to 8kHz; my point is that it moves around in the audible treble range, not the ultrasonic range.

I wonder if this could be related: https://www.audiosciencereview.com/...368-has-high-pitched-tone-at-all-times.49375/

 

[ThatSoundsGood]

Yeah, that's a bummer. My Hypex amps definitely don't do that so there is clearly something wrong with it. I can't help with the electronics side, but it obviously needs to be repaired.

 

[AcousticTheory]

I don't think I'm an expert using the AP analyzer, so I wouldn't trust myself to take a "real" SINAD measurement, but I was able to bring that measurement into the panel, and into a 4-ohm speaker load at 5W/1kHz I was able to record a SINAD in the upper 70's - maybe low 80's with this module, and that seems quite low. I had some help setting up the amp for measurement using the AUX-0025, so the real-time FFT in my video is 'real'.

On the topic of a repair, I bought a DIY amp second-hand, so there's no warranty whatsoever. If everything had been perfect, I'd be groovy, but clearly it's not. Hypex doesn't have any obligation to speak to me because this is an OEM module, but my next step is going to be to reach out to them. Maybe they would like this amp module back for investigation. If not, I might be able to repurpose it to be a subwoofer amp; I've got stereo passive subs. These modules are available here and there, but they usually go for just a little less than $500, which is a bitter pill to swallow; for that price I can have two Icepower 300AS1 in Ghent cases fully built out, and they're not as 'cool' as the Hypex NCore MP modules with their post-filter feedback and very high performance design in other areas, but the 300AS1 still sounds pretty decent. So I guess I'm going to have to think for a bit about what to do. I bought the NC252MP amp to have something portable I could take places, or use in a second system; the NCx500 monos are still my best amps in technical performance and in sound.