I'm using a Meanwell EPP-200-27 SMPS to power a TPA3251 based amplifier, see here for pics. As shown in the pictures, I was not using the auxiliary output, only the main 27V output for the amp. Today I connected a LED (with a current-limiting 2k resistor in series) to that aux power.
With the LED connected, there is a noticeable hiss/whine coming from the power supply. If I remove the LED, the hiss/whine diminishes considerably. It's still there, but quiet enough that it can only be heard with the cover off the case and standing right by the amp. But with that LED plugged in, the noise is hard to ignore, even with the case lid on and standing far away from the amp.
Is this a "feature" of the power supply, or have I made a mistake in implementation somewhere? Any way to reduce the noise? I can power the LED with the main 27V supply, but that auxiliary supply is lower voltage and just seemed more appropriate for a power-on indicator light.
Thanks!
With the LED connected, there is a noticeable hiss/whine coming from the power supply. If I remove the LED, the hiss/whine diminishes considerably. It's still there, but quiet enough that it can only be heard with the cover off the case and standing right by the amp. But with that LED plugged in, the noise is hard to ignore, even with the case lid on and standing far away from the amp.
Is this a "feature" of the power supply, or have I made a mistake in implementation somewhere? Any way to reduce the noise? I can power the LED with the main 27V supply, but that auxiliary supply is lower voltage and just seemed more appropriate for a power-on indicator light.
Thanks!
Technically it's the "fan supply", see the datasheet. 12v/0.5a +/-15%. (See the "function" line in the middle of the table on the first page of the DS; also connector CN101 on the last page of the DS.)
Edit: See attached pic, I circled the fan supply in red (nothing plugged into it in this picture).
Edit: See attached pic, I circled the fan supply in red (nothing plugged into it in this picture).
Attachments
Last edited:
I understand your surprise that 5mA loading can cause such an effect.
I would try to give it a 470uF capacitor across the LED and see if such smoothening of the load should help.
If not, try to see if it is the simple loading of the 12V output that causes the instability. First the 2K resistor alone, then a 220 Ohm resistor alone.
I would try to give it a 470uF capacitor across the LED and see if such smoothening of the load should help.
If not, try to see if it is the simple loading of the 12V output that causes the instability. First the 2K resistor alone, then a 220 Ohm resistor alone.
FauxFrench, thank you for the ideas. The 470uF cap across the LED had no effect either way as far as I could tell.
Then I went straight to a 102R resistor (technically two 51R in series). That actually resulted in a noticeably louder hiss than the LED.
Just for kicks I also attached a 12v/0.16a computer fan, since that's what the header is designed for. Same result: fan turned, but the hissing was quite loud (similar to the 100R power resistor).
Then I went straight to a 102R resistor (technically two 51R in series). That actually resulted in a noticeably louder hiss than the LED.
Just for kicks I also attached a 12v/0.16a computer fan, since that's what the header is designed for. Same result: fan turned, but the hissing was quite loud (similar to the 100R power resistor).
This is something the manufacturer should have noticed.
You can of course move your "power ON" LED to the main output and you have little instability. To start redesigning the supply is hardly worth the effort.
The 12V output is generated by a tertiary winding (shown in the block diagram) and the whining could be caused by switching of the rectifier diodes when they start carrying some current. Could they have used Schottky rectifiers for lower voltage drop? If the tertiary winding has got importance leakage inductance, it may start a resonance with the rectifier current spikes. If so, the tertiary winding should have been better made or the rectifier diode current spikes damped, eventually with snubbers.
At least, not an impressive product in particular because the intended use (a fan) causes clearly audible discomfort.
You can of course move your "power ON" LED to the main output and you have little instability. To start redesigning the supply is hardly worth the effort.
The 12V output is generated by a tertiary winding (shown in the block diagram) and the whining could be caused by switching of the rectifier diodes when they start carrying some current. Could they have used Schottky rectifiers for lower voltage drop? If the tertiary winding has got importance leakage inductance, it may start a resonance with the rectifier current spikes. If so, the tertiary winding should have been better made or the rectifier diode current spikes damped, eventually with snubbers.
At least, not an impressive product in particular because the intended use (a fan) causes clearly audible discomfort.
Last edited:
I don't suppose you (or anyone else) has a direct email address?
When I try submitting via their Technical Service web form, I hit submit, then it just hangs on "Please wait...". I tried on both Chrome and Firefox, same thing. Can't tell if it's going through or not.
Edit: I also sent emails to info@meanwell.com and info@meanwellusa.com. We'll see...
Last edited:
Give them a few days. If they don't get back to you, try contacting TRC Electronics. They're a distributor of Mean Well and might have a direct email address.
Tom
Tom
While I wait to hear back from Meanwell, I have a bad feeling this may be a "feature" of the EPP series: I have another one of these PSUs, the 24 volt version (EPP-200-24). I just tested my LED on its fan header, same result: whining/hissing. I can of course put the power on LED on the main output. But it does take away from what I thought was a near perfect match for the amp (small size, high efficiency, PFC, not expensive).
It's pretty typical for power supply vendors to have a sort of "burst mode" operation at light load, in order to meet power efficiency regulations. This burst mode operation can result in effective audible switching frequencies, resulting in the types of hiss you're hearing. If you were to load down the power supply to probably 15-20% of rated output, the noise should go away.
Off-the-shelf power supplies typically have some drawbacks like this that aren't terribly desirable for audio applications.
Off-the-shelf power supplies typically have some drawbacks like this that aren't terribly desirable for audio applications.
If I can solder cleanly, I might do this. Otherwise, I'll just go with the main power output.
I think you're on to something here. Meanwell got back to me, one of things that was said was, "there is a chance the whining noise caused by low loading on the main output."
I'm using an AC meter (Kill-a-Watt) to see how much AC power I'm drawing: 4 to 5 watts / 20 to 21 volt-amps. For a 200W power supply, I think this counts as low or light loading. The amp could theoretically max out the power supply. But for my purposes, I'm certain I'll never get anywhere near that. My speakers are reasonably efficient (87 dB IIRC), and my typical listening levels are well below 90 dB (according to a free SPL app on my phone). Not precise measurements by any means, but I'm certain my actual power needs are generally in the single-digit range.
In fact, that 4-5W is what is being drawn from the AC supply. This doesn't include the efficiency of the MeanWell supply itself at light load, which won't be terribly high. So probably from the output of the MeanWell, you may only be drawing 2-3W.
For audio applications, disabling light load bursting, ensuring high electrical current capability, but not necessarily thermal capability are desirable traits of power supplies. Since audio is "crest-y", the power supply doesn't need to provide sinusoidal output power forever, so if it's a little under-designed in that regards, that's acceptable.
Any off-the-shelf AC-DC power supply isn't likely going to have these characteristics, since they're designed to be used with just about any electronic application.
It's not unusual for converters to be designed for standard* operation from 10% of maximum load upwards. Standard* means continuous conduction of the output inductor and fairly steady pulse width modulation. Below 10% loading the PWM will vary according to the load and input voltage, with the control loop behaving in an, ideally, predictable manner. The manner in this case is probably a fast hiccup or lower frequency of operation. You have probably got it into an acoustically noisy region. The solution might be just to add a little dummy load. Personally I only use off-the-shelf SMPSU's for class A amplifiers.
So does it also mean my Meanwell LR150-36 is also as per design? when it is connected to a Class-D for nearfield listening, this Meanwell produce whinning, cld be a couple-of-meter distance.
Can you hear the PSU? If YES, add a dummy load resistor until the noise goes away. If NO, there's no problem.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.