Just for the purpose of testing I used a crappy chinese 30V switching supply adjusted down to 26.6V.
You can see increasing the power supply voltage reduces the higher order harmonics.
The harmonic seen around 40kHz is purely due to the switch mode supply.
You can see increasing the power supply voltage reduces the higher order harmonics.
The harmonic seen around 40kHz is purely due to the switch mode supply.
Good information Pico. Thank you.
So there is a reason for better 24 V sound vs 19 V.
Have you changed only the lower FET?
Best regards
Ernst
So there is a reason for better 24 V sound vs 19 V.
Have you changed only the lower FET?
Best regards
Ernst
Top and bottom of one channel.
I'll go ahead and purchase some 24V supplies, and hopefully they turn out to be as good as the Asus supply.
I'll go ahead and purchase some 24V supplies, and hopefully they turn out to be as good as the Asus supply.
I got a 24 VDC Meanwell supply rated at 14.5 amps for powereing 4 ACA cards that I've used with two cards for several months now and it works very well. I put the supply in a separate box and am using Powercons for the umbilical.
I haven't heard and\ ACA with a good linear supply, but that Meanwell powering the ACA cards is clearly the best amp I've had in my system. I've several more months of work to get to my finished amps and will do a full report when done.
Skip
I haven't heard and\ ACA with a good linear supply, but that Meanwell powering the ACA cards is clearly the best amp I've had in my system. I've several more months of work to get to my finished amps and will do a full report when done.
Skip
Last edited:
No arguement.
Most musically enjoyable amp I've heard, comes closest to sounding like real musical instruments.
The only caveat is you need appropriate speakers.
Last edited:
Exactly 24V + Papa Mod
I pulled out a decent lab supply and set it at 24V
THD is roughly the same but higher order harmonics have been reduced compared with 19V supply as previously observed with the 26.6V supply.
Vds across Q1 was set at 12.5 V
Bias is around 1.35A with Papa mod.
I pulled out a decent lab supply and set it at 24V
THD is roughly the same but higher order harmonics have been reduced compared with 19V supply as previously observed with the 26.6V supply.
Vds across Q1 was set at 12.5 V
Bias is around 1.35A with Papa mod.
I'll probably settle on a Bias current around 1.15A, and see what temperature the heatsinks get to.
Cheap 24V switching supplies arrived.
Not really happy with the noise at 50kHz.
I have some Roland Keyboard supplies to test tomorrow. If they are no better I'll build my own regulated supply.
Not really happy with the noise at 50kHz.
I have some Roland Keyboard supplies to test tomorrow. If they are no better I'll build my own regulated supply.
Scope images are interesting certainly.
BUT... Is there audible difference from your speakers ?
IMO ..All info is suspect, if your speakers are not the the Equal of Nelson's 🙂
Apples to bananas is an irrelevant comparison... imo.
BUT... Is there audible difference from your speakers ?
IMO ..All info is suspect, if your speakers are not the the Equal of Nelson's 🙂
Apples to bananas is an irrelevant comparison... imo.
Maybe his speakers are objectively better than Nelson's speaker?
Maybe he used a turntable instead of a digital system?
What if?
Maybe he used a turntable instead of a digital system?
What if?
I've got no idea what you're on about.
Most non-sensical thing I've heard in my life.
There is no speaker in the chain here. I'm purely measuring the performance of power supply and the amp into a 8 Ohm resistive load.
The aim here is to find a decent cheap 24V switching supply.
Last edited:
I find that the switching supplies are superior to a linear supply any day if you want the ultimate low noise, no hum PSU. The trick is to use a high quality name brand 19v SMPS brick from HP/Dell/Etc. then use a DC to DC step up converter. Now you can dial in any voltage to 35 or 45 in some models. Mains hum will be non existent and use a CRCRC to filter out any residual hash. Bypass the big caps with low ESR x7r. I have *Never* been able to achieve a 5dB bump above -130dB noise floor for any linear PSU I have made. And I have tried very hard. Just can't be done - at least not with normal linear supplies with a trafo. The reason I say major namebrand SMPS brick is that they are engineered to be low noise. I tried cheap no names ones and they are very noisy, except for a 24v 5amp LED PSU I got for $17 once. It was just as clean and I ran it through a cap multiplier to drop to 20v and knock off residual 60Hz hum. Usually the DC step up gets rid of the hum.
Look at what the measurements I have for a SE Class A running at 28v and 1.25amps. This is with a HP 19v 4.65amp supply feeding a DC step up followed by CRCRC. This is a headphone amp normally driving maybe 50mW into 50ohms and I was doing a stress test by pushing 8ohms to 1000mW.
This is 8v peak to peak into 8ohms. Look at the absence of any mains related peaks:
With a linear trafo, bridge, usual CRCRC, the hum generating peak at 60Hz is around -80dB (50dB higher).
Look at what the measurements I have for a SE Class A running at 28v and 1.25amps. This is with a HP 19v 4.65amp supply feeding a DC step up followed by CRCRC. This is a headphone amp normally driving maybe 50mW into 50ohms and I was doing a stress test by pushing 8ohms to 1000mW.
This is 8v peak to peak into 8ohms. Look at the absence of any mains related peaks:
With a linear trafo, bridge, usual CRCRC, the hum generating peak at 60Hz is around -80dB (50dB higher).
Last edited:
That's a nice result, about the same as my Asus laptop supply.
I can do better than that with my regulated supply but it really comes down to cost and time to build your own.
If I can find a 24 V switcher for under $50, with performance like that I'm certainly going to use the switcher.
If it costs over $100 it's starting to get in the territory of building my own.
I have another idea that hasn't been tested yet. I might give that a try.
I can do better than that with my regulated supply but it really comes down to cost and time to build your own.
If I can find a 24 V switcher for under $50, with performance like that I'm certainly going to use the switcher.
If it costs over $100 it's starting to get in the territory of building my own.
I have another idea that hasn't been tested yet. I might give that a try.
Last edited: