Just one word to the tantalum caps, maybe someone will use them (I think about it): they do not cope well with high pulse currents, they cannot tolerate overvoltage so choose one with appropriate voltage, and the superimposed AC voltage must not cause a polarity reversal.
On the bench tonight is my old LuFo Lite amp. I had a broken halogen lamp (resistive load) on it for a while and repaired that. While I was at it, I had some unused Mark Johnson Scourge front end boards waiting for a home. These are JFET input with BJT output stage unity gain buffer pushing an Edcor 1:5 step up transformer for voltage gain. Normally used in a Pass VFET amp (which I am already using the stock input stage).
I am powering the Scourge with a 30v SMPS and the LuFo lites are each powered with a 24V 5A SMPS (although it’s only running a little under 1A current).
I hooked up my XKi foamcore soeakers with Dayton RS100-8 drivers and I have to say, they sound quite nice. I can remember why I got into audio with fullrange drivers and small Class A amps. A lot of fun.
The lamps are very bright and a bit annoying so the best thing is to turn the amp into a lamp with a lampshade. You really only use this amp when you are in the room - sort of like a lamp. Turn off when leaving.
I am powering the Scourge with a 30v SMPS and the LuFo lites are each powered with a 24V 5A SMPS (although it’s only running a little under 1A current).
I hooked up my XKi foamcore soeakers with Dayton RS100-8 drivers and I have to say, they sound quite nice. I can remember why I got into audio with fullrange drivers and small Class A amps. A lot of fun.
The lamps are very bright and a bit annoying so the best thing is to turn the amp into a lamp with a lampshade. You really only use this amp when you are in the room - sort of like a lamp. Turn off when leaving.
Last edited:
On the bench tonight is the LuFo Lite amp and some FFT measurements. The setup is using a Focusrite 2i4 and Victor’s low distortion 1kHz oscillator. I’m also looking at the noise floor of the Jhofland Pearl 3 phono stage at the output of the LuFo Lite.
Here’s the setup:
Here is FFT with 355mV rms 1kHz sine wave - this amp has lots of distortion - in the 1.5% THD range but it is dominant 2nd order and monotonically descending. Very SET tube like and thick and sweet sound.
Notice that there is some hash near 3-5kHz. It’s quite low but that is a very sensitive region to the ears. I suspect it may be coming from the SMPS interaction and this produces low frequency beat notes that are much lower than the Class D switch frequency.
I plugged in the SnubWay Noise Defender mains SMPS line noise filter into the outlet strip containing all the PSU’s for the LuFo Lite.
And here is the FFT with the SnubWay installed:
Note the reduction in the “grass”’near 3kHz to 5kHz. The SnubWay is cleaning up the. Noise floor of the amp.
Next is to try measuring the noise floor of the Jhofland Pearl 3 phono stage with MC cartridge installed.
Here is noise floor without the SnubWay:
Here is with SnubWay installed:
Here’s the setup:
Here is FFT with 355mV rms 1kHz sine wave - this amp has lots of distortion - in the 1.5% THD range but it is dominant 2nd order and monotonically descending. Very SET tube like and thick and sweet sound.
Notice that there is some hash near 3-5kHz. It’s quite low but that is a very sensitive region to the ears. I suspect it may be coming from the SMPS interaction and this produces low frequency beat notes that are much lower than the Class D switch frequency.
I plugged in the SnubWay Noise Defender mains SMPS line noise filter into the outlet strip containing all the PSU’s for the LuFo Lite.
And here is the FFT with the SnubWay installed:
Note the reduction in the “grass”’near 3kHz to 5kHz. The SnubWay is cleaning up the. Noise floor of the amp.
Next is to try measuring the noise floor of the Jhofland Pearl 3 phono stage with MC cartridge installed.
Here is noise floor without the SnubWay:
Here is with SnubWay installed:
On the bench is the same test now looking at the direct MC Phono preamp outputs on the FFT analyzer. This is Jhofland’s Pearl 3 discussed earlier. This is a different power outlet strip containing a laptop SMPS and an LED SMPS along with other desktop accessories. But the three SMPS from the previous test are not connected to this strip.
Here is comparison of the effect of the SnubWay Noise Defender in and out of this outlet strip. The green trace is offset for clarity.
Here is comparison of the effect of the SnubWay Noise Defender in and out of this outlet strip. The green trace is offset for clarity.
This is how the SnubWay is used in practice. Just plug it into the outlet strip to get cleaner audio noise floor.
Here was the setup for just measuring the noise floor of the MC Phono stage.
Technical details: 256k length FFT acquired at 192kHz Blackman Harris 7 window function with 16 averages. Focusrite 2i4 USB audio interface and REW software. 100kHz 8th order passive brick wall filter on FFT input to prevent data aliasing from high frequency content. FFT scale: 0dB FS is 2.2vrms
Here was the setup for just measuring the noise floor of the MC Phono stage.
Technical details: 256k length FFT acquired at 192kHz Blackman Harris 7 window function with 16 averages. Focusrite 2i4 USB audio interface and REW software. 100kHz 8th order passive brick wall filter on FFT input to prevent data aliasing from high frequency content. FFT scale: 0dB FS is 2.2vrms
Sorry this is a commercial product and not public. I can say that it is a parallel 5 LRC filter with front end MOV/gas discharge lightning protection circuit. The circuit was designed using LTSpice for attenuation of 10kHz to 1MHz noise spikes in mains.
On the bench is the new A40 Class A amp pre-production sample getting built. We are discovering lots of little issues as we go and that’s the normal part of why we do this. To build a comprehensive list of fixes that need to be applied for production. Nothing major. Work on the front on the front panel PCB is progressing well. The VU meters have a place to plug in along with the main power switch and mute switch. All this connects to the main SFPP/SSR board on the amp floor board via a single ribbon cable with MicroMatch connectors. A big thanks to Keantoken for taking on this big job!
In the main chassis we can see the A40 amp boards proper mounted on the heatsinks and the SFPP/SSR board in the floor:
Here is a closer look at the amp boards:
In the main chassis we can see the A40 amp boards proper mounted on the heatsinks and the SFPP/SSR board in the floor:
Here is a closer look at the amp boards:
The Snubway is looking like a cool product! Maybe it can be the new must-have item for people using unnecessarily powerful class d amps with huge switching mode power supplies inside. I.E. everyone on audiosciencereview 
(and me, sometimes).
nice to see the A40 move closer to production!
(and me, sometimes). nice to see the A40 move closer to production!