PS for Pi?
Not sure if this is appropriate for this thread or if I should start a new one. My question concerns using the low voltage positive supply section of the black board for a linear power supply for a Raspberry Pi/DAC combo. The Pi recommends a 2A PS and the DAC sits on top of it and gets its supply from the Pi, so I am thinking maybe 3A max @ 5V. BJT or IFR? Any other issues to consider? I already have board (I used other two parts for +/- 25V.)
Thanks
Not sure if this is appropriate for this thread or if I should start a new one. My question concerns using the low voltage positive supply section of the black board for a linear power supply for a Raspberry Pi/DAC combo. The Pi recommends a 2A PS and the DAC sits on top of it and gets its supply from the Pi, so I am thinking maybe 3A max @ 5V. BJT or IFR? Any other issues to consider? I already have board (I used other two parts for +/- 25V.)
Thanks
BIB as is is not reliable for >1A. You would need slower big MOSFETS like IRFP9230/230 on big sinks to run 3A. But from what I see here https://learn.adafruit.com/embedded-linux-board-comparison/power-usage the Pi demands are lower. Measure your DAC demands also. A 0.1R in series with its now PSU feed could indicate by voltage drop across it and Ohm's law.
I have the Pi2 which is a quad core @ higher clock rate and more RAM than A/B and I have read reports that it "stumbles" on less than 1800mA (without the DAC) so I guess I'll have to look elsewhere.
Thansk
There was a Linear Audio Magazine survey with thorough tests by Jack Walton
The Salas regulator was an SSLV1.1 with couple of Leds & resistor Vref, just 100uF Vref filter capacitor, and 0.15A CCS setting
http://linearaudionet.solide-ict.nl/sites/linearaudio.net/files/V4 JW F5.pdf
http://linearaudionet.solide-ict.nl/sites/linearaudio.net/files/V4 JW F7.pdf
The Salas regulator was an SSLV1.1 with couple of Leds & resistor Vref, just 100uF Vref filter capacitor, and 0.15A CCS setting
http://linearaudionet.solide-ict.nl/sites/linearaudio.net/files/V4 JW F5.pdf
http://linearaudionet.solide-ict.nl/sites/linearaudio.net/files/V4 JW F7.pdf
LEDs as references
One remark: Many LEDs have superbly low noise. They also have a significant temperature coefficient of the forward voltage Vf, which for some is a decent match to the Vbe of a silicon bipolar. This has been exploited by many in current generators, with a resistor in the emitter and the base tied to the LED, with the resulting collector current fairly constant with collector voltage and temperature. However the noise in the current is limited by the current noise in the emitter resistor, along with the shot noise in the base current and the thermal voltage noise of the base spreading resistance. The delta V, Vf -Vbe, is a bit low for best noise performance.
The quietest voltage reference I've found to date is when it is derived from a cascoded or multiple-cascoded JFET arrangement as a ~zero-tempco current generator. The reference voltage can then be developed across a parallel R-C and buffered. The low-frequency excess noise in particular can be exceedingly low, much better than any commercial voltage reference.
The somewhat-impractical aspect of such a reference is due to the wide parameter variations in JFETs, which require one to test a sample and make adjustments for low tempco.
One remark: Many LEDs have superbly low noise. They also have a significant temperature coefficient of the forward voltage Vf, which for some is a decent match to the Vbe of a silicon bipolar. This has been exploited by many in current generators, with a resistor in the emitter and the base tied to the LED, with the resulting collector current fairly constant with collector voltage and temperature. However the noise in the current is limited by the current noise in the emitter resistor, along with the shot noise in the base current and the thermal voltage noise of the base spreading resistance. The delta V, Vf -Vbe, is a bit low for best noise performance.
The quietest voltage reference I've found to date is when it is derived from a cascoded or multiple-cascoded JFET arrangement as a ~zero-tempco current generator. The reference voltage can then be developed across a parallel R-C and buffered. The low-frequency excess noise in particular can be exceedingly low, much better than any commercial voltage reference.
The somewhat-impractical aspect of such a reference is due to the wide parameter variations in JFETs, which require one to test a sample and make adjustments for low tempco.
They all are repeatedly, but since its long to dig now go to the bib builds thread. Has pdf on first post. http://www.diyaudio.com/forums/power-supplies/192625-sslv1-1-builds-fairy-tales.html
The loop areas aren't tight and long distance bypassing isn't productive, but it should work electrically. As for correct routing I am not sure, can't be following it in detail with a fast look. Just check everything twice.
P.S. Where are the Leds for CCS? Using one Vbe transitor current control voltage?
P.S. Where are the Leds for CCS? Using one Vbe transitor current control voltage?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.