The simplistic Salas low voltage shunt regulator

Status
Not open for further replies.
Jump to Latest
#4,003
A diode drops ~600mV to 700mV depending on the current passing the diode.
A series pair will drop ~ 1.2V to 1.4V.

The diodes acts like a shunt regulator, much the same as a Zener Diode shunts away excess current.
The CCS sets the total current flowing to the load and flowing through the shunting diodes.

If the CCS is set to a current that exceeds the load current then the diodes shunt away the excess current and as a result, drop a voltage, Vf. Vf can be set by adjusting the CCS current. Vf is your output voltage to feed the load.
 
#4,005
AndrewT said:
A diode drops ~600mV to 700mV depending on the current passing the diode.
A series pair will drop ~ 1.2V to 1.4V.

The diodes acts like a shunt regulator, much the same as a Zener Diode shunts away excess current.
The CCS sets the total current flowing to the load and flowing through the shunting diodes.

If the CCS is set to a current that exceeds the load current then the diodes shunt away the excess current and as a result, drop a voltage, Vf. Vf can be set by adjusting the CCS current. Vf is your output voltage to feed the load.
Click to expand...

The question is what is the noise level of a pair of diodes compared to a single LED?
 
#4,014
Salas said:
A string has more impedance to filter easier by the cap and staggers the dissipation between pieces which is a good thing. If with one fat zener, add 100-220R in series to help the filtering. They don't add up noise linearly so its not that bad using more. The cap should be bigger than when with a resistor ref (you did well with the one you used), the zeners are livelier noise sources (like pink type noise), unless we talk TC compensated defence contract style ones which are silent, though hard to come by. Any pics?
Click to expand...

Hi Salas, listening better, the sound is rolled off at hi frequencies, the cymbals are too silky and beyond the soundstage, like there was a gently low pass filter
in the hi treble region.
I'm using a led string (5x12v) bypassed with 680uF.Some years ago, working on
a CCS, I tried to bypass a single zener(was a 12v one) with different values of capacitance and I noted subjectively that when the cap was big,there was more
bass and viceversa.
Today I made some try; I used a 220R+ led string and bypassed all with a Pan
220uF. Now the sound it's a bit better, the noise is a bit less, but the timbre it's not correct again.Finally I changed the cap with a BG 10uF. Now the sound is more vivid and real and the noise it's the same with the 220uF cap.
Question:bypassing a zener, different values of capacitance can change the
timbre( i.e. the freq response) of a circuit?

PS , I'm talking of listening impressions, I made no measurements.
 
#4,015
Because you talk electrolytics and Zeners, yes there are differences for noise level, its shape, the filtering capacitors values, their THD, bandwidth. With LEDs & 220R the capacitance value can be substantially lower than with a Zener. Depends on the audio circuit's gain and PSRR a lot also. If 10uF was enough in the particular application, you can even try a plastic cap.
 
#4,018
It was the fifth harmonic of your clock if I remember correctly and was there with the previous psu too. When you are at 100MHZ region it can be picking by the probe's gnd crock inductance even. I don't think that the 1.2R can produce anything that high on itself anyway. Did you try ferrite beads on suspicious places?
 
#4,019
Yes, I did that too. Both regular and smd ferrite beads were put on the test. I read the article provided by ikoflexer (big thanks!), and found this is not an easy task to just find out the cause. Due to my lack of knowledge, I ordered 'Electromagnetic Compatibility Engineering ' - a book by Henry Ott. It seems there are certain techniques to deal inteference problem. I hope I can solve it one day.
 
Status
Not open for further replies.
Top Bottom