Hi Andrew
Thank you for the very clear explanation.
What is your opinion about a precision reference like LM329 ? The pdf states a dynamic impedance of 0.8ohm.... Can it be used for vref in this case ? What about tjhe value of the filtering cap in this case ?
I concur Andrew, bloody marvellous explanation!
Though may I ask, what effect the dynamic imp has on the regulation? Does it "react" quicker to give a smoother output or holds the output voltage more stable under temperature variations?
Look forward to an informative answer. Thanks for putting up with those of us that don't know all this stuff!
Though may I ask, what effect the dynamic imp has on the regulation? Does it "react" quicker to give a smoother output or holds the output voltage more stable under temperature variations?
Look forward to an informative answer. Thanks for putting up with those of us that don't know all this stuff!
With TeaBag's GB being dispatched soon, I though it was time to start working on my BOM.
The goal is to power my DD turntable with the BiB.
The TT needs ~200mA at normal spinning and ~400mA peak when starting rotating, all @21V.
Because the load is both digital and inductive, I think that going hot-rod would be a nice idea. Maybe as high as 1A?
The input DC is going to be 30V (2x12V secondaries @120VA)
According to the excel sheet, the CSS should be 3leds and R101 0,5R at 0,5W
Since the voltage is on the high side, I guess that a zener (20V 0,5W ?) filtered with lytics should be a better choice for the reference, along with a 1K trimmer for tweaking.
Unfortunately Q106 results in dissipating almost 17W and will need a substantial size of heatsink. So I really need your opinion on if hot-rodding would really be useful in my case. Dropping the current to 600mA results in ~9W at Q106 which again is high but really more manageable.
Thanks a lot for your time
Dimitris
The goal is to power my DD turntable with the BiB.
The TT needs ~200mA at normal spinning and ~400mA peak when starting rotating, all @21V.
Because the load is both digital and inductive, I think that going hot-rod would be a nice idea. Maybe as high as 1A?
The input DC is going to be 30V (2x12V secondaries @120VA)
According to the excel sheet, the CSS should be 3leds and R101 0,5R at 0,5W
Since the voltage is on the high side, I guess that a zener (20V 0,5W ?) filtered with lytics should be a better choice for the reference, along with a 1K trimmer for tweaking.
Unfortunately Q106 results in dissipating almost 17W and will need a substantial size of heatsink. So I really need your opinion on if hot-rodding would really be useful in my case. Dropping the current to 600mA results in ~9W at Q106 which again is high but really more manageable.
Thanks a lot for your time
Dimitris
Is ok this formula:
Rd = 26mV @ 25ºC / If
For example dide 20mA will be:
Rd = 26mV / 20mA = 1.3 ohms
For a resistor will be:
Rd = 26mV / (I:V/R across Resistor)
26mV is given by the formula where
k is the Boltzmann constant k=1.38e^-23 joules/Kelvin
T is the absolut temperature in Kelvin
q is the electron charge q=1.602e^-19 coulomb
This yields approximately a Vt of 25mV @ 20 or
26mV @ 25
Rd = 26mV @ 25ºC / If
For example dide 20mA will be:
Rd = 26mV / 20mA = 1.3 ohms
For a resistor will be:
Rd = 26mV / (I:V/R across Resistor)
26mV is given by the formula where
k is the Boltzmann constant k=1.38e^-23 joules/Kelvin
T is the absolut temperature in Kelvin
q is the electron charge q=1.602e^-19 coulomb
This yields approximately a Vt of 25mV @ 20 or
26mV @ 25
Some makers publish graphs. Most LEDS like that are about 20 Ohm Z at nominal current. Regarding references its a big chapter, and this is a builds thread. Very fast, dynamic impedance of a reference is not only about absolute minimum, but is also about matching to any current source and the error amp circuit. Its about stiffness, harmonic, and random noise, lending to easy filtering etc.
For 15Vout now I'm using a string of 8 red LEDs at Vref & sounds amazing, can be improved with:
Vbe 0.6V
2 x LM329DZ 6.9V = 13.8V
1 x 1N4007 0.6V = 0.6V
Total 15V
For 45Vout now I'm using 2 1.9V LED, 6.8k R303, 5K trimmer. Other parts jumper. As can be a lot of LEDs I want to reduce at maximum the static resistance to do that I have on hand:
4 x LM4040DIZ 10V
10 x LM329DZ 6.9V
4 x LM336Z 5V
Wich option will be the best for 45Vout
1st option
0.6V vbe
3 x LM329DZ 6.9V = 20.7V
2 x LM4040DIZ 10V = 20V
2 x red LEDs 1.9V = 3.8V
Total 45.1V
2nd option
0.6V vbe
2 x LM336Z 5V = 10V
2 x LM329DZ 6.9V = 13.8V
2 x LM4040DIZ 10V = 20V
1 x 1N4007 0.6V = 0.6V
Total 45V
3rd option
0.6V vbe
2 x LM336Z 5V = 10V
2 x LM329DZ 6.9V = 13.8V
2 x LM4040DIZ 10V = 20V
1 x 1N4007 0.6V = 0.6V
Total 45V
Vbe 0.6V
2 x LM329DZ 6.9V = 13.8V
1 x 1N4007 0.6V = 0.6V
Total 15V
For 45Vout now I'm using 2 1.9V LED, 6.8k R303, 5K trimmer. Other parts jumper. As can be a lot of LEDs I want to reduce at maximum the static resistance to do that I have on hand:
4 x LM4040DIZ 10V
10 x LM329DZ 6.9V
4 x LM336Z 5V
Wich option will be the best for 45Vout
1st option
0.6V vbe
3 x LM329DZ 6.9V = 20.7V
2 x LM4040DIZ 10V = 20V
2 x red LEDs 1.9V = 3.8V
Total 45.1V
2nd option
0.6V vbe
2 x LM336Z 5V = 10V
2 x LM329DZ 6.9V = 13.8V
2 x LM4040DIZ 10V = 20V
1 x 1N4007 0.6V = 0.6V
Total 45V
3rd option
0.6V vbe
2 x LM336Z 5V = 10V
2 x LM329DZ 6.9V = 13.8V
2 x LM4040DIZ 10V = 20V
1 x 1N4007 0.6V = 0.6V
Total 45V
Vref cap depends application as stated in the BiB manual: Clocks, MC pre Lytic rest: lineamps. buffer, MM pre Film.
Dynamic impedance:
LM336Z 5V 0,6 Ohms
LM329DZ 6.9V 0,6 Ohms
LM440DIZ 10V 0,8 Ohms