High Current Regulator Design with 7806

[iruraz]

Hi All,

I intend to make 6V high current regulator with 7806. In 7806's datasheet, there is reference design for high current. To find R1 and Io values, I need to know what "BQ1" is. I couldn't understand it. Can anyone explain in these expressions what "BQ1" refers to? Thanks in advance.

Regards.

7806 's datasheet:

http://www.fairchildsemi.com/ds/LM/LM7805.pdf

**broken link removed**

 

[keith1200rs]

I would guess it is the beta of Q1 or hfe.

Keith

 

[KerimF]

Io should be known by knowing first the equivalent DC resistance (R_load) of the load:
Io = 6 / R_load

BQ1 is the ratio Ic_q1 / Ib_q1

At high currents, we can approximately say:
Ic_q1 = Io

Unfortunately BQ1 is not constant. It decreases quickly near saturation (Vce < 1V for example) and at high currents.
Practically, we assume BQ1 as an intermediate value, says 50, and we let Vce_min not less than 1.5 V.

For the supply to be reliable, Vin_max should not be exceeded at the input and it is calculated from the dissipated power of Q1, then checked for the dissipiated power of LM7806.

Pmax_q1 = Io * ( Vin_max - Vout ) + Ib_q1 * Vbe_q1

Vin_max could be known if Pmax_q1, Io, Vout, Ib_q1 and Vbe_q1 are estimated based on the whole circuit board.
Pmax_q1 is related to the heatsink of Q1.
Io, as we saw, depends on the maximum load.
Vout is 6V here.
Ib_q1 is approximately equals to Io/BQ1 hence Io/50.
Vbe_q1 varies a little with the current Ib_q1 (from 0.75 to 1 V). It could be assumed as 0.9 V.

After calculating Vin_max, it is better to check the dissipation power of LM7806:
Pmax_reg = ( Ib_q1 + Vbe_q1 / R1 ) * ( Vin_max - Vbe_q1 - Vout )
Here another value is added (R1). The resistance of R1 is not critical as long Pmax_reg is within limit (also depends on LM7806 heatsink, if needed).
Usually I start with R1=33R. If I find out that, for the actual Vin_max, Pmax_reg is relatively low I can decrease R1.

Please note that designing a circuit properly is possible only when the external parts connected to it are well known. Otherwise we have to estimate and approximate Therefore if you notice a step on the above analysis being not logical to you, please don't hesitate to point it out. You know... I may do a silly mistake anytime

 

[iruraz]

@keith1200rs, @KerimF thanks for your help...