LM317 tracking pre-regulator and current regulator

[yagoolar]

Based on tracking pre-regulator circuit form National and this link I expect to achieve 1A regulated current. But in reality ripple voltage is around 0.2V RMS.
The voltage across second reg is around 2V and does not change if one or two red LEDs are connected between Vout2 and Vadj1.
Sims in LTspice are ok.

[yagoolar]

Two setups were tested - same result.
Below there are LT1086, just for comparison - LM317T were used for both

[peranders]

Do you plan to built it also?

[yagoolar]

I have built it already. "Reality" is the key.
I added 8mF+8mF to voltage doubler (VD), so results in 12,7mF+12,7mF . There is visible improvement in ripple. So I should go 20mF+ for VD, I think.

[yagoolar]

Here it is. On the left there is the load resistance of 4 33ohm in parallel

[sofaspud]

The voltage across second reg is around 2V and does not change if one or two red LEDs are connected between Vout2 and Vadj1.

Are you sure about that?

[yagoolar]

Originally Posted by sofaspud

Are you sure about that?

When 4,7mF caps were in the doubler - 100%. And it was odd. Changing R7 did not help.

[sofaspud]

The drop across reg2 should be 1.25V + the voltage drop across the LEDs. So it should definitely change 1 LED vs 2 LEDs. Unless something is wrong.
This topic hits close to home, because the below circuit is next in my circuit build queue:

[yagoolar]

"Unless" is the keyword.

[sofaspud]

OK. So if the drop is 2V across reg2, what is the drop across R7?