I had to reduce the R1 negative tup to 2.7R. So that the voltage does not decrease when the load is connected
Last edited:
One thing is there is some Vbe voltage loss in the start for Q2 due to transistor temperature coefficient physics. Because you use SMD probably more pronounced. That steals some current from the Vref and decreases Vout until thermally settled. Another thing is you need to allow enough spare current beyond load, say 100mA more. With R1=3.9R you had 230mA limit (0.9V/3.9R) and 170mA load (17V/100R).
I looked carefully. In the store I was given instead of BC850 - (1G) BC847(((/ I think it's not scary.
That's +60mA spare, it better was more. I don't know about Q1 Vbe ~0.9V, seems bit high or maybe right, the standard reg uses through hole BC327/337.
How much rectified DC you got on the input?
This board is for experiments. I have BC327/337, BC546/556 I can try to solder it carefully in the evening, for the purity of the experiment.
You must have at least 5V higher on C1 than your Vout target for good stability. If you don't, use higher VAC transformer or target a lower Vout.
I replaced the transformer with a more powerful one ( today I connected a transformer with a low V A). Now the AC voltage is 18.5 V, on C1 22.5V, the voltage on R1 is 0.56 V.
Now you have the usually expected voltage range on R1. By the way your transformers show relatively high regulation % loss under load. And/or its your bridge diodes loss. How much Vf drop on each diode?
How to check it correctly? The multimeter in the diode check mode shows 483. In the on state, the voltage on the diodes is 8.33 V. MUR120 diodes