Looking for general help in understanding regulator design

[Toke]

Nice.

The 50 Ohm to 70mA looks a bit odd, I thought the 7805 would regulate current to get 5V across the resistors?
That would give you 100mA at 50 Ohm.

 

[comradmax]

I am not very clear on why I get what I get either... The regulator I used is the TO-92 little guy, and there is a circuit for current regulation in its datasheet, but it's diff - the input cap connects to ground directly. What I used was in the other datasheet, the one for the TO-220 guy, it shows all pins connecting together and then through LD to ground.

There is a formula in there, Io=Vxx/R1 (neglecting Iq for now), which is what you saw as well, and it gives 100mA with 50Ohm. However, in the TO-92 datasheet there is also a formula, Io=Vo/R1. Vo~4.2V, so Io=84mA, maybe a little closer to the true value, but I didnt actually measure Vo...

I think that somehow when operating in CC mode the chip forgets about that Vxx=5V and does whatever gets current to be the desired value. I feel experimentation is the key to DIY driver design

 

[comradmax]

Killed another PHR I tried the mini-driver with a test-load and it gave 70mA, then tried it with the LD while still in the bread-board - same thing. The driver ended up being just a tad too big for the inside the aixiz after I put one layer of electrical tape... I even got the module to screw back together, but when I powered it on again I got 140mA, then 400mA! Extracted the driver and it showed the same 400mA on the test-load, so somewhere between soldering to the LD and stuffing the driver inside the module I messed it up. Live and learn, I now have 3x LOC sleds to kill I am sticking to 'outside-the-tiny-module' stuff going forward:beer:

 

[Toke]

I had a few accidents before realising that a LM317 have output to case.
That means no lasing and very high tailcap currents.
No diodes were killed.