Awesome build, texas. I think these kinds of lamps are an ideal application for the Hi-CRI neutral 219. I need to give this a whirl.
What makes it so too low a of resistor value won’t allow you to dial up to full brightness (and vice versa)? That seems counterintuitive (I’m in a circuit analysis 1 class this fall)?
If you read the Datasheet for the PT4115E, you’ll find that there is an internal 200K pull-up to the internally regulated 5V supply. Reading further tells you that the DIM pin can be controlled by an analog voltage with 0.5V corresponding to OFF and 2.5V corresponding to Full On.
A little math, and you’ll find that a 180K pot plus 20K series resistor gives you almost these exact values. Unfortunately, 180K pots are not readily available, so we have to use a 200K pot and another resistor.
Using standard values, this is about as good as it gets… at full “off”, the DIM pin sees .45V (So it’s definitely off) and at full “on” the DIM pin sees 2.511V (So it’s definitely on)
This assumes we can trust the datasheet on all three points:
Hey, it’s a data sheet… with three lies in one! With the driver that I am playing with it looks like full off is with around 20K and full on is around 125K.
Also, the chip is supposed to do 1.2A When I tweaked the sense resistor to do that, the chip went into thermal overload protection mode after a few minutes (the light starts blinking off). This could be due to the cheap Chinese driver boards not having a good thermal path out of the chip. It seemed to run a 1A just fine.
Overall, it seems to be a nice little driver chip. I haven’t killed one yet… it even survived hooking the power supply to the LED output terminals. For a cheap little easy to use buck converter, it’s hard to beat.
Thanks texaspyro for answering my question, but this is all way over my head. If you get around to adding your dimmer, please post what you came up with. I’d love to give this mod a try. I’ll also be on the lookout for a driver with a dimming function already built in ready to go.
I got in a 100K audio taper pot. Series’d it with a 22K fixed resistor. It works, but the minimum resistance is a bit too high to fully dim the light. I fully dims until you bring your hand near the pot/wiring, then the light flickers.
I think the audio taper pot gives a better dimming profile than a linear taper one. A 150K pot without a series resistor is probably your best bet for dimming this chip. 125K pots are had to find and tend to be expensive.
Does switching to a 20K series resistor ruin the top end? If it’s that close (only flickers when you get near the wiring) then surely just a LITTLE more drag to ground would do it…
It would normally, probably not… but, I measured the 100K pot that I had and it was 92.5K I am pretty sure that I am losing a bit on the top end with the combo I tried because the measured max driver current was a little lower than normal. I will probably try a bigger resistor. I kinda like having the thing on dim all the time. It is like a night-light and provides some useful light when the room is dark.
I finally got around to installing the dimming pot/switch/knob. A friend of mine printed me a bushing (George Bushing) with his 3D printer. I superglued George into the old switch hole. I used a Philmore 100K audio taper pot with switch with a 30K resistor in series with the pot. Gives a nice dimming range.