[WIP] 17mm *9x* 7135 for dual-PWM or white/red: A17-L81

9x7135 driver w/ 8+1 configuration for STAR w/ Dual-PWM configured. Similar to other Moonlight Special compatible drivers.

Several users have asked about running a red LED for moonlight on white LED lights. This requires a second channel be available on top of the driver, something my A17-PZL driver did not offer (channel 2 is on the bottom). This driver features a new top layout to allow the second channel to be on top of the driver. Leave the driver intact for standard Dual-PWM use, but for red-moonlight use please cut on the dotted line.

This driver is fairly similar to my A17PZL driver. Here are a few things of note:

  • No room for a Zener footprint, but… a mildly interesting twist is that I’ve used the traditional location for C1. That means that a Zener may be stacked on C1.
  • Due to the extensive use of traces under the GND tab, additional silkscreen has been added in order to minimize the chances of scratches etc happening during building.
  • Smaller, rounded 0805 pads are in use here.
  • L+ is in the center w/ a 1.3mm via. L- is on the outside with a 1.3mm via.
  • L2- is only for use after trace-cutting on the dotted line and features a 2mm diameter solder pad on top of a somewhat larger pour (approximately 2x-3x the surface area).

For other notes and discussion, please refer to the A17PZL thread. While the entire thread may carry some relevance for this driver, the initial WIP release of this driver was in post #87.

A17-L81 v009


I don’t see why not… oh, oops. Now that you mention it, I forgot to run a DRC…

… right, so there were a few unacceptable DRC errors. I cleaned those up, v009 is now in the OP. As I was saying, I don’t see any reason not to order them. I might expect some minor calibration differences, as with other 7135 driver variants.

Note that if anyone chooses to do a Zener mod on this driver and still wants to use red for moonlight it would probably be prudent to wire 3 or 4 red LEDs in series…

I’ll order some tonight.

Speaking of calibration, I really need to test the lvp of all of the drivers I’ve made recently. I think one of them was shutting off around 3.6v

Dont forget that to be able to have the red LED turn off above moonlight FW mods must be made. TK has a perfect FW for this (from what I understand, absolutely no offense to her in any way but I dont like the way she does her FW, again nothing is wrong with it and it seems to work fine for her and others, it’s just very different that the way I was taught and I can do what I need with STAR)

You can also check the STAR thread, post 973 to see how to mod standard STAR_ dual_pwm to work with this.

What don’t you approve of?

It’s not that I dont approve, as I say I’m sure it works, the way she organizes her code is just pretty different from the way I learned and it’s hard for me to read and modify for my needs.

Again no disrespect to her or her methods, I’m not at all implying her way is wrong [because it’s not], it’s just different from what I do. If someone that doesnt do any coding wants to use this it may be easier for them to mod standard STAR than to switch to her’s. Pretty much all I’m getting at is that you cant load a standard version of STAR_off_time on this (or any other) driver and have it be able to run a 2nd channel like you would need for a red/white build but that’s no fault (of yours or her’s or anyone elses), it’s just different, it’s never been what STAR was ever intended to do so it takes a little work.

Now that Wight is gone, can anyone see a problem with this design? My copy goes direct drive regardless of firmware, and I can’t see any shorts in my build.

Edit: I cut the trace and tried the 2nd led- pad and that works fine, so the problem is either with the main led- pad or somewhere else in the primary circuit

Bump to the top…

Anybody with any ideas?

Sounds like a shorted main Led- pad to ground, its pretty close on the top side and bottom of the via. Might have a retaining ring touching it ?

It’s not even in the light yet, just a test rig.

wight mentioned something about the traces running under the 7135 ground tabs, that’s the only thing I can come up with. When I get home today I’ll try pulling off all the 7135’s and look for exposed traces under them.

Just an update in case anyone reads this thread later on:

I completely stripped the board down to a single 7135 on the main circuit and the light still comes on, so obviously the problem is in the pcb.

I swapped all of the same parts to a new pcb and it works correctly, so it appears the problem was with Oshpark, not wight’s design.

With the center pad so close to Vdd it wouldn’t take much to lock it high. Was it direct drive or just high? Direct drive is usually led- shorted to gnd bypassing current regulator chips. Stuck on high us usually pwm trace shorted to B+.

At first I thought it was DD, but the brightness went down as I removed chips, so it must’ve just been stuck on high.

Yesterday I managed to assemble one of these, the firmware took some tweaking to get it to fit onto the ATTiny13A but finally got it to work. Modes as follow: Red, Low, Med, High.
It is only a prototype, just to test the concept. Need to tweak the firmware a bit more to get the most out of the Red LEDs but they probably won’t make it into the next reincarnation.