thefreeman’s HDR Anduril 2 high efficiency drivers - update : FWxA boost driver

Some news!

I received the last parts Wednesday (at last), yesterday I made one board, first time reflowing anything else than an LED but it turned out well :

This morning I tested the driver, only the hardware by bypassing the MCU :

And it works! The output voltage/current is stable, here is a quick output voltage waveform with the scope I just bought :

The only issue is that I got less current than expected, with 44mV ref voltage and 9.2mΩ (7+mosfet resistance) I should get ~4.78A, but I only get ~3.7A. After a quick check the real ref is ~43mV and the mosfet resistance is ~4.4mΩ which explains the difference. Swapping the resistance with a 5mΩ one should do the trick, but I’m worried about the ON resistance variation of the mosfet between parts.

Next I need to adapt Anduril for it.

:slight_smile:

I just received 21mm boards from Oshpark

They’re prettier.

1 Thank

Following

I’m not sure how I’ve missed this thread! Well done, thefreeman! I can’t wait to see this develop more. And of course, I’m happy to see the usage of the 1-Series AVRs.

Congratulations! I cant wait to see how this turns out…

How does your driver regarding these characteristics? If behaves well regarding these qualities (from dtr’s laser shop), your driver could be a good choice for LEP driver swaps if more e-switch LEPs pop up in the future.

An ideal power supply for a laser diode has the following characteristics:
• Current source
• Transient suppression (also low noise)
• Independent clamping current limit
• Slow start / ramping the current signal during switch-on
• Output overvoltage protection
• Input undervoltage detection
• Output short-circuit / interruption detection
• Shorting output during driver off status for ESD protection
• No undershooting of the output voltage at switch-off of the laser, so that a negative voltage over the laser diode cannot occur.

Thanks.

I got a bit of a scare yesterday, I couldn’t read the device in microchip studio. Couldn’t sleep while trying to figure out what was wrong, then it clicked, I ordered the older programing key :person_facepalming: , despite that you told me you changed it, I completely forgot that, I must have had the older key already in my oshpark projects. Thankfully it didn’t fry the chip (5V on UDPI, UDPI on GND, GND on +).

For now I just tested your TCD sample code.

Aren’t laser diodes low voltage ? The present driver uses a boost converter for voltages higher than li-ion.

• Current source : yes.
• Transient suppression (also low noise) : switching converters are noisy, maybe a linear driver would be better suited for this.
• Independent clamping current limit : sometimes there is an adjustable switch limit but this isn’t the case with the TPS61288 (fixed limit).
• Slow start / ramping the current signal during switch-on : sure that’s possible, in hardware (it’s already the case, but can be increased) or software.
• Output overvoltage protection : yes.
• Input undervoltage detection : most switching ICs have an undervoltage lockout, sometimes adjustable, on the TPS61288 it’s 1.9V, can be done in software, anduril have a default LVP of 2.8V (or 2.7 ?).
• Output short-circuit / interruption detection : don’t know how to do that.
• Shorting output during driver off status for ESD protection : idem.
• No undershooting of the output voltage at switch-off of the laser, so that a negative voltage over the laser diode cannot occur. : not sure about this either.

I made a quick efficiency measurement, this is with a current clamp so it’s not very precise.
Vin : 3.96V Iin 6.36A Pin = 25.19W
Vout : 6.40V Iout : 3.73A Pout = 23.87W
Efficiency = 94.8%

This is with a reverse protection PFET, without it it could increase by 0.5~1, and with a bigger inductor with lower DCR like a XAL7070 maybe ~1 (more at lower input voltage)

Good. :slight_smile:

Probably a long way off but if you end up making these in any quantity I will need to purchase at least one. It seems a 12v e-switch boost driver is kind of a unicorn, I need one for an FT02 host with the GT-FC40 LED, to have one with Anduril would be even better.

I will probably make some, what is the size and available height ?

I think it is 25mm in diameter, might be 26mm, I can check when I have it in hand. In terms of height, I can check that also but I don’t think component height will be major limiting factor, it does use a boost driver stock and the battery tube has a lot of extra space too. Thank you for responding, and for your work on this project, very exciting.

I would need a few for KR1 and M21C-U... just saying...

KR1 has a tail switch with signal tube (as you probably noticed :smiley: ), it would require a different PCB. Bigger PCBs are easy though, just need to increase the diameter.

Oh man that would be AWESOME if you could pull that one.

KR1 with GT-FC40 would be an amazing EDC. My tint snobbery has become so bad I can’t tolerate my W2 KR1 anymore, so I already have a good host, and would take one of those too if they end up being made!

And then the FW21 Pro after that. I’m going to try and one to use an ANGIE optic. Think of it as a spiritual KR4S. If this driver also supports 6V, I’d run 4 parallel Nichia B35AM.

My thought exactly!

Present size is perfect for Convoy M21C-U (22mm) which is also great host

The issue with the KR1, as pointed out by loneoceans, is the shallow driver cavity which cannot fit a 3mm tall inductor such as a XAL7030, going with a 2mm one would severely impact the current capabilities, to fit a 7030 inductor he milled a cutout, which isn’t an easy thing to do if you don’t have the necessary tools.
I don’t know exactly what is the available height, but if it’s arround 2.5mm then we could use a 0.8mm PCB, with a spacer (simply a solid wire) to gain 0.8mm, or maybe even a bit more depending on how much is needed, and what the tube assembly allows. This is a much simpler modification.

Although as somebody pointed me out (JaredM I think) maybe the 0.8mm wouldn’t be as strong and might need potting, or maybe just a pad between the inductor and body would sufficient.
I have some 0.8mm 22mm PCB coming in the mail so I’ll see, they’re for a D4v2, which I still don’t have actually (maybe somebody has a dead one for me ? :smiley: ) , I’m guessing also has a shallow driver cavity.

But about developing another PCB specifically for the KR1, I don’t know, it depends what Loneoceans does I guess, as he already has one for it, but hasn’t published it yet.

Another light with inner body tube, with a shallow cavity too I imagine, has somebody disassembled it ?