There has been some discussion about this driver around the forum, but I thought it deserves its own thread. This is a linear driver which achieves constant current by operating a MOSFET in its linear regime. This driver should not be confused with the other convoy constant current driver which has temperature protection. That driver is not configurable (besides changing the sense resistor).
This driver has 3 modes which are customizable. g_damien summarized its operation:
The UI and mode customization make this driver pretty unique. The current at 100% mode is set by the sense resistor and comes stock set at around 7.5A. If used with a higher Vf single emitter the driver will act as a “direct drive” driver in 100% mode with the current determined by the Vf and resistance in the circuit, though the driver resistance is probably more than a good FET driver. Barkuti experimented with altering the max current:
What I really like about this driver is that the current can be adjusted in the UI and the modes are true constant current; not like a typical 7135 driver where the low modes are PWM and so are only regulated as long as the 100% mode is regulated. I had a closer look at all the possible modes by measuring the current with a clamp meter. The lowest mode is 1% mode and is about ~0.07A. Then up to about 1.4A the current can be adjusted in 0.07A steps. From 1.4A to 7.xA the current can be adjusted in ~0.35A steps. So in general, it is adjustable in 1% steps to about 20, then in 5 steps up to 100%.
The driver uses an attiny 13A, so I wonder if it’s possible to adapt other FWs including e-switch FWs.
Thank you for bringing this up, I agree with you I really like the out put from this driver but yes the UI is not in my opinion user friendly, it also requires a reverse switch to operate properly. I have one in a M21A host with a SST40 de domed works great one click instant high, but that is it (M21A comes with a 10 amp forward switch). I sent a message to Toykeeper and asked if she would look at the possibility of other firm wares, but she did not seem interested. I hope that others with alot more knowledge would take a look.
Thanks for bringing this up.
You can use a forward switch but I agree it is awkward. However, this is similar with all clicky switch UIs that I know of; they work better with a reverse clicky switch.
I guess it won’t be hard to adapt other FWs to this driver since it uses a familiar MCU. But IMO this FW really offers configurability that most common FWs don’t; the ability to choose the precise mode levels from the UI. Dr Jone’s FW on the H17F driver is the only other FW I know of that has this capability.
I also wish there was Andruil port. I wish so much…
But is it so easy?
How is the FET controlled?
I suppose the software side of things is different from simple PWM of a pin.
If that’s the case, changes may not be so simple.
Some of the members here were speculating that these convoy drivers could be using a chip similar to the QX7138. My understanding is this chip senses the current using the sense resistor then controls the external FET gate voltage in a feedback loop, and this chip can be signaled to adjust the current using a PWM input.
If that’s the case then maybe it’s not as hard…and maybe someone capable will pick this task. The sad thing is that there aren’t many people like that.
An interesting thing about this driver is how small it is.
With FET on MCPCB and 1634 MCU it could fit some really small devices.
I’ll take a guess on this driver (not the other one linked in OP): The feedback loop to the FET is the output from an OP-AMP that has the voltage over the current sense resistor as one input, and PWM output from the MCU going through an RC filter as the other. If that’s the case, then the output is simply controlled by PWM.
I have this driver and like it. It’s powering my uf-20 zoomie xhp50.2 3v. Makes a lot of power in a small, under $7 USD shipped package. With my Vapcell vtc6 rewrap it gets crazy bright. Drives the sst40 really hard (gets the host really not-shas no temp throttling or stepdown). More powerful than the A6 driver I had in another light. The only thing I don’t like is the UI. If someone could find a way to get Andruil on it it would be perfect. If Simon puts Biscottti on these ones like he is the others, I’d get more.
Yup, that’s right… linear MOSFET controlled by an OP-AMP. That J36 is OPA336 in SOT23 footprint.
The QX7138 is nice and Convoy uses it in the H1 (headlamp). I’m also using it in some custom drivers. My UT01 driver PCBs using it should show up tomorrow. But the QX7138 is only spec’ed at 3 amps. And it is not used in this new Convoy driver.
I really need to learn more about this OP-AMP setup. I might have to order one and reverse engineer it, then publish an Oshpark board… Maybe even a Anduril-capable one. Time will tell. My list of projects is getting deeper.
I’m not up to date with other’s firmware. Does Biscotti run on a ATtiny13A? If not, this driver would need a different MCU.
I have a basic understanding of it. I’ve designed a driver with it, except that I use DAC output instead of PWM through RC filter. What I don’t understand is if this OP-AMP solution would be any better than a current sense amplifier feeding voltage drop from sense resistor to ADC input of MCU. I intend to experiment a little on it, I’ve got a second design of the board on the way that uses this solution so I can compare them both.
All the FET needs to regulate output is different voltage levels in the gate pin, so in theory current sensing is not even needed. The good thing about my current sense amplifier design is that I don’t need it to regulate output. All I have to do is adjust DAC output to FET gate and see what happens. Once I’ve got that sorted I can use the current sense amplifier to read the current flowing through LED. I’m sure I’m missing something though, I’ve seen this OP-AMP solution before. Amongst others, Loneoceans uses it on his GXB172 driver and he certainly knows what he is doing. There has to be something about it that I don’t understand. Anybody got any pointers? Hopefully I can figure it out once I get my boards from OSH Park.
Same problem here… the queue only grows with time, not the other way around.
There are Convoy drivers that have Biscotti (I think these ones) New Firmware US $5.77 | 7135x3 / 7135x4 / 7135x6 / 7135x8 with new firmware AliExpress
Well, then Biscotti runs on the 13A. As far as I know he only uses 13A. But honesty, I should shut up with statements like these, I really don’t know as I haven’t touched other drivers than my own except when ripping them out of lights I’ve bought.
I checked the photo and it looks like that driver is using the 13A. Hard to see but that’s what it looks like. Maybe that’s not the actual photo of the driver though? I don’t know enough about how programs works to know whether that mcu runs complex firmware.
Just an update regarding FW on this driver. A user here shared some of his mods/developments of this driver for lights he is selling. His custom FW on the attiny chip is described in the linked thread.
The mode levels are very customizable like the stock FW, but the UI is more traditional, while still being full featured. For example single tap advances the mode, and double tap goes to the previous mode. I like the sound of this feature, and I don’t think I’ve seen it before. It is similar to in bistro where a longer tap reverses modes.
Unfortunately the FW is not shared, but it’s an example of what can be done with the driver.
But the QX7138 is only spec’ed at 3 amps. And it is not used in this new Convoy driver.
