Convoy boost driver problem

A while back I purchased an Amutorch SD3 with the intention to swap an SST-70 with a boost driver. After some repeated failures (this be my 3rd try) I got the light to ‘work’, albeit momentarily. Shortly after turn ON it flickers and dies. Many false positives as sometimes would work a few minutes, other times a couple of seconds.
A pic to better visualize the culprit:

Would the circled resistance have a solder bridge? With a Uni-T SMD tester (UT116-C ) I measure ~120 kOhms. Side note: I’m posting a link as I’m dumbfounded as to how it can measure directly on the PCB - maybe with some frequency modulation?

I originally suspected either a faulty pathway to the driver, hence the solder points along the edge to the extension ring. The bridges seem broken but they do show continuity. The tube scratches the blobs in a distinct pattern.
Or perhaps my cleaning solution, which is caustic, left some residue and is partially giving the wrong signal to the MCU. I tried my best in cleaning and wiping up any residual solution, but it may be futile if under a component.

There are nicks on that “SV???” component just below, and some on the MCU (not clearly shown on this picture).

I’m at the lower end of the learning curve;
Any help would be appreciated…

I am totally not the person to ask for driver details but I've been trying to learn a little. That SV package looks like it's the boost regulator. Seems like a lot of material missing but I don't know if that would matter for operation (would affect heat though). Not sure about that resistor...wonder if it has a value or is just a jumper. Hopefully one of these folks who know their stuff will be along to help you better. Hopefully whatever you cleaned with hasn't corroded traces or shorted anything...maybe a bath in isopropyl alcohol is in order?

Edit: looking at what I think is the same driver in Simon's store. Hard to tell but it kinda looks like that resistor is connected to ground and then the one of the legs on the SV boost. I would assume (assume!) that resistor plays an important role then and should be replaced if you can find out what it should be. Then again, it might be more cost effective to buy a new driver if you need to hunt down and purchase more than one component for replacement. I find it really difficult to solder these teeny things (shakier than I used to be, hard to see them well, and accidentally toasting them).

Hey, thanks for a response Correllux.

I was about to order another driver but my perseverance prevails. I have about all the necessary tooling to replace that resistance*, but as the Uni-T DMM shows it 120kOhms, not bridged. And, the boost regulator does seem to work as the light does occasionally work, all modes.

*I lack the experience with the 0603 packages - damn small. I’ll have to remove and replace, if it is .060” x .030”. I don’t have any 0402. And knowing my luck, the problem will persist. So I’m turning to the more knowledgeable folks here on the forum, maybe someone would have some insight?

Reading a bit. In a roundabout way this is one datasheet that came up although the chip markings aren't in it (which seems very common). Sounds like that resistor is quite important and you may be seeing over-voltage protection kicking in. If I'm way off track, my apologies.

I guess the trick here is knowing what the output voltage was set at so you can pick the resistor in a formula.

That’s quite nice to chew up some data. So, basically a voltage regulator for the boost. As you mentioned, I wouldn’t know what the design voltage was set at.

As I said, the Uni-T SMD DMM does show a resistance. If it was to be different, there’s hardly any way of knowing without a functional driver to compare to.

When I received the SMD DMM, I did test across some known resistors and capacitors mounted on a PCB. I was surprised to have it read the correct values. Typically, you need to remove the component for assessment.

So I replace the res. with what the Uni-T DMM gives (120 kΩ) - may be lucky…

Addendum: This was my 3rd driver. The first one toasted with the reflector shorting the MCPCB leads (didn’t see that one). The second came disfunctional - maybe me when testing? Both were from the LHT store and not the Convoys. This one may go to the “trial & error” bin.

I hope one of these Smart People comes along to confirm or deny any or all of this. :) I suppose you could try messaging Simon to see if he could get the value or maybe a schematic from his driver supplier. It'd be fun to actually be able to fix it and understand the whole thing better. I'm amazed sometimes at what all is happening in these little chips-of-convenience, let alone an entire circuit design.

That thought had crossed my mind. But the simpler way is to fork out some $ for the 4th attempt - albeit takes 3 weeks till delivery.

Meantime I might go on a dissection of each component, starting with the Feedback resistor, then re-flowing each of the components (perhaps one desoldered as I removed the center brass stub post for the + battery?).

Keep us updated please. I'm curious to see how things end up.

The Convoy V20 driver in your photo above looks exactly like this one from Simon's store on Aliexpress:

Convoy V20 driver

I can't help you diagnose the individual components and those 0603 or smaller parts are difficult to work with, especially with over being over 60 years old and having unsteady hands these days.

My solution would be to throw money ($12.66 for two) at the problem and order two replacements, and have those on the way as backup in case you can't get the driver in your photo working. But that's just me.

Good luck!

Update: replaced the feedback resistor (FB) and for the moment, the light works. No flickering.

@ max: The driver is a 4.8 A Convoy boost driver which is marketed for XHP70.2. But as I too am on the older side, my eyesight and hand coordination are not what it once was. So undertaking a small replacement is a feat. But your suggestion of buying a couple, just in case one does fail in my modding, will be my rule. I can always cross-reference a component’s design value, else swap.

@ Correlux: Many thanks for your input. The datasheet of the MIC2295 (AKA SVAA) boost regulator does concur. Although I removed the FB, it’s too small to be pronged with the SMD DMM. I tried a 120kΩ 0603 and with some adjusting, it sits well. The spring pad goes a bit underneath but doesn’t seem to make contact - luck?

I don’t like my edge bridging, but I put much effort into this light and for all its problems, I’ll leave it as such - wouldn’t want to create another hurdle.

The SC-70 package with ”SVAJE” (pics on Convoy’s store) on it is an Op-Amp, no idea which one exactly but that resistor is connected to its output and then to the FB pin of the boost regulator on the other side (TPS61088), 120k sounds like a plausible value, the exact value isn’t critical but if shorted or unconnected the driver wont work properly.

See the shematic of this driver it’s a different driver but the principle is the same, the resistor is R9.

Edit : wait no that doesn’t correspond to an sc-70 package op-amp, rather an SOT-23 one.


So there is a schematic of this driver. Much more complex than my knowledge, but the R9 @ 100kΩ would be within the tested value of 120k. I’ll be testing the light tonight.

I’m trying to figure out how that resistor is behaving. It does measure on the board, but breaks down in use. It is too small to actually have a look (40x magnification) if the solder may have bridged it, or it’s a residual of my cleaning solution. I never used this solution before, but my go-to flux left some splatter everywhere and just didn’t dissolve with alcohol (it somehow changed properties with the residual flux originally on the board). Just now got some Kingbo RMA 218 flux that i’ll be putting to some use and verifying if it has less splatter. My bad as I used quite a lot of heat to remove the brass battery post. Above 350ºC the flux makes for some problems.


So the 120kΩ resistor would be too much, more like 10k?

I just got the S11 with that driver in it. I like the driver, do not understand it, but like it. Can someone tell me what the gold colored component is? Must dissipate a lot of heat?

Didn’t you say you already tested with a new 120k resistor and that it works ?

The driver I linked is the xhp35 driver, it’s based on the MP3421 boost regulator, but as I mentionned the principle is the same.

No 120k sounds adequate, 10k would be too low, likely anything between 47k and 150k will work.

I was also curious about this thing that looks like a piece of brass ?
It is located under the boost regulator (see the vias), so yes maybe you are right and it is for better dissipation.

Oh OK. From that diagram, pin 3 of the op-amp goes to a 10k.

Yes, I did change it and holds for the few minutes I tested it, all modes.

So I’m good. And asking as mattlward, what is the gold (coloured) component?

Gold component? I ques its heat sink for boost converter

Man, that is a tiny boost converter!

That resistor that I pulled is not much wider than the trace - .020”. I can’t get it between the test prongs for a measurement.

Under the scope, it looked fine and dandy, a bit of edge solder but nothing as a bridge nor a crack on the oxide film. I really suspect there are wee little demons working against me.

Nevertheless, I ordered a couple of spares drivers as I still have my doubts about the fix. At times, the clicky goes into the next mode memory from OFF then immediately to ON.


Some of my lightbox values:

Low - 22 lm -> ~1%
Med - 390 -> 14%
High - 1400 -> 51%
Max - 2720 -> 100 %

Firstly, my box is not calibrated and I’ve been getting lower than supposed values on all the blasters.
Secondly, the driver will not give the same % as the XHP70.2 - different Vf.

But at 1400 / 2700 lm, I’m happy. The SD3 is rather low mass and I doubt could dissipate much more heat. I thought I had taken some current and voltage readings prior to re-assembly, but I’m mistaken with a parallel universe – the CULPM1.TG

Its on other side…

Good...glad it's working, or ish. And I'm glad smart people showed up. :)

All of this component wizardry fascinates me but it didn't take long to realize just how complex everything is - including the fact that different package formats can be capable of very different things as thefreeman pointed out! Funny, I was thinking of posting a new thread asking about those little brass squares, too. Hadn't seen those before until some of Simon's new drivers started sporting them. I assumed they were heat spreaders but I haven't had one in hand/out of a light to take a closer look.

Thanks for posting the link to the driver review, thefreeman...hadn't seen that one before as I've browsed past driver stuff here. So much to sift through and learn!