FETs and gate resistors - scope images

does anybody happen to have screen shots or images that were posted by comfychair? trying to learn more about the whole C1 cap placement decision on the current iterations of the BLF 17DD.

Is this what you are after?

thanks that does help me with the history. I was specifically curious about how noisy the Vcc line gets with the current C1 cap placement since the screenshots are missing. From what I can tell in the comments of this thread there may be noise/spikes with the current C1 cap placement but it was deemed acceptable. Also, I’m still wondering why the gate resistor was deleted if it could have helped with the spikes as well. Guess I need a scope and more studying :party:

I’ve been having an issue with a driver built from a wight board running star firmware. When running a triple XPL at high current I’ve been having issues with the driver flickering and shutting down. High current being bypassed springs with LG HE2 cells. Nonbypassed springs or tamer cells don’t seem to have this problem so I was looking here for clues :bigsmile:

I need a scope…

From memory, the revised C1 placement cleaned up the spike greatly.

That’s interesting, I’m having the same problem with one of the 17mm DD MTN drivers, which is pretty much the same as Wight’s board.

Does your problem driver also switch off if you run it on low for a few seconds?

I know the design of these boards are proven, as there are hundreds of them out there, & I’ve built plenty before this one, but it’s still a pain when you can’t put your finger on a particular issue…

Weird - I'm using a wight FET+1*7135 driver with the SIR800DP in a triple XPL X6 getting 10 amps with all heavy bypass wires, etc. Basically works well but I got one noticeable issue and it's a blink when changing modes to moon mood, fron turbo or not. RMM recommended trying a resisotr off the FET gate, like a 100K or so -- didn't try that yet.

yeah the problem will happen in low it will flicker a bit and then shut off a lot of times. really only in a high current situation though. I can’t remember if it was medium and high too, but definitely low.

my parts are from RMM so I think this driver is a NXP PSMN2R4-30YLDX FET. have some SIR800DP but haven’t tried those yet. I don’t think the FET choice is an issue though.

funny you should mention the gate resistor because that did seem to be a good thing from this discussion, so I was trying to understand why it was removed from the design in the first place. parts count reasons?

I actually added a 200ohm resistor (replaced the existing diode) and a zener along with a second 10uF cap in parallel with the zenner (also all known as the zener mod - but still running as single cell). C1 still has 10uF cap as well. This was discussed earlier in the thread as a possibility of smoothing out the Vcc line.

To be honest I haven’t been able to get my light to act up again after doing the zener mod with this single cell driver. It was pretty easy to reproduce the behavior before I added the zener. I would like to try the gate resistor at some point too (without the zener). I will probably build another driver and keep tinkering. I’ll update if my driver acts up again.

It was removed because it did nothing to have it there, at the time. I was involved in the FET drivers early on, when comfy was leading us to the FET farm, FET fame, FET nirvana, errr something like that... I l think the theory of the resistor was good, but practically had no effect. I've built a lot of FET drivers up, from Nanjg's at first, then the OSHPark versions of the BLF DD 1.0, C_K's, and wights. Never seen a problem til this little blink issue - all 3 FET+1x7135 wight drivers I built have it.

I’m tempted to see if I can salvage any of the old pics from this thread from archive.org or other caches. I found the data very useful, and would like to be able to refer to it again in the future.

lol FET nirvana. Gotcha, well thanks for clearing up the history a bit. I appreciate your insight on this, especially since you’ve probably built a million more drivers than I have :bigsmile: Please be sure to post any updates, I would be interested to hear what you find.

yeah I tried wayback but that didn’t seem to work. that info sure would be useful.

comfy had some great ref material - illustrated pics, drawings, etc. Boy, hope you can find backups - that would be great.

I've never seen one of the [properly built] MTN drivers glitch up, not once. You're not running zener mod, right?

No, not doing the zener build.

I was hoping you were reading this thread, in case you had come across a similar issue, & could point me in the right direction of what may be the cause.

I’m using the components (resistors, cap, 13a, diode) from new Nanjg’s, & the board & fet were from your store.

The only component I haven’t swapped out on the problem board is the fet, so that may well be the answer.

Hmm, I don't think it would be the FET. Maybe the Nanjg's capacitor isn't up to snuff. I would suspect that first.

Here is the full thread, with pics. Oops - looks like I’m having a connectivity issue right now so I can’t finish renaming this at the moment. For now you can go here and click on the link with a size next to it.

Nice, thanks Wight. That took me back.

Wow, great work wight! Never thought of doing that. Think TK posted links to comfy's pics before, but putting it together like you did is really, really nice!!

It's all just html after all, so can be captured with the pic links changed - I assume that's what you did -- pretty cool.

That’s exactly what I did Tom E: I pulled up the thread in 300-post view instead of my normal 30-post view, pruned out some unnecessary stuff, saved it, then fixed the links to refer to the copy of comfy’s pics I host. I used Notepad for the fixing.

When comfychair disappeared for a week or two I became concerned. After comfy was gone for 1 month I scraped a full set of the images from comfychair’s server, along with the BLF forum threads - just in case. Sure enough, it wasn’t very long after that when comfy’s image server went away. That’s where TK’s copy came from - I passed it along around the same time I was leaving. I was hoping that TK & SB would be able to effectively do a “search and replace” on comfy’s URLs here to get the threads healthy again but it seems that they were not able to do that. For FF users a greasemonkey script would probably be the most simple solution, but I use Chrome. I guess Chrome has similar things now, so maybe someone could whip up a script? That would be great.

@FmC - It definitely took me back too. Comfy turned in a lot of good work over here… and is one of those missing people I can’t help worrying about when they cross my mind.

+1 wight, I've often thought about him too, hoping he didn't get frustrated with us. me being one of the "us" .

I think SB has been too busy to work on database edits. Perhaps eventually that can still be done, though.

Yeah, messing around with changes inside the post DB for BLF isn’t something you want to jump into without a chance to be very certain that you aren’t damaging something. TK is aware of this, but for those who aren’t “in the business”: we must realize that if SB were to make changes and review them but overlook a subtle problem for a few days/weeks - we could potentially be stuck with the problems. Backups don’t necessarily fix certain types of pervasive damage to a DB which is in constant use.

Anyway in the meantime I have a much less dangerous workaround. It’s a userscript for greasemonkey/tampermonkey, as I mentioned earlier. Turns out that it’s very easy to write this script since there are no complex URL transforms involved… so I whipped one up! It should fix any embedded images which comfychair previously hosted as well as any direct links to those images. Unless I misunderstood something (it’s my first userscript) this should only operate while you’re browsing on BLF. Here are some quick/rough instructions on how to use it:

  1. Install Greasemonkey for Firefox or Tampermonkey for Google Chrome.
  2. Either ‘install’, ‘import’, or whatever this userscript.
  3. If you don’t get an option to install the script simply copy the contents into a new userscript and save it from within your __monkey extension.
  4. In either case, put a checkbox next to that userscript in your usescript manager so that it’s on all the time. You’ll see a little icon in your browser when you’re on BLF to show that the script is running.

EDIT: I’m running the script in Chrome using Tampermonkey. I’ve just tested it on a few of comfychair’s other threads - seems to be doing exactly what we’d want.