If the Zener isn’t properly connected your MCU will fry. The question is, how soon. You might hook it up for testing and it seems to work fine, assemble the light and the first time you use it find it only runs a minute or two and dies. It would seem that there are tolerances that allow different reactions from different parts, so sometimes it’s instantaneous, other times it’s not. But 8.4V SHOULD fry the MCU nearly immediately as it’s several V higher than it’s rated for.
To the best of my knowledge, the FET driver will still need access to the power line, as the MCU needs input at <5V. So the board itself has to see the battery positive as well. Since your battery contacts the driver board, and you’re taking power from that contact direct to the LED, you should be fine. The MCU will be getting it’s power from the traces in the board. The FET works off of the MCU PWM signal and ground. It’s not in the positive line.
Edit: My first builds with the FET were done using modified Qlites. I found the power trace that runs under the MCU was not up to the task and started drilling a hole through the board, bringing a 22ga Silicone wire directly from the spring pad to the LED +. This worked perfectly and alleviated the trace blowing out. So what you’re doing should be tantamount to the same thing and should work fine as well.
I say all of this only through my own experience building the drivers and modifying lights, of which I have done many. Might be good to get confirmation from someone that is better versed in electronics before you proceed. (my memory has been known to have holes in it)
The FET does need to be connecting to BATT+ somewhere in order to power the MCU, but it doesn't care if it's from the spring side or from the LED+ driver pad, because they're the same trace. But yes, it does need to be connected somewhere.
I've found (as have a few others) that the MCU will sometimes last for 30 seconds to a few minutes running on around 6-8 volts, but it won't last long at all. The MT-G2 doesn't care about the zener mod; it's the MCU that does.
Okay guys, thanks.
I check all my soldering several times with a DMM, so hopefully I don’t fry an MCU. I wasn’t sure if there was a good way to know if the Zener is functioning. My poor old eyes really take a beating soldering those zener’s and resistors.
I was just about to finalize a homemade contact board that only sent the bat+ directly to the LED and decided at the last moment to check on that. Good thing I did.
Check for continuity between:
The zener diode side with the line and BATT+ The zener diode side without the line and GROUND
If you've got that, you should be good to go.
I finally got some really nice teflon 26 AWG wire, it wasn't cheap for 26 AWG but it is SUPER thin and can take the heat. I stuck the soldering iron on there a 720F and held it for about 20 seconds...nothing happened. 19 strand so it is still pretty flexible. Website looks like it momentarily went down, so I'll post the link later.
I’m not following you Richard. One side of the Zener connects to ground, but the other doesn’t connect to bat. I can see that its soldered to the bottom component OK, and my DMM tells me so. I don’t get any continuity tone from Bat. I don’t think my other Zener mods do either.
I love teflon wire. Especially for the smaller diameters. I’ve been using it almost exclusively lately.
What line is everybody talking about?
The end of the capacitor nearest the outer ground ring is ground. The end of the capacitor nearest the 4th star doesn’t connect to positive. I’ve checked this on a bunch of Qlites.
C1 bridges between Vcc(after D1) and ground. There is a via in the trace from C1 that comes out on the other side of the board under the mcu connecting to the trace between D1, pin 8, and R1. The Zener diode that gets soldered piggyback on C1 shunts excess voltage limiting the voltage seen by the mcu. The 200 ohm resistor that replaces D1 limits the current that can pass through the Zener to ~40mA without dropping the voltage to the mcu too much.
The Zener mod should work with 3s LEDs and 3s cells but the leakage current through the Zener diode will be greater unless a greater resistance resistor is used.
In either case low voltage protection is lost unless further mods are done to the board.
The line is often but not always seen faintly on the top of D1. The board has a corresponding line to show orientation but when there isn’t a line on the diode you have to look through the PDF data sheets for that diode to see how the cathode(line side) relates to the printing on the diode.
Ouchyfoot, if you are checking continuity between C1 and the spring pad it won’t work because of the diode D1 blocking the measurement. Try it by touching one probe to C1 and the other to either:
Pin 8 of the mcu.
One end of D1.
One end of R1.
Or try reversing the probes. It’s not a zero ohm connection to the spring pad because of the voltage drop across D1.
Thanks Rufus.
I checked, and have continuity to ground at one end, and to the last pin of the MCU on the other. I guess it’s OK.
I never realized there was a difference between the ends of the Zener. I’ve just been soldering them onto the capacitor any old way. They’ve all worked, so I must have been lucky. I still can’t make out a line anywhere.
These are the zener’s I’ve been using. http://ca.mouser.com/Search/ProductDetail.aspx?R=MMSZ5229Bvirtualkey51210000virtualkey512-MMSZ5229B
Hmmm. I just looked at mine through a jewelers loop and it’s D4T. I have the D nearest the ground. I guess that means its on the wrong way and needs to be done over.
In all the Zener tutorials I’ve read, no one ever mentioned there was a difference between the two ends. Does that mean I have to pull out my previous Zener drivers and check them, or would they have blown up by now?
It’s a good thing I don’t get too embarrassed about asking dumb questions or I’d be having lots of trouble in the near future.
I’d say they would have blown up by now, but IIRC sometimes you do not put much runtime on your hotrods? In that case it might be best to check before something fries.
I don’t think I’ve seen any very good zener mod tutorials, and I’m pretty certain I haven’t seen any that explained the concept. (no offense to anyone who’s written a tutorial!)
I posted this in a few other threads, but here's my zener mod tutorial. You can apply the same principles to the FET drivers, since you know which end of the capacitors is connected to ground.
The working principle behind it is basically this:
1. Anything above 6V is too much for the attiny13a MCU.
2. We use a 4.3v zener diode, which will burn off any voltage above 4.3v along with a 200 ohm resistor to limit the amount of current/heat generated by the zener diode. Since the amount of energy required by the MCU is small, this solution works well.
Well, I’m certainly happy that I learned my left foot from my right. After all this, I now understand more than I did before. I now know where to check for connectivity, even though it still registers when the Zener is on backwards. I can see ” the line” now. I always start with the stars in front of me, so if I can read D4T, I’m good. If it’s upside down, I’m bad.
Thanks for all the help everyone. I’ve got a new Zener in place, but my eyes are way too tired to test the darn thing tonight.
Richard. Now you know why I never order just one thing from you. I always have to take incompetence and inexperience into account. It’s always good to have backup.
I’ve looked all over RMMs sight and somehow I never saw those instructions. I’ve been using a hodgepodge from various threads. I’ve got it bookmarked now.
It’s easy for me to remember that “the end with the D is +” although I’ll check for “the line” too.
I agree. I understand how the zener mod works and where to put the components, but I sure didn’t find it on BLF! Every time I’ve searched BLF for a tutorial or explanation of the zener mod I’ve come up with lackluster results. RMM’s tutorial document is a big step up.