This one uses pin 6 and This one you can’t tell since the traces are under the blue wire. Still, the 7135’s should tell you which is the pwm pin. With a 7135 at the top of the board, left is output, center is ground, right is pwm input. Output should connect to led- pad, ground may connect to ground ring though many just use the larger tab, pwm input is the only one that connects to the mcu.
Hi,
Yes, that is what I figured, i.e., the rightmost pin of the 3 pins of the 7135 is the input, so that is the pad I soldered the AOD510 Gate pin to, and I can ohm that back to pin 5 (I think) on the MCU. I’ll check that again tomorrow, but I know that that was the pad I soldered the Gate pin to, so I’m puzzled why I don’t seem to be getting modes.
Jim
BTW, I noticed that comfy’s original post/thread (howto: Build a Nanjg-092) mentions moving a capacitor. Is that still necessary? Maybe that’s why I’m not seeing modes?
Since the pics are gone from his original post, I can’t see which capacitor he was referring to, or where it needs to be moved? Anyone have any idea about that for a 8x7135 board?
EDIT: BTW, re. the question of which MCU pin goes to the FET Gate input, the pic in this post (#61, comfychair-inspired quick+cheap+lazy single-sided 17DD FET-driver (poor man's nanjg92)) seems to show air wire connection from MCU pin #5 to the FET Gate input.
EDIT 2: I think I give up. I was going to try a different FET, and ended up pulling one of the original7135 pads off, so I put the new FET on another 7135’s pad, standing up instead of flattened, and after that, I still get no modes. There was comment in comfy’s original thread about some NANJG firmware using different PWM frequency not working (not because of the different frequency but because it used different control method or something), so I’m starting to think that the NANJG I have is one of those.
Try stacking an extra cap on C1 or adding one between B+ and ground as close to pin 8 as possible.
Hi,
I haven’t tried that, but I found something (else) that is really puzzling me. I put a scope across the FET Gate and the board ground ring and when I power the board on, it is just constant. I’ve tried clicking a switch, but it stays constant. In other words, it doesn’t look like there is any PWM type signal. Just as a sanity check, I put the same scope across the same pins on an EAST-092, and I see the expected signal (e.g., different frequency square waves for different modes.
I’m trying to decide if I should sacrifice another driver and/or FET ??
Was the board tested before you did the swap? It’s possible you have a bad or mis programmed mcu. I’ve fried a few boards myself without understanding how or why. Keep trying and keep looking up threads that date from around this time as this was discussed for a few months.
Yes, this was a new “generic” NANJG 8x7135 from FT (this is the one with the mode groups via clicking in low mode) and I tested it before removing the 7135s.
One problem is the pics in the original (comfy’s) NANJG-92 thread OP are gone. Plus I’m still confused about something: Even though the pics are gone, I followed that original thread, and I could have sworn that the original mod had the 3 pin FET upside down and the 70n02/3 and AOD510 FETs were mentioned in the thread, so it seems like the AOD510 like I have should be mounted upside down.
Jim
EDIT: The other thing that makes it hard to check/remember is that the link to the FET in the OP of that thread is no longer existing also :(…
EDIT 2: Can someone who has this working post some pics, especially showing which cap to remove and where to put the cap?
I’ve spent the morning going back in time and found This post by WarHawk-AVG (post 44) that mentions it. I’ll look some more to see if I can find some other info. I’d suggest you google the different fets and compare gate pin locations, that’s what determines FET orientation.
Ohaya, do you recall This post or wight’s response ? A bit earlier in the thread he also mentioned that he thought it possible that a higher value cap should be used to absorb the spike but basically if it ain’t Atmel, it ain’t cookin’.
Hi,
Here’s some pics of my testing board.
Showing the top, AOD510 soldered to other 7135 position because pulled pad off other position:
Top view (you can see the missing solder pad):
Bottom view:
No, I don’t remember that post specifically - wish I had done this back then when comfy and wight were around :(…
Re. Atmel, do you mean that if the MCU isn’t Atmel that this FET conversion won’t work? I have to check that. As I said, the board I’m using is one of the “generic NANJG” drivers from FT. I was hoping to use that because it does mode switching by clicking rather than using the stars.
EDIT: That post you mentioned was about whether the MCU from a NANJG could be used on the BLFDD board, rather than about NANJG-92.
EDIT 2: One of the things that is really bothering me is why when I scope the FET pins, I get no (or constant) signal? In other words, it’s like the MCU is not working. Either that, or there’s a short to something like +V that I’m not finding
If the FET is This one then your hook up looks right. I don’t know why a different mcu should make a difference except in not being able to reflash it but WTHDIK. It may be that the different mcu is unable to charge the gate on the FET or it may be fried. To test that you could reinstall a 7135 to check it. On boards where I killed 7135’s the mcu was still able to drive the led in different modes but all were variations of moon.
Is the white wire soldered only to the ground ring or is there a via under the tip of it that connects to the led- trace?
Was just about to reinstall 7135. If that works I’ll try w a different type of nanjg, from illumn.
I think the conversion should work but not if you switch the mcu to a different board with different I/O hook ups and you wouldn’t be able to reprogram it.
Lots of members I miss too, those and others.
Hard to tell from the pic but to me it looks like the white wire might be soldered to both gnd and led- which would bypass the FET and lock it in DD.
It is another driver than the 105C, but with the AK-47 I have an original NANJG version that uses a different Attiny pin for the 7135gate than a generic version sold by Banggood. So the lay-out is not always preserved.
Perhaps a bit useful, here's a stereo picture and a stereo gif that I posted in comfy's thread, they are still there, these are the original 105C board an Vishay FET from the OP of comfy's thread:
RBD, djozz,
If I remove the FET and solder one of the 7135 in, and move the emitter - wire to the normal place on the driver, I get modes.
If I remove the 7135 (again) and solder in FET, and move the emitter - wire to the FET tab, I get no modes.
The white wire is only there to make it easy for me to test, i.e., I hook the - of my bench supply to that white wire and the + from my bench supply to the spring to test. It’s only soldered to the ground ring on the battery side of the driver board (I’ll double check but that was my intention)…
EDIT: I think that I’m concluding that the boards I have don’t work with FETs. The MCU has no markings so I can’t tell if it’s an Atmel.
Hah! So the original mod DID use a 3-pin FET (looking at your images), right? Now I know I’m not going crazy!
But, do you know which FET was used, because in the original the FET was positioned with the ground tab on top (upside down) so I think the Gate pin for whatever FET was used must be the opposite side/end from the AOD510?
That sounds indeed like this FETdoes not work well with the 105C gate-signal, in other words: that goes way beyond my expertise. Rufusbduck, RMM or dthoang know more about FET's.