Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
Status
Not open for further replies.

LukeKnepp

Member
Hello everyone,

I have attached a snip image of a design I am using and have been using for almost a year and a half now...
Here is the story on the board.

(Jump down to circuit explanation if you want)
Last year, I made a few hundred using the MP2357 regulator, but since it went out of stock, I switched to the AOZ1280CI, and made about 2000 pieces last year, with a different board layout.
The schematic looked identical to the one attached, I was just using a WH148 style PCB mount right angle potentiometer with the detent and everything...

Then this year, I switched to a higher quality upright PCB mount potentiometer, and kept all the other same components...
I made 460 pieces of this exact board in the course of 3 batches, and the success rate was almost 100%, 1 got shorted out and blew up, and 1 other one did not shut off, but 458 working out of 460 is what I would call pretty good.
Then, I hired another worker, jumped into this board, wanting to get a lot of stock made, so I did the rest of those 1000 boards(540 pieces) all in one batch.
I had him test the boards one at a time for the correct output voltages(Variable from 5V to 10V with the potentiometer) and we just had the input voltage set at 16.5V.
The testing procedure went as follows....
1. Power up the board with 16.5V
2. Turn the detent on and make sure the output is 5V and the load runs(load is a very small fan)
3. Turn the potentiometer to maximum and make sure the output is 10V.
4. Turn the input voltage down to 15.7V and make sure the board shuts off.
5. Turn the input up to 16.5V and power cycle, making sure it restarts.

All of the boards passed our testing, so I sent them to be assembled onto the fans.
When they tested the fans, about 1 in 4 worked, and the rest of the boards had the AOZ1280CI blown!!!
After some testing, the ones that blow all work at 16.5V and 17V but as soon as you get to the 19V and 20V range, about one in four blow up when you start them up!
All of the previous boards were tested with the same inputs, and 3000 have been in use for more than 6 months, and no problems so far....

The only explanation I can come up with is something in the assembly process....
Like could the oven temperature too high(I had boosted it a little because of older solder paste) be an issue?
Any other ideas?



Circuit Explanation -
The input source is about a 15V Min 20.5V Max 18V nominal lithium ion cordless tool battery(DeWalt, Makita, Milwaukee, etc.)
The power comes into the circuit board into J3, goes through the potentiometer detent(for complete power cutoff when the detent is switched off) and then goes to the net called VIN.
VIN goes through Q1 for low voltage protection, then to the input of the regulator circuit.
The feedback of the regulator is adjustable with the potentiometer for variable speed on the fan.
The 6.2V Zener diode(D2) and the resistor divider(R10/R9) sets the low voltage cutoff's protection point to about 15.9V.
The T2/T3 latch makes the cutoff stay off when activated, until power is cycled, and C6 holds the latch open for startup, but anyway, I'm not worried about the low voltage part, the only thing you need to know is that it works and is quite reliable.

The trouble starts at the input of the regulator(I know this because when testing, I shorted the source and drain of Q1 and got the same results)
I took a blown board, removed C1, C2, C3, L1, and D1, and they are all still functional, it blew just the regulator...
I also tested multiple with Q1 shorted, so it has nothing to do with the low voltage shutoff part of it....



Fan Board V2_2.PNG

Fan Board_V2.2.png
Fan Board_V2.2B.png
 
After a closer look, the Input pin on the AOZ1280CI actually burns up. All the other pins and connections are still fine, the Input pin is completely gone!
 
UPDATE!!!!!

It is the input capacitor(C1) that is causing the issue.
I’ve been making these boards with this input capacitor this year.

TDK Corporation Part # C3216X7R1H106K160AC
https://www.digikey.com/en/products/detail/tdk-corporation/C3216X7R1H106K160AC/9991386


In the past, I had used this one from Samsung Electro-Mechanics.
Samsung Electro-Mechanics Part # CL31B475KBHNNNE
https://www.digikey.com/en/products/detail/samsung-electro-mechanics/CL31B475KBHNNNE/3888447


I got those back out this morning, and replaced the TDK input cap with the Samsung cap on 4 boards.
All four boards acted EXACTLY the same!
They worked absolutely perfectly all the way up to 20.8V input, everything PERFECT!
At 20.9V, after a power cycle, there was no output voltage.
It never works again after there is no output voltage, and if I turn it up 1 volt or so and power cycle, then it blows just like the others all do at 17.5V…

So, there is something that makes so the 10uF from TDK pops the IC at 17.5V, and the Samsung 4.7uF doesn’t until 20.8V, but it is not just lower capacitance because I put two 1uF caps in parallel for the input cap for a 2uF 50V, and it also popped the IC at the lower voltage….

Technically, the 4.7uF would work, since the battery voltage never gets above 20.3V or 20.4V anyway, but it is awfully close for my liking….

Could it be something with ESR or ESL?
 
Look at the input pin on a scope as you ramp up the voltage.

The only thing I can think of, if it does it with no load, is resonance in the caps.
Multilayer ceramics often exhibit piezoelectric effects, and it could be "ringing" and producing voltage spikes???
 
If the input inductance is too high (like you have long test leads), it resonates with the input ceramic capacitor and if this capacitor has a low ESR, the input voltage to the circuit can ring up to twice VIN due to the lack of damping in the tuned circuit. However, with the voltages you talk about, these should be less than 50V which is the rating of the capacitor. is the chip blowing up and taking the input cap with it since the chip is only rated to 26V. You can simulate this in LTspice.
 
Hi...With a 1.8 MHz most extreme recurrence, the AOZ1280's 100-ns least

on-time is somewhat energetic for a 24 V - > 5 V converter, so I'll stay with the AOZ1282, which is 150 ns min and 540 kHz max.

https://www.7pcb.com/
 
Last edited:
Status
Not open for further replies.

New Articles From Microcontroller Tips

Back
Top