Author Topic: Transient positive voltage spikes killing MOSFET drivers (Read 26211 times)

MarkM · « **on:** July 21, 2015, 06:38:13 pm »

I'm switching two of these N channel MOSFETs in parallel with a TC4422a driver @ 60-100hz using an Atmega328 PWM pin on a custom PCB on battery power(2s-3s lipo pack). I started with the TC4427 driver and they died randomly. Now the TC4422a is dying randomly. My buddy used his scope and noticed something very strange. Hoping someone can shed some light on it. The load is resistive(30a for this test or so), but it has some inductive properties, as it's a coil of resistance wire. Flyback diode is in place for the load.

The issue I believe is the positive voltage spikes that are high enough to go over the max of the driver(18v), but only on the 12v rail from the battery. The usual spikes are around 20v and as high as 30v. The output from the driver to the gate looks great as well as the actual load itself. I attached pics so you could see them. The 12v input is the battery going from loaded to unloaded due to switching. It looks to me like the positive spike is where it goes from loaded to unloaded.

The 12v input from the batteries also powers a linear regulator that is working perfectly even while switching. The rest of the board is functioning as it should as well.

It would seem that a TVS diode would work, but my required clamping voltage(ideally around 15v) is too close to the max operating voltage. Is there any other solution? Would really like to figure out what's going on here.

Please let me know if you need more info.

MOSFETS:(2 in parallel)
http://www.digikey.com/product-detail/en/0/568-11557-6-ND

DRIVER:(0.1uf and 4.7uf bypass caps)
http://www.digikey.com/product-detail/en/TC4427ACOA713/TC4427ACOA713CT-ND/3622383

Simon · « **Reply #1 on:** July 21, 2015, 06:50:19 pm »

Sounds a bit odd. Try a zener at a couple of volts over the battery voltage. You could also try a nice bypass capacitor or a few in parallel, That might help deal with sudden load changes in the system

T3sl4co1l · « **Reply #2 on:** July 21, 2015, 06:52:45 pm »

Two problems:

1. Add series gate resistors. This controls the risetime at turnoff, which limits how much trouble you get into with inductances (increasing switching loss instead). At 100Hz, you probably don't need a proper gate driver at all; a level translator would suffice (e.g., just a few BJTs to switch 12V instead of 5 -- or use logic level FETs?). That is to say, by the time you're looking at 1kohm gate resistors, your peak gate current can only ever be 10mA or so, and a freaking several-ampere driver becomes absurd.

That TC4xxx thingy can drive gates in the tens of nanoseconds. That's over a million times faster than your waveform! It's a wonder you saw a spike at all on that scope shot -- the transients might be < 0.001 of a pixel wide, so it's very lucky to see them at all.

The random appearance or absence of spikes is a consequence of aliasing on the scope -- quite simply, most of the time, it doesn't happen to sample the input voltage at the exact moment the peak is its tallest. To correctly resolve them, you must zoom in on the rising edge, as tightly as possible, probably with a timebase well below 1us/div. Then you will see all the spikes and ringing in their full, horrendous glory.

2. You don't show the complete clamp circuit.
(2a. Good on you to put in a clamp circuit at all!)

With an even slightly inductive load, the instant the transistor(s) turn off, drain voltage rises, and keeps on rising (past whatever +V is on the far side of the load). The diode must be placed as close to the transistor as possible, so it can do its job as soon as possible. But a diode alone doesn't solve anything; +V needs to be stable, too. So you also need bypass caps, as close as possible to the diode and transistor. For something in the power switching range, you might be looking at an electrolytic >= 100uF.

And one possible gotcha:
3. TC4xxx are only rated to 18V, so any supply spikes are a hazard. I recommend a TVS (at least P6KE or SMAJ size, unidirectional, 12 or 15V nominal rating) and good filtering (you have 4.7uF among others, so that's not too bad; some electrolytic bulk might be handy, but may not be necessary).

Note that, when a transistor dies, it typically goes three-way-shorted, which means drain voltage getting to the gate. Which usually destroys the driver. If it's not dying, the overshoot and spikes, and ground bounce, and stuff, can cause problems, not just to the drivers but the rest of the circuit. Layout is a concern; ground loop; etc.

Tim

MarkM · « **Reply #3 on:** July 21, 2015, 07:03:38 pm »

Hi, thanks for the replies. My buddy did add an extra bypass cap and it seems to help supress the spike a bit, but just one of those spikes above 18v and the driver is destroyed. Would the zener be fast enough?

I don't have any type of clamp on it now. It's a wonder the board has been working as long as it has been.

It is a very low frequency and I really appreciate that you think the driver isn't needed, as I didn't think so either. My original design used a simple current limiting resistor from the MCU pwm pin to the gate of a MOSFET, but I switched to a 3.3v system from 5v and Rds(on) is just too high. My original thought was to use some kind of N channel -> P channel MOSFET array as a buffer with current limiting resistors on the PWM pin from the MCU and the gate pin on the MOSFETs. I found one earlier this morning.

http://www.digikey.com/product-detail/en/SI4559ADY-T1-E3/SI4559ADY-T1-E3CT-ND/1656955

This will definitely handle the spikes.

Do you think a series gate resistor will help using the TC4xxx driver now? It's probably worth a shot.

Simon · « **Reply #4 on:** July 21, 2015, 07:06:54 pm »

putting a resistor in series with the gate as Tim suggested will probably help a lot. You can try a proper TVS but as you have pointed out the manoeuvring room may be limited.

PSR B1257 · « **Reply #5 on:** July 21, 2015, 07:07:32 pm »

Quote

With an even slightly inductive load,

And/or layout. Depending how inductive* the GND-loop is and how close the driver is to the MOSFETs it leads to more or less trouble.

*) At high frequencies (i.e. fast transients) the ohmic component of a wire link becomes less and less dominant over the inductive component.

Simon · « **Reply #6 on:** July 21, 2015, 07:16:23 pm »

Quote from: PSR B1257 on July 21, 2015, 07:07:32 pm

*) At high frequencies (i.e. fast transients) the ohmic component of a wire link becomes less and less dominant over the inductive component.

As I found when I made a basic switchmode supply of my own, it was so bad the mosfet oscilated and was less efficient than a linear regulator...

Zero999 · « **Reply #7 on:** July 21, 2015, 08:08:17 pm »

Quote from: T3sl4co1l on July 21, 2015, 06:52:45 pm

Two problems:

1. Add series gate resistors. This controls the risetime at turnoff, which limits how much trouble you get into with inductances (increasing switching loss instead). At 100Hz, you probably don't need a proper gate driver at all; a level translator would suffice (e.g., just a few BJTs to switch 12V instead of 5 -- or use logic level FETs?). That is to say, by the time you're looking at 1kohm gate resistors, your peak gate current can only ever be 10mA or so, and a freaking several-ampere driver becomes absurd.

That TC4xxx thingy can drive gates in the tens of nanoseconds. That's over a million times faster than your waveform! It's a wonder you saw a spike at all on that scope shot -- the transients might be < 0.001 of a pixel wide, so it's very lucky to see them at all.
Tim

I agree. At such low frequencies, the old 4504 could be used to drive the MOSFETs directly.
http://www.onsemi.com/pub_link/Collateral/MC14504B-D.PDF

MarkM · « **Reply #8 on:** July 21, 2015, 08:12:27 pm »

My buddy added a 1k series gate resistor and the spikes only go to about battery voltage + 1v! Not bad at all! He's also using one 0.1uf and 2x 10uf in parallel for bypass caps close to the driver now. I still wouldn't be comfortable using it like this without some sort of clamp, though.

The other option is using that MOSFET array I linked to instead of the driver. It wouldn't need to be clamped. Any thoughts?

I'd rather not use a large electrolytic cap if I can avoid it, but a stable system would make me happy.

EDIT:I should've said that I'm really trying to get Rds(on) as low as possible, hence the reason for battery voltage at the gates.

krish2487 · « **Reply #9 on:** July 21, 2015, 09:00:43 pm »

Please dont such random values like 1K.
A gate drive design is an interesting topic in and of itself.

Long story short - The series gate resistance, the input gate capacitance and the driver supply voltage decide (for oversimplification purposes) decide how fast the FET turns on, turns off and what the rise time slope will be like.

If you are only operating in the 100 Hz range, then it is a lesser concern but a concern nevertheless. The risetime of the gate decides how long the FET behaves as a resistor - the shorter the time in this region the less power the FET dissipates as wasted power. (Again gross oversimplification but makes the point). The FETs hardly dissipate any power during on time (very low RDSon) and during off time (no drain current except leakage current). Hence any wasted power in the form of heat is during the high-low and low-high transition.

Decide how much time you would like the transistor to spend in the transition stage (generally 2-5 % of the period is a good point to start), appropriately size the resistor to charge the input gate capacitance in the calculated time and choose the nearest value.

Secondly, Were the measurements taken in 1x or 1:10 attenuation? For measurements like this the probe should always be in 1:10 mode.
The scope probe capacitance and the long, flying crocodile clip of GND are good enough reasons to cause the FETs to blow up. The probe is directly parallel to the gate - source terminals -> It is a load for the driver.

MarkM · « **Reply #10 on:** July 21, 2015, 09:10:09 pm »

The 1k value would limit the current at 12v to 12ma. The value will end up being an issue, as this board is designed to work with 7v-12.6v input, but it looks great at 12ma. Here is what the output looks like with a 1k resistor. Even with the right value resistor, the real issue to me is still clamping the driver VDD to 15v or so. Not sure about the best way to do that.

I'm not sure what he had the probe set to. I'll ask him.

krish2487 · « **Reply #11 on:** July 21, 2015, 09:22:19 pm »

Hmm, Actually the reason that a spike occurs on the drain of the FET is during the off time, for the reasons you mentioned.

The inductive load present will not allow the electric field to collapse instantly when the drive is removed. Hence the voltage build up on the drain
aka Spike. The faster you try to turn off the device, the larger the spike on the drain. The 1k resistance is proof. The 12V, 1K, 4nF capacitance charged and discharged the gate at such a slow rate that the inductive spike is heavily damped.

Now, it also means that we can now use a seperate resistance - diode combination for turn on and turn off.
Something like this

-----|>---/\/\/\/----
-------- ------------
-----<|---/\/\/\/----

in the place of the 1K. The top branch is for turn-on and the bottom for the turn-off.
You can start with 22E for the turn on and 47/100E for the turn off. Then add a snubber capacitor directly across the Drain Source of the mosfet, as close as possible. Say something like a 1uF or a 2.2uF. That should damp the spikes considerably.

MarkM · « **Reply #12 on:** July 21, 2015, 09:36:06 pm »

I have a flyback diode in parallel with the load. There's no spikes on the output or between the driver output and MOSFET gates even without the 1k resistor(which is odd, because the driver isn't clamped). The problem is the battery voltage going into the driver. The issue is the positive spikes, I believe. VDD was seeing spikes of up to +30V and killing them. I'm also using two of these MOSFETs in parallel.

Here's a pic of what the battery voltage going to the driver VDD looks like now. I should've posted it before. If you compare it to the last one, there's a massive difference.

krish2487 · « **Reply #13 on:** July 21, 2015, 09:48:27 pm »

I apologize, I misread that you were blowing up the FETs themselves.

Try decoupling the Battery using a diode in series between the battery and the driver VDD pin. Any low forward voltage drop diode should do.

Then after the diode use a ferrite bead in series and then a tank cap (10 uF Electrolytic) || filter cap (0.1uF MLC) across the supply pins, as close as possible. That should clamp the voltage appearing across the driver. Adjust the filter caps as required 0.1/0.2/0.47 etc.

PS : The spikes never occur in between the driver output and the gate. And yes, use the series gate resistor between the driver and the FET. 10/22/47 E preferably.

MarkM · « **Reply #14 on:** July 21, 2015, 09:59:59 pm »

Thanks. I'm still new to electronics, so I might not be very clear on things sometimes.

What does the diode and ferrite bead do? I was hoping for some kind of solution like a TVS diode, but there's not one that exists that meets my requirements. Someone said you can use a zener diode to supress the spikes as well.

krish2487 · « **Reply #15 on:** July 21, 2015, 10:07:17 pm »

The diode "decouples" the battery from the driver. I.E it still supplies the voltage and current required but any noise on the driver is not "loaded" onto the battery and vice versa (to a certain extent).

The ferrite bead smoothens out any such noise and ripples on the supply. Similar to a inductor but a very low value inductor.

Actually a TVS is a zener meant for power applications. And no, a zener is a wrong solution.

T3sl4co1l · « **Reply #16 on:** July 21, 2015, 10:14:19 pm »

A TVS is a perfectly legitimate way to clamp spikes; the question is whether it's best for the task (and given other constraints, like cost or size).

If you already have a well-bypassed +V rail on your circuit, then a clamp diode is far easier, cheaper and more reliable. If you don't, a TVS may well be the better option.

Such is the case with my electronic fuse:
http://seventransistorlabs.com/Fuse/
A TVS absorbs spikes from even the most violent disconnect events: peak currents up to 110A, for microseconds at a time. With an undefined load voltage, the clamp is simply chosen to be somewhat above the nominal rating, so it doesn't interfere with normal operation but protects the transistor even in fault conditions.

Tim

krish2487 · « **Reply #17 on:** July 21, 2015, 10:16:06 pm »

Duh moments!!!

TVS / Zener is a wrong solution for this case.
Not as a generalization.

MarkM · « **Reply #18 on:** July 21, 2015, 10:30:40 pm »

@Tom

What would be considered as a well bypassed voltage rail? There's a 10uf MLCC on the input for the 3.3v regulator and we tested with these MLCC caps on VDD of the driver:

0.1uf + 4.7uf
0.1uf + 10uf
0.1uf + 2 10uf in parallel

Is the clamp diode what krish2487 is talking about?

@krish2487 Thanks. I'll keep that in mind. Cool stuff.

PSR B1257 · « **Reply #19 on:** July 22, 2015, 04:35:52 pm »

Quote

What would be considered as a well bypassed voltage rail?

First of all, you need a proper layout, as I mentioned.
If you have an inappropriate layout, which is causing voltage spikes, it is more or less meaningless to try to clamp these spikes.
It is in general not a good idea to medicate (so to speak) the effect, rather than the reason.

So, what does your layout look like?

However this

Quote

He's also using one 0.1uf and 2x 10uf in parallel for bypass caps close to the driver now.

is how it should be done. But if the return path to the battery and/or the connection to the FETs are inappropriate, it dosen't help a lot.

Regarding the oscilloscope:
To verify the amplitude spikes you should set the trigger above the supply voltage, set it to a fast timebase (i.e. high sampling rate) and have it in single shot mode.

MarkM · « **Reply #20 on:** July 22, 2015, 08:46:39 pm »

I absolutely agree with you. I would much rather have a properly designed board. The design that's having problems is most likely the layout, but I didn't think about that before. I've designed some very basic PCB's before without issues, but as you'll see these have issues. I'd actually rather have someone that knows what they're doing design the PCB if you know anyone. Would rather have it done right the first time instead of these issues. If you could give me some tips though, that would be very much appreciated as well. I have the firmware 99% done for the project and it usually works great until the driver dies. Works for up to a week sometimes.

First off, the huge pads are all stitched, but you can't see in this 3d render. You can see there is a perfect place to put the driver between the battey + and - pads in the old design. That was my beginner mistake and being in a rush, as I widened the space between them to move the 3.3v regulator on the bottom of the board when I could've also moved the driver closer to the FETs. I also didn't even think about the driver return path. The gate trace is also way too long and could've been shortened in the old design. Weird thing is there's no ringing or any other issues at the gates. Maybe that's due to the low frequency?

The new design is one that I sent off 4-5 days ago and is untested. I switched to the tag connect footprint instead of the ICSP header, switched to another driver(which wasn't needed) , added pads on the board for SMD switches, switched some passive components to 0603(also wasn't needed) and moved the driver closer to the FETs. The new design is untested.

Thanks for your time.

MarkM · « **Reply #21 on:** July 22, 2015, 09:25:31 pm »

I've been searching a bit more, and I'm starting to think the inductance from the battery wires could be the issue or at least add to it. The spike is random though, so that doesn't make much sense. I attached a pic from another driver datasheet.

My battery wires are 14g and about 5 inches long total.

nuno · « **Reply #22 on:** July 22, 2015, 10:48:48 pm »

Do you have a schematic?

krish2487 · « **Reply #23 on:** July 23, 2015, 12:06:15 am »

Cable inductances as you mentioned are mostly an issue with a switching power supply. With a battery, it is not really a issue.
Besides, proper LC filtering and decoupling near the Driver IC is more than sufficient to offset and stray transmission inductances.

[size=0px]'ve been searching a bit more, and I'm starting to think the inductance from the battery wires could be the issue or at least add to it. The spike is random though, so that doesn't make much sense. I attached a pic from another driver datasheet. [/size][/size][size=0px]My battery wires are 14g and about 5 inches long total.[/size]

PSR B1257 · « **Reply #24 on:** July 23, 2015, 03:22:06 pm »

Well, the layout without all designators is a bit odd...

Quote

The spike is random though, so that doesn't make much sense.

What doesn't make much sense? My advise regarding the oscilloscope? It makes perfect sence, BECAUSE it is (at lest appears to be) random. As already mentionend by T3sl4co1l, the oscilloscope could give you absolutely wrong readings (if it runs in continiuos mode).

MarkM · « **Reply #25 on:** July 23, 2015, 06:54:17 pm »

I see what you mean. I finally had a chance to test myself. The spike isn't random at all. It happens every time the MOSFETs turn off. I tested how the board was before any modifications and I saw a spike of 18.5v every pulse. I added a 1k series gate resistor and the spike went down to about 13.5vabout 1v over battery voltage or so. I used the board for awhile over and over and the MOSFETs didn't heat up at all. I also added a 470uf electrolytic capacitor and saw no change at all, which was odd.

I know the proper thing to do is use a gate resistor for each MOSFET and to calculate the correct value. Impossibe to add two resistors on this board revision, but I'll fix in the next one. I'll also test a lower value resistor after calculating, but even limiting it to 10-12ma is working just fine @ 60hz. No issues. Side note, I'm trying to figure out how people are using low resistance for the series gate resistor in an acceptable sized package. I'd like to go 0805 max. Maybe 1206 would work, though.

I attached some pics showing the spikes after adding the gate resistor. My scope sucks and is reading a bit higher than it should has well, so the spikes are probably only about 13.3v.

If you have any other ideas, please let me know. The results so far are acceptable with the gate resistor to me. If I get a chance, I'll show you guys the board in action.

PSR B1257 · « **Reply #26 on:** July 23, 2015, 07:57:23 pm »

Quote

I also added a 470uf electrolytic capacitor and saw no change at all, which was odd.

The battery has a very low ESR and the wires to the board are also quite short, therefore the voltage drop at the input terminals of the board is quite low too.

Try this with wires ten times this long (or even longer), with and without a capacitor and you'll most likely see a difference.

Quote

I finally had a chance to test myself. The spike isn't random at all.

There you go...

Quote

Impossibe to add two resistors on this board revision

I wouldn't be too concerned about a single resistor for both MOSFETs, not to mention the resistor-diode combination for different charge and discharge times. For this application your one resistor is just fine.

MarkM · « **Reply #27 on:** July 23, 2015, 08:26:43 pm »

Thanks! I'm trying to see if I'm calulating this correctly. From what I calculated, the 1k resistor with this setup should be just fine.

10v worst case Vgs with 3s input using a driver with dead battery
1k ohm series gate resistor
2 MOSFETs in parallel with 25nC gate charge(rounded up to 30nC each)
Frequency = 100hz
Period = 10ms

10v/1k= 10ma of current

60nC/10ma = 6µS(0.006ms) rise/fall time.

Time the MOSFETs spend in linear region each period:
0.006ms * 2 / 10ms=0.12%

^rise + fall time divided by period. Is that correct? Even limiting it to 10ma seems perfectly OK to me.

MarkM · « **Reply #28 on:** July 25, 2015, 05:07:54 am »

Well, the driver is still dying, but for some reason I didn't realize that it's not while it's running. I've only killed a couple so far myself. The drivers are actually being killed by power on transients.

krish2487 · « **Reply #29 on:** July 25, 2015, 07:46:41 am »

The calculations look good.

You have started another thread on the PSU transient spikes on the driver.
I ll just post the reply to it here, because I have already suggested a solution for it on this thread.

Diode decoupling and putting a LC filter inbetween the battery and the driver Power pins.
If you use a slow diode like 1N400x then you can more effectively isolate the fast acting transients from propagating to the driver chip.

As I have mentioned earlier, the spikes occur during the turn off. The same property of an inductor (no instantaneous rise / fall of current) allows you to place a inductor(ferrite bead) in series with the battery and reduce the spike.

If you are able to see the spikes on a oscilloscope, then you can see the difference it makes with and without a ferrite bead.

T3sl4co1l · « **Reply #30 on:** July 25, 2015, 12:16:28 pm »

Quote from: krish2487 on July 25, 2015, 07:46:41 am

Diode decoupling and putting a LC filter inbetween the battery and the driver Power pins.
If you use a slow diode like 1N400x then you can more effectively isolate the fast acting transients from propagating to the driver chip.

Uh...

MarkM · « **Reply #31 on:** July 25, 2015, 04:40:02 pm »

Hi, thanks again. I was thinking my question was different enough to create another post, but you're right, it comes back to the same thing. I'm researching now.

EDIT:I forgot to post. My buddy was intentionally sending voltage spikes of 25v into the MOSFET driver and it wouldn't die. I think the issue could be inrush current.

MarkM · « **Reply #32 on:** July 26, 2015, 08:45:31 am »

Unfortunately, I can't capture the power on spike with my crappy wannabe scope. I did run some tests with my board, though. I added in a rocker switch on the input from the battery pack and a 100µf electrolytic capacitor on the input from the battery as well. I sat there and switched on and off over and over(even very quickly) for literally 20+ minutes. The driver never died. Keep in mind this is still the TC4427 on this board.

I took the big cap off the board and the driver smoked the first time I turned it on! I put another driver on and forgot to add the cap for testing. The driver smoked the second time I turned it on. It's a bit of a bummer, as I don't have anymore, but I'm going to order more and really test it for a long time. It's almost confirmed. Needs more testing and to be scoped properly.

The other thing I did is I rigged the TC4422a to test since I have some. At first I added it to a breadboard with the same value ceramic caps soldered to 2.54mm pin headers with the same rocker switch setup. I ran the test over and over for awhile with the bulk cap. It didn't die.... I took the big cap off and it still didn't die. Bummer... I didn't test as long as the first one, but I tested for a good 5 minutes over and over.

I ended up soldering the TC4422a with jumpers on my board, as the pinout is different from the TC4427 and test like that. I still couldn't kill it without the cap for awhile, but I eventually did after 5 minutes or so. I don't think the 4422a is quite as susceptible to the inrush current/power on spike. There's no reason for me to use this driver, though.

I'll update this thread when I confirm this and hopefully mark it solved. Thanks again for your help!

Zero999 · « **Reply #33 on:** July 26, 2015, 09:53:10 am »

Quote from: MarkM on July 26, 2015, 08:45:31 am

Unfortunately, I can't capture the power on spike with my crappy wannabe scope

I notice you're using Rigol 'scope. What model? Rigol is generally a good brand, especially for the price. Have you hacked it to achieve the full bandwidth?

MarkM · « **Reply #34 on:** July 26, 2015, 05:07:57 pm »

Those are pics from my buddy's scope that's helping me test. I'm not sure which model it is, but he said something about hacking it. He captured the power up spike on his. I have one of those little Hantek USB scopes.

Zero999 · « **Reply #35 on:** July 26, 2015, 05:20:08 pm »

Quote from: MarkM on July 26, 2015, 05:07:57 pm

Those are pics from my buddy's scope that's helping me test. I'm not sure which model it is, but he said something about hacking it. He captured the power up spike on his. I have one of those little Hantek USB scopes.

Those USB 'scopes are shite.

MarkM · « **Reply #36 on:** July 26, 2015, 05:27:16 pm »

Haha! For sure. I found a guy local with a Hantek DSO5102P for $200. I might just go for it. I need to be able to test myself.

Zarhi · « **Reply #37 on:** July 27, 2015, 04:39:44 am »

Quote from: MarkM on July 21, 2015, 06:38:13 pm

I'm switching two of these N channel MOSFETs in parallel with a TC4422a driver @ 60-100hz using an Atmega328 PWM pin on a custom PCB on battery power(2s-3s lipo pack). I started with the TC4427 driver and they died randomly. Now the TC4422a is dying randomly.

Same situation two years ago. LED pwm controller, PIC16F1509, 2xTC4427A, 4xIRF1010NS, 78L05. Mosfet drivers randomly dying at power on. Powered from 12V/7Ah lead-acid battery, small 12V 'wallmart' type adapter, ATX PSU, professional 12V/500W constant voltage LED power supply... no difference: tc4427 randomly dies at power on ( start heating itself ). I replace about 20 tc4427a, none of them survives more that 10-15 power on. Does not matter LED's are connected to mosfets, or not. Interesting was that if tc4427 survives power on, they work infinite time ( over an week ), including multiple cpu programming.

Series resistor between tc4427 and irf1010ns does not help. Decoupling also does not help. Just for test I made another pcb with TC4468. And no problems at all. I never was able to burn TC4468, even with inductive load.

Final pcb was made without mosfet drivers at all. PIC was able to drive irf1010ns at 976Hz and 3906Hz just fine, power loss was acceptable.

MarkM · « **Reply #38 on:** July 27, 2015, 08:29:50 am »

Thanks for the tip! Glad to see I'm not the only one.

T3sl4co1l · « **Reply #39 on:** July 27, 2015, 01:20:02 pm »

I did mention earlier that the supply voltage ratings of these drivers are quite low.

If you're going to keep blowing chips, and never going to take the time to figure out what's actually going on, I can't be of much help.

If you're interested in doing some scoping, we can probably find why they're dying.

Specifically, power-on transients sounds like it might be related to lead inductance ringing with bypass capacitance.

Tim

MarkM · « **Reply #40 on:** July 27, 2015, 09:43:32 pm »

Hi, the thread is a bit rough to follow.. My buddy has a Rigol that he's using. I've been using a POS hantek 6022be, which I can't capture the spike with. I took a drive and scored a hantek DSO5102P last night for $185, so I can actually test myself now. I didn't feel like I was just 'burning chips', because there was a point to it and I couldn't do anything else myself.

This issue is the ceramic capacitors charging when the battery is plugged in. That's 100% confirmed. My buddy tested the tc4427 on a bare board with and without the capacitors. Without the capacitors the tc4427 didn't die and there was no huge spike when turned on. Add caps and there's a huge spike and the tc4427 is dead. Actually, you can test this with caps by themselves and see the same thing. https://www.pololu.com/docs/0J16/all

Zarhi said he/she had the same issue with the tc4427 dying like that and it's most likely the same issue. The TC4422a was very hard to kill for me even with the caps. I'm for sure ditching the TC4xxx either way, though. I'm looking at cheaper options like the ZGD3003, which should be fine. For some reason this driver isn't very popular, and I don't know why. Any thoughts?

As far as actually fixing the issue, I've been thinking about it and reading quite a bit. I think the best option would be to use a MOSFET to limit the inrush current. What are your thoughts on that?

http://www.eet-china.com/ARTICLES/2000DEC/2000DEC13_AMD_AN5.PDF?SOURCES=DOWNLOAD

MarkM · « **Reply #41 on:** July 27, 2015, 09:57:15 pm »

Yes, that's definitely the issue.

steverino · « **Reply #42 on:** July 27, 2015, 10:35:57 pm »

Just curious, the referenced LT application note concluded "To obtain optimum transient characteristic, the input circuit
has to be damped." and then showed three different simple damping solutions. Will these not work in your circuit?

MarkM · « **Reply #43 on:** July 27, 2015, 11:13:28 pm »

The Motorola app note? I'm going to have to read both a few more times to let all of the calculations sink in... I'm honestly not even sure if this is really needed for my board. The little arduino boards use a similar LDO to power them with similar input caps, but I've never seen one with any kind of inrush current limiter. I've plugged and unplugged battery packs over and over while testing with zero issues even after a long time. My board is basically an arduino UNO with MOSFETs, a MOSFET driver and an OLED display on the board.

The other MOSFET driver is good up to 40v(ZXGD3003). The max spikes that was seen was 25v or so. Would you add some kind of inrush current limiter in this situation?

MarkM · « **Reply #44 on:** July 28, 2015, 12:32:27 am »

Thanks. The 10x AL cap is actually what I tested so far, but not with the scope that I just aquired. The driver didn't die, so it definitely helps. There may be room on the board to add the big cap and use the other driver.

I think adding the big cap with the other driver would be a reasonable solution.

MarkM · « **Reply #45 on:** July 28, 2015, 03:27:11 am »

Man, you lost me on those pics. Care to elaborate on that?

I got a chance to capture some of the spikes myself and test some things. My setup is simple. I have a 10µf ceramic cap soldered to 2.54mm pin headers, a rocker switch, XT60 connecter, 3s battery pack and a load resistor on a breadboard. Check out the first pic. The next pic is with a 47µf electrolytic capacitor. After that is a 220µf. I lost my 100µf cap somehow... It looks like a 100µf cap should do just fine for my purposes from the results. 220µf would work as well if I could find one small enough. Any thoughts?

I need to sit down and learn how to use the scope. It wouldn't show the max values for each trigger. I'll figure that out. Loving it so far.

nuno · « **Reply #46 on:** July 28, 2015, 08:45:43 am »

If you twist your power cable plus and minus, or put them more apart, does the ringing frequency of the 1st test (only the 10uF ceramic cap) change?

Looks like the wiring inductance is resonating with the input cap. The resonance frequency of such arrangement (series LC) is given by: Fr = 1 / (2 x pi x SQRT(L x C))

A rough estimate from your 1st scope shot could be 5us period, with C = 10uF and solving for L we have 5us = 2 x pi x SQRT(L x 10uF) => L = (5us / (2 x Pi))^2 / 10uF ~ 63.4nH which sounds realistic for a "piece" of wiring.

In AcHmed99's circuit, with 4.7uH inductance and 39uF cap, the formula above gives ~11.7KHz for the resonance frequency, which is spot on with his simulations.

T3sl4co1l · « **Reply #47 on:** July 28, 2015, 02:30:31 pm »

Quote from: nuno on July 28, 2015, 08:45:43 am

If you twist your power cable plus and minus, or put them more apart, does the ringing frequency of the 1st test (only the 10uF ceramic cap) change?

Should be. Mind that, if you're thinking of a solution along these lines -- this isn't a help, because who's going to give a crap about where they are? They're just wires... Hey, I hooked it up and now it doesn't work...

Better to embrace the effect (free filtering..?), and damp it to an expected worst-case level (say, several meters of loose wiring -- a few uH, probably).

You can also use a TVS to clamp the spike.

If you're curious, this is what a short circuit at 30V looks like, when you switch it into 150mm of twisted pair:

20A/div. The leading edge is about 0.2uH inductance, and whatever resistance was around (it's curved a bit). The hump and cutoff are due to the current limiting and fault protecting action of the electronic fuse being tested.

What's not seen during this transient is the voltage: it dips down (putting the full 30V across the wiring), then the current limiter pulls out of saturation and the voltage overshoots. A TVS diode protects the current limiter, clamping the flyback at 50-60V.

You can do a similar thing for supply protection -- and should -- by adding a TVS across the supply, at the connector or at the first large capacitor. This also affords surge and reverse protection.

Tim

nuno · « **Reply #48 on:** July 28, 2015, 06:01:37 pm »

Quote from: T3sl4co1l on July 28, 2015, 02:30:31 pm

Quote from: nuno on July 28, 2015, 08:45:43 am
If you twist your power cable plus and minus, or put them more apart, does the ringing frequency of the 1st test (only the 10uF ceramic cap) change?

Should be. Mind that, if you're thinking of a solution along these lines -- this isn't a help, because who's going to give a crap about where they are? They're just wires... Hey, I hooked it up and now it doesn't work...

I'm not

. It's just an easy way for the OP to confirm that the effect is indeed this wiring-cap resonance.

Personally, and since I've met this effect while simulating "power systems" ^*, I would just add input capacitance, also making sure the resonance frequency is far away from any "power" oscillations (such as PWM signals) used in the system.

^* Ideally we want caps with as low ESR (and ESL) as possible here. Caps are charged/discharged in a controlled way.

T3sl4co1l · « **Reply #49 on:** July 28, 2015, 06:52:53 pm »

Quote from: nuno on July 28, 2015, 06:01:37 pm

Personally, and since I've met this effect while simulating "power systems" ^*, I would just add input capacitance, also making sure the resonance frequency is far away from any "power" oscillations (such as PWM signals) used in the system.

^* Ideally we want caps with as low ESR (and ESL) as possible here. Caps are charged/discharged in a controlled way.

Depends on what "power systems" include, but this can be undesirable as well. Startup and turn-on transients can very easily excite unintended resonances, so it's always good design to make sure all resonances are well damped. If nothing else, you can't always guarantee that the only inputs to, or outputs from, the system will be fully under control of the control circuitry. E.g., say you slew-rate limit the circuit, so it can't experience startup transients under normal conditions. But someone can always short-circuit the output, forcing a transient there. This can extend to, yes, rather unlikely stimuli (e.g., a diode fails shorted, discharging the filter network?), but the "best practice" conclusion is that, such situations should be avoided, by design, in the first place.

For an analogy, it's the difference between a car that drives absolutely smoothly, versus one that has some stupid rattle or vibration when the engine is at a particular RPM and load. Or suspension that rides over bumps just fine, but sways just a little nauseatingly (a common mode - differential mode problem).

What I like to do with DC link filters is, put on enough electrolytic bulk as needed for filtering / ride-through, and enough film capacitors to handle the high frequency switching ripple and to get a nice low loop impedance going into the inverter (which may have additional inductance added intentionally). The electrolytics don't have vanishingly small ESR, so they act to dampen the film caps (which, depending on how they are connected by traces or pours, may act essentially like one solid bulk capacitor, or may exhibit some lumped-equivalent transmission line behavior; in either case, ESR plays a role in damping it).

You generally want to avoid multiple levels of bypasses, because you don't need it, and you don't want it. Placing small ceramics right at the inverter results in an additional LC loop, which probably isn't well damped because sqrt(L/C) will be higher than the surrounding impedances, and a suitable amount of series inductance
is desirable in the inverter.

Tim

MarkM · « **Reply #50 on:** July 28, 2015, 09:18:20 pm »

I really appreciate all of the great info! Thanks for the replies. I'm going to go through everything and research. Some of this is over my head ATM.

nuno · « **Reply #51 on:** July 29, 2015, 10:48:46 pm »

The application I was simulating was 5 - 10KW DC motor control (<80V). High currents and high ESR -> heat on the electrolytic caps.

I've seen the "double cap level" recommendations for this application, but I could see in simulation that, under certain conditions, the bulk eletrolytics can resonate with the ceramic/film ones; the problem with this is increased RMS current on the caps. And it's hard to actually measure this (at least at my hobby level), so how can you be sure it's not happening?

I've also seen some designs with only film, but that translates to too little capacitance to leave me unworried about battery - DC link caps loop inductance.

Due to the power involved in this application is hard to protected from all cases, and so in my limited knowledge, I prefer to keep things simple and add tolerance.

T3sl4co1l · « **Reply #52 on:** July 30, 2015, 01:27:13 pm »

Quote from: nuno on July 29, 2015, 10:48:46 pm

The application I was simulating was 5 - 10KW DC motor control (<80V). High currents and high ESR -> heat on the electrolytic caps.

Yikes! Pile on the caps, indeed!

Quote

I've seen the "double cap level" recommendations for this application, but I could see in simulation that, under certain conditions, the bulk eletrolytics can resonate with the ceramic/film ones; the problem with this is increased RMS current on the caps. And it's hard to actually measure this (at least at my hobby level), so how can you be sure it's not happening?

If they are resonating, there is voltage somewhere, and there is current somewhere [else]. The voltage ripple need not be much because the impedance will be low, but the current can be detected easily with an inductive probe.

If you can guarantee the correctly damped combination of ESR and impedance by design, that's probably fine.

Quote

I've also seen some designs with only film, but that translates to too little capacitance to leave me unworried about battery - DC link caps loop inductance.

Due to the power involved in this application is hard to protected from all cases, and so in my limited knowledge, I prefer to keep things simple and add tolerance.

For something like that, about all you can do is pile on electrolytics -- the voltage is too low for films anyway, but you might consider the low-voltage equivalent of them, which is aluminum polymer (equivalent in terms of ESR time constant and energy density). With such low impedances, it's fairly unlikely that you'll get resonances, but it can still be checked in the simulator, assuming some typical wiring inductances and ESR. And maybe even PCB resistance..

And for the battery end, you can also invoke some rough estimate of battery ESR (which depends on all sorts of things, such as charge state and ionic diffusion, worse for some types), which might not help much, but you can also look at using TVSs to clamp what spike remains (of course, it needs to be a bit larger than your average SMAJ size TVS... have fun with that?).

Tim

nuno · « **Reply #53 on:** August 04, 2015, 11:18:59 pm »

Quote from: T3sl4co1l on July 30, 2015, 01:27:13 pm

Quote
I've seen the "double cap level" recommendations for this application, but I could see in simulation that, under certain conditions, the bulk eletrolytics can resonate with the ceramic/film ones; the problem with this is increased RMS current on the caps. And it's hard to actually measure this (at least at my hobby level), so how can you be sure it's not happening?

If they are resonating, there is voltage somewhere, and there is current somewhere [else]. The voltage ripple need not be much because the impedance will be low, but the current can be detected easily with an inductive probe.

Only if you mean one of those utterly expensive "proximity" current probes, because otherwise there isn't space for any other kind of (current) probe. The voltage ripple is indeed very small, and the frequency is high (extra problem for those special current probes).

Quote from: T3sl4co1l

If you can guarantee the correctly damped combination of ESR and impedance by design, that's probably fine.

Problem may be getting the specs from the caps.

Quote from: T3sl4co1l

For something like that, about all you can do is pile on electrolytics -- the voltage is too low for films anyway, but you might consider the low-voltage equivalent of them, which is aluminum polymer (equivalent in terms of ESR time constant and energy density). With such low impedances, it's fairly unlikely that you'll get resonances, but it can still be checked in the simulator, assuming some typical wiring inductances and ESR. And maybe even PCB resistance..

I'll have to look more carefully at polymers, because at the 1st look I got the impression that they are too "sensitive" compared to electrolytics.

Quote from: T3sl4co1l

And for the battery end, you can also invoke some rough estimate of battery ESR (which depends on all sorts of things, such as charge state and ionic diffusion, worse for some types), which might not help much, but you can also look at using TVSs to clamp what spike remains (of course, it needs to be a bit larger than your average SMAJ size TVS... have fun with that?).

Yes, just a little larger, lol. From the simulations, the worst thing is battery - controller (DC link caps) wiring inductance.

T3sl4co1l · « **Reply #54 on:** August 05, 2015, 01:58:20 am »

Quote from: nuno on August 04, 2015, 11:18:59 pm

Only if you mean one of those utterly expensive "proximity" current probes, because otherwise there isn't space for any other kind of (current) probe. The voltage ripple is indeed very small, and the frequency is high (extra problem for those special current probes).

Expensive?

Two turns on the end of some RG-174, so what?

Quote

I'll have to look more carefully at polymers, because at the 1st look I got the impression that they are too "sensitive" compared to electrolytics.

How so?

The capacitance is stable, the ripple, ESR and lifetime are specified, and the breakdown voltage is conservative.

Apparently, they breakdown much like film caps do, self-healing at about double the rated voltage.

The only downside seems to be the wear mechanism, which leads to failure (ESR rising I think???) at high temperatures. The culprit is ingress of humidity, which degrades the polymer. As this is a single order diffusion mechanism, lifetime should be quite good at lower temperatures (which are easier to achieve, thanks to the low ESR).

I don't use them all that much, mainly because their ESR is too low!

I don't have many applications where electrolytic ESR (or an estimate thereof) isn't enough. But they're perfect for a really low impedance application like yours. If you don't mind the price of course...

Tim

nuno · « **Reply #55 on:** August 09, 2015, 12:54:03 pm »

Quote from: T3sl4co1l on August 05, 2015, 01:58:20 am

Expensive?
(...)
Two turns on the end of some RG-174, so what?

Ah, that... I was thinking about those that look like a pen.
I see what you mean, there will be high current in this application (so the source signal is strong) and we just want to see if it's ringing so no need to have an absolute amplitude measurement... makes sense, thanks for the tip!

Quote from: T3sl4co1l

The capacitance is stable, the ripple, ESR and lifetime are specified, and the breakdown voltage is conservative.

Apparently, they breakdown much like film caps do, self-healing at about double the rated voltage.

The only downside seems to be the wear mechanism, which leads to failure (ESR rising I think???) at high (...)

Ok, I'll have a better look at them.


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Transient positive voltage spikes killing MOSFET drivers (Read 26211 times)

steverino

Share me