Author Topic: Gate drivers ICS Dying( UCC37322 and UCC37321 )  (Read 1110 times)

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #20 on: March 19, 2020, 07:49:45 PM »
As I said before the Slayer exciter is just for providing the feedback signal which is then picked up by the antennae.( It is being powered from by bench power supply so completely isolated from the driver .)

And I feel the volatage on both of the waveforms on the gates of the two mosfets are same .
And yes I admit there is a slight damping happening that could be fixed by reducing the series resistor value with gate from from the gdt by a little bit .
And by tantalum I mean the ones used for decoupling of the ICS
So instead of it can I use ceramic and film caps of 100 nf coz I don't have uF caps in a usable size .( I have film mkp type 1uf cap but obviously it's too big )
« Last Edit: March 19, 2020, 07:53:00 PM by prabhatkumar »

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #21 on: March 19, 2020, 08:27:08 PM »
Oh right ok, I forget it uses antenna feedback. btw I edited my previous post on the other page.

The decoupling caps across the IC can be lots of random ones in parallel, that's what I did. It's just there as a low esr energy reservoir for the fast peak current draw until the main filter capacitor can catch up across the PCB parasitic inductance.
« Last Edit: March 19, 2020, 08:31:56 PM by John123 »

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #22 on: March 19, 2020, 08:45:04 PM »
Ok so now what do I do next after adding 2 more 104 caps ??
Like just tell me some way in which I can be sure not to burn my mosfet drivers again while testing . And I hope my core which I used for the gdt is doing okay
Edit :now I know what might be the reason of killing the mosfet drivers i.e overcurrent at the output might cause it as John123 mentioned. But I have small question which might be stupid but please excuse I am a newbie and still learning . You said that the max current is given by 8.7 v / 1.47 ohm. But then should it not be the rms value taken as the voltage here ?
« Last Edit: March 19, 2020, 08:53:19 PM by prabhatkumar »

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #23 on: March 19, 2020, 09:12:45 PM »
It's the peak currents which matter too with semiconductors as high peak currents lasting just microseconds can destroy them. The RMS current ie the average current draw measured from the supply will always be much less, for example I've got a flyback driver which can draw a maximum average current of 8-10 amps from the supply but whilst it's doing that the drive mosfet switching the primary coil will see 20+ amps peak.

If you're happy to move onto the next testing step then here's what I'd do next, supply the gate drivers with at least 12v (to fully turn the mosfets on but put a resistor in series to limit the current if anything were to go wrong again, something like 5-10 ohms a few watts.


Now connect the bridge up to a low voltage source and limit the current somehow with a resistor or power supply current limit, again a low ohms power resistor is fine.

Without connecting anything to where the primary coil connects take a look at the gate waveforms, do they look ok with voltage applied to the bridge? are the chips or limiting resistors getting hot?


The idea with these resistors and low voltage is to prevent damage if anything starts misbehaving and save the chips.
« Last Edit: March 19, 2020, 09:20:56 PM by John123 »

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #24 on: March 19, 2020, 09:24:24 PM »
Thanks for the reply once again !! The thing is that I am currently driving the circuit with a 9 volt transformer . So I get around +10-10 volts across the gdt but if I use a 12 volt transformer the peaks are aas high as 18 volt which might even destroy the ucc drivers once again . And I am fine with mosfets being driven in that high resistance linear region for now . I will see on that issue later . So I will proceed as you said and update about the status tomorrow only if mosfet drivers don't fail .

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #25 on: March 19, 2020, 09:26:18 PM »
Thanks for the reply once again !! The thing is that I am currently driving the circuit with a 9 volt transformer . So I get around +10-10 volts across the gdt but if I use a 12 volt transformer the peaks are aas high as 18 volt which might even destroy the ucc drivers once again . And I am fine with mosfets being driven in that high resistance linear region for now . I will see on that issue later . So I will proceed as you said and update about the status tomorrow only if mosfet drivers don't fail .

Oh is that a 12v ac transformer or is 12v the DC rectified and smoothed value? What is the voltage of that 12v power supply unloaded? Sometimes these wall warts only put out their name plate voltage at the rated current.

Urm yeah I guess its ok to keep them in the linear region for testing then, just don't pass too much current through them drain to source (it has to do with the mosfet safe operating area too)!
« Last Edit: March 19, 2020, 09:28:37 PM by John123 »

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #26 on: March 19, 2020, 09:28:23 PM »
When I connect the 12 volt ac transformer, then with a bridge rectified and a 2200uF capacitor on the output the unloaded voltage is as high as 16.2 volts
And I am using a 9 v ac transformer currently so loaded voltage is around 12.1 volt dc

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #27 on: March 19, 2020, 09:32:43 PM »
When I connect the 12 volt ac transformer, then with a bridge rectified and a 2200uF capacitor on the output the unloaded voltage is as high as 16.2 volts
And I am using a 9 v ac transformer currently so loaded voltage is around 12.1 volt dc

Ah yes that's normal, it's the filter capacitor charging up to the peak of the AC waveform.


Oh crap I just noticed in the datasheet of the gate drive chips you're using:

My cheap solution in this case would be to put a couple of diodes in series with the +16v to drop the volts a bit and then add a 15v zener diode/tvs across the supply rails to clamp any voltage spikes from destroying the chip.
« Last Edit: March 19, 2020, 09:34:37 PM by John123 »

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #28 on: March 19, 2020, 09:34:09 PM »
That's the sole reason I went with a 9 volt ac transformer at the first place now after blowing few chips

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #29 on: March 19, 2020, 09:37:11 PM »
That's the sole reason I went with a 9 volt ac transformer at the first place now after blowing few chips

Ok I must of glanced over that then sorry, yeah anything above 16v (including brief transient spikes) will blow the gate drive chips. Put a few diodes in series to drop the max volts a bit if you've got any laying around. Then put a zener/TVS across the supply pins near the IC to clamp any voltage spikes.

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #30 on: March 19, 2020, 09:39:35 PM »
Can you just draw a rough diagram sort of to tell me how to put the zeners in the antiparallel configuration for clamping the voltage at 15 volts ( I don't have 15 volt TVS ).

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #31 on: March 19, 2020, 09:43:20 PM »
Can you just draw a rough diagram sort of to tell me how to put the zeners in the antiparallel configuration for clamping the voltage at 15 volts ( I don't have 15 volt TVS ).

What value of zeners do you have spare? If not have you got any LEDs and how many regular diodes have you got?

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #32 on: March 19, 2020, 09:45:01 PM »
I have an assortment of 1 watt zeners ..so no issue with the values I hope I could get one from that ..and also have leds and plenty of fr107 and 1n4007

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #33 on: March 19, 2020, 09:46:08 PM »
What voltage zeners do you have?

Here:

It's just a case of dropping the extra volts and clamping spikes on the rail by placing the 15v zener right at the supply pins of the gate drive IC's.
« Last Edit: March 19, 2020, 10:11:52 PM by John123 »

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #34 on: March 19, 2020, 10:26:29 PM »
Thanks a lot for this . Also do I need to put additional freewheeling diodes at the outputs of the ucc driver as stated above ?

Now all I want is that mads replies and also justifies maybe if there are certain issues with his design or is it just me experiencing these weird issues which no one or maybe very few people encountered beacuse I have done quite a lot of research and then finally thought to take help here as I couldn't find anything which related to the dying of driver chips which I am encountering
« Last Edit: March 20, 2020, 06:05:20 PM by Mads Barnkob »

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #35 on: March 19, 2020, 10:43:15 PM »
Well I always put the freewheeling diodes in on mine, doesn't hurt anything and only helps to protect the output stages of the chips.

I don't think there's anything wrong with Mad's design, it's just the gate drive chips can't take anything above 16v even for a moment. And that transformer was putting out more than that plus brief dv/dt induced voltage spikes.

That's why the 15v zener/TVS should be literally soldered across the supply pins of the IC to catch any fast spikes, I have a feeling the extra capacitors you installed across them might also be helping to dampen transient voltage spikes on the supply rails (especially if it made them run cooler).

To scope the supply rail spiking use the little ground spring instead of ground wire, it may only be for a very short period like in the nanoseconds and have very fast edges. Regular switchable x1, x10 scope probes should be on x10 to get maximum bandwidth (unless you've got some fancy probe).
« Last Edit: March 19, 2020, 10:54:24 PM by John123 »

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #36 on: March 20, 2020, 05:03:53 AM »
So now I will do the recommended changes and be back again hopefully if everything is right .And yah I am aware of that spring method to measure the noise but then I think I will just go ahead with the modifications proposed. Also another thing which might give some clue is that whenever the gate driver ICS died the schmitt trigger (c40106)also died with it. So does it have to do something with this or just a matter of coincidence ?
PS : Previous to this , I was giving supply to the driver by the bench power supply which had regulated 12 volt with current limiting so it rules out the chances of ICS dying due to spiking .
« Last Edit: March 20, 2020, 06:08:54 AM by prabhatkumar »

Offline John123

  • High Voltage Experimenter
  • **
  • Posts: 89
  • Karma: +2/-0
  • All the gear, no idea.
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #37 on: March 20, 2020, 11:57:13 AM »
You could have the the most expensive and best power supply in the world and the voltage spikes will still appear on the circuit board, it's down to physics and the parasitic inductance of the various wires and PCB traces.

The gate drive chip outputs are switching really fast with fast edges, even with capacitors close the parasitic series inductances effectively creates a boost converter on the supply pins of the chip. These spikes won't show up on the power supply, as far as your power supply is concerned there's a steady 12v across its output terminals. it's the reason why commercial circuit boards are often plastered with capacitors of all shapes and sizes many in parallel with each other (especially computer motherboards).

Some videos on the subject, the chips being used doesn't matter as its the same across the board but just gets worse and worse with faster speed switching and currents (like the 9 amp peak currents of the gate drive IC's).
https://www.youtube.com/watch?v=9EaTdc2mr34
https://www.youtube.com/watch?v=VfCu-siq0-Y (at 6:00 when he removes the ceramic capacitor, that's when the effects of parasitic inductance kicks in).
https://www.youtube.com/watch?v=1xicZF9glH0

Offline prabhatkumar

  • High Voltage Enthusiast
  • *
  • Posts: 46
  • Karma: +0/-0
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #38 on: March 20, 2020, 12:29:27 PM »
I agree to John123 that spiking can be present . Niether is my my power supply a fancy one like all rigol and keysight or any other popular manufacturer of good power supplies . Mine is cheap one ( but still I feel it's expensive beacuse my budget very restricted on such things) .
Another issue which arised now is that after doing the mods ( adding freewheeling diodes across the outputs of the chips ( I used uf4007 ) , is that now my half bridge current through the primary of the gdt looks somewhat funny . Beacuse what happenes now is that the current through the gdt goes higher in amplitude than it it goes below the zero reference line , which I think might cause problems so till now also I didn't attach a supply to the bridge. I also couldn't fit more decoupling capacitors as the space is now completely taken by the freewheeling diodes. I have 100 nf film and 1uf ceramic across each of the ucc chips
 Please tell me if I do need to add more ) . I will put the waveform right here .
I did the mods by referring this schematic as told in the forum above ( I also used the 1 ohm resistor between vcc and 12 volt and ground and 12volt , so is that causing any issue ) .

I sincerely request you all to put views on why is the half bridge signal turning on longer for one side while not so for the other side ?. Is it normal so that i can proceed with connecting a load and low voltage power supply to the bridge with a resistor in series to reduce the current.If not normal then please put some solutions and also why is happening so
« Last Edit: March 20, 2020, 06:05:57 PM by Mads Barnkob »

Online davekni

  • High Voltage Engineer
  • ****
  • Posts: 270
  • Karma: +14/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #39 on: March 21, 2020, 04:25:49 AM »
The 1-ohm resistors in the ground (pin 3) gate driver leads are a big problem.  Current spikes during switching can easily move the ground high enough to switch the logic state seen on the driver's input.  Ideally there's a good ground plane to which the drivers both connect, along with the circuitry driving the inputs.
David Knierim

High Voltage Forum

Re: Gate drivers ICS Dying( UCC37322 and UCC37321 )
« Reply #39 on: March 21, 2020, 04:25:49 AM »

 


* Recent Topics and Posts

post Re: TVS diode selection for 400v transistor
[Electronic Circuits]
davekni
Today at 06:35:13 PM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
Today at 05:00:11 PM
post Re: TVS diode selection for 400v transistor
[Electronic Circuits]
John123
Today at 04:30:55 PM
post Re: TVS diode selection for 400v transistor
[Electronic Circuits]
SteveN87
Today at 01:16:47 PM
post Re: Conical pulse skip DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Acid Byte
Today at 07:23:21 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ritaismyconscience
Today at 06:13:52 AM
post Re: A few (stupidly basic) questions
[Beginners]
MRMILSTAR
Today at 05:31:27 AM
post Re: synchronous motors up for bid
[General Chat]
klugesmith
Today at 02:55:41 AM
post Re: A few (stupidly basic) questions
[Beginners]
davekni
Today at 02:23:23 AM
post Re: Skm100 coil
[Dual Resonant Solid State Tesla coils (DRSSTC)]
oneKone
Today at 02:22:33 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
Today at 02:10:26 AM
post synchronous motors up for bid
[General Chat]
thedoc298
Today at 02:05:39 AM
post Re: DRSSTC design questions + project progress
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
Today at 12:12:15 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
Today at 12:06:18 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ritaismyconscience
March 29, 2020, 10:51:57 PM
post Re: Corona lockdown thread, tell about your situation
[General Chat]
Mads Barnkob
March 29, 2020, 10:04:21 PM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
March 29, 2020, 09:59:16 PM
post Re: Conical pulse skip DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
March 29, 2020, 09:45:05 PM
post A few (stupidly basic) questions
[Beginners]
curtislemay
March 29, 2020, 09:32:01 PM
post Re: TVS diode selection for 400v transistor
[Electronic Circuits]
davekni
March 29, 2020, 09:08:39 PM
post Re: Personal Protective Equipment reuse / Covid-19
[General Chat]
klugesmith
March 29, 2020, 05:44:32 PM
post Re: Personal Protective Equipment reuse / Covid-19
[General Chat]
rikkitikkitavi
March 29, 2020, 03:18:22 PM
post Conical pulse skip DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Acid Byte
March 29, 2020, 11:48:49 AM
post Re: TVS diode selection for 400v transistor
[Electronic Circuits]
John123
March 29, 2020, 07:52:58 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ritaismyconscience
March 29, 2020, 12:51:38 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
March 29, 2020, 12:47:38 AM
post Re: No current flowing through half bridge (kaizer sstc 3)
[Solid State Tesla Coils (SSTC)]
davekni
March 29, 2020, 12:43:24 AM
post Re: Personal Protective Equipment reuse / Covid-19
[General Chat]
johnf
March 29, 2020, 12:34:14 AM
post Re: GDT keeps on killing IGBTs
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ritaismyconscience
March 29, 2020, 12:06:03 AM
post Re: Personal Protective Equipment reuse / Covid-19
[General Chat]
rikkitikkitavi
March 28, 2020, 11:53:57 PM
post Re: Corona lockdown thread, tell about your situation
[General Chat]
Mads Barnkob
March 28, 2020, 10:27:34 PM
post Re: Does not name a type even though I made a typedef
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
T3sl4co1l
March 28, 2020, 10:15:20 PM
post Re: Personal Protective Equipment reuse / Covid-19
[General Chat]
johnf
March 28, 2020, 10:08:35 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic Circuits]
T3sl4co1l
March 28, 2020, 10:08:16 PM
post Re: Personal Protective Equipment reuse / Covid-19
[General Chat]
johnf
March 28, 2020, 09:53:39 PM
post Re: Personal Protective Equipment reuse / Covid-19
[General Chat]
rikkitikkitavi
March 28, 2020, 08:34:30 PM
post Re: No current flowing through half bridge (kaizer sstc 3)
[Solid State Tesla Coils (SSTC)]
prabhatkumar
March 28, 2020, 07:33:57 PM
post Re: No current flowing through half bridge (kaizer sstc 3)
[Solid State Tesla Coils (SSTC)]
davekni
March 28, 2020, 07:06:19 PM
post Re: Does not name a type even though I made a typedef
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
SteveN87
March 28, 2020, 06:45:21 PM
post Re: No current flowing through half bridge (kaizer sstc 3)
[Solid State Tesla Coils (SSTC)]
prabhatkumar
March 28, 2020, 06:30:27 PM
post Re: Does not name a type even though I made a typedef
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
klugesmith
March 28, 2020, 06:27:49 PM
post Re: No current flowing through half bridge (kaizer sstc 3)
[Solid State Tesla Coils (SSTC)]
davekni
March 28, 2020, 05:54:40 PM
post Re: Does not name a type even though I made a typedef
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
March 28, 2020, 04:47:45 PM
post Re: Does not name a type even though I made a typedef
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
SteveN87
March 28, 2020, 03:27:54 PM
post Re: Smoking power resistor wattage overhead
[Electronic Circuits]
petespaco
March 28, 2020, 03:10:04 PM
post Re: Does not name a type even though I made a typedef
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
March 28, 2020, 01:29:53 PM
post Personal Protective Equipment reuse / Covid-19
[General Chat]
profdc9
March 28, 2020, 10:20:04 AM
post Re: No current flowing through half bridge (kaizer sstc 3)
[Solid State Tesla Coils (SSTC)]
prabhatkumar
March 28, 2020, 07:05:29 AM
post Re: g10 disk
[General Chat]
MRMILSTAR
March 28, 2020, 04:50:34 AM
post Re: TVS diode selection for 400v transistor
[Electronic Circuits]
davekni
March 28, 2020, 04:44:17 AM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
SimplePortal 2.3.6 © 2008-2014, SimplePortal