Flyback diode problem (heat) in boost converter

Thread Starter

DanDare

Joined Apr 29, 2021
11
Hello!
I need to decide what I should understand/learn to go ahead with my boost converter and stop killing my MOSFETS in the wild.
I'm not a engineer (at least not one of the 'right type') and electronics is just a hobby.
I have a boost converter that's basically scheme.jpg (MOSFETS are IRF640).
board.jpg is my testing board.
The flyback diodes are SBL3040 https://www.vishay.com/docs/88732/sbl3030p.pdf

I have PWM at fixed duty cycle and adjusted (hopefully) it's frequency for the best efficiency (higher possible output DC voltage with the fixed load) and it's around 16KHz

If I leave circuit running without the flyback diodes the MOSFETS gets hot (i cant touch them anymore) after 30 seconds.
If I leave it running with the flyback diodes the MOSFETS are almost just warm after 90 seconds but the robust SBL3040 pair gets hot (i cant touch them anymore).

scope1.jpg shows signal at MOSFET gate with the SBL3040 pair
scope2.jpg shows the same gate signal without the SBL3040 pair

Now I'm curious about:
Are the built-in MOSFETS diodes making them hot running the circuit without the external diodes?
What I need to go ahead investigating/learning to have this system equalized? Or maybe my approach to this is just wrong?

I often see in circuits of this type a capacitor and resistor connected in each of the primary coils. Possibly a filter but I dont know what are they doing or how to calculate them. Like in this example choppercaps.jpg
Perhaps that's what I need? Then again I cannot understand what they are doing.

Any tips/suggestion is appreciated!
Thanks!
 

Attachments

Papabravo

Joined Feb 24, 2006
16,773
You switching frequency seems "low" for this application and your drive circuit with a "big" resistor and a zener diode is also suspect. I think you need to discard this circuit and begin a search for something better, or at least easier on the components.
 

Thread Starter

DanDare

Joined Apr 29, 2021
11
You switching frequency seems "low" for this application and your drive circuit with a "big" resistor and a zener diode is also suspect. I think you need to discard this circuit and begin a search for something better, or at least easier on the components.
Hi there. Ugh I'm sorry I have an error in the schematics, the 300R resistors are in between TL494 and totem pole driver. Actually the link between totem pole and fet gates are direct, no resistors. The zener is just for mosfet gate protection and 12V ones is what I have in hands but I see 15V would be better (considering my VCC is 12V), other than this there are problems about the zener? I've though it would not affect driving efficiency (or at least not if they were 15V ones?).
 

Papabravo

Joined Feb 24, 2006
16,773
Inaccurate schematics are more of a problem for you than for me. How do you expect to debug or troubleshoot what is going on in a circuit that has no accurate representation. Your chances of getting anyone else to hep you are also substantially reduced. If you copied this circuit from somewhere on the internet and you're having trouble I would contact the author. If you just threw a bunch of parts up against the wall and said: "oooooh that looks good", then sad to say you're on you own. I don't know what you think you've made, but it obviously is not "engineered". Go back to square one and actually learn something for a change, instead of wasting time on things, that may look cool, or sorta work, that you don't understand.

Maybe try working with a simulator, as a learning aid.
At least you can make more elegant schematics with one.
 

Thread Starter

DanDare

Joined Apr 29, 2021
11
"Remember the human - Be courteous when replying to others"
I'm pretty sure a bot would be more constructive than you while answering. Perhaps you're a bot, with selected AI that learns (very well) the obnoxiuos part of the internet. That being said, perhaps the problem is this not being the right community to participate (for me). At least this wasn't the impression I had. My suggestion is go back square one about how to "waste" your time on internet.
Anyway there are so many quite basic and relevant answers to my question that could be made than randomly citing a gate's zener. Also quite funny you concluding the PWM frequency is "too low" not even knowing what's the chopper transformer. Yes I'm totally newbie in power electronics, basic stuff to learn. So for any other people reading through this, my conclusions for now for this experiment:

1- We don't have a flyback diode in a push-pull topology, normally. https://electronics.stackexchange.com/questions/83833/irf44n-and-push-pull-regulator-topology
2- Leakage inductance is what causes significant spikes (5x or more Vin), poluting everything, causing extra load to mosfet's, and noise.

I started trying to have the spikes mitigated with zener but not really nice because it must dissipate too much energy. Started trying to think a way do have a simple RCD Snubber but really not sure it's suitable for push-pull topology (suggestions?).
 
Last edited:

Ian0

Joined Aug 7, 2020
3,186
You need an output inductor.
without it, the MOSFETs see the output capacitor as their load, and at high frequency, a big capacitor looks like a dead short.
Is your output rectifier suitable for the application? A Bridge rectifier intended for 50Hz won’t do. You need fast- or schottky diodes.
 

MrAl

Joined Jun 17, 2014
8,473
Hello!
I need to decide what I should understand/learn to go ahead with my boost converter and stop killing my MOSFETS in the wild.
I'm not a engineer (at least not one of the 'right type') and electronics is just a hobby.
I have a boost converter that's basically scheme.jpg (MOSFETS are IRF640).
board.jpg is my testing board.
The flyback diodes are SBL3040 https://www.vishay.com/docs/88732/sbl3030p.pdf

I have PWM at fixed duty cycle and adjusted (hopefully) it's frequency for the best efficiency (higher possible output DC voltage with the fixed load) and it's around 16KHz

If I leave circuit running without the flyback diodes the MOSFETS gets hot (i cant touch them anymore) after 30 seconds.
If I leave it running with the flyback diodes the MOSFETS are almost just warm after 90 seconds but the robust SBL3040 pair gets hot (i cant touch them anymore).

scope1.jpg shows signal at MOSFET gate with the SBL3040 pair
scope2.jpg shows the same gate signal without the SBL3040 pair

Now I'm curious about:
Are the built-in MOSFETS diodes making them hot running the circuit without the external diodes?
What I need to go ahead investigating/learning to have this system equalized? Or maybe my approach to this is just wrong?

I often see in circuits of this type a capacitor and resistor connected in each of the primary coils. Possibly a filter but I dont know what are they doing or how to calculate them. Like in this example choppercaps.jpg
Perhaps that's what I need? Then again I cannot understand what they are doing.

Any tips/suggestion is appreciated!
Thanks!
Hello,

First, i could not open your pdf for the diodes.

Second, it is interesting to see that you have placed diodes across the MOSFETs that is a finer point of converter designs that a lot of people who never worked in that area professionally do not think of doing because the MOSFETs have built in reverse diodes. At first it appears redundant, but it's not. So i wonder where you got that idea from.

Third, if there is leakage inductance in the transformer (which there most likely is) then you may not need an extra inductor on the output to soften the current getting to the output caps.

Fourth, the output bridge needs to be a Schottky or other high speed diode type you cant use regular rectifier diodes like 1N5400 or a bridge rectifier made for line frequencies as it will eat up too much energy when the frequency is higher. If you already have this that's great but if not that's the first thing i would change.

Fifth, it looks like the reason for the transistor diode heating could be because of the spike energy. When a transistor turns off that terminal of the primary will try to shoot up higher than +12v and that means the other terminal of the primary (connected to the opposite transistor drain) will try to go extremely negative causing the opposite diode to conduct. The heating is caused by the amount of current that was flowing in the primary prior to turn off and the time that the current flows in the diode. The current is hard to control because it has to be high enough to feed the output, so the only other thing you can do is try to limit the time the diode conducts. I would think this could be minimized by upgrading the drive circuit. 330 Ohms for the gate drive is surely not a good choice. MOSFETs have very very high static input impedance but very very low dynamic input impedance meaning to get them to turn on and off fast you need very low impedance drivers. A good starting point is 1 amp gate current to turn on, and 1 amp gate current to turn off, and keep the dead time as short as possible (if you can, although PWM may still be a possibility). To get that level of current you would normally use a dedicated MOSFET driver chip made just for that, but if you want you can try a NPN + PNP dual voltage follower (in switch mode). The MOSFET driver chip is the best choice however as it ensures fast turn on and fast turn off.
If you still have problems you might even try slowing down the turn off while speeding up the turn on of the MOSFETs. This would be done using an asymmetrical driver but you have to be careful not to turn off too slow or the MOSFETs themselves will heat up too much with or without extra diodes.
 
Last edited:

Thread Starter

DanDare

Joined Apr 29, 2021
11
Hello Ian and MrAl, many thanks for the insights and guidance.
"We must conquer the temptation to assign bad motives to people who disagree with us.". Thanks for the reminder as We should try not falling in bad temptations. On that note, sorry for any English "drifting" here.

- "... at high frequency, a big capacitor looks like a dead short." Yes I've set TL494 for a really soft "soft start" but observation still aplicable anyway. I must go learn how to decide what/size inductor.
- "... Is your output rectifier suitable for the application?". I'm aware fast Schottky being the right option. I'm usually lost in diodes datasheet, some don't even mention speed. Best I could get here (down the street) was UF5408. At least it have 'Soft Recovery Ultrafast Plastic Rectifier' in the title :). I will try to discover more about this one specific, if enough for the task as I cant decide this yet. Will try to get Schottky ones anyway.
- "... i could not open your pdf for the diodes." Strange, it opens for me. It's "Dual Common Cathode Schottky Rectifier". Actually used half part of it on the scheme. Saved from old ATX SMPS.
- "... it is interesting to see that you have placed diodes across the MOSFETs" At least my conclusion is built-in diodes are not sufficient in power applications (or at least, questionable) generally speaking. Anyway, guy in the stack exchange asnwer I linked said "The body diodes of the MOSFETs are never forward-biased" in a push-pull topology. Should I just not include them? Ir that's not right conclusion?
- "... it looks like the reason for the transistor diode heating could be because of the spike energy" Agreed, the spikes are really big (112V for Vin=12V measured in Vds on the scope) this whole difference in Vds must be consumed anywhere. Not sure I'm right but my conclusion is it's being consumed in the SBL3040 diodes (in subtle ways) or directly in the MOSFETs if diodes are not there. My empiric experimentattion with/without the diodes show this at least. Leakage inductance is a new term for me.

For now I've attached a 500nF/250V in between Vds on just ONE of the sides. It really did ease out the spikes in 80% margin for the mosfet, and about 70% for the other. But i'ts was just a wild try. Well there's some hope that this will do it, hopefully :)

Thanks for the all MOSFET driving insights they are really enlightening. I can see many tests I could try following you guidance. I've got many (9) IR2110 but really want to taste/try tunning this driving from discrete so I can exercise https://www.ti.com/lit/slua618 , the Totem Pole in this case. Even more being just low-side switching in this case is a good opportunity for this.
Yes switching time is key either case.
Again sorry I have the 330R placed wrongly, they are before totem pole not after.
Papabravo is right as no excuses to not go to LTspice (or anything). Anyway for now what I have is the basic push-pull. I have it in LTspice but partially. Still need to get it right about representing the chopper transformer to go ahead (in todo list). LTspice archive is attached now.
Now I see I have some quite bad choices on my PCB, "sneaking" fet drain paths being one of them, among others. I made it really quick but I need other layout, even for this epxeriment. PCB generall overview attached now and Diptrace archives. Sorry if you look in schematics in Diptrace as is really ugly to look, it needs a clean (in todo list).
 

Attachments

Ian0

Joined Aug 7, 2020
3,186
if there is leakage inductance in the transformer (which there most likely is) then you may not need an extra inductor on the output to soften the current getting to the output caps.
True, but if there is leakage inductance between the two halves of the primary it would cause the switch-off spikes on the MOSFETs. Best to have the transformer with as little leakage inductance all round, and then add a known amount of inductance after the rectifier.
Too much leakage inductance will lead to poor regulation.

(Unless, of course, it is an LLC resonant design, which is another story entirely!)

As this is being run open-loop, the dead time will determine the minimum amount of inductance. Allowing 20% ripple current:

ΔI = Imax/5 (because it's 20%)
V= output voltage
Δt = dead time
minimum L=V/(ΔI/Δt) = V.Δt/ΔI

If you are driving MOSFETs, I would recommend the SG3525 instead of the TL494, because it has proper push-pull MOSFET drivers built in. Apart from that feature, the two devices are quite similar.
 

MrAl

Joined Jun 17, 2014
8,473
Hello Ian and MrAl, many thanks for the insights and guidance.
"We must conquer the temptation to assign bad motives to people who disagree with us.". Thanks for the reminder as We should try not falling in bad temptations. On that note, sorry for any English "drifting" here.

- "... at high frequency, a big capacitor looks like a dead short." Yes I've set TL494 for a really soft "soft start" but observation still aplicable anyway. I must go learn how to decide what/size inductor.
- "... Is your output rectifier suitable for the application?". I'm aware fast Schottky being the right option. I'm usually lost in diodes datasheet, some don't even mention speed. Best I could get here (down the street) was UF5408. At least it have 'Soft Recovery Ultrafast Plastic Rectifier' in the title :). I will try to discover more about this one specific, if enough for the task as I cant decide this yet. Will try to get Schottky ones anyway.
- "... i could not open your pdf for the diodes." Strange, it opens for me. It's "Dual Common Cathode Schottky Rectifier". Actually used half part of it on the scheme. Saved from old ATX SMPS.
- "... it is interesting to see that you have placed diodes across the MOSFETs" At least my conclusion is built-in diodes are not sufficient in power applications (or at least, questionable) generally speaking. Anyway, guy in the stack exchange asnwer I linked said "The body diodes of the MOSFETs are never forward-biased" in a push-pull topology. Should I just not include them? Ir that's not right conclusion?
- "... it looks like the reason for the transistor diode heating could be because of the spike energy" Agreed, the spikes are really big (112V for Vin=12V measured in Vds on the scope) this whole difference in Vds must be consumed anywhere. Not sure I'm right but my conclusion is it's being consumed in the SBL3040 diodes (in subtle ways) or directly in the MOSFETs if diodes are not there. My empiric experimentattion with/without the diodes show this at least. Leakage inductance is a new term for me.

For now I've attached a 500nF/250V in between Vds on just ONE of the sides. It really did ease out the spikes in 80% margin for the mosfet, and about 70% for the other. But i'ts was just a wild try. Well there's some hope that this will do it, hopefully :)

Thanks for the all MOSFET driving insights they are really enlightening. I can see many tests I could try following you guidance. I've got many (9) IR2110 but really want to taste/try tunning this driving from discrete so I can exercise https://www.ti.com/lit/slua618 , the Totem Pole in this case. Even more being just low-side switching in this case is a good opportunity for this.
Yes switching time is key either case.
Again sorry I have the 330R placed wrongly, they are before totem pole not after.
Papabravo is right as no excuses to not go to LTspice (or anything). Anyway for now what I have is the basic push-pull. I have it in LTspice but partially. Still need to get it right about representing the chopper transformer to go ahead (in todo list). LTspice archive is attached now.
Now I see I have some quite bad choices on my PCB, "sneaking" fet drain paths being one of them, among others. I made it really quick but I need other layout, even for this epxeriment. PCB generall overview attached now and Diptrace archives. Sorry if you look in schematics in Diptrace as is really ugly to look, it needs a clean (in todo list).

Hi,

A few points of interest...

The body diodes would conduct when the opposite transistor turns off.
Basically each side of the transformer with it's associtaed transistor forms
a boost converter which acts like one for a short time and that is one way
to explain the spike that appears when a transistor turns off. That high spike
couples to the other primary winding and causes that other end to go more negative
than ground, thus forcing the diode to conduct. It amy only be for a short time
though it depends on the dead time and/or PWM pulse time.
If the spikes are too high you may need to add snubbers and that's very typical
for a converter.
Usually we dont use caps across any CE of any transistors because it stresses
both the transistor and the cap. A snubber is different as it also incudes
a diode and a bleeder resistor so it only absorbs spikes not normal switching
voltage energy.
On the PC board, the source conenction should be tight with a separate lead for
the driver return line. That will help stop unwanted ringing which can also
cause more diode heating.
Leakage inductance is the equivalent transformer internal inductance that acts
like either an inductor in series with the output or in series with the input.
One of the main advantages is no need for extra magnetics which beings in
extra cost and weight as well as construction issues if you wind your own.
 

MrAl

Joined Jun 17, 2014
8,473
True, but if there is leakage inductance between the two halves of the primary it would cause the switch-off spikes on the MOSFETs. Best to have the transformer with as little leakage inductance all round, and then add a known amount of inductance after the rectifier.
Too much leakage inductance will lead to poor regulation.

(Unless, of course, it is an LLC resonant design, which is another story entirely!)

As this is being run open-loop, the dead time will determine the minimum amount of inductance. Allowing 20% ripple current:

ΔI = Imax/5 (because it's 20%)
V= output voltage
Δt = dead time
minimum L=V/(ΔI/Δt) = V.Δt/ΔI

If you are driving MOSFETs, I would recommend the SG3525 instead of the TL494, because it has proper push-pull MOSFET drivers built in. Apart from that feature, the two devices are quite similar.
Hi,

Does he have the option to change the output transformer and thus the leakage inductance?
Many people do not have that option.
Even so, an external DC inductor brings up other problems such as added cost and also
that emans there will be DC current in the inductor all the time. Almost all of the
converter we shipped out (long time ago) had leakage inductance in the output transformer
because an extra inductor would be large and be quite heavy in those high power
applications.
Also, the inductor current must flow though the diodes and thus through the secondary so
there will still be back emf getting to the primary.
 

Thread Starter

DanDare

Joined Apr 29, 2021
11
True, but if there is leakage inductance between the two halves of the primary it would cause the switch-off spikes on the MOSFETs. Best to have the transformer with as little leakage inductance all round, and then add a known amount of inductance after the rectifier.
Too much leakage inductance will lead to poor regulation.

(Unless, of course, it is an LLC resonant design, which is another story entirely!)

As this is being run open-loop, the dead time will determine the minimum amount of inductance. Allowing 20% ripple current:

ΔI = Imax/5 (because it's 20%)
V= output voltage
Δt = dead time
minimum L=V/(ΔI/Δt) = V.Δt/ΔI

If you are driving MOSFETs, I would recommend the SG3525 instead of the TL494, because it has proper push-pull MOSFET drivers built in. Apart from that feature, the two devices are quite similar.
Nice. Yeah I've checked SG3525 and want to experiment with it. It also requires less components around. By the other hand as far I'm concerned driver gate output should be the closer as possible to the mosfets to avoid parasitic inductance (ringing) so a external driver seems positive in this aspect (than placing the chip inside a turbulent place?). Well, so nice there are different options to choose for a best fit case by case :) My reference at this moment for this is https://toshiba.semicon-storage.com... parasitic oscillation,as shown in Figure 3.4. 'MOSFET Parallening (Parasitic Oscillation between Parallel Power MOSFETs'

Hi,

A few points of interest...

The body diodes would conduct when the opposite transistor turns off.
Basically each side of the transformer with it's associtaed transistor forms
a boost converter which acts like one for a short time and that is one way
to explain the spike that appears when a transistor turns off. That high spike
couples to the other primary winding and causes that other end to go more negative
than ground, thus forcing the diode to conduct. It amy only be for a short time
though it depends on the dead time and/or PWM pulse time.
If the spikes are too high you may need to add snubbers and that's very typical
for a converter.
Usually we dont use caps across any CE of any transistors because it stresses
both the transistor and the cap. A snubber is different as it also incudes
a diode and a bleeder resistor so it only absorbs spikes not normal switching
voltage energy.
On the PC board, the source conenction should be tight with a separate lead for
the driver return line. That will help stop unwanted ringing which can also
cause more diode heating.
Leakage inductance is the equivalent transformer internal inductance that acts
like either an inductor in series with the output or in series with the input.
One of the main advantages is no need for extra magnetics which beings in
extra cost and weight as well as construction issues if you wind your own.
Thanks for detailed explanations. Not sure I understand "source conenction should be tight with a separate lead". I can understand why it should be tight but not about separate. For noise mitigation?
Someone else (outside this forum) suggested me a 5R resistor and a capacitor across the drain lines to mitigate the spikes. Something that I'm going to test. It looks like a tradeoff between how much mitigation X how much power to dissipate. Of course a propper snubber sounds like the most right option that im considering too (if I'm able to get it right).
 

MrAl

Joined Jun 17, 2014
8,473
Nice. Yeah I've checked SG3525 and want to experiment with it. It also requires less components around. By the other hand as far I'm concerned driver gate output should be the closer as possible to the mosfets to avoid parasitic inductance (ringing) so a external driver seems positive in this aspect (than placing the chip inside a turbulent place?). Well, so nice there are different options to choose for a best fit case by case :) My reference at this moment for this is https://toshiba.semicon-storage.com... parasitic oscillation,as shown in Figure 3.4. 'MOSFET Parallening (Parasitic Oscillation between Parallel Power MOSFETs'



Thanks for detailed explanations. Not sure I understand "source conenction should be tight with a separate lead". I can understand why it should be tight but not about separate. For noise mitigation?
Someone else (outside this forum) suggested me a 5R resistor and a capacitor across the drain lines to mitigate the spikes. Something that I'm going to test. It looks like a tradeoff between how much mitigation X how much power to dissipate. Of course a propper snubber sounds like the most right option that im considering too (if I'm able to get it right).
Hi,

There are at least two connections to the MOSFET source terminal.
1. The heavy ground lead that carries the main power current which gets turned on and off.
2. The light gauge signal ground lead (from the driver circuit),
Now if the main power current (that gets turned on and off) flows though any part of the ground signal lead it modulates the gate to source voltage which when the MOSFET swtiches it could caues it to turn on and off several times before it settles to just one state (on or off). If the signal ground lead is separate, it goes to the source alone so none of the high current can flow though any part of it.
If the signal ground lead is not separate going to the source then that means is must have connected to the power ground lead somewhere before the MOSFET (not directly at the terminal) so when high power current flows it is flowing through the same wire that the signal current is flowing so the signal sees an increase (and later decrease) in current, which of course means that both the resistance of the shared wire and the inductance of the shared wire cause voltage drops. The resistance causes a set amount of voltage drop that starts when the MOSFET switches on and ends when the MOSFET turns off. The inductance of the shared lead causes a voltage drop spike that starts when the MOSFET turns on and decreases some short time later, then goes negative when the MOSET turns off then goes back to zero some short time later. The inductance is the most problematic and causes oscillation but the resistance doesnt help any either.
Keeping the leads separate avoids both of these effects that cause oscillation during part of the switching period which causes ringing which causes more power dissipation in both MOSFET and diode.
So the MOSFET source gets a heavy gauge lead direction from the input capacitors right to the source, and the driver circuit gets a separate lead from the input capacitors to the driver circuit, and then a separate lead directly to the MOSFET source. So there ends up being two separate leads connected directly to the source terminal of each MOSFET.
"Directly" here means as directly as possible. Sometimes a PC card is mounted right on the MOSFET heatsink so there may be only one short lead to the MOSFET, source, but then the separate lead connects to the PC card. If there is a PC card you can still keep them separate though just be running two separate leads from the PC card to the MOSFET source terminal.
So this is quite simple really, see the attachment. The big cap is the input capacitor bank.

MOSFET_Wiring-1.gif
 

Bordodynov

Joined May 20, 2015
2,906
I developed special models of windings, cores (linear and non-linear). Sometimes I need to control the value of B. This can be calculated using external, non-linear sources for calculations. But I am lazy to do it every time and I made a core model in which the calculations are done internally. I derived values of H and B. This allows you to derive the hysteresis loop if you want. But because of additional calculations in such a core, the simulation is longer. In the future I replace the core with a simpler one (without H and B). This speeds up the calculation. You can have a look at my model file for the transformer. It's easy there.
I felt that insulation was unnecessary. But if the power source were a 110 or 230 volt household power grid, then yes. In that case, galvanic isolation is needed.
 

du00000001

Joined Nov 10, 2020
77
Hello!
I need to decide what I should understand/learn to go ahead with my boost converter and stop killing my MOSFETS in the wild.
I'm not a engineer (at least not one of the 'right type') and electronics is just a hobby.
I have a boost converter that's basically scheme.jpg (MOSFETS are IRF640).
board.jpg is my testing board.
The flyback diodes are SBL3040 https://www.vishay.com/docs/88732/sbl3030p.pdf

I have PWM at fixed duty cycle and adjusted (hopefully) it's frequency for the best efficiency (higher possible output DC voltage with the fixed load) and it's around 16KHz

If I leave circuit running without the flyback diodes the MOSFETS gets hot (i cant touch them anymore) after 30 seconds.
If I leave it running with the flyback diodes the MOSFETS are almost just warm after 90 seconds but the robust SBL3040 pair gets hot (i cant touch them anymore).

scope1.jpg shows signal at MOSFET gate with the SBL3040 pair
scope2.jpg shows the same gate signal without the SBL3040 pair

Now I'm curious about:
Are the built-in MOSFETS diodes making them hot running the circuit without the external diodes?
What I need to go ahead investigating/learning to have this system equalized? Or maybe my approach to this is just wrong?

I often see in circuits of this type a capacitor and resistor connected in each of the primary coils. Possibly a filter but I dont know what are they doing or how to calculate them. Like in this example choppercaps.jpg
Perhaps that's what I need? Then again I cannot understand what they are doing.

Any tips/suggestion is appreciated!
Thanks!
R16/R17 and C5/C6 are the key to your issues:
the freewheeling diodes (whether intrinsic or SBLs in parallel) cannot provide the continuity of the current path during switchoff. Thus your MOSFETs resp Schottky diodes are heavily loaded. (No small wonder that the Schottkys survive at all.)
These 2 R-C combinations might be replaceable by simple diodes (cathode at 12 V, anode at the end of the coil) - you can try to put the SBLs there.
I'd expect this to work like a charm.
 
Top