Drive signal issues with half bridge

Z

zeus_threat

Member
Hi Microsim, the current schematic was tested with 100V and it worked but with 220V it blew up the fuse after and the mosfets 30 seconds or so. Most proabaly cross conduction due to remaining charge on the GS not switching off completely i suppose due to those steps in my waveform and that bad current shape indicating the same issue. There was a transformer buzzing noise too due to the problem i have mentionned above that was most probably driving the GDT to saturation. Thats why am insisting so much on help to waveshape the GDT secondary properly.
 
MicrosiM

MicrosiM

Administrator
Staff member
New smps

Zeus

Then you can make a new one using a NEW board like that if you like.

I want you to look for toroidal cores for GDT, and show me all you can get.

I will show you schematic for driver + Help step by step.

I think you have oscilloscope. things will be easy.

Also, I dont like that transformer at output, can you get ETD34? or ETD39? or ETD49?

Thanks
 
Z

zeus_threat

Member
Hi Microsim, i don't doubt that a toroid core could be better than an EEL16 though i have seen EEL16 used in very high quality PSU's as gate drive like silverstone strider series. The EI-33 core is suited for the power and frequency i am targetting ~140W @ 50KHz and BMAX~2000G, though this core has been used to powers of 250W and more. I don't see the need right now to go for an ETD core. All these cores have specs that need to be abide to, Bmax V/S frequency and so on, surely some are more popular in the DIY scene like the ETD series but IMHO its a question of matching the part specs to the design and i think mine does.
 
MicrosiM

MicrosiM

Administrator
Staff member
That might be true.

But I am still trying to help you. and you are no longer interested

I offered you help to start making one.

Thanks
 
Z

zeus_threat

Member
Hi Microsim don't get me wrong i do appreciate your efforts you are providing and it is really helpful but as i mentionned i would like to try get my this one working even at least get the wave shaped correctly it will be a win. Its more a learning project first of all. Dumping this project for the time i've been struggling with this would be a little bit like losing battle. If its a new one based on the same components i.e. TL494 + toroidal GDT this time, i don't mind considering but dropping to go to SG3525 + IR2110 or an IR2153 is something am considering to go after later on.
 
Last edited:
MicrosiM

MicrosiM

Administrator
Staff member
Smps

zeus_threat said:
Hi Microsim don't get me wrong i do appreciate your efforts you are providing and it is really helpful but as i mentionned i would like to try get my this one working even at least get the wave shaped correctly it will be a win. Its more a learning project first of all. Dumping this project for the time i've been struggling with this would be a little bit like losing battle. If its a new one based on the sample componets TL494 + GDT i don't mind considering but dropping to go to SG3525 + IR2110 or an IR2153 is something am considering to go after later on.
Click to expand...


For me, I do prefer to use the GDT, and I didn't ask you to use IR2110 either. I wanted to debug your circuit from the beginning.

And start every thing from 0, like winding transformers and making a new board, same as old one, but with some help from me.

Your driving circuit is not like mine, some slight changes needed.

Any way, If you are ready to make a new board + Re winding transformers, Let me know!

So we can get started! :D
 
Z

zeus_threat

Member
Okies sorry for misunderstanding lets get going i have double almost all components cores, ic transistors. We can start from veryyyyyyyyyyyyyyy basic, the calculations? Before we get serious what do you think? It will be a good way for me to check things that went wrong as well if thats ok with you?
 
MicrosiM

MicrosiM

Administrator
Staff member
Start

zeus_threat said:
Okies sorry for misunderstanding lets get going i have double almost all components cores, ic transistors. We can start from veryyyyyyyyyyyyyyy basic, the calculations? Before we get serious what do you think? It will be a good way for me to check things that went wrong as well if thats ok with you?
Click to expand...

Zeus

See attached schematic. Use a NEW board, Build this section first without GDT and the Mosfets and high voltage section.

Only the SG3525 + 4 Transistors, and show me wave at output of transistors when you finish while make sure your voltage is +15VDC.

Build that section in good layout on the PCB.

Cores you have on hand are new or old ones? kindly show me clear pictures

Thats the first STEP! :)

GDT_Driver.jpg
 
Z

zeus_threat

Member
Micro my friend sorry to tell you that but SG3525 is a very rare component here as is KA3525 plus ZTX are impossible to find but 2SAXXX and 2SCXXX we have chances. As controller ic i will have to keep going with TL494. Is there any particular reason for goign with the SG3525?
 
Z

zeus_threat

Member
Waveform switching times fixed

I've fixed my waveform at least for the switching times. I will post an updated schematic. Will the step on top of my signal cause any issues? ADS00001.jpg. Am thinking of using a zener to flatten it. Thanks
 
Z

zeus_threat

Member
Am maybe no flattening on second thought i fear the interaction of zener on rise and fall time. Now gate and ic signal are properly in time
 
Z

zeus_threat

Member
Hi Microsim i could add a zener in the secondary gate drive to flatten the signal to 10V completely and dissipate that step, but it could disrupt my timing. What do you think?
 
MicrosiM

MicrosiM

Administrator
Staff member
Gdt

Zeus

You told in first post

(The gate drive transfo is an EEL16 , an AL of 810, 1:1 turns ratio giving 100 turns primary and secondary.)

Is that TRUE?????

100 Turns??? :eek:

In reference to this :- http://www.feryster.pl/polski/eel16.php?lang=en


I am not sure if your core will do the Job Zeus.

I am not sure, AL/nH of 3000 at least is needed for best results
 
Z

zeus_threat

Member
Hi Micro yes right 800. In parallel i am comparing toroid core v/s price ir2110 in ranges of 5 or 10 and shipping cost to decide which would be a more interesting approach.
 
MicrosiM

MicrosiM

Administrator
Staff member
zeus_threat said:
Hi Micro yes right 800. In parallel i am comparing toroid core v/s price ir2110 in ranges of 5 or 10 and shipping cost to decide which would be a more interesting approach.
Click to expand...

1 Mistake with SMPS, Or blowing Mosfets / IGBTs will make your IR2110 fly to heaven.

I dont like those chips, thats why I am using GDT into my commercial SMPS.

So, Think carefully
 
You must log in or register to reply here.
Top