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

DCDC-pushpull drawbacks

Status
Not open for further replies.

dark

Member
Hello forum,
I am thinking to make an isolated convertr using pushpull as shown below schematic.Please advise is there any drawback in using such a topology . I am making my own because board space / cost is constraint beside it has to meet explosion proof standard .The voltage I try is 3V to 3.25 volt isolation . I can see the regulation is rather poor because of no feed back . Is there any other problems that could arise with such a topology.

Thanks for your time.
PS: topology shown for reference only
 
Last edited:

Hero999

Banned
Here's a simpler version that has the advantage of being self-oscillating.

Another problem with these simple circuits is high voltage pulses generated by the coil destroying the transistors, especially on lighter loads.
 

Attachments

dark

Member
Here's a simpler version that has the advantage of being self-oscillating.

Another problem with these simple circuits is high voltage pulses generated by the coil destroying the transistors, especially on lighter loads.
Thanks for reply , can you please provide link to transformer design for these , and how to derate these converters.

-Adi
 

MrAl

Well-Known Member
Most Helpful Member
Hi,

Sometimes transistors explode in converters because something goes
wrong and the transistor draws a huge current. Because of this a fuse
is needed to cap the max energy available to the transistors in case
of a failure. It also prevents circuit traces from blowing off the board
because of excess current too.

You need to study transformer design a bit as you need some leakage
inductance to help with the flux 'ratcheting' effect.
If you cant find anything else, try OnSemiconductor's site. They have
excellent power converter app notes and other documentation.
 

dark

Member
Hi,

Sometimes transistors explode in converters because something goes
wrong and the transistor draws a huge current. Because of this a fuse
is needed to cap the max energy available to the transistors in case
of a failure. It also prevents circuit traces from blowing off the board
because of excess current too.

You need to study transformer design a bit as you need some leakage
inductance to help with the flux 'ratcheting' effect.
If you cant find anything else, try OnSemiconductor's site. They have
excellent power converter app notes and other documentation.

Thanks , I will check the Onsemi site . Can one replace a EI core transformer with a Potcore ,I wonder are these compatible if we get similar inductances? .what to look for .

-Adi
 

MrAl

Well-Known Member
Most Helpful Member
Hi,

Potcores work pretty nicely for converters. I would definitely look into that.
I like the ease in which the windings can be made too, compared to a
toroid which is difficult.
 

tcmtech

Banned
Most Helpful Member
The basic push pull oscillator circuit is very adaptable and can take surprisingly rough service if the components are properly sized.
I have used that basic circuit with the addition of one or two additional capacitors in parallel with the primary and or secondary windings.
It can work with 50 -60 cycle iron core transformers or toroids but the ferrite or potted type cores are better for higher frequency use and do help pick up some added efficiency too.

When done right I have been able to get that circuit to actually work reliably with air core coil setups at several 10's of megahertz using cheap transistors!
And using vacuum tubes it will easily do 100's of Mhz.
 

dark

Member
Hi,

I understand these type of converters has to be treated well , I plan to oscillate it at around 50KHz , Can you please advise how to select toroids for this and how to keep turn ratios. Application notes specific to this will be very help full.

Thanks
 

Hero999

Banned
The transformer in this circuit I posted is not critical. It will always oscillate at the optimum frequency. If you use a low frequency mains transformer it will oscillate at a low frequency and if you use a high frequency ferrite it will oscillate at a much higher frequency.
 

colin55

Well-Known Member
1. How many milliamp do you require?
2. With your +5v -5v design, both outputs must be loaded for the outputs to deliver current.
 

Mr RB

Well-Known Member
Does the output NEED to be isolated? Or do you just need a -5v supply?

Since you seem to be daunted by winding the transformer there are very simple ways to get a -5v supply using a cheap 2-wire pre-wound inductor.
 

dark

Member
Hi,

Hero999 said:
The transformer in this circuit I posted is not critical. It will always oscillate at the optimum frequency.
If you use a low frequency mains transformer it will oscillate at a low frequency and if you use a high frequency
ferrite it will oscillate at a much higher frequency.
But I guess at higher frequency we require less turns?


colin55 said:
1. How many milliamp do you require?
2. With your +5v -5v design, both outputs must be loaded for the outputs to deliver current.
I require around 250mA . Do you say if the o/p isnt loaded the converter will fail as soon as its powewred up.


Mr RB said:
Does the output NEED to be isolated? Or do you just need a -5v supply?
Since you seem to be daunted by winding the transformer there are very simple ways to
get a -5v supply using a cheap 2-wire pre-wound inductor.
Yes , Thats the issue . The O/P has to be isolated and I require isolated supplies .


Thanks
 

Hero999

Banned
Yes at higher frequencies less turns are required.

The disadvantage with higher frequencies is that switching loss becomes more of a factor since there's a limit to how fast the transistors can switch. Hysteresis and eddy losses in thre transformer also contribute.

The advantage is higher frequency is generally better, smaller transformers cost less and fewer turns mean lower resistive losses.

The circuit also oscillates at a higher frequency at lower loads.
 

colin55

Well-Known Member
That means you cannot use the supply above as it will be unbalanced.

Use the principle of the following circuit and add another winding to get -5v.
And add another winding for the +5v if you want it to be isolated. Put the windings on an old 10mm antenna rod - 15mmm long.
 

Attachments

Last edited:

dark

Member
That means you cannot use the supply above as it will be unbalanced.

Use the principle of the following circuit and add another winding to get -5v.
And add another winding for the +5v if you want it to be isolated. Put the windings on an old 10mm antenna rod - 15mmm long.
The schematic attached , is a non isolated and I dont even know calculating turns properly for this topology.

BTW what is meant by "transistor induced shoot through"

PS : I dont have any hard-n fast rules the negative side could be provided more current if that causes an unbalance . Power availablity is a no problem . What I want is isolation . I cannot use a commercial product as I in the next step I will work out for a appropriate certification .


Thanks
 
Last edited:

Sceadwian

Banned
Why can't you use a commercial product? Certification is for the entire unit.
 

dark

Member
Why can't you use a commercial product? Certification is for the entire unit.
Hi , My client needs ATEX approval and some custom voltages .
Like 5V to 5.25V(200mA) isolation. 3V to 3.3V(250mA) etc . Also the size has to be too small.
 
Last edited:
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top