problem with PWM signal

[maicael]

Hi,

A ferrite pot core comes to mind here. They are easier to wind too. But what are you going to use this for and how much power?

BTW 20kHz isnt that high but yes steel laminations are not a good idea.

i am on a team where we are building inverters using different techniques like the H-bridge and centre tap technique.Actually i am the novice there.
we are experimenting on different ideas.
Any way i want to wind a 12-0-12/220V /1000W centre tap transformer that will be switched at the mentioned frequency but i cant use iron core because of high eddy current losses and other forms of losses involved with iron cores operating at high frequencies.
if you have any useful info on ferrite pot core as you mentioned i would really appreciate.
2)I am also looking at LC filters so if you have a pdf or link on how this filters are wound i will be grateful.
i do have some notes though but any addition would do.

 

[MrAl]

Hi,

You seem very interested in learning much more about this, more than one or two posts worth, so i would have to recommend that you check out the "Magnetics, Inc." site where they have a ton of information and design guides for designing magnetic devices like transformers and chokes. There's a lot of information there that should be able to help you out.
For example, in one paper they talk about core material vs operating temperature where the operating temperature has an effect on the efficiency, so they get pretty involved with their discussions.
They also have wire size tables taking into consideration skin effect and stuff like that, so you should find at least something that helps. I think they even have design software now too.

 

[maicael]

Hello,
is there any advantage to powering an IC through a regulator as to powering it directly from power supply.
For example using a 7812 to power an IC instead of powering it directly form the 12V battery source.
Is there any advantage to it like when battery voltage begins to drop or something like that.
i see it in some circuits.
Also how many IC's can a one regulator power at the same time.

 

[alec_t]

A 7812 regulator needs more than 12V for its supply. Check the datasheet.
As for your second question, that's like asking 'How long is a piece of string?' Think about it. Which version of the 7812, which ICs, operating temperature range ....?

 

[MrAl]

Hi,

Some IC's work better with a regulator, some dont need it. It depends on the IC part number.

The 7812 takes a higher voltage and regulates it to 12v. It needs a voltage that is maybe 2v higher though, so that's 14v. It can not boost a voltage lower than that however. There are regulators that can do that but they are more complex and more expensive.

 

[maicael]

hello,
Is a totem pole driver suitable to drive mosfets in push pull configuration which will inturn drive a centre tap transformer as i do not want to drive the mosfets directly from the logic IC it self.
Are there any other better drive circuits for push pull config apart from using a real mosfet driver IC's.

 

[audioguru]

I do not see a schematic of your circuit in this thread so I do not know what needs to drive the Mosfets. The comparators? Do the comparators have a +12V supply?
How much gate voltage do your Mosfets need to fully turn on?

 

[alec_t]

"a real mosfet driver IC" would provide dead-time control to prevent shoot-through, i.e high-side and low-side FETs on simultaneously for an instant. Depending on switching speed and FET spec a simple totem-pole might not allow gate charge to change rapidly enough to prevent some shoot-through.

 

[MrAl]

Hello again,

Here is a quick driver schematic. This is only to drive the logic level MOSFET. Any required delays would have to be built into the code.

 

[maicael]

hello,
to be honest this question was more out of curiosity as i was browsing online and saw simple astable IC4047 used to drive FETS in a push pull manner and there was this talk about using an extra driver circuit so i got curious as to why it is needed in the first place and i saw some reasons after doing some research which is why i asked in the first place.

 

[maicael]

Hello again,

Here is a quick driver schematic. This is only to drive the logic level MOSFET. Any required delays would have to be built into the code.

so this circuit wont drive a power mosfet like the IRFP260N since logic level mosfets require only abut 5V to fully turn on

 

[MrAl]

Hi,

Well it depends on the circuit that actually drives this two transistor circuit. If it is driven directly from a microcontroller pin that only can reach up to +5v then yes, but if you have a drive that is as high as +12v then it can do it. Or alternately if you add a common emitter inverter using another transistor and collector resistor then you can drive this circuit with that, and drive the single transistor with the microcontroller pin, as long as you have a power supply source voltage that goes up to the required level to drive the MOSFET in question (such as +12v available).
I'll add it to the schematic in a minute.

The circuit is shown in the attachment. Note that the Vcc source depends on the MOSFET you intend to drive.

 

[audioguru]

hello,
to be honest this question was more out of curiosity as i was browsing online and saw simple astable IC4047 used to drive FETS in a push pull manner and there was this talk about using an extra driver circuit so i got curious as to why it is needed in the first place and i saw some reasons after doing some research which is why i asked in the first place.

The CD4047 has output current that is too low to drive the high capacitance gate of a Mosfet quickly. Its outputs have no "dead time" so both the upper and lower Mosfets are turned on for a moment shorting the power supply.

 

[maicael]

Hi,

Well it depends on the circuit that actually drives this two transistor circuit. If it is driven directly from a microcontroller pin that only can reach up to +5v then yes, but if you have a drive that is as high as +12v then it can do it. Or alternately if you add a common emitter inverter using another transistor and collector resistor then you can drive this circuit with that, and drive the single transistor with the microcontroller pin, as long as you have a power supply source voltage that goes up to the required level to drive the MOSFET in question (such as +12v available).
I'll add it to the schematic in a minute.

The circuit is shown in the attachment. Note that the Vcc source depends on the MOSFET you intend to drive.

you said it depends on the circuit that actually drives this two transistor circuit.
well i just saw an SG3524 IC which they say is popular so can this IC drive this two transistor circuit and then inturn drive the irfp260n power fet which is not a logic FET assuming the power source is 12V.

 

[MrAl]

Hello again,

I'm pretty sure the SG chip you talk about can drive the two transistor circuit. It must be able to put out around +12v and it should be able to do that. You just have to check the data sheet to see how to get it to do that. That chip also has dead time adjustment i think.

 

[alec_t]

If you're using the SG chip why do you need the 2-transistor circuit?

 

[MrAl]

Hi alec,

That's a good question, and i believe the answer is because the SG chip is not that good at driving a MOSFET only bipolars. Thus the SG drives the external bipolars and they in turn drive the MOSFET. Been a while since i looked at this chip though so i'll check again now.

It looks like it can be done but it wont be as good as using a two transistor driver. The SG chip only has two bipolar outputs, and i dont think we can use them as a totem pole driver. You could look at this in more detail though and see if you can find a way to do it.

Using one of the two transistors as the output drive, we'd have to use a pullup or pulldown resistor, which is not as good as an active low and active high which is what we get with the two bipolar driver.

 

[alec_t]

You're right, there's no active pull-ups on the individual output bjts of the SG3524, but the two bjts can be configured as one push-pull (totem-pole) driver. From the datasheet:-

The outputs may be applied in a push-pull configuration in which their frequency is one-half that of the base oscillator, or paralleled for single-ended applications in which the frequency is equal to that of the oscillator.

So ok for a half bridge without external transistors.
The SG3525A has two totem-pole output stages rated at 400mA peak source and sink for driving mosfets directly.

 

[MrAl]

Hello again alec,

Yes that looks very promising. It looks like the output can drive up to 400ma so it wont be exactly as good as external transistors, but it's not bad really. I'd go for it. The other little catch is that the output active pullup is not just one transistor follower, it's one feeding the other, so there's more voltage drop. So with a +12v power supply we might only be able to count on +10v or maybe +10.5, but that's still not bad and if that's not enough just up the positive rail to +15 or whatever is required. I think that should be a very workable design for frequencies say 100kHz and below, maybe even up to 200kHz except in the most demanding applications.

 

[maicael]

hello,
i am thinking of using a combination of 2SD882 and a BD140 transistor in a totem pole configuration to drive MOSFETS.
i just want to know if they will be very compatible since it is what i can find here for now so far.