Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Query on about a 12VDc-310VDC smps unit

Status
Not open for further replies.

Iykeeh Ezeonyekwelu

Junior Member level 2
Joined
Jan 15, 2014
Messages
23
Helped
1
Reputation
2
Reaction score
1
Trophy points
3
Activity points
169
Hello Guys..... Been a very long time.
I wish to design a 12VDC-310VDC SMPS unit. I picked SG3525 to provide 150khz PWM on each leg of IRFB3207. My question is, Can the FET go as high as 150khz?. My transformer is ETD44, from the datasheet, it gives approx. 500W at 100khz. My guess is that at 150kHz, its can sustain 500W and still have a little headroom.

- - - Updated - - -

it's a push push topology amd i intend to use 2FETs per leg.

- - - Updated - - -

it's a push push topology and i intend to use 2FETs per leg.
 

Use a proper Mosfet gate driver,
And calculate the proper turnon and turnoff network.

Keep us posted about your progress.
 

I was thinking that SG3525 have enough current to efficiently drive two IRFB3207 per leg.
 

At lower frequencies it might.
At 150 khz it will have difficulty driving a pair of fets on each side during the Miller plateau.

I suggest that, based on fet data sheet values, that you calculate how much gate charge you have to transfer, and then with the proposed frequency, that you calculate the average current.

You'll be surprised.
 

Yea, about that.... forgive my ignorance cos i don't know much.
please, can you detail me on the frequency, current and gate charge relationship. Thanks alot.
 

Hi,

Usually every
* Mosfet manufacturer
* Mosfet gate driver manufacturer
* SMPS IC manufacturer
should have documentation about this on their internet sites.

Klaus
 

What i don't know is how to calculate to get the gate current that can efficiently switch the FET's at hundreds of kilohertz.
Something like a formular that can compute some parameters and give the current required.

- - - Updated - - -

This is my first SMPS unit..... Current Economic state does allow for errors.

- - - Updated - - -

This is my first SMPS..... the current economy does not give room for errors.
 

Hi,

every MOSFET datasheet contains the gate_charge_information.
This is charge is needed for every switch_ON of the MOSFET.
Depending on drive strength the gate current waveform may be short and high, or wider and lower. But the charge stats the same.

It is no DC current. Indeed it´s just some kind of spike.
So if you ask for gate current vs frequency you must know that this value can only be an "average" current.
And this average current is: charge x frequency. But this "average" value often is not of interest, because it has nothing to do with switch_on_time.

Klaus
 

Thanks for your quick response but you are like going in circles. How do i detetmine the required current for the FET cos all i see from the datasheet is gata charge.
 

Hi,

you are like going in circles
A forum is meant to help you to do your job, but we don´t do your job.
You have all the informations you need. The formulas are given in the recommended documents.

Do you want us to calculate the values for you? This is not our job!
But in case someone here wants to calculate the values, then you have to provide the necessary data according the document´s formulas.
You don´t even tell if you are interested in peak current or average current.

Klaus
 

Sorry about the way i sound...... I don't know much about gate charge and drive current relationship. If there is a formular for it, it might b useful to me. I just don't want to jump into what i don't know.
 

Hi,

It´s all explained very well in the documents / application notes. Read them.
I can´t explain it better.

Klaus
 

... How do i detetmine the required current for the FET cos all i see from the datasheet is gata charge.

Current is noting but rate of flow of charge. MOSFET gate is more like a capacitor rather than a resistor. This charge (say Q) must be deposited on the gate for the FET to turn on.

You are going to turn on the fellow 150,000 times every second. So the total charge to be delivered to the MOSFET gate per sec is 150000*Q and this is the average current. This number is useless because this is a rather small number.

Now every cycle (150000/sec) has a period of 6.66 us. You want the FET to turn on fast: 1/100 of this time or less. So the turn-on time should be at least 66ns. Please see whether your MOSFET can turn on fast *enough*- this is key.

So your peak current should be at least 100 times the average current. Please keep 2-5 times this value as margin.

At lower frequency, you should not slow down the turn on period because that will cause too much of dissipation.
 

Thank you bery kindly my friend.... Now it's starting to make sense to me. lemme start doing some home work. I will keep checkin for further inputs. Any contribution is appreciated, No knowledge is a waste.
 

Thanks a lot Guyz.... u guyz rock!. Sorry 4 my long absence, am coming work with school so time is more of a serious issue. Av made the circuit using SG3525 oscillating at 80khz, all seemed but i tried quite a couple of ir2110, 2112....all failed, couldn't get a genuine part. I resorted to "Old school" to drive my irf3205 FETs, i used D882 & B772 with emitter resistors of 4.7R and base resistors of 100R for drivers, a Gate resistor of 10R to fets. To keep "Zap" low, I decided to start off with small Core of 75W with a Ac=0.4cm. My input voltage is Vmax=15V, Vmin=10V, Npri-6T. I made the secondary to have three Voltages: 6V=3T,15V=9T,18V=11T. The system started up, without smoke, Zero Noise ....all wer kool.
Woooooooow.... I took a 12mins Dance.
Hahahahaaaa.....First trial without smoke.
I picked my feedback from the 6V point as i shunt it with a 1k resistor to ground, The secondary and primary shared the same ground, I thought it wouldn't matter much since no lethal voltage is present, though I have 2200uf capacitor on the FETs ground and transformer primary centre tap.
Now the problem....
With proper feedback finetune....all sec voltages were good....but as the Battery voltage starts to drop, the secondary voltages starts to increase at a rate of 10mV battery drop to 100mV secondary voltage increase. I can't figure out wer the "Jiggy" is coming from. I thought it wise to consult mi guardians Before I start suspectin a counterfeit SG3525.

- - - Updated - - -

Happy New Year to all My Very Good People of EDABOOARD....God's Blessings this year I pray!
 

Feedback is an important concept, not only in electronics but also in other areas for example, social sciences, economics, biology, human interactions, psychology and what not. It means, very crudely, looking at the output and trying to adjust the input with an objective to keep the output constant. Here in this case you need to sample one voltage, and compare with some reference and tell the SG3525 to make adjustments. The real question is how?

First thing first: you have three outputs and you want to look at all of them or watch and control one and the others will hopefully behave themselves. If you want all the three to be monitored (but you can have only one feedback signal) you need to add the three voltages (output values) with suitable weights (you may want 6V to be more stable but also want to take care of the other two). I see that the voltages you have mentioned are not exactly in the same turns ratio (3T:9T:11T :: 6:15:18) but that is a different story.

Next thing is giving feedback: SG3525 expects some feedback but I have not seen the datasheet for SG3525 and therefore cannot comment. Use good rectifier (Fast with sufficient current rating) diodes for all the output voltages. Decide the load current and select the filter capacitor. Yes, select good filter capacitors and good resistors to device the voltage (sample a part of the output DC to be used as feedback).

Why the output voltage increases if the input battery voltage decreases?
 

I can't tie the problem to anything...am thinkin of attaching regulators to the outputs to keep it fixed, but just worried.
please, I need help.
 

am thinkin of attaching regulators to the outputs to keep it fixed...

This is not unusual; you need a feedback loop in which a part of the output voltage is sampled and sent back to the input circuit. This value is compared with a set value and the controller IC regulated the current to the transformer based on that.

It is not good for the input and output share the same ground. Hence the feedback is commonly made via an optoisolator. The optoisolators are designed to separate the signals so that the need for a common ground is eliminated.

But finally I have not seen your circuit so far.
 

Yea...
about feedbacks, I saw few circuits online with outputs sharing the same ground with general ground, perhaps, because the output voltages ar not lethal, like an SMPS I saw with dual 35VDC output, the feedback from this output wasn't isolated.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top