Hi all

I've spent many hours trying to design a circuit that fits my need but still didn't reach a good design..
The circuit will have a voltage source as input with voltage 24V DC, the voltage could reach 30V and maybe 32V. The input voltage decreases down to 12V and even less.

Now, the circuit have an output voltage that doesn't exceed 24~26V and disconnects when the input voltage becomes less then 12V. The output voltage could be fixed at any voltage between 24 and 12 or can change according to the input voltage and the circuit... The main point is to have the output voltage not exceed 24~26V and of course not less than 12V or say >15V. The output current should be about 2A, better to be more than 2A.

I thought about looking for an smps dc-dc converter than does what I need, but didn't found such thing in the market.
I tried to make a circuit with several diodes that drops about 6V when the voltage is >24V, and using a comparator circuit with relay... it was an unstuck one.
I also thought about using a linear regulator as 78XX with power PNP transistor for current boosting... but this way will dissipate a lot of power especially when the input voltage is high.
SMPS converter along with a low voltage cutoff circuit is the best way.. but designing smps is a not easy.

Regards,
Hazim

 

How about using a 12V battery to supply voltage to the load, and use the input voltage to supply power to charge the battery. It would be a UPS.
The only problem then is to have an idea how long to run the load if the input drops below about 15V.

 

I drew something together during my lunch break.

See the attached picture.

You still need to do some maths for the voltage divider and zener diode resistors and voltage/current ratings for the MOSFETs.

These are comparators with open drains ! Not Opamps.
Ex. LM2903
The upper one blocks the output voltage when the input voltage is higher than 24V.
The lower comparator is for undervoltage detection.

 

That sort of works, Praondevou - small problem though, the Vgs of the P-ch MOSFET is exceeded if the supply voltage > 20v.

But, I think that Hazim wants a method to keep the output within 13v-24v and stay on, regardless of the input voltage.

That would involve a noninverting buck-boost SMPS; somewhat difficult to build in Lebanon where supplies of components are very limited.

Hazim, can you obtain some broken computer power supplies? Might be able to do something with those relatively cheaply, but we would need for you to tell us what parts you have. It would be difficult for us to list what you need to buy, as in my country we have a huge variety available.

 

Thanks all. praondevou, what SgtWookie said is right. I don't need to block the output voltage when the input exceeds 24V, but I've learned some issues from your circuit.
SgtWookie, I can get many cheap old computer power supplies, and I've some now. I have a wide collection of n-channel power MOSFETs and have many of 2N7000, but power p-channel MOSFETs are harder to get, I have and can get IRF9530 (and smd APM4953. I search here for available parts: http://www.ekt2.com/EKT/prodList.asp It's bad that they don't have categorizations for the transistors. With some changes, can't I use a relay instead of the power MOSFET in your circuit?

 

But, I think that Hazim wants a method to keep the output within 13v-24v and stay on, regardless of the input voltage.

oops, I misunderstood the requirements.

 

What I am thinking of is you building an SMPS using salvaged parts.

But, it would help a great deal if we know ahead of time what you have, or can easily buy.

That website is really atrocious 189 pages of MOSFETs, no specifications, and no really good way to get to where you want to be.

 

Yes this is a big problem I face. That is the main and biggest supplier for electronic parts and this is how there website is... I usually look for transistors that fits my need using vrt or by googling then check what's available using the site's search. I think that the noninverting buck-boost SMPS is what I need and I have saved the attached circuit few days ago, the circuit have several problems that can be managed to fit my need but the main problem is the IC.

What main parts you want me to look and check for availability?

 

Why don't you open up some of those PC power supplies that you have there, and start documenting what ICs are in them?

Let's try to focus on what you DO have, rather than what you cannot get.

 

Right. The IC used is the SMPS controller KA7500B. Here is its datasheet:
http://www.fairchildsemi.com/ds/KA/KA7500B.pdf
I don't think this helps by the way. Other SMPS /PWM controller ICs available are UC3842, KA3525 and SG3524.

 

So......

 

Anyway it's not an easy circuit to design...

 

I haven't forgotten about you. What I'm having problems with is the simulation models for these regulators. I'm running a dual-core Pentium at 2.5GHz, and it still takes a very, very long time to run simulations.

 

Ah.. Thank you very much.

 

Hi again.
We can do it in another way. A step-down dc-dc smps converter to convert the input voltage if it was <24V to some lower voltage, say 22V (or 24V), current should be about 1A at minimum. If the input voltage is 12<Vin<24 then the circuit could pass the input directly to the output plug without any conversion, and we put a cut-off circuit to switch off the output from the input if source voltage decreases under 12V.
So the main point will be the smps step-down converter.

Regards,
Hazim