[PIC] Car Cooling System PSU Design Related

[baileychic]

Is this okay now?

 

[BradtheRad]

Diode D6 only allows charging to the lower battery. It should be placed to stop current backflow to the 12V circuitry, but allow current to the 24V circuitry.

 

[KlausST]

Hi,

There are still a lot if issues..already discussed.

A new one is Q5.
What does it do? It disconnects GND to U1 in a live system. GND is it's reference...most voltages specified in the datasheet refer to that voltage. Not a good idea to disconnect it?
And what do you want to aachieve? To ON/OFF the 12V? Then why not simpky use pin5 of U1? Its meant exactly for that.

Klaus

 

[baileychic]

Q5 was for turning On/Off 12V supply. I will better use ON/OFF pin of LM2596 to achieve it. I will post a new circuit soon.

This is the new circuit.

Check this and reply what needs to be changed.

The 24V from batteries initially run the DC-DC Converter and get +15V which drives the +12V (FAN POWER) and also the +5V for MCU. The MCU after On can control the FAN power.

The Peltier Power control MOSFET is a logic level Mosfet and again MCU will control it.

Discard the PSU-1.0pdf file.

Consider only PSU-1.1.pdf file.

I have used better buck regulator for 5V power to minimize power loss due to LCD backlight and dropping of 15V of DC-DC Converter to 5V using LM7805. The new 5V regulator can provide 5V@3A.

Now, what else needs to be fixed?

The new current requirement is 15V@25A and so a 24Vin to 15Vout@30A DC-DC Converter is being used.

 

[BradtheRad]

Consider only PSU-1.1.pdf file.

Sorry, your attachment still appears to be 1.0.

 

[baileychic]

Sorry, I had forgotten to attach the file. Here it is.

Now, my question is only related to the D6 diode on the battery GND circuit.

If I connect that diode as shown by BradTheRad in post #22 then how can I connect the Battery negative terminal to the GND circuit of Solar Panel / DC-DC Converter input side circuit?

 

[KlausST]

Hi,

in post#22:
* the batteries can be charged by the left hand side circuit.
* the batteries can supply the load on the right hand side.
...it works..


in your schematic:
* Battery B2 can be charged, but never supply a load with power.
This makes B2 useless.

Klaus

 

[baileychic]

The issue is, I don't know whether the DC-DC Converter's input and output side Hands are common or not as I have not yet purchased it but I guess it has a common GND.

Then if I connect the bottom diode as in post 22 circuit then how can I connect the bottom batteries' negative terminal to the Solar Panel / DC-DC Converter input circuit as their GND is connected to say GND of post 22's battery driving circuit's GND.

 

[KlausST]

Hi,

we don´t know either.

I still stay with my post#9

Klaus

 

[baileychic]

See this circuit and tell me if this is correct that is battery -ve terminal and SolarPanel/DC-DC Converter input side circuit. The Batterie's 24V is applied to DC-DC Converter input side circuit.

Regarding power dessipation in MOSFETs and Disodes is this correct?

Rd(mosfet) = I * I * Rds(on) (if rds on = 3/1m Ohms) then 25A * 25 A * 3.1m Ohms = 1.937W

Regsrding diode power dessipation Pd(diode) = Vf * I = 1.2V * 25A = 30W? Is this 30W true per diode?

 

[baileychic]

Reply soon. This is a student project and the student has to submit the project next week. He still has to get the PCB layout designed and manufactured and test the project on hardware and submit it. It is the student's Final year project. There will be Christmas and new year holidays for the PCB manufacturers and he has to get PCB manufactured and assembled before that.

 

[KlausST]

Hi,

Reply soon.

I tend to ignore posts/threads containing such sentences. We all have to do our job, and we don´t need to be stressed in our spare time.
If it´s a student´s project, then still the students are responsible to keep the timeline.

******

I still can´t see how this whole circuit should and can work.
There still are a couple of mistakes....and the intended function is (pratly) not clear.

Issues:
* D1, D2 function unclear
* SW1 is useless
* Q1 circuit or device or drive circuit wrong. Did you check simulation?
* "+15V SENSE" can´t work accurately. Did you check expected values vs simulated values? @ min / max values?
* R14, R15 function unclear
* check L2 value
...I did not check complete schematic/function.

For each node consider highest and lowest voltage levels.

*****

Rd(mosfet) = I * I * Rds(on) (if rds on = 3/1m Ohms) then 25A * 25 A * 3.1m Ohms = 1.937W

The calculation generally is correct. I didn´t check datasheet / values.
Additionally expect wiring loss and expect switching loss.

Regsrding diode power dessipation Pd(diode) = Vf * I = 1.2V * 25A = 30W? Is this 30W true per diode?

The calculation generally is correct. But where does Vf = 1.2V come from? Which diode / datasheet?


Klaus

 

[baileychic]

Hi,


I tend to ignore posts/threads containing such sentences. We all have to do our job, and we don´t need to be stressed in our spare time.
If it´s a student´s project, then still the students are responsible to keep the timeline.

******

I still can´t see how this whole circuit should and can work.
There still are a couple of mistakes....and the intended function is (pratly) not clear.

Issues:
* D1, D2 function unclear
* SW1 is useless
* Q1 circuit or device or drive circuit wrong. Did you check simulation?
* "+15V SENSE" can´t work accurately. Did you check expected values vs simulated values? @ min / max values?
* R14, R15 function unclear
* check L2 value
...I did not check complete schematic/function.

For each node consider highest and lowest voltage levels.

*****


The calculation generally is correct. I didn´t check datasheet / values.
Additionally expect wiring loss and expect switching loss.


The calculation generally is correct. But where does Vf = 1.2V come from? Which diode / datasheet?


Klaus

D1 and D2 are to block the reverse flow of +24V of battery into the Solar Panel.

Whare is SW1 in PSU-1.2.pdf?

The switches in the Battery circuit was only for simulation because if batteries are connected and battery charging voltage graph is plotted then it shows 12V and 12V battery voltages due to the Battery models property.

The Q1 works as needed I guess. You can see the plotted graph. If it is wrong then what is wrong in it?

Why do you think +15V sense can't work? What is the reason. I didn't simulate that in the above Simulation file. I just simulated that circuit with a variable PSU model and that works.

R14 and R15 will not be used in the final circuit even though PCB layout will have the option to insert them. They will not be populated by default. That is only for simulayion to get the graph shown.

L2 is not actually 1H. I forgot to rename it. I will calaculate that and it will be that is both L1 and L2 will be between 47uH to 100uH.


These are the Power MOSFETS and diodes that will be used.

What will be the exact Pd of diode at 15V 28A?

https://www.mouser.in/ProductDetail/IXYS/IXFH270N06T3?qs=YJXtFlaHA7a5qa9uARCJZA==

https://www.mouser.in/ProductDetail/IXYS/IXTP76P10T?qs=cvHLLyFtoE1JQxA7edhB%2BQ==

https://www.mouser.in/ProductDetail...PbsnyyCs1rOfK%2BN29jdflOz9fH%2B%2BbXOcR8SsQ==


Consider this cleaned up circuit.

The main issue is I am sure that the input and output side GNDs of the DC-DC Converter will be common. Then how should I connect the -ve terminal og bottom Battery to DC-DC Converter input circuit/Solar Panel circuit that is the GND of the DC-DC Converter input circuit / Solar Panel circuit will get connected to GND of low side mosfet.

See new graph. All voltages are fine. Even +15V and +15V sense.

 

[KlausST]

Hi,

D1 and D2 are to block the reverse flow of +24V of battery into the Solar Panel.

Ok, then why two diode?
You want to stop draining out the battery? Then aren´t you concerned about the "loop" back from battery to the DCDC converter? It will drain out battery, too, because it tries to generate 15V to charge the batteries from the 24V coming from the batteries.

Whare is SW1 in PSU-1.2.pdf?

Section C1

The Q1 works as needed I guess. You can see the plotted graph. If it is wrong then what is wrong in it?

Q1 seems to be a P-Ch MOSFET. now just mind the diode inside the symbol. The diode is always in forward direction. The diode (and thus the MOSFET device) is conductive. Always. As soon as you swith ON Q2 there will be short circuit. Overcurrent, overheat, fire?

Why do you think +15V sense can't work? What is the reason. I didn't simulate that in the above Simulation file. I just simulated that circuit with a variable PSU model and that works.

I didn´t say it does not work. I said it does not work "accurately".
Why: Already explaind in this thread.
I asked to compare the expected value with the simulation value.
In case the simulation shows the expected value, then I recommend to make a simple test with real parts.
I just pointed you to the problem I see. If you are satisfied with the results, then everything is fine.

The main issue is I am sure that the input and output side GNDs of the DC-DC Converter will be common.

If they are isolated then you are free to connect them or not. (If you connect them, then the quastion arises why you use an isolated one, and not a cheaper non_isolated one)

Then how should I connect the -ve terminal og bottom Battery to DC-DC Converter input circuit/Solar Panel circuit that is the GND of the DC-DC Converter input circuit / Solar Panel circuit will get connected to GND of low side mosfet.

Not sure what this means.

Klaus

 

[baileychic]

Okay, I can remove the D2 diode if it is not necessary. I put that for double protection that is for both +ve and -Ve lines.

Regarding +24V of Battery into DC-DC Converter In which charges the Battery, it will not because when there is no Solar Power or when Battery is fully charged the Battery charging is cut-off that is Mosfets Off.

I am sure the DC-DC Converter is not isolated one. I don't need an isolated one.

The issue is if I have to follow circuit in post #22 then how should I connect the -ve terminal of the Battery (bottom battery) to the DC-DC Converter GND? because (D2 omitted). The GND of input/output of DC-DC Converter will be common and the GND of Battery charging Low-Side MOSFET will also be common that is GND (all connected together) then I can't connect -ve terminal of bottom battery to GND.

How Q1 will short? I am using PWM. The MOSFETs are complementary. So, should I use non-enhancement type MOSFET that is one without internal diode?

Edit:

I see that Q1 shorts.

Shall I replace Q1 and Q2 with 2x N-Channel Power Mosfets and Half-Bridge Driver usign IR2184 then I get complementary output with shorting issue solved?


I solved the MOSFETs issue. Now there is no short. The signals looks fine. Please check it once.

- - - Updated - - -

Will I get +24V battery out voltage into DC-DC Converter Input with PSU-1.4.pdf circuit? Check the -ve terminal of the bottom battery and low-side MOSFET GND and SolarPanel/DC-DC COnverter input GNDs and reply.

 

[KlausST]

Hi,

Regarding +24V of Battery into DC-DC Converter In which charges the Battery, it will not because when there is no Solar Power or when Battery is fully charged the Battery charging is cut-off that is Mosfets Off.

If you say so....I will not question it again..

then I can't connect -ve terminal of bottom battery to GND.

Why not? I don´t understand the problem, but maybe because I don´t understand the whole circuit/function.
You use a simulation tool. Does the simulation show any problem?

Shall I replace Q1 and Q2 with 2x N-Channel Power Mosfets and Half-Bridge Driver usign IR2184 then I get complementary output with shorting issue solved?

What do I know...
One drawback of the IR2184 driver is: It can´t work up to 100% duty cycle. Don´t know if you need this.

Maybe you just wanted to interchange D and S of Q1 (P-Ch). I don´t know. But still there may be a crossbar problem during swtiching.

--> use the simulation tool. First decide what behaviour you expect from your circuit (I often use pencil and paper), then run the simulation and check whether the voltages and currents are as expected.

Klaus

updated:

Will I get +24V battery out voltage into DC-DC Converter Input with PSU-1.4.pdf circuit? Check the -ve terminal of the bottom battery and low-side MOSFET GND and SolarPanel/DC-DC COnverter input GNDs and reply.

Aah, that´s the idea of D6!
OK, D6 prevents cuurrent flow back from battery to DCDC.
But it also prevents any other (load) current flow out of battery to GND.
It makes (what you call) -ve termial to float (w.r.t. GND) in times when battery is not being charged. This makes battery voltage reading via BAT SENSE at least erronous, if not impossible.

In most systems one wants GND continously connected to all devices. Thus you can rely on GND refernce all the time.

Check whether a diode in the +ve path is possible.

btw: Schematic V1.4 does not solve the Q1 problem.

 

[baileychic]

I will go with IR2184 and in the 1.4.pdf it uses a 50% duty PWM and that is what I will use to charge the batteries with 3 to 4A max charging current. Please check the graph in the simulation. It looks okay to me. Are there any other issues? Will any Battery short to GND when low-side MOSFET is On? That is during PWM. The simulation doesn't show any errors.

1.4 file was wrong.

1.5 file is th enew one with IR2184. Please check ther signals on the graph especially the Battery charging signals.

- - - Updated - - -

Just removed the D6 diode and the signal looks fine. Why do we need D6? When Solar Power is not available then SD pin of IR2184 will be used to turn Off charging and +24V of Batteries will be fed into DC-DC Converter to get 15V output.

See this 1.6 pdf. Is this okay now? All fine? Sense GND fine now?

 

[KlausST]

Hi,

See this 1.6 pdf. Is this okay now? All fine? Sense GND fine now?

Don´t know if all this is according your requirements, but now there is at least a better chance for it to work.

Klaus

 

[baileychic]

Okay. But what was the purpose of D6 diode in earlier schematics following from post #22? If Batteries can charge without it then why it was included? We were already using Complementary or one switch On at a time charging device.

Check this updated circuit. It is same as previous one but with current sensors added.

5A current sensor for Battery charging current monitoring and 30A sensor for load current monitoring.

Load current will never exceed 28A.

I am measuring only current of one battery charging because if I try to measure both batteries charging current on the BAT CHG line then when MOSFETs switch the current reading will change sign but if that is a better method then I will take abs() value of the current reading to always get +ve current reading.

What do you say?

There is one new issue. I am getting fatal simulation error related to 12V regulator. What is the issue with it. All others simulate well. If I disable 12V regulator then it simulates fine.

- - - Updated - - -

Should I use heatsinks for these diodes at Vf = 1.57V and Id = 30A?

https://www.mouser.in/ProductDetail...PbsnyyCs1rOfK%2BN29jdflOz9fH%2B%2BbXOcR8SsQ==

 

[KlausST]

Hi,

About D6 ... I can't answer this, because I'm not the designer if the circuit...

Should I use heatsinks for these diodes at Vf = 1.57V and Id = 30A?

This makes 47W of power dissipation.
With a small copper area on the PCB you maybe get a thermal resitance junction to ambient R_th_ja of 30K/W.
Multiply it with the power dissipation of 47W the you get a temperature rise of 1450°C.
OK this is not realistic, because with 1450°C the thermal radiation reduces the total thermal resistance significantly.
So let's consider a temperature rise of 800°C.

If you don't want it to get that hot ... you should consider to use a heatsink. A big heatsink.

Klaus