Circuit validation & improvement

[programbot]

Hello,
i have a scenario at hand, i need to drive an RC plane with a specific battery pack that is also size limited. the specified battery has a high internal impedance thus 0.01 ohm whereas i have to draw aprox. 50 amps (that decreases down to 33 amps when the plane attains its intended speed). i loose lots of terminal voltage, which i thought of solving by putting up more series cells (as battery heating is not an issue). how ever the electronic speed controller wouldn't let me go higher then 18 series NIMH cells. which i am trying to solve using the following circuit. the circuit works, but i would like opinions and suggestions about it. if someone can please. thank you.

 

[Audioguru]

RC planes use lightweight Li-PO batteries that can provide 50A directly without dropping the voltage.
Your circuit has an output voltage of 24V but the 0.7 ohm resistor in series with the motor limits the current a lot.
The transistors and resistor waste a lot of battery power making heat.

RC planes use brushless motors with 3 wires, not what you show. Maybe you show the current limited 24V powering the ESC?

 

[programbot]

first kindly ignore that 0.7 ohm resister, second that motor over there is supposed to be the ESC couldnt find a model for that so i placed a motor there. the motor and resistor both are taken out of the circuit and an ESC was connected in its place. (i placed that resistor because i first had to test the circuit on a much smaller dc motor so if something goes wrong the damage might be minimal)

 

[Audioguru]

My RC airplanes use the entire Li-Po battery voltage for the ESC, not a much lower regulated voltage that wastes battery power making heat.

 

[programbot]

well yes regarding LiPos, unfortunately i am stuck with NiMH or NiCD. the rc airplane i made this for, is scheduled to fly in a competition that disallows the use of any other battery pack except NiXX

 

[programbot]

true that it does produce a lot of heat, but given the restrictions in place, if its flown with the lowest IR battery i can find it wont raise an inch from the ground. however placing this in between the battery and ESC it flies even does a few maneuvers.. about the lowered voltage, the ESC and motors i usually find are at max rated 6S i.e. 22.2 v thats why i had to down limit the voltage to 24v. (there are some higher rated ESCs and motors but i am unable to find one suitable for my task)

 

[Alec_t]

The LM358 output voltage can only get to within ~1.3V of the +ve rail. With that, and the voltage drops in Q1, Q2 and Q3 you are dropping around 3V in total. Have you considered replacing those transistors with a single p-channel MOSFET?

 

[Audioguru]

I understand why you want to use more battery cells in series than you need. They will make heat and steady power. A 34V battery can provide regulated 24V at full current until the battery is dead but a 24V battery provides full current only for a few minutes then the current drops as the battery runs down. My RC airplanes with a Li-PO battery can go vertically like a rocket until half the battery charge is gone. But the airplanes can fly normally with reduced maximum power until the battery is almost dead.

Your voltage regulator gets hot because it is linear. You should use a switching voltage regulator (using PWM) that does not get hot then the power will be the same but battery will last longer.

 

[programbot]

The LM358 output voltage can only get to within ~1.3V of the +ve rail. With that, and the voltage drops in Q1, Q2 and Q3 you are dropping around 3V in total. Have you considered replacing those transistors with a single p-channel MOSFET?

Could you please recommend a mosfet? Max current support in this configuration is 80 amps. In my opinion the only voltage loss that matters will occurs on Q1, as when its turned on collector voltage will be equal to battery voltage, from that subtract the loss on the Q1 that'd be the emitter voltage. Correct me if I am wrong. In any case I think using a mosfet would be better. Btw thanks for pointing it out I hadn't considered the loss of voltage.

 

[programbot]

The LM358 output voltage can only get to within ~1.3V of the +ve rail. With that, and the voltage drops in Q1, Q2 and Q3 you are dropping around 3V in total. Have you considered replacing those transistors with a single p-channel MOSFET?

Could you please explain this statement, 'The LM358 output voltage can only get to within ~1.3V of the +ve rail' what I understand is that the lm358 can output a maximum of 32.7volts aprox with a 34v supply. Is that what you mean?

 

[programbot]

I understand why you want to use more battery cells in series than you need. They will make heat and steady power. A 34V battery can provide regulated 24V at full current until the battery is dead but a 24V battery provides full current only for a few minutes then the current drops as the battery runs down. My RC airplanes with a Li-PO battery can go vertically like a rocket until half the battery charge is gone. But the airplanes can fly normally with reduced maximum power until the battery is almost dead.

Your voltage regulator gets hot because it is linear. You should use a switching voltage regulator (using PWM) that does not get hot then the power will be the same but battery will last longer.

Unfortunately I can't use a lipo battery, I am restricted from doing so by rules of the competition. The lm7812 isn't getting hot, I am talking about the mj11028 Darlington pair, it gets hot like hell.... tried touching it a few times :-(

 

[Alec_t]

Max current support in this configuration is 80 amps.

Isn't it the ESC which handles the high current?

what I understand is that the lm358 can output a maximum of 32.7volts aprox with a 34v supply. Is that what you mean?

Yes.

 

[programbot]

Isn't it the ESC which handles the high current?

Yes.

The circuit is placed between the ESC and the battery, thus allowing the ESC to see only 24v aprox. Hence all the 80 amps current must also pass through the mj11028 when it is in ON state.

 

[Alec_t]

The MJ11028 is rated for 50A, not 80A! No wonder it gets hot .
I can't recommend a particular MOSFET, but other members here probably can.

 

[programbot]

The MJ11028 is rated for 50A, not 80A! No wonder it gets hot .
I can't recommend a particular MOSFET, but other members here probably can.

Well thats a blunder on my behalf gotta find one with higher rating.... anything else you find odd about my little circuit ???

 

[Alec_t]

Just wondering, how big a heatsink do you have for Q1? Dropping ~10V at 80A means you have to dissipate ~800W as heat somewhere .

 

[Audioguru]

I think it is the battery and its wiring that heats with nearly 800W, not the power transistor. He said that at full power the battery loses terminal voltage therefore he added battery cells in series to make up for the loss.

The voltage regulation is used to reduce the maximum battery voltage but pass the very high current when it is needed. The motor in the airplane dissipates 24V x 80A= 1920W and might be red hot at full power.

 

[programbot]

I think it is the battery and its wiring that heats with nearly 800W, not the power transistor. He said that at full power the battery loses terminal voltage therefore he added battery cells in series to make up for the loss.

The voltage regulation is used to reduce the maximum battery voltage but pass the very high current when it is needed. The motor in the airplane dissipates 24V x 80A= 1920W and might be red hot at full power.

Audioguru has pointed it out, the Darlington just drops 1.4v @ 80amps for static run of the plane which how ever drops down to 30 amps as linearly over a period of 5 seconds in which the plane attains the max speed. Which roughly means 112 watts at the beginning gradually droping down to 42 watts loss after the 5th second. However as audioguru pointedout the battery looses over 800 watts as heat, which I suppose is unavoidable given the scenario. However the battery has been placed in such a way that the spaces between each cell and an opening for high speed wind cools it down. The highest recorded temperature for the battery is 79 degree Celsius full battery discharge on static run. I can't deal with that heat however is there a way to reduce the heat produced in the Darlington???

 

[programbot]

The motor in the airplane dissipates 24V x 80A= 1920W and might be red hot at full power.

The motor is rated 2000 watt. It doesn't get much hot (touchable)

 

[Audioguru]

The datasheet fore the MJ11028 darlington shows that its maximum saturation voltage loss is 3.5V at 50A and at 80A it is much higher. At 80A it might be 5.4V max which is a dissipation of 80A x 5.4V= 432W which will destroy it even with a perfect heatsink. Use two or more darlingtons in parallel to share the heat.