Remote Control Car:H-Bridge Vs Simple Switch DPD

[kthouz]

Hello everybody!
I am a beginner in electronics and mechanical but i really want to know more in that field.
Now, i want to make a remote control car with a high speed possible.
I have already made one with a 3Volts DC MOTOR controlled by a PWM (LM555) and potential-meters, the power supply of the IC's is 9V, the directions of the motor is controlled by a switch SW-DPDT-MOM-8 connected to the motor. It works, it can run but it is slow. Now, i want to upgrade my car or even make a new one that can work with a much higher speed. I have read topics about "DC Motor control" and i got some on a motor controlled by a H-Brige. The problem here is that these H-Bridge (as i saw) are mostly controlled by PIC (Microcontrollers) and i didnt learn any thing about Microcontrollers. Now my questions are:
1. Is it possible to use an H-Bridge without using Microcontrollers. (My car is not yet to a wireless one). Here i want to know, if it is possible that i connect my PWModulators to the H-bridge. hence the H-bridge circuit should replace the switch that i used.
2. Which is most efficient between H-Bridge and Switch in controlling my car and boosting its speed.
To increase the speed of the last car (the one with PWM and switches), i saw that it is done by changing the capacitors on the oscillator.
3. Wat should be then the max best frequency (and the capacitance) if i want to have a car with about 60-75MPH?
Thanks in advance for yout help.

 

[fleem]

1. The answer is "yes" in theory but "no" in practice. Here are the caveats:

a) If the motor were 100% efficient (i.e it had zero internal resistance) and the H-bridge had zero on-resistance and you had a big cap across the batteries (so inefficient battery charging/discharging weren't a problem), then it would be fine to drive the motor with an H-bridge that always had two of its gates turned on (and thus either positive or negative voltage was always applied to the motor, but with varying duty cylce). But all the various inefficiencies are there when the PWM duty cycle is 50% and the car is stopped! That would run down those batteries quickly and for no good reason.

But for the sake of discussion, if you still decided to go with PWM, there are other caveats:

b) The PWM signal applied to one set of diagonal gates of the H-bridge must, of course, be inverted.

c) You'd need to make sure there is never even a brief time that all the H-bridge gates are turned on, because then you'll short your power. If the H-bridge gates are fast and the PWM signals switches fast, then you might be able to keep that on-time negligibly short. Otherwise you'll need to make sure there's a short delay when turning one set of diagonal gates off before turning the other set on. That would take some delay circuitry.

d) You may need to apply some voltage bias to the signals driving the H-bridge gates (check the fully-on gate voltage and fully-off gate voltage of those gates--which are probably power FETs, I presume).

2. I'm not positive what you are asking here, but a good H-bridge made with good power FETs can be pretty efficient. What makes them inefficient is the on-resistance of the power FETs compared to the typical run-resistance of the motor. For example, if the run resistance of the motor (which varies with speed and load, BTW) is substantially higher than the on-resistance of the two switched-on series FETs in the H-bridge, then the efficiency will be fairly high. The efficiency is the ratio of the motor resistance to the FET (both of them in series) resistance.

3. You put a big cap across the motor only when the driving circuit goes high-impedance (all gates turned off) at various points in time. The cap prevents coil fly-back when the voltage is removed (a coil with current going through it will produce a huge spike of voltage when the switches are turned off, and the cap keeps that voltage down and slow, so that power isn't wasted). Putting a big cap on a PWM signal that never goes high-impedance is a no-no because it would be like a short-circuit each time the PWM signal switched. You WOULD want to put a LITTLE cap there to reduce radio interference, but we're talking about .01uF. In any case the freq should be just a little higher than the freq that makes a noticeable jitter or buzz in the operation of the car. Say just barely ultrasonic (about 15KHz).

Now what you CAN do is drive the H-bridge gate pairs separately such that when the motor is stopped all four gates remain off, and the duty cycle of the pertinent pair of gates increases as the motor speed needs to increase (and the other pair remains off). but that's not a simple PWM signal--its two PWM signals in phase. But in this case you need to make sure the big cap across the motor does a good job smoothing the flyback, but the gates don't switch polarity so fast that they are fighting the cap voltage.

 

[Topher925]

1. Sure you can. All you need to do is generate a PWM with changeable pulse widths. You can do this with a 555 timer and some external components, but a uC is going to make your life a lot easier.

2. The switches will be the most efficient at controlling the motor at WOT. However, anything less than WOT the H-bridge should be more efficient.

3. I don't understand your last question. The speed of the car is not dependent upon the frequency or capacitance of anything. Your H-bridge may like a particular frequency or period of PWM but it will for a range of periods.

 

[kthouz]

3. I don't understand your last question. The speed of the car is not dependent upon the frequency or capacitance of anything. Your H-bridge may like a particular frequency or period of PWM but it will for a range of periods.

Actually, as i read (if i did it understand) is that for a signal input in pin5 of 555 (LM or NE555), the output from pin3 depends on the frequency of the oscillator (on the 555 trigger). My problem is then how? I tried to make a simulation with my Proteus and i remarked that for some voltage in pin5, a small frequency on the trigger will produce a great output voltage on pin3. Is there any mathematical or physical rule that i can use for accurate calcuations?

 

[vk6kro]

The best speed for your motor is the one it was designed for. If you put more than 3 volts on a 3 volt motor it may not burn out immediately, but it won't last as long as it should.

So, if your motor is meant for 3 volts, try it on 3 volts and see if it is fast enough for you.
If it isn't, H-Bridges are not going to help you. You need a different car.

H-Bridges are really a convenience to avoid you having to switch the motor polarity to go backwards. But how often do you do that? They have a "stopped" position in the middle of the control and go in opposite directions as you rotate the control away from centre.

If you put a power FET in series with the motor and drive the FET with pulse width modulation, the FET will turn on and off rapidly and control the power to the motor. Full power is when the motor gets 3 volts all the time.

The output voltage of a 555 is always the same. It is a square wave, but the ON-OFF time of the square wave can vary. If you get a CMOS 555, the output voltage will be nearly the same as the power supply voltage or it will be zero depending on which part of the square wave you are looking at.
There are 555 designs that should give you full speed control of the car from stopped to full speed.

.

 

[kthouz]

First of all, thanks for your help especially to this last post. it is really helpful. But i still have some few to ask again:

So, if your motor is meant for 3 volts, try it on 3 volts and see if it is fast enough for you.
If it isn't, H-Bridges are not going to help you. You need a different car... .

Actually, i think i am going to use switches instead of H-bridge because i don't find the control of that one not accurate. I need my car to stop immediately when i command it to stop (with a potentiometer for example).

...If you put a power FET in series with the motor and drive the FET with pulse width modulation, the FET will turn on and off rapidly and control the power to the motor. Full power is when the motor gets 3 volts all the time...

I think i will use power supply of motors different from the power supply of the other control's circuit (IC, buffers, ...) so that they can get their fully 3V. Am i right? What about dealing with what size, type, friction factors,... of the wheels that i am going to use?

..It is a square wave, but the ON-OFF time of the square wave can vary.

So what are the factors that make that period (frequency) to change?

If you get a CMOS 555, the output voltage will be nearly the same as the power supply voltage or it will be zero depending on which part of the square wave you are looking at.
There are 555 designs that should give you full speed control of the car from stopped to full speed.

Can you please indicate me the what are those 555 types that i can use (their part number, i mean)? and i will try to find them at my local electronics shops?

Now tell me about those resistances and capacitors that are mounted with the 555. Their values have an effect on the output frequency, don't they?

When I am done with my circuit, i will try to post it here and let you see.
Thanks again!

 

[vk6kro]

Have a look at the following page:
http://www.dprg.org/tutorials/2005-11a/index.html

I think that answers most of what you asked.
I did a Google search for "555 pwm" and got a lot of circuits.
They probably all use variations on the same circuit.

The one above gives quite a good explanation of how it operates and shows how to drive a MOSFET with it. If it isn't clear, please ask.

If you find a circuit you want to try, you could just get all the bits it specifies in the description.
The CMOS 555 goes by several names. See
http://en.wikipedia.org/wiki/555_timer_IC
for a list of them. But if you did need one, you could just ask at the store and they would know what you wanted.
CMOS 555's make designing circuits for the 555 easier, but if you are building someone else's circuit, you need to follow what they say to use.

 

[kthouz]

The best speed for your motor is the one it was designed for. If you put more than 3 volts on a 3 volt motor it may not burn out immediately, but it won't last as long as it should.

I read that page and i found exactly that only duty cycle can change in function of the input signal in pin5.
Now suppose i make two similar cars and manage such that their circuits' motors receive all their fully 3V. My question is then, what can i do such that one has higher speed than another. That is actually the hint that will help me in boosting my car.
Thanks

 

[Stingray]

Now suppose i make two similar cars and manage such that their circuits' motors receive all their fully 3V. My question is then, what can i do such that one has higher speed than another. That is actually the hint that will help me in boosting my car.
Thanks

Change gear ratios, reduce weight, and minimize friction with rotating parts. But you're not going to go very fast on a 3 V motor regardless.

If you do decide to buy a more powerful motor (one meant for hobby-grade RC cars), the current demands are very high. It is not simple to build an appropriate control circuit from scratch. This is especially true if you want it to fit in a small space. It's not just the electronics, either. With more powerful motors, you'll need AWD to keep traction. A sophisticated suspension and proper differentials also become necessary to maintain control at higher speeds.

 

[vk6kro]

How are you supplying power to these cars?
I picture a track where the cars follow a fixed path like a figure 8 and they run in grooves in the track. If so, there would be parallel tracks and you could control the cars individually.

You understand duty cycle OK, so you would see that if both cars get their 3 volts all the time, they would go at the same speed.
The only way around this would be to run them on 6 volts with a 50% duty cycle and have a "turbo" mode where you could give them a higher duty cycle for a short period.

There are much bigger model cars that you have probably seen that you take out on a quiet road and zoom around with them. These have large batteries in them and only run for 15 minutes or so before they need a recharge. They are expensive, though.

There are also gasoline powered cars which can cost thousands of dollars.

I would stick with your present cars for a while and learn the techniques of pulse width modulation.

 

[kthouz]

Up to now, Here is the circuit that i have done using Proteus Professional 7.1. this is a car which goes in one direction only (forward). to make it going backward, I'm planning to put a double switch between motors and those capacitor and diode that protect the motors.

When i turn on the switch, an i measure sent/received signals, the 555 (oscillator) is working properly but for the PWModulators i am getting no signal. I don't know why?

Just have a look on it and tell me what do you think about it. Just help to debug it please!
Thanks

 

[vk6kro]

Pin 3 of the 555s seems to be earthed. It must not be earthed.
It should go the the gate of a FET.

If those circular objects are motors, the bottom end of them is earthed so they will run at full speed.
Also, the 555 should not be used directly to drive a motor unless it is a very low current one. Car motors probably draw at least an amp which would be far too much for a 555.

Have another look at the circuit on the following page:
http://www.dprg.org/tutorials/2005-11a/index.html
Notice however that they use pin 7 as an output, which is unusual but OK in this circuit.
In most circuits involving the 555 the output is pin 3.

 

[kthouz]

It should go the the gate of a FET.

So you mean that i should put an FET between pin 3 and the motor?

Also, the 555 should not be used directly to drive a motor unless it is a very low current one. Car motors probably draw at least an amp which would be far too much for a 555.

the i put an amplifier after pin3?

And can you tell me a bit more about the "turbo" that you mentioned in the previous reply?

 

[vk6kro]

So you mean that i should put an FET between pin 3 and the motor?

Yes, a FET can handle larger currents than a 555 can so you use the 555 to control a FET.
Like this:

http://www.dprg.org/tutorials/2005-11a/2005-10a-004.gif Then I put an amplifier after pin3?
Yes, the amplifier is the FET. In this case the output is pin 7 though.And can you tell me a bit more about the "turbo" that you mentioned in the previous reply?
You would start with 6 volts across the motor and FET in series. The motor is only rated for 3 volts.
In the above diagram, you could put a 100K resistor in series with the 100K pot on the left so that you only got 50% duty cycle maximum normally. You could then put a switch across the resistor to get 6 volts across the motor all the time. This would effectively double the voltage on the motor (and exceed its rated voltage) but you might like to have that ability for a short burst. Actually, both resistor and pot should probably be 50K to keep the circuit behaviour the same.

 

[kthouz]

I want to replace that FET by a BJT what type of BJT should i use. Because i guess most of FETs heat up!

 

[vk6kro]

I want to replace that FET by a BJT what type of BJT should i use. Because i guess most of FETs heat up!

You could use a bipolar transistor. Probably look for a Darlington pair type as you would need a lot of gain. There are lots of them, but it needs to be a power type that can carry the current you need. Get one that can carry 5 amps or so, just to be safe. It will probably need a heatsink. Darlington pairs are two transistors in the one package to give extra amplification.

Power FETs heat up less than BJTs do. What makes them hot is the voltage across them when they turn on. This is quite a bit lower in a power FET than in the same size BJT, so they don't get as hot. But BJTs are fine. Both may need heatsinks, depending on the current your car draws.

If you can't find any Darlingtons, you could try taking the output (from pin 3) of the above circuit via a 1 K resistor to the BJT and see if you get enough current through the motor.
There might just be enough with a 6 volt supply and a transistor with a gain of 100. Otherwise, you will need two bipolar transistors connected like the Darlington above.

 

[kthouz]

I was somehow absent these last two days because i was preparing my tests. but my i still thinking of my car, i didn't give up. Talking about:

You would start with 6 volts across the motor and FET in series. The motor is only rated for 3 volts.
In the above diagram, you could put a 100K resistor in series with the 100K pot on the left so that you only got 50% duty cycle maximum normally. .

I am not getting this well, if we put another resistor in series with 100k pot, we increase the equivalent resistance of the circuit and decrease the intensity and voltage of the trigger since much power will dissipate in those resistors. This will have an effect to limit the max duty cycle to 50%. Is this right?
Then why do we have to decrease the duty cycle instead of increasing it? I ask this because i thought that as more is the duty cycle the more the output signal seem to be continuous and this should be the more efficient. (Or i didn't understand better about the duty cycle).

However, i simulate the circuit as you see on the graph and it works better than the other i did before. What remains is to make it and check about the heat and noise.
That circuit shows only one motor to control one wheel. I am now going to make another for the second wheel such that the control of my car will be as you see on the graph.
And i used 1N4007 because my software doesn't have 1N5818.

 

[vk6kro]

I was somehow absent these last two days because i was preparing my tests. but my i still thinking of my car, i didn't give up. Talking about:


I am not getting this well, if we put another resistor in series with 100k pot, we increase the equivalent resistance of the circuit and decrease the intensity and voltage of the trigger since much power will dissipate in those resistors. This will have an effect to limit the max duty cycle to 50%. Is this right?
Then why do we have to decrease the duty cycle instead of increasing it? I ask this because i thought that as more is the duty cycle the more the output signal seem to be continuous and this should be the more efficient. (Or i didn't understand better about the duty cycle).

However, i simulate the circuit as you see on the graph and it works better than the other i did before. What remains is to make it and check about the heat and noise.
That circuit shows only one motor to control one wheel. I am now going to make another for the second wheel such that the control of my car will be as you see on the graph.
And i used 1N4007 because my software doesn't have 1N5818.

Because we put 6 volts on a 3 volt motor, we avoid damage by running it at 50% duty cycle. So, instead of a 100 K pot we put a 50 K pot and a 50 K resistor to limit the pot's control to only 0 to 50%. There is no danger of overheating such large resistors in a low voltage circuit. (See my comment about using 50 K resistors in that last posting. 47K is OK.)
Now so you have a "turbo" effect you put a switch across the 50 K resistor to give you control of the power up to 100 %. Just don't do it for too long or your motor might get too hot. Your diagram shows this quite well.

 

[kthouz]

Good then now i can understand. I am going to improve that circuit (introducing diodes and capacitors for protection and mount the circuit). I will post my observations of course.

 

[kthouz]

Hello again!
Here is now what i have done as remote control car. If anyone has a suggestion on what i have to change please, you're free.
And thanks a lot to all you who have been helping me. you've been very helpful for me.

But i do still have q question:
What can i do to control those relays (SW2 and SW3) at the same time? Is there any electronic method (device) that i can use? As far as mechanical it is possible, i have joined them all together on the joystick.

 

[vk6kro]

I think we decided it was a 3 volt motor so why are you running it off 12 volts?

The relay coils can just be connected in parallel if you like, but won't that mean both cars behave the same way? That is, they both go forwards or both go backwards.

 

[kthouz]

actually i put 12volts and protect motors by capacitors and diode, i think it works. Doesn't it?
And about that relay i think it can work because both motors have to turn is the same direction always. Only their different speed control the steering mechanism so i think there is no problem to use it.

 

[vk6kro]

actually i put 12volts and protect motors by capacitors and diode, i think it works. Doesn't it?
And about that relay i think it can work because both motors have to turn is the same direction always. Only their different speed control the steering mechanism so i think there is no problem to use it.




The diode only protects the FET. You can still cook your motor with 12 volts on it.
Any more than 25 % duty cycle would be too much.

I though you had two cars. OK, I guess you can just join the two relays together or use one relay with 2 sets of switches (contacts) on it.