6A voltage regulator problem

[toxicmouse]

I am trying to run 2 DC motors at 6A, 8V combined. I have a 12V car battery to supply the juice, but I seem to be having problems with voltage regulation.

Being in the middle of nowhere, there are no voltage regulators rated at 8V 6A. By connecting 8V 1A voltage regulators in parallel will I get the same result? Is there another way of reducing the voltage from 12V to 8V? maybe using diodes to get a voltage drop?

thanks a lot.

 

[westaust55]

Years ago you could buy the 7805 in a 10 Amp T03 case version. The 78H05 was a 5V 5amp regulator. I'll see what I can track down on the 10Amp version.

For now another method is to use a parallel transistor to boost the current rating. See the attached but use a 2N2955 (15 Amp rated) as the transistor in T03 case format with a VERY LARGE heatsink.

A thought: do you know how to determine the size of heatsink required? This has been discussed on this forum. Try a search if need be.

 

[InvaderZim]

You might want a DC-DC converter instead of a regulator. The regulator will just burn (x volts times y amps) watts off as heat. A converter (aka SMPS) will do the conversion at around 80% efficiency. It's a more expensive design, but will eliminate some of that wasted heat and get more out of the battery, possibly saving you the cost of big heat sinks and battery issues. A lot of off the shelf converters have a variable output, so can be made to do 8V. Or if they have a 12V output (pretty standard), you can hack the feedback resistors to lower the voltage to 8V.

 

[BeanieBots]

Have to agree with switch-mode approach. 6A*4V=24W is a heck of a lot of heat to get rid of.
Buck regulators are the simplest of all sms designs to make and there are plenty of sms chips available to help out.

For your original question. No you CANNOT parallel voltage regulators to get more current. They all have slightly different voltages. The one with the highest voltage will try to source ALL the current even it's only a millivolt higher than the others. You could put a "current share" resistor on the output of each one but then the regulation would not be as good.

Have a look at the LM317 application notes. It shows how to make a current boost for a linear type regulator.

 

[boriz]

Motors and other ‘power’ devices are normally chosen at the same time / in complement with the battery / power supply. You are finding out why!

Seriously consider re-assessing your design to allow direct connection of motor and power. Regulators are handy for the logic / low power circuit components, but of limited use for the more power hungry components. And inefficient too. Something you should definitely be concerned about when working with batteries.

 

[Andrew Cowan]

Put the motors in series?

 

[gbrusseau]

The LM396K and LT1038 are 10amp regulators

 

[westaust55]

Put the motors in series?

I think not feasible as each motor is 8V so in series needs 16V. there would be a serious drop in output torque if running at 75% voltage.
Also could be one on each side of a robot for steering.

Certainly, as others have suggested, use of a switchmode power supply will be more efficient than a linear regulator based on a 78xx type device. It may come down to efficiency versus availability and cost in the end.

 

[goom]

Why not use pulse width modulation (PWM)? Essentially this would be a speed controller, and by limiting the duty cycle to about 60% would give a maximum average current of 8A. You could brew your own "speed controller", but would need some careful design and component selection. There are plenty of MOSFETS capable of handling the current, but a suitable gate drive would be essential.
If you want steering, then, presumably, you need speed control anyway.
You could even use a commercially available hobby speed controller fed with pulses using the SERVO command.

 

[boriz]

Bloody good point Goom. Obvious really. PWMing the motors means you can connect them directly to the 12v supply. 8v is 66% of 12v.

 

[westaust55]

PWM on electric motors

While recognizing that many are using PWM from their PICAXE chips to control LED’s, Servos and motors, there is an issue with PWM and electrical insulation in the real world. LED’s are one thing but motors with insulation materials are a different issue.
PWM creates high dv/dt (voltage rise) leading to voltage spikes. This is well known in the electrical industry where most currently available AC variable speed drive controllers use PWM output. This gives a near sinusoidal current through the effects of the motor inductance but a series of square wave voltage pulses to the motor insulation. In many cases, the start of each group of pulses has a single voltage spike roughly 1.8 time the normal voltage level.
To overcome these voltage spikes, motor manufacturers now produce motors with insulation twice the nominal voltage rating to reduce insulation failure in motors.

Thus to operate an 8V rated motor on 12V DC with PWM control may be leading to early motor insulation failure. Might be one thing if you are playing around with cheap toy motors but if using more expensive motors and looking for a longer life from the motor, these spikes may be something to consider.

Not having access to an oscilloscope at home, that might be an opportunity for other members to look at the waveform when using a PICAXE with PWM into a motor and see how the voltage output looks with respect to over voltage spikes. Some posted waveforms would be good for all.

 

[goom]

In my (very limited) experience, AC motor speed controllers are often Variable Frequency Drives (VFD's). They provide a sinusoidal voltage, but vary the frequency for speed control.
You may well have a point with regard to spikes and insulation tolerance. However, there are comminly available small "hobby" motors that draw 40A or more, with PWM control. Maybe they are just built to take such "abuse". Burnt fingers are a common problem if one touches a high performance motor after a run in an RC car or boat.
I would suppose that there are plenty of noisy spikes with a brushed DC motor under normal operation. Small capacitors across the motor terminals are often fitted to reduce high frequency spikes, and are definitely recommended where electronics are in the vicinity.
Perhaps a 12V zener diode across the power leads to the motor would be a good idea?

 

[westaust55]

Variable Speed Drives (VSD and Variable Frequency Drives (VFD) are one and the same. Believe while here in Oz we use the term "VSD", in the US and Canada it is "VFD". In reality both are VVVF (Triple V – F's) as they are Variable Voltage, Variable Frequency converters.
As mentioned, the output current is "nearly sinusiodal" which reduces harmonics into the motors and thus reduced extra heating, but the voltage is PWM generated from a series of voltage pulses.

Yes even with model railways they put a small capacitor across the brushes in the locos to reduce RFI noise getting into other household electrical/electronic appliances.

For the VSD/VFD's you can buy dv/dt filters to remove the large “spikes”. To add small capacitor across the supply terminals of a motor will help reduce the magnitude of the "spikes" above the normal DC voltage level. The capacitor will effectively integrate the square wave and give a curved leading edge. Likely no good for servo’s though where the pulse timing relates to position.

 

[toxicmouse]

Thanks a lot for your replies. I will try out some of your ideas to find something that works.

the 12V supply is a deep cycle 50Ah battery and it is powering an RC helicopter through a thin wire. THe battery pack on the helicopter is supposed to be 7.2V. It is not a conventional use for an RC helicopter but I am learning to fly and it is too time consuming to recharge batteries (even with three battery packs), so a fixed line approach makes sense. The 7.2V battery has the same mass as 10m of cable, so I am a bit limited on range but all I am trying to do is hovering and basic moves.

so burning 24W in a voltage regulator is not too serious, I am still going to get a few hours flying time.

 

[BeanieBots]

Please be carefull if this is a model helicopter. The motors in these helicopters are usually run at more than twice their recommended rating to keep power/weight ratio down. The reason this can (to some extent) be done is because a 'typical' flight time will not exceed 5 minutes with at least a 15minute pause before the next flight.
I bet somewhere in the manual it says allow at least 10 minutes between flights. There are "umbilical" heli's available but even they state that you should not fly continuous for more than 15 minutes. Ever tried more than 10 minutes in the air? Never mind the cost of batteries, it's the cost of vallium and therapy that keeps me out of pocket.

I fully understand your frustration. Especially when learning.
You've just built up the courage to get airborne, just getting into a nice stable hover and then the battery dies. Sorry my friend, that's just part of the joy of electric helicopters. If you are serious about the hobby, then a second heli would be a better option. It also saves the frustration of being grounded when practicing those inverted landings
You WILL crash badly at some point and it will be when the shops are shut and don't have the parts you need until next week.

Even Jamie Robertson and Kurtis Youngblood keep their flights to about 3 minutes.

 

[boriz]

“powering an RC helicopter through a thin wire”
“10m of cable”

6A through 10m of wire thin enough to not cause significant drag might provide all the (resistive) voltage drop you need

Now that you have explained the application, PWM is probably out of the question. It would very likely interfere with the RC circuitry.

This is not a good plan. A trailing wire (even a thin one) will complicate the helicopter control significantly! Much better to get several battery packs, charge them all and swap them on site. You can get (say) an hour of flight with maybe 5 or 6 packs. That should be plenty of stressful practice for one day.

 

[BeanieBots]

Get a good flight sim and pracrice on that. Limit actual air time to ablout 20mins/day or you'll fry your nerves and not progress.
FMS is free but it's flight physics are a bit too easy. You can use PPjoy (also free) to connect the buddy Tx to your PC via the microphone socket.
HeliSimRC is another freebie. Not very good graphics but the physics are very realistic. It also supports variable "simulation time" to help improve your reaction times.
Better still, pop onto the RCgroups forum and get some advice some from some exceptionally skilled people on the subject.