Electronic H Bridge Losses

[Derek Dickson]

Hi I am a newbie to this site.Looking forward to conversing.
I have a small robot which I am working on.

There is a small electronic H bridge PCB electronic circuit which drives
the movement of the robot back and forth.
The supply voltage to this H bridge PCB is 4.8 volts (3 new xAAA batteries).
The voltage at the motor is 4.3volts.
There is a loss of 0.5 volt across the H bridge PCB.
When I bypass the H bridge the robot is much more powerful and allot more fun.The voltage to power ratio on the gearmotor is certainly not linear.the 0.5 volts make a massive difference.

Is there away of building a near loss-less H bridge PCB so the full voltage is applied to the motor through an electronic H bridge ? Current draw to the planetary gearmotor is about max 0.4 amps ( locked rotor current).

Also is there a way with out an oscilloscope to check if there is any PWM on the H bridge that may be responible for the 0.5 v drop ?

 

[Harald Kapp]

Hi Derek, welcome aboard.

A H-bridge will always incur some loss. By carefully selecting the transistors and the driver circuit you can minimize them, but not to 0. To be able to help, we would need a schematic of the circuit, incl. component values/types.

To check for PWM, set your voltmeter to AC. If there's PWM, it should register a voltage. If there's only DC, it should show 0V.

If the speed of the motor is variable (not just back - stop - forward), it is most probably controlled by PWM.

Harald

 

[Derek Dickson]

Thanks Harald, Awesome information, yes I am registering a PWM via you method of setting the multi meter to AC
So the next question is can I place some thing in line with the L and R control outputs from the controller chip to the electronic H bridge to negate the PWM control ?Sounds like a new thread.
Cheers Derek

 

[Harald Kapp]

Derek, better stay within this thread, as the questions are related.

What do you mean by "can I place some thing in line with the L and R control outputs from the controller chip to the electronic H bridge to negate the PWM control"? When you manipulate these control lines, care has to taken that the "on" signals from both L and R never overlap. otherwise you'd create a conducting path through both "legs" of the H-brifge and quickly destroy the transistors.
If you want to minimize the losses within the bridge, as stated in post #1, you will have to analyze the circuit in detail, find the reason for the voltage drop (presumably it's the transistors) and see if you can find components with less loss.

Relevant information for looking at this issue is:

• What is the current through the motor at full speed?
• Which transistor types are used?
• Can we look at the PWM signal?
• What type of controller is used and are there any chances to modify it?

I know you said you have no oscilloscope but a look at the waveform is invaluable. Apart from the voltage drop you measure another possibility for not reaching top speed can be a less than ideal PWM signal. If if does not reach 100% duty cycle at full speed, tweaking the controller to output such a signal could be an option, provided the controller is tweakable.

Using "better" transistors requires us to know the current and the type of transistors used (bipolar, MOSFET?) to look for an equivalent type with less voltage drop at the currents involved.

 

[Derek Dickson]

Hi I,m working on a little project (well curious) . A small battery operated robot I purchased .I want to negate the PWM on the controller chip Left and Right base or gate supply,. So as to have as much of the full batter power delivered to the motor as possible.The motor currently at full speed still has a fair bit of PWM. The controller chip is not accessible.
Is there a simple way of placing a circuit in line with the Left and right gate supply from the controller circuit so as to negate the pwm put still have the left and right control from the controller ?.I realize this will also negate any possibility of any speed variation and permanently be at full speed.The one set of batteries also powers the control circuit.
Haven't got any other hard values at this stage.The motor is fully capable of full supply voltage.Any circuit ideas ?

 

[Harald Kapp]

The B772 is with high likelihood this PNP transistor.
The D882 is with high likelihood this NPN transistor.

Knowing the current could help find equivalent transistors with less Vce at the current involved. But for a bipolar transistor a Vce of ~0.25V (1/2 of your measured 0.5V voltage drop because two transistors are involved for each direction) at high motor currents is not bad. Finding one with less Vce is not going to be easy.

By "negating the PWM" to obviously want to create a steady state signal so the motor turns either left or right at full speed, right? I can't envision a simple way to do this. A monoflop could be used to strecth the PWM pulses such that the output is steady. But this will not resolve the issue with the voltage drop across the transistors. It will further crreate problems with a change in direction because the stretched signals will then overlap for a short period which in turn will create a short circuit across the h-bridge "frying" your transistors in an instant. Another circuit would be required to remedy this issue...

I have two suggestions:

• Live with the reduced speed.
• Use a higher voltage power supply to increase the speed.

For a higher supply voltage you can either use a 4. batttery or use a step-up (boost) converter to create an e.g. 5V supply from your 3.8V batteries. E.g. something similar to this (rated for 500mA current so the question returns: what is the current drawn by the motor?).

 

[Derek Dickson]

Thanks Harold.The running current for the motor is 0,3amps no load

 

[Derek Dickson]

Yes so I'm going to take you up on your suggestion Harold and go for the step up boost converter it's well with in the lock rotor current of the motor sounds like it could work.
Will let you know how I fair in a month or so or when ever it gets here.
Thanks great stuff.