Another H-Bridge Topic

[coastergeekrtc]

First off, I know there is a H-Bridge topic out there already...made by useless. I have read every post, but since his motors are much ligther duty that mine, most of the suggestions in that topic don't help me. Thus I made this topic to not impose on his...

My motor is the Traxxas Titan 500...better known as the Mabuchi RS-550VC. 1.2A no load, 85A stall, 6-14.4V. Now in the R/C car there are two 7.2V batteries in a series to get 14.4V...to run two of these motors...I'm only using one battery to run one motor. I set up the system it will run in and measured resistance: 0.7 ohms, current: about 8A spike and then settles between 3-4A, and voltage: 8.1V. Now using Ohms Law...the curent actually spikes to almost 12A. So my H-Bridge needs to handle this 12A spike. And any noise from the motor could be huge at these ratings, so I am going to need to use diodes to dirrect the current away from the MOSFETs. So I loked around and found many schematics, but this one really seems to have everything...only thing is that I'm not sure what the National DS0026 (below) is for.

H-bridge
<A href='http://www.williamson-labs.com/images/h-bridgecontroller-n-02.gif' Target=_Blank>External Web Link</a>
National DS0026
<A href='http://www.williamson-labs.com/images/h-bridgecontroller-n-logic-02.gif' Target=_Blank>External Web Link</a>


Edited by - coastergeekrtc on 21/06/2006 11:54:25

 

[Technical]

Some FETs cannot be driven directly from a microcontroller output pin. The purpose of the DS00026 is to 'amplify' the microcontroller output to a value large enough to drive the FET.

 

[jodicalhon]

Well, I'm out of my league here, but I have a question.

In the second of your links, which shows the PWM pulses to the DS0026, the control pulses for inputs A and B, (and C and D), are in phase, although they rest at opposite levels, either high or low. Don't the in-phase pulses on the same side of the bridge mean disaster?

P.S. I WAS going to have a quiet couple of days, but now I see I'll be goo-goo-googling to learn more about mosfet h-bridges!

Edited by - Jo C on 21/06/2006 13:28:24

 

[coastergeekrtc]

From what I see, both A and D are high...which enables braking. As for Technical, thats was what I thought the DS0026 was for, but wasn't sure. Now will this amplify the signal enough when I use the picaxe 28X?

 

[jodicalhon]

Thanks coastergeekrtc. I have no problem with the 'resting' levels and can see the braking effect. I'm confused by the in-phase PWM. Perhaps I'm misunderstanding the schematic, not having read the description. Anyway, as per my P.S., I'm off to look at other examples! Good luck with your project - I've been following your other thread and am quite interested.

Jo

 

[Technical]

The DS00026 should work fine on a PICAXE output.

 

[ljg]

You may save yourself some time and a lot of heartache if you just get a premade product. I'm using this product:

http://www.pololu.com/products/pololu/0205/

with the same motors you are using and have had no problems.

One caution:
Athough the chip is nominally a 30 Amp unit, you are better off figuring about 6 Amps continuous unless you add some heat sinking and maybe a fan. I added a stick-on copper heat sink intended for PC memory chips and found the chip stays cool. I have a provision to add a small fan, but I haven't found it necessary at the speeds I run my robot (fast walk).

 

[coastergeekrtc]

That is much better...much more professional and just what I needed. I'm just suprised I didn't find it...I have searched a lot for h-bridges. But just to make sure this will work, I am going to need to figure out the current when I brake the motor from close to full speed. Now how do I do this? And I still need to amplify the PWM signal, right?

Edited by - coastergeekrtc on 21/06/2006 23:16:55

 

[ljg]

the Pololu unit just needs TTL level inputs, so you can control it directly with a Picaxe.

As to going from full forward to full brake-- that's bad practice with any H-bridge. About the only thing you could do worse is to go from full forward to full reverse.
You don't drive your car that way, do you?

The fix is easy. Go from full forward, turn the unit off for a few tens of milliseconds until the pulse from the magnetic field collapse dissipates, THEN brake.

Edited by - Larry on 22/06/2006 03:00:27

 

[ljg]

BTW, I didn't mean to suggest that the Pololu h-bridge was the only solution out there, it was just one I used.

Take a look at:

http://www.bdmicro.com/rx50/

or
Go here:

http://www.newmicros.com
and look at their h-bridges.

Both great products.

 

[coastergeekrtc]

^Little bit of a misunderstanding. What I meant is that the motor will be running close to full speed, shut off and left to coast, then braked like you are saying. But right as I enable braking, the motor will not have lost much speed, so the h-bridge will be braking the motor from close to full speed. Didn't know if I need to calculate the current going through the motor at this point to see if it is les sthan the max current the h-bridge can take. And thanks for the other options as well...but the Pololu unit did seem to fit my needs perfectly...especially the price.

Edited by - coastergeekrtc on 22/06/2006 05:01:51

 

[whizzer]

As a matter of interst for anyone still considering building their H-bridge from discrete components, there are a couple of aspects of the design shown in the first web-link referred to by coastergeekrtc which may or may-not be desirable.

1. The circuit is not a complementary-symmetry design, i.e. all the power mosfets are N-channel devices -rather than a mix of P and N channel types (or, in equivalent BJT designs, a mix of PNP &amp; NPN devices).

2. All four inputs to the H-bridge employ capacitive coupling which is then followed by DC-restoration diodes. This is one way of developing a design which allows all the output devices to be of the same type without using any level-translating driver stages. However a design like this does require that the control signals come from a low impedance source, hence the use of the DS0026 driver chip.

Another characteristic of this design is that the control signals must be in the form of pulses in order to be successfully passed through the coupling capacitors. (So just having steady-state DC levels at the inputs will actually prevent any of the output stages from conducting). This leaves only the pulse-width, and phase relationships between the input signals to control the motor -and also explains why this circuit will only work with pulse-type signals such as PWM.

Hope this info is of some use
Regards

 

[ljg]

Other than real quick pulses, what motor guys try to figure is stall current, that is, the current going through the motor when it is prevented from moving. With the two Traxxas motors in series, I think is was around 12-14 amps. As long as the motors don't go from full forward to stop instantly, I don't think you can exceed that for long periods.
I've never been able to stall the motors even when I tried (I heve the lowest available gearing in the transmision)

And as long as you heat sink, you should be able to dissipate the heat generated for small bursts. The Pololu bridge runs a lot cooler than the ESC that came with the car.

What kills the bridge is heat. As long as you don't melt it, you should be fine.Just be realistic with what you ask the thing to do.

 

[coastergeekrtc]

Well I'm sticking this in a box, so I will get a large heatsink and possibly a fan. Do you just attach the heat sink with some glue (epoxy or something strong)?

As far as what I'm asking the motor to do, just a simple run close to full speed, coast for a few tens of milliseconds, brake, reverse at a slow speed, stop. And I am only using one battery pack (8.1 volts go into the motor when running) so the current through the motor won't be as high as it would be if I was using the usual two battery packs. Speaking of battery packs, what power adapter would I look for to replace the pack?

 

[whizzer]

Heat sinks are usually held in place against the device they are cooling by means of clamp / bracket arrangement or something similar. To get good thermal transfer into the heat-sink it&#8217;s a good idea to use a <i>thin </i> layer of special heat-sink paste which is available from most electronics shops. Just be aware that this white thick goo usually contains beryllium and ideally you shouldn&#8217;t get any on your fingers while applying it.

I&#8217;ll pass on the power adapter question -as this would be specific to your country of residence.

Regards

 

[ljg]

The heat sink I used was similar to this:

http://www.scan.co.uk/Products/ProductInfo.asp?WebProductID=51796

they are little copper slugs with fins that fasten with thermally conductive tape - normally on PC memory cards. There are also sinks that fasten with thermal epoxy, but that seems messier.

The ones I used were actually twice the size shown, so mine used four of the slugs per card and were just the right size.

Just peel and stick.

The fan was scavenged off the back of a defunct PC DVD reader that was headed to the recycler. It measured about 1&quot; square (25mm for the inch-challenged types)

As I said, I really haven't found the need to use it, so it stays unconnected.
Mine ran at 5 volts- some are rated for 12 volts. I think either would work.

One last type to look for is a laptop processor heat sink and fan combo- if you can find the right size. I didn't.

I have no advice for a power supply, other than I would probably use an old SLA car battery and a charger to ensure enough current supply.

 

[ljg]

sorry for the double post.

fat fingers.

Edited by - Larry on 22/06/2006 08:19:46

 

[coastergeekrtc]

I really don't have to have a power adapter, but I would like it if I didn't have to keep recharging the battery I use to run my R/C car (the one with the Mabuchi motor I'm using). Batteries stump me. This battery is a group of 6 Sanyo RC-1700 Ni-Cd cells in series. Each cell is said to be 1.2V, making the entire pack 7.2V yet when I measure voltage, it is really 8V. The capacity for a cell is 1750mAh, but the group is really 1700mAh. Now the current through the motor at 8V under no load is almost 11.5A according to Ohms Law (motor has a resistance of 0.7 ohms (will capacitors from the leads of the motor to the casing screw up my measurements?)). But when I measure the current at no load with a multimeter, I get 1.2A, giving a motor resistance of 6.7 ohms according to Ohms Law. So should I look for an 8VDC 12A power adapter or an 8VDC 1.2A power adapter? Or, I have an adapter that is DC 14.5V 1.5A that seemed to run the motor just like the battery (same speed).

Edited by - coastergeekrtc on 22/06/2006 09:31:01

 

[jodicalhon]

whizzer wrote: <i>As a matter of interst for anyone still considering building their H-bridge from discrete components ...snipped... Hope this info is of some use </i>

Very useful! Thanks whizzer.

Another thread where I've learned a lot!

 

[BeanieBots]

Motors do not obey Ohms Law in the way you are thinking.
When the motor is stationary, it will look similar to a resistor (actually an inductor but let's not complicate things just yet) and obey the law. However, as it starts to spin up, it is a coil rotating in a magnetic field which acts like a generator. The voltage produced by this effect is in opposition to the voltage applied by the battery. Thus, the EFFECTIVE voltage is reduced and hence less current flows. The net result is that the battery 'thinks' the resistance increases as the motor speed increases.
If the motor was perfect (no friction or electrical losses) the final speed would be exactly proportion to the applied voltage because it would increase in speed until the battery voltage was the same as the generated voltage at which time there would no current.

As for batteries, the 1.2v is just a nominal value. When flat, they are about 1.1v and when fully charged they are about 1.45v.
Just like the motor, they are not perfect and also have a resistance of their own. Often it can be ignored because it is very low when compared to the resistance of the load. However, when you drain them at a rate that will make them flat in just a few minutes, this resistance becomes more significant and you will see a volts being 'lost' inside the battery. That is also why they get hot. I squared X R = Power = heat generated.

 

[coastergeekrtc]

Well that makes sense...learned about that in physics, can't believe it didn't cross my mind. So then what do I base my needs on for replacing the battery with a power adapter?

 

[coastergeekrtc]

Well just to let people know, I figured out the power supply problem. Well someone else did. I asked my friend to ask his dad about it, he is an electrical engineer. He says that you need to know the voltage and current the motor is running at under normal load. This was 8V and 3.7A, so I need a power supply around that range. Just a little bit of info if any of you wanted to know.

Now that problem is with the power supply. How could I amplify the current...4A does not seem very common.

Edited by - coastergeekrtc on 28/06/2006 08:02:45

 

[flyingnunrt]

The initial current to accelerate the motor up to normal running speed will be somwhat greater than the no load running speed current.
If your power supply cannot supply the demand then it will be slower to accelerate up to this speed.
The nicad battery pack can supply in excess of 30 amps for a short duration so it can easily accelerate the motor.
I reckon you will need a power supply with at least a capacity of 10 maybe 15 amps. Unless of course you run a power supply, or charger, in addition to the batteries.

 

[BeanieBots]

I agree that you will need a supply capable of around 10A.
Supplies of around 4A are often fused for safety, so your starting current would pop the fuse.
A high quality switching power supply that has built in current limit and all the other bells and whistles would be OK but VERY expensive.
I would opt for the suggested battery/charger option but be careful that the charger does overcharge the batteries.
NiCd and NiMh are not really designed for that type of use. Lead Acid batteries would be better because they can have a 'float' charge that is permanently left on. When fully charged, they simply stop drawing current from the charger.