H-Bridge short protection

[davidd31415]

Hi,

I have 2 H-bridges being used to actuate 2 separate DC motors (forward and reverse on both) using a total of 8 electro-mechanical relays.

There are 2 outputs per H-Bridge/Motor combination.

The DC motors share a common so 3 outputs are connected to the 2 motors and a jumper is connected between the unused output and one of the other outputs.

During system startup all 8 relays close. If the power supply connected to the H-bridges is on and the jumper is in place a short is made and a fuse blows.

I am looking for a way to prevent this from happening, any suggestions? I was thinking about using 2 SSRs to exclusive OR two of the H-bridge relays that may be connected to jumpers but in this case I am worried about the SSRs being destroyed before the fuse blows in the case that a system operator shorts two wires while connecting motors...

Thanks,

Dave

 

[nbucska]

use one DPDT relay for polarity.

Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[davidd31415]

If anyone could explain what he said in a few more words, I'd appreciate it.

 

[itsmoked]

Using electromechanical relays to create the H bridges. argh!

On system power up all the relays are necessarily in the same(unpowered) states resulting in nasty shoot thru. (of course)

If you are going to do it this way David you really should have a relay that turns on the Motive power to the H bridges after control power has come on and the H bridge relays are in their correct states. You also have to design for the main power supply going away while the system is running.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[VEBill]

What nbukska was probably hinting at is that you should wire the system so that the power supplies are connected to the fixed contacts in the relays, and the motors to the moving contacts in the relay. Thus, because the PS contacts are tied down in the relay, the power supplies can't short out across the relay contacts.

Then you'll need another relay (as mentioned by Mr. Smoked) to kill the power because the first relay only selects direction and can't make the system stop.

 

[electricuwe]

The type of switcing device used is not the problem. Basicly you can make such a kind of ciruit working properly with any kind of switch when you do it right.

The basic problem is the two motors sharing one common. If you want to operate the motors in all four possible combinations:

fwd/fwd
fwd/rev
rev/fwd
rev/rev

you need either to wire four cables or two separate power supplies. Otherwise one of the combinations will lead to a short circuit regardless of the swicthing devices used.

 

[nbucska]

It is easy to build an electronic H but it is more diffucult to protect against all possible part failure.
An electro-mechanical H built with a DPDT relay can't short the supply.


Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[IRstuff]

I would suggest that you redesign your bridge drive to not close all 8 relays at power on. What's up with that? Is that a glitch, or intentional?

Why would you want that? Even if you didn't blow up the fuse, powering all the relays would certainly reduce the life, if nothing else.

TTFN

 

[analogkid2digitalman]

I's the default power on state of data acquisition/control hardware. Discussion at:

http://forums.ni.com/ni/board/message?board.id=250&message.id=25457

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"It's the questions that drive us"
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

 

[IRstuff]

not obvious. The responder there asks about NO ouputs, which ought not energize on power up.

TTFN

 

[ScottyUK]

Some of the hybrid driver IC's incorporate a desaturation detection circuit which initiates a driver shutdown if the switching device is driven 'on' but the voltage across the switch is abnormally high as occurs during a DC bus shoot-through condition. These drivers are usually aimed at medium-to-large IGBT switches, but the principle would apply to a FET too. Fuji and Powerex are two suppliers I'm aware of.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[electricuwe]

To me it is useless to continue to discuss about the kind of switch unless the basic error in the design is solved. See my previous post.

 

[ScottyUK]

Electricuwe,

That's a very valid point. If the motors share a common then a full bridge becomes pointless. A half bridge run from dual polarity supplies would allow bidirectional control. A capacitor centre tap for the half bridge would be viable in a single DC rail is available. The OP seems to have lost interest anyway.



----------------------------------

Sometimes I only open my mouth to swap feet...

 

[davidd31415]

Still interested, thanks for the responses- I'm trying to understand all the responses and the problem is on the back burner for a few weeks.

It's good to see my explanation was understood, that did not seem to happen in the thread I posted in NI's forums.

Since the two motors being switched always share a common and only one power supply is available I am considering either an output card that does not initialize all relays to closed or a relay to switch the power supply on/off.

 

[itsmoked]

Would it not be more reasonable to add a piece of wire and stop the sharing?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[davidd31415]

The system tests components that contain 2 motors wired with a shared common. I would have to disassemble the parts and bypass the wire harness to do that (not ideal for testing).

That does seem like the best approach, otherwise!

 

[electricuwe]

So consider the second alternative:

use two isolated power supplies

 

[VEBill]

I missed the 'common wire' the first time through.

Similar to what was just posted by 'electricuwe', use a garden variety split (+/-) power supply with the common connected to the motor common. Then you can connect either or both motors to either polarity and run the motors any which way you want in any combination.

You'll need one SPDT switch for each motor.

 

[davidd31415]

The power supplies run over $2k; made the other options more attractive. The first engineer at NI I spoke with didn't think the card could be initialized with the relays closed but someone more familiar with the card caught my forum post over there- pull-down resistors will do the trick, I'm all set on this.

Thanks all.