Society of Robots - Robot Forum

Electronics => Electronics => Topic started by: airman00 on August 29, 2009, 11:23:40 AM

Title: Cheaper High Current Drivers
Post by: airman00 on August 29, 2009, 11:23:40 AM
It just occurred to me that for high current situations ( like motors) it might be cheaper to use Solid State Relays.
http://futurlec.com/RelSS.shtml

For example a 40 amp motor controller - four relays - would be 17.90*4 = $71.60 . And since they are solid state and not mechanical you can pulse width them as much as you want.


To be clear: you could make a cheaper motor driver using MOSFETs , but in that case you would need additional components - MOSFET motor drivers, as well as a custom PCB. Using solid state relays allows you to make a motor controller without any custom boards and its much simpler. A MOSFET motor driver requires some know-how.
Title: Re: Cheaper High Current Drivers
Post by: Daanii on August 29, 2009, 03:47:35 PM
So you would use the four relays to make an H-bridge?  Clever idea, but is the switching frequency fast enough? 
Title: Re: Cheaper High Current Drivers
Post by: airman00 on August 29, 2009, 03:50:36 PM
From the datasheet:
Max Turn on time = .5ms   Max Turn off time = .5ms
if you do your pulsing correctly you should be able to have proper speed control.
Title: Re: Cheaper High Current Drivers
Post by: Daanii on August 29, 2009, 07:29:31 PM
I notice these relays switch AC power.  Does pulse width modulation work to shape an AC power line to drive a motor?  I thought pulse width modulation for motors was just for DC power. 
Title: Re: Cheaper High Current Drivers
Post by: SmAsH on August 29, 2009, 07:34:44 PM
im not too sure, but most relays can switch both...
Title: Re: Cheaper High Current Drivers
Post by: Daanii on August 29, 2009, 08:36:21 PM
im not too sure, but most relays can switch both...

That makes sense.  No reason why a relay that could handle AC power could not handle DC power.  Using relays to make an H-bridge is a clever idea.  Not something I had ever thought of. 
Title: Re: Cheaper High Current Drivers
Post by: mack33 on August 29, 2009, 10:00:59 PM
From the datasheet:
Max Turn on time = .5ms   Max Turn off time = .5ms
if you do your pulsing correctly you should be able to have proper speed control.

I think I found the specific data sheet that you were talking about. It appears that the max turn on time and turn off time are 10ms not .5ms.

Is this the correct data sheet? http://futurlec.com/Relays/SSR40A.shtml

If I am correct, I'm pretty sure it's to slow.
Title: Re: Cheaper High Current Drivers
Post by: airman00 on August 29, 2009, 10:03:21 PM
Strange, Datasheet has different info. I attached it.
Title: Re: Cheaper High Current Drivers
Post by: Soeren on August 30, 2009, 06:51:32 AM
Hi,


Why don't you break out your calculator for a quick run down of the available data, when you think you've found the Holy Grail, instead of starting a discussion that's a waste of time, since it's so easy to see that it's not doable?

The AC SSRs cannot switch DC!
The DC SSRs on the bottom of the page (priced $21.90 and $27.90) can switch DC but with the 0.5ms max on/off switching times.
That means that the max. frequency to design with is 1kHz and then there's nothing left for PWM.
To do useable PWM on them, you'd need to settle for a PWM frequency of around 10Hz tops, to make the switching times negligible.

IOW, you cannot use these SSRs for H-bridges (unless you're trying to build a Senior Citizen Clone  ;D).
Title: Re: Cheaper High Current Drivers
Post by: Daanii on August 31, 2009, 10:53:56 AM
Why don't you break out your calculator for a quick run down of the available data, when you think you've found the Holy Grail, instead of starting a discussion that's a waste of time, since it's so easy to see that it's not doable?

Better to put out an idea for others to think about than to stay silent.  Even if the idea doesn't work, it is clever. 

Transistors did not always switch at nanosecond rates.  Relays have a long way to go to work fast enough, as you point out, but who knows what the future will bring.  These solid-state relays are pretty remarkable compared to their mechanical forebears. 
Title: Re: Cheaper High Current Drivers
Post by: Soeren on August 31, 2009, 04:12:15 PM
Hi,

[...] Even if the idea doesn't work, it is clever.
What clever is there over something that doesen't work? (Besides that you just created a new oxymoron).


Transistors did not always switch at nanosecond rates.  Relays have a long way to go to work fast enough, as you point out, but who knows what the future will bring.  These solid-state relays are pretty remarkable compared to their mechanical forebears. 
Sure, and perhaps mankind will travel to another planet someday, but I find it utterly pointless telling that once we hunted with sticks and that we don't know whether we will hunt with target locking LASERs someday. I tend to work with what's here and now, as it's what's here and now!
Title: Re: Cheaper High Current Drivers
Post by: Razor Concepts on August 31, 2009, 04:22:19 PM
You are in a life or death situation. Kidnapped by evil people, you are trapped. The only way to escape is to create a small robot that is laden with explosives, which will travel to the evil people and detonate. The problem is, there are no MOSFETs to be found. You panic. You die.

Or, you remember this thread, and find some DC SSRs lying around. You remember that the PWM frequency issue, so you lower it. The robot works, the evil people die, and MacGyver sends you an autographed picture.


Title: Re: Cheaper High Current Drivers
Post by: Soeren on August 31, 2009, 10:29:02 PM
Hi,

You are in a life or death situation. Kidnapped by evil people, you are trapped.
Trapped, so (worst case) it has to be a potential life or death situation.


The only way to escape is to create a small robot that is laden with explosives, which will travel to the evil people and detonate. The problem is, there are no MOSFETs to be found. You panic. You die.
Nah, I use a paper clip or similar to make a contact B+ to M+ since there's no reason for going slow when you deal with baddies.


Or, you remember this thread, and find some DC SSRs lying around. You remember that the PWM frequency issue, so you lower it. The robot works, the evil people die,
Had the evil people been stupid enough to leave DC-SSRs when sweeping for MOSFETs, they probably can't kill a worm, so no worries ;)
So I don't even have to break open the SSRs to get to the MOSFETs inside and neither do I have to kill the baddies.


and MacGyver sends you an autographed picture.
Only if I don't kill anyone I believe - not that I really want a piccie of McGuyver, the master-wuss of them all, autograph or no ;D

But he's got more pressing things to ponder anyway...
(http://www.tuaw.com/media/2006/03/mcguyver2.gif)
They can't spell either ;D