Old dog needs a few new tricks.

[trooks]

A friend shook up my retirement by getting me interested in helping him develop an idea he has for a product. He is the one with the college degree and it is Civil Engineering. We hard wired a basic system just to get 'proof of concept' to satisfy ourselves. A hard wired system has minimal utility for a small amount of uses. My job is to develop a distributed network of units to work in unison.

What I have come up with mates Parallax BasicStamp2 units to XBee RFmodems to create individually programmed units in a network. The system has to be RTOS command and control so I wrote it in Parallax PBasic. I have the communications and timing programming done and checked out except for some small details.

Those small details are LEDs driven my 3.3 V logic. Those LEDs need to be changed into 12 V automotive door lock/unlock actuators. In my working life I had the title of Field Service Engineer which meant I could fix anything someone had designed and built but my capacity for design has so far released a lot of the 'magic smoke' from a lot of components. You have to realize that for the last half of my career I can not even remember the last time I had to pick up a soldering iron.

This old dog needs to learn one more trick to get this job done and start making payments on a house on easy street. What is the best and most reliable way to quickly get 3.3 V 10 to 20 ma to operate an actuator capable of producing 15 ft lb push/pull? Please point me to articles, books and components that will work best at this task.

I promise to wear a NutsVolts Cap and/or Tee Shirt once we start recording demos. My partner and I might even become the next 'Quack Dynasty. I will be the old goat in the background fiddling with the equipment that they let be on camera from a distance.

trooks
in the woods of Southern Appalachia

[Lenp]

Hi, and welcome!
I would suggest using a MOSFET with is easily interfaced to logic level signals. Do consider a damping diode to suppress the reverse voltage spike when the solenoid is deenergized.
There are many examples online of workable circuits. I have switched to the Picaxe chip and their site contains a free download of their manual which has many interfacing ideas that would apply to all micros.
Look at http://www.picaxe.com then 'manuals' then "interfacing'.
Need more help? Kick it back!

[trooks]

Len,

Thanks for the quick howdy and first lead on a direction to look.

I have come across 'picaxe' mentioned quite few times in my poking around the web. I went with Parallax since early on one of my 'relearning' electronics projects had been a couple of Sumo-Bots and was delighted to have the 'Stamp' packaged in a way that cut down a lot of potential problems interfacing to my design.

I am not as quick a study as I once was but I will check out picaxe.com and get baack to you.

Thanks again,

trooks

You get experience when you didn't get what you wanted and it is better than getting nothing<G>.

[haklesup]

what you need to bone up on is the interface electronics between the digital controller and the outside world. To drive a motor powerful enough for your job you need to level shift the 3.3V voltage logic signal to be a high current and whatever voltage the motor needs (usually a lot more than 3.3V)

An NPN transistor may do the trick but a MOSFET based circuit would be more efficient (less hot) and may even have fewer components. However to drive 12V door lock motors, maybe all you need is a relay network where the relay coils can be closed by 3.3V and the 12V stuff is all on the contacts side)

Depending on the relay you try to use, you may be faced with the same problem of level shifting so shop around for a sensitive relay coil that will close at 3.3V.

[dyarker]

... but a MOSFET based circuit would be more efficient (less hot) ...

often, but not automatically.

for example -
A bipolar junction transistor with a collector-emitter saturation voltage of 1V will drop 10W with a 10A load. A field effect transistor with a drain-source on resistance of 0.1 Ohms will also drop 10W with a 10A load.

Many MOSFETs have lower Rds-on. You need to check the spec sheet and do a little math before picking a particlular BJT or FET.

If going with relays, choosing a 12V coil with heavy enough contacts would be better. A set of contacts with current and voltage you will take about the same power to move and hold regardless of coil voltage. Which means the 3.3V coil needs a lot more current than a 12V coil. Plus you'll need to drop 14V - 3.3V = 10.7V somewhere in the nominal 12V environment. Easier to use small transistors (BJT or FET) to drive 12V coils. This makes directly driving 12V (nominal) load with power transistor look better. All design trade-offs that need thinking about.

Cheers,

[Lenp]

Consider that a relay capable of long life switching that inductive load will likely require more current than a logic gate can supply
so some form of driver interfacing will be required. Also look into using an optocoupler to interface a MOSFET. Inherent level shifting protection, and reliability are some of the plusses.

[haklesup]

I guess so, I just looked up coil current on signal relays and ~50mA to 60mA is typical while even 74ACT logic has outputs for 25mA loads. A solid state relay might do it. 41 came up on digikey with operating voltage in the 3V to 15V range.

However the transistor level shifting is not hard once you get it

[trooks]

While checking my bank statement I realized they are no longer taking out money for my student loans. I checked and realized that they were due to be paid off. I got no notice that they were paid off but the money stopped leaving my bank account.

With all the wonderful feedback I have gotten and the new-found wealth<G> I will do the actuators 3 ways and see which I like best. I will be using an opto-isolator, Darlington transistor drive and a two level relay setup.

I will set up each actuator up to push a button that feeds back to a counter in the program. I can then run three 'debug' sessions simultainous that will display each counter. I will check on then at intervals for any 'funny smells' or hot to the touch developments.

This could take a few weeks to get all the parts and check with my partner for him to set up what he thinks will be 'valid' test of the actuators.

Gosh I love when a plan starts coming together.

Thanks much Y'all,

trooks

"Do not look directly into the lazer with your remaining good eye."

[gerty]

Are you planning on just one type/model year of vehicle? they vary from year to year and many different types. another thing to consider, if the vehicle is remote equipped a small signal to the vehicles ECU will lock the doors for you.
Mid TN here...

[trooks]

gerty,

I appreciate the input but this project only uses automotive actuators and has nothing to do with autos.

The solenoids have been scavenged from wrecked/junk autos. My partner and I have way more time to hunt and scavenge that we have money to buy these costly items.

Each is mated to a microprocessor and an RF modem to be part of a coordinated system to perform a specific task. I am having to 'guesstimate' the current that each will require to be able to provide 15 ft lbs of push/pull. My main concern is that the relay contacts be 'hefty' enough to do repeated operations without getting pitted and corroded to the point of not providing reliable contact.

We are likely doing overkill initially because of being terrified of getting a demo set up for potential investors and having the thing not be perfect. I am also looking at driving the actuators directly via darlington transistors. Since there is no way to know the orientation of the solenoid at any given time all power must be planed for it to be firing vertically and springs have to me strong enough for the return even if it is firing down.

All we need is one or two good demos and then we will have money to hire a real engineer to fine tune the design and set it up for ease of manufacturing. Then my partner and I both get nice motor homes and travel around the country doing sales demos and maybe a TV show now and again<G>.

trooks