Current limiting /diodes or discreets? 4

[Generalx5]

what is the common way of supplying low currents to an IC or microchip without having to tap into a secondary low supply source. I want to run an IC right from the high amp source, I've looked at Current limiting diodes and wondered if those little diodes could handle high input currents. Whats the key to supplying low currents? for microchips, h-bridges, mos-Drivers, logic inveters. etc. Often those things are rated for only a few miliamps.....any higher would do significant damage. Please let me in on the secrets to obtaining low currents.

 

[Skogsgurra]

Hello!

The secret is not in the current. It is in the voltage.

So, if you have some CMOS circuitry that are supposed to run off a 12 V supply voltage and consume arount 10 microamps, you can (in principle) connect that circuit directly to a 12 V bar battery that is capable of delivering hundreds of amperes for cranking. No probs there.

But, the standard way of doing it to

A Fuse the current (or, as it is, really, voltage source) where you connect to it.

B Add some filtering to avoid getting noise from other circuitry connected to same battery.

C Provide overvoltage protection. This is a must in mobile applications where a load dump can take the battery voltage up to 50+ volts for a few milliseconds. But long enough to fry anything electronic.

D If your voltage source is DC has wrong voltage, put a switcher in between so that you get the right voltage. You can also use a linear voltage regulator, but then the voltage source must be a couple volts more than your circuit needs. If you have very low current consumption and a decent difference between your available voltage and needed voltage, then a simple resistor/zener combination does the trick. Simple, cheap, reliable.

E If you have an AC source (mains, for instance) then you need some sort of power supply unit (PSU). There are ones that range from little wall-warts with a few volts at a few milliamps to large units with multiple outputs and capable of supplying hundreds of amps.

But, you can forget about constant current diodes. They are handy as electronic components. But don't use them to reduce voltage to micros and other electronic devices.

Google "power supply" for loads of info.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Skogsgurra]

"bar battery" = "car battery". There are some more typos, but they are more "transparent".

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[richs]

Hiya-

I just gave Skogsgurra a star for his post. "Bar" batteries aside I concur wholeheartedly.

I might strongly suggest that you research the 3 terminal voltage regulators out there. Common sizes range from supplying 100 Ma to 1 A to 3 A sizes, are relatively foolproof with proper bypassing (read the data sheets), and although not the most efficient of operating modes (they are what we call linear), they are much easier for a relative newcomer to electronics to use, rather than going through a switcher power supply design.

You do have to watch both the voltage input to the device (most are around 35V max. or so) and you also have to limit the power dissapated by them. So, depending upon the source voltage you have (assuming that it is at least 2V higher than the output voltage desired, usually 5V) you might want to put a low value series resistor in series with the voltage regulator to dissapate some of the power. For higher wattage situations, heat sinking is required.

As a start, you might want to google the data sheet for
the 7805 voltage regulator. This is a 1A device that comes in a somewhat small package that is easily heat sinked and available for very little money. In fact, you *MIGHT* even be able to find it in a Radio Shack store still. If not, there are many mail order houses that stock these.

For a variable voltage power supply you might want to look at the LM317 or it's varients. With a couple of resistors, this provides a regulated supply of adjustable voltage. With a potentiometer (variable resistor) you can have a range of voltages for your efforts.

The mention of wall warts is a good one. I use them all the time and look for them at goodwill and garage sales. They can be had for a buck or so and will supply current to a project from the AC mains safely and without too much mechanical construction.

So, with this information in hand, and a printout of the above mentioned data sheets, you can be well on your way to getting the proper amount of voltage (and current) to your desired equipment.

Best of luck and I hope this helps!

Cheers,

Rich S.

 

[Generalx5]

Well, here is the data sheet on the chip im using. Its an H-Bridge controller.

http://www.national.com/ds/LM/LMD18201.pdf

The typical supply current is 13mA, as it is a costly IC, I don't want to risk frying it. But what about LEDs? Those 10mA LEDs needs current regulation when pluged into a 9volt battery, I still think I would have to do this with the IC.

 

[logbook]

To power a 2V LED from a 9V battery use a series resistor.
If you need 10mA then there is 7V across the resistor (9V-2V). 7V/10mA = 700 ohms. Use a 680R. Power is V*V/R = 7*7/680 = 0.07W, so a 1206 surface mount or 1/4W wire ended type will be fine.

 

[Skogsgurra]

OK,

That's a somewhat different bird. It ain't TTL, it ain't CMOS and tain't one of those new 3.3 or even lower voltage ICs. It is a - shall we say - medium voltage power circuit.

What you need is a DC motor power supply, like 42 V. The new car electric system voltage. The LMD18201 can work directly with that supply. What was said earlier about fusing and overvoltage protection still applies.

For the LEDs, a series resistor is what usually is used. If you have a red LED, it will drop something like 1.6 V at 10 mA. To connect it to a 5 V supply, you need a series resistor equal to (5 - 1.6)/10 = 340 ohms. Nearest standard resistor is 330 ohms.

Other LEDs have other voltage drops, but they are all somewhere between 1.3 V (IR) to more than 2 V (UV). Connecting an LED to 42 V may be be dissipating too much power to be comfortable. Ten mA @ 40 V is 400 mW. Not impossible, but on the high side.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Generalx5]

But then, if small LEDs rated at 10mA requres current regulation, wouldnt the IC need regulation as well? So runing the IC dirctly off the power rail is a bad idea then?

 

[Skogsgurra]

No!

You have to debrief yourself.

The LED has an internal voltagedrop that is around 2 V. If you apply more voltage, it will hug all current there is. And probably explode.

An IC is not built like that. It is built to run off a certain voltage. It then consumes a rated current. Just like a normal incandescent lamp does. You dont use series resistors for normal lamps. Do you?

You can trust us. We know about these things. You don't.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Generalx5]

Awesome =D Thanks so much. I'll just be putting in some caps at the terminals for filtering and a 10k resistor to pull down the outputs in case of any failures.

 

[Skogsgurra]

OK. It is sometimes difficult to get out of one's line of thought. I hope I didn't kick too hard

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Generalx5]

Its all good, Im going to start etching now hahaha. Thanks everyone!

 

[Generalx5]

I Just remembered something, Will this CHIP work with PWM signals that look like this: 1ms is full reverse, 1.5ms is neutral, 2ms is full forward.

The specification data on page 6 of the data sheet says there are two types of PWM signals that are capable. I don't know if the first type is the one that matches with what my Microcontroller has.

 

[Skogsgurra]

No, that seems to be one of the RC servo control standards. Not what you shall feed into an H bridge.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Generalx5]

Ive just tested the PWMs, I have a servo that works with RCs and I've plugged in into the PWM controller that I have with 8 PWM outputs. I've programmed it so each output is differnt but ranges from 0-255, then I powered up the controler and tried the servo on it. It works! I don't know why, but the RC servo worked with the PWM outputs on my Controller.

 

[Skogsgurra]

All the better!

But, are you really using that output with 1, 1.5 and 2 ms? That would probably give the motor a 50-100 % speed range if running 1-quadrant and 0-100%, but no reverse, if running 4-quadrant.

Or do you use a pure PWM output, going all the way from zero to hundred percent duty cycle?


Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Generalx5]

My controller is a pure 0 - 100% PWM controller. I programmed it so that the output is at 127. The RC servo I plugged in responded correctly to the neutral position. At 255, the arm swings fully right, and at 0, the arm swings fully left. I believe this is the type of PWM used in RC applications? If that is the case, then the LMD18201 should be able to take the PWM from my controller to good use?

 

[Skogsgurra]

Sounds OK then. What is the carrier frequency? Do you hear any whining sound from the motor?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Generalx5]

I opened up a servo and checked the frequency, the motor does make whining sounds, but only when there is force applied on the arm as the motors are attempting to restore the PWM indicated position. The motors in the Server are pulsed anywhere from 30hz to 400hz...havnt gotten anything higher that that. But just to be sure, are RC PWM signal ranging from 1 - 2 ms? and in between those two are variable magnitudes?

 

[Skogsgurra]

I do not know much about those RC servos. But there is a potential risk that your application will be noisier than you want it to be if you run such a low frequency.

There is also the probblem with armature current ripple. If your frequency is lower than, say, ten times inverse of armature time constant, you will probably have brushes sparking and low brush and commutator life. Small H bridges can easily be taken up to around 20 kHz. That will give you a silent motor and zero current ripple.

Talking about servos. I have been using a very nice little micro lately. It is small, low cost and has a prototype board that costs around USD 25 in single quantity. The board contains processor, memory, crystal, 5 V and 3.3 V regulators and lots of breadboard area. The language is somewhat similar to C, but less brackets and parentheses. Very easy to get going and very suitable for lots of small control tasks. The company even has a suitable motor and a drive to go with it. Have a look at

http://www.parallax.com/detail.asp?product_id=32212

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...