• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Power supply for automotive project

Status
Not open for further replies.

Axe_Murderer

New Member
I've built a project that is now in my car. It works great when the car is not running, and works jusy "okay" when the car is running. I have some temperature sensors (LM34) that produce pretty wild variations in read-out only when the car is running. Because of this I'm feeling pretty sure that the problem lies with noise/ripple on my power supply (feel free to disagree if you have an idea).

The power supply I've built is the usual design I use, a pair of caps flanking a voltage regulator. Since this is going into my car I used a pretty big cap on the input side, 2200UF 35V, with the usual tiny cap on the output side, a .010UF. I'm thinking that maybe I didn't put a big enough cap on the output side to deal with the noise? Or maybe the input side is too big and not buffering high frequency noise well enough? Obviously I don't know so that's why I'm checking with you folks.

If you have a suggestion on how I can improve the ripple rejection on my p/s I'm all ears!

Thanks.
-Mark L

(P.S. the project web page, not really populated much yet: www.markdepot.com/multi )
 

stevez

Active Member
Just a thought - can you substitute a small battery to see if the circuit behaves any differently?? Your power supply might be fine. This might help verify.

There can be some very high and fast spikes on auto power systems. Some kind of transient suppression on the input side of the regulator might be useful. This could be a smaller value cap more suited to the application (low ESR I think, maybe tantalum??).
 

Axe_Murderer

New Member
I *could* run it with a battery, but I don't think there wouldn't be any point. The only parts the device interfaces with the car are the power and one wire going to the O2 sensor, which is supposed to fluctuate and never goes above 1v anyway. None of the other parts go to ground or connect to any other part of the car electronically. I will however test that when I get home.

A little more info as I drove to lunch today. Twice the pic (pic controls the whole thing) froze up and stopped running the program until I removed power. Thoughts on that? Sounds like the same instability to me.

I'll post a schematic of the device later, I still haven't made one yet. There are a few extra parts to the p/s I haven't described yet, that being a few diodes to provide overvoltage and reverse voltage protection. I don't see them effecting this but you never know (er, well a dope like *me* never knows).

Thanks.
 

BeeBop

Active Member
I *could* run it with a battery, but I don't think there wouldn't be any point.
The point stevez was trying to make was that you could *temporarily* run it off a battery, to isolate your problem. I tend to agree. The noise could be comming from your ignition system, or other place. Running your circuit from a battery would show you if the problem was, in fact, your power suppply.
Regards,
Robert
 
H

hyedenny

Guest
Hi - make sure the MCLR pin on the pic is tied high with a 4.7k resistor. Then make sure you have a bypass capacitor (eg .01uF) very close to the PIC between Vcc and Vss. Maybe you need a snubber circuit with a cap, a resistor, and maybe even an MOV to smooth spikes. Also, make sure that you have at least 10,000 uF for every amp that your power supply is delivering, assuming its a linear power supply. Above all, make sure you have a clean path to ground! If none of this works, disable the alternator to eliminate that noise as a culprit. How close is the LM34 to the ignition and/or the alternator? Are you using a pic with onboard A/D?? Ive done this same thing for motorcycles with success, so I hope this helps!
 

Axe_Murderer

New Member
Thanks for the help everybody, here is today's update.

When running off battery power it works as it should. I've documented my power supply circuit for your scrutiny.

I don't have a bypass cap on the pic, but I've never needed one before. I don't know what a "snubber circuit" or MOV are.
I haven't left any input pins floating.
The regulator is an L78S05.

Thanks again for you input. I only know what I've taught myself and there are huge gaps in my knowledge.
 

Attachments

H

hyedenny

Guest
Hi,
Dont forget about your voltage drops across the diodes!! Also check your 2200uF capacitor polarity.
A simple passive RC "snubber" circuit is just a resistor and capacitor in series (sometimes also inductors and diodes) to dampen resonances of the parasitic elements in a power circuit. In other words, it absorbs voltages from inductive loads like motors.
An MOV (Metal Oxide Varistor) absorbs large voltage spikes such as those which might occur from high voltage ignition sources, lightning, etc.
 

BeeBop

Active Member
I'm inclined to change this first, because this could have a lot to do with it:
I don't have a bypass cap on the pic, but I've never needed one before.
I have seen PICs running strangely without this. It'll take you two minutes, and cost two cents, and it's worth it to get this out of the way.
Then test it again; there may be several issues at the same time, which always makes chasing down problems a lot more fun.
Regards,
Robert
 
H

hyedenny

Guest
HI,
Did you build that circuit exactly like its indicated in the schematic?? Check the polarity of that 2200uF cap!!!! Also, the diode on the low side of the circuit will fool the regulator into thinking its one diode drop away from 0 volts... Your output will NOT be 5 volts!!
 

Axe_Murderer

New Member
Sorry, diagram is as noted except that the cap is installed correctly.

As far as my output not being 5volts, well my multimeter disagrees quite strongly. :)

Will try the bypass cap and let you know the results.

Are there any suggestions on improving the power supply to ensure ripple is eliminated? Or is it good enough?

Thanks again.

-Mark
 
H

hyedenny

Guest
Im sorry, I guess it WILL have 5 volts - as long as the chassis ground is separate from the project grounds. The 7805 and caps should smooth any ripple you have as long as the current stays relatively low. Definitely use a bypass cap close to the pic, and pull MCLR high! good luck.
 

Axe_Murderer

New Member
Alrighty. .01uF cap installed right next to the PIC. No change in behavior.

I noted something interesting though, that being the farthest LM34 has the largest variation in reading. Following that trend the next farthest LM34 has the second largest variation in reading. Because of this I'm inclined to suspect some radio style interference and plan on implimenting some measures to test that according to the LM34 PDF. Will let you know how it goes.

Note however that this doesn't explain the PIC freezing up during use, but one thing at a time I guess, right? :)
 
Status
Not open for further replies.

EE World Online Articles

Loading
Top