Unusual power supply

A

Andres Rodriguez

New Member
I am working on a POV application where the Picaxe will be in a PCB that will rotate. The battery (11.1 or 7.4V) will be fixed. The contacts will be brushes (from an RC car motor) rubbing on two separate discs of .025" copper. The PCB containig the Picaxe will include a 7805 regulator.
Here are the questions:
  1. Am I out of my mind?
  2. What size capacitors should I use before and after the 7805?
You could ignore question #1, my wife has answered that one :rolleyes:.
Thank you in advance for any suggestion.
Andres
 
M

MPep

Senior Member
Hi Andres,

Typical how wives always make you feel just great about yourself!!! :D

For the input capacitor, try 10uF. You may get noise problems though because of the brushes. You may even need a pi-filter, 10uF capacitor feeding through to an inductor (choke), and then to another 10uF capacitor.
For the output try 100nF or more. Depends on the current drawn.

MPep.
 
B

BeanieBots

Moderator
I'd go with Mpep's suggestion for more 'normal' circumstances but I'd suggest larger values for your application. It really depends a lot on how much current your circuit will draw.
Having the brushes between supply and regulator gives the advantage of letting the regulator remove the large amount of noise but 10uF might not be enough, especially if the source is only 7.4v. Depends on the regulator, some won't even work below 8v.

I'd suggest as much C as practical with the space available on the regulator input. A minumum of 10uF per mA of load current and 10uF as well as 100nF on the output but check with the regulator datasheet to make sure it will be happy with that. Some don't like high capacitance low ESR on their outputs. (most require it).

To reduce brush surge currents, also fit a resistor in series between supply and the cap on the regulator input. Again, the value is largely determined by your load current but assuming your load is only a few 10's of mA a few 10's of ohms would be fine.
 
B

boriz

Senior Member
I’m with BB.

Low ESR will be handy in this app. After the brushes but before the regulator. A neat trick is parallel capacitors. Two similar capacitors in parallel will have twice the capacitance and less than half the ESR of just one (because lower cap values tend to have lower ESR values). So I’d favour two 470uF caps in parallel rather than a single 1000uF for example. Or even 3* 330uF. A small resistance between the brushes and the caps is a good idea. 10R? The voltage dropped across this resistor + the v drop across the regulator + the regulator vout (5v) must always be less than the battery terminal voltage, even when the battery is fully loaded and partially drained.
 
M

manuka

Senior Member
This may not provide enough current for your unspecified POV (assumed Persistence Of Vision) needs, but I've seen some very successful photovoltaic based "optical" powering via a small rotating PV, a suitable fixed lamp + directing lens/mirrors. Efficiencies will only be ~5-10% at best, meaning a significant light source would perhaps be needed for just a few LEDs demand, but the electronically the setup becomes simple & noiseless.

EXTRA: Don't forget "noise free" mercury contacts- although this is now pretty historic given toxicity issues.
 
Last edited:
A

Andrew Cowan

Senior Member
Voltage regulators are very good power supply smoothers. What kind of currents will you be drawing? Make sure you clean the copper discs regularly (with alcohol?) What kind of speeds will it rotate at?

A
 
Last edited:
A

Andres Rodriguez

New Member
First, thank all of you for the recomendations.
I enclosed the schematic, I estimate the the maximum current will not exceed 300mA and the maximum on any pin will be 20mA
 

Attachments

B

BeanieBots

Moderator
Couple of concerns.

What's going on with that 10R/100uF 11.1v bit of the circuit?

The suggestion was to have a lowish value R (10 ohms) between the power in and the cap on the input side of the regulator.

300mA total from the 28X1?
Maybe somebody can give the exact figure but to the best of my knowledge without looking it up, the maximum total is around 100mA with a max of 20mA on any one pin. Please check.
 
A

Andres Rodriguez

New Member
BeanieBots said:
Couple of concerns.

What's going on with that 10R/100uF 11.1v bit of the circuit?

The suggestion was to have a lowish value R (10 ohms) between the power in and the cap on the input side of the regulator.

300mA total from the 28X1?
Maybe somebody can give the exact figure but to the best of my knowledge without looking it up, the maximum total is around 100mA with a max of 20mA on any one pin. Please check.
Click to expand...
I see my mistake regarding the 10R/Capacitor.
Regarding the total current, I thought that the limit was on the maximum per pin. If the total max is 100mA tha means that I can ony drive 5 LEDs at any given time. OOPS:mad:
 
Peter M

Peter M

Senior Member
If you use transistors/FETs to buffer the outputs, so that you can still drive all your LEDs, you may need to lower the 10R resistor a little.

Depending on the drop out voltage of your regulator, as 10R @ 300mA is 3V.

Taking this from your fully charged lipo of 12.6V is 9.6V, but by the time it gets down to 9.6V (depending on what you beleive is the correct cut off voltage for these?) you only have 6.6V.... maybe a little close to loss of regulation?
 
B

BeanieBots

Moderator
The use of a darlington will make the volt-drop much worse than using basic transistors due to the way a darlington works. (Typically 1.2v.)
 
B

BeanieBots

Moderator
Yes, that's the beast. Just bear in mind that it is a darlington, so it will lose about 1.2v unlike a FET which will pull down to very close to 0v.

If your LEDs are blue or white they will have volt-drop of around 3.5v plus a darlington drop of around 1.2v gives a total of around 4.7v.
Not much headroom for the current limiting resistor if the supply is 5v.
Using transistors (or the darlington array) would however allow you to source the LEDs from the higher voltage input to the 5v regulator which would make constant current drive much easier but take note of Peter M's comments about battery voltage drop.
 
Last edited:
J

jglenn

Senior Member
Just a couple of unsolicited comments. The slip ring idea will be too much trouble. Copper is soft and will wear out quickly, not to mention corrode and oxidize. Slip rings should be the last resort.

Try a rotary transformer. A ferrite core might be messy, so shoot for a lower freq than 20-100KHZ, perhaps 100-1000Hz. Wind two coils on small spools with a hole big enough for a steel bolt, perhaps 1/4" dia. You could probably use a 555 driving a fet to generate the drive, sine would be better though. A simple bridge rect and cap on the output winding will give you gobs of DC, if the coils have the right guage, turns, and flux strength.

For driving leds, try SOT-23 smd fets, easy to solder, can drive a couple amps I think.
 
Peter M

Peter M

Senior Member
The slip ring idea will be too much trouble. Copper is soft and will wear out quickly, not to mention corrode and oxidize. Slip rings should be the last resort.
Click to expand...
they seem to last ok vehicle alternators. Usually replace a set of brushes and bearings before the slip rings wear out. Granted these are probably bronze, but electric drills etc. do use copper.

And mercury may be all but gone in many an apps., but its alive and well for this application, havn't checked out prices, but check them out here
http://www.mercotac.com/

.
 
A

Andres Rodriguez

New Member
I have used the slip ring method previously (see photos) but it was just to power some LEDs in the propeller and inside the spinner. It worked just fine for that application. Noise was not an issue and after about 100 flights there is no noticeable wear (the propeller turns at about 9000 RPM).
In the current application I will be using the same motor but there will be two separate rings. Because I will be powering a Picaxe the big concern is the electrical noise. With the suggested addition of Capacitors in parallel I hope that it will work.
 

Attachments

Peter M

Peter M

Senior Member
I have used slip rings as well, I run audio through it (not hifi mind you) for a demo of the Dopler effect.

Mine is simply a piece of fibre glass PCB with a couple of continuos circular tracks and a pair of large drill motor brushess rubbing on the face.
There is no noteable crackling from the speakers hanging off the other side.

Sorry don't have any photo's
 
W

westaust55

Moderator
copper/bronze sliprings and carbon brushes

Something else to think about.

Most larger industrial electric motors have bronze or stainless slip rings.
These motors run at speeds like ~750rpm, 960rpm and 1440 rpm.
Bronze will be similar to copper but do not have the hardness data to hand.

The usually "carbon" brushes used in these above mentioned motors are quite a lot harder than the slip rings. Some brushes have a combination of carbon and copper and they are harder still.

All that stops the bronze slip rings wearing out rapidly or suffering from "threading" (grooves) is running the current density in the brushes hot enough to put a layer of carbon on the slip rings (often called a skin or patina).

With your motor runing at 9000 rpm if no carbon layer develops on the copper tracks of the PCB be prepared for rapid copper wear or damage.
 
B

boriz

Senior Member
Or…

Perhaps you could generate the power you need ‘locally’ with a stepper motor. The steppers case and wiring rotate with your POV device, but the shaft is mechanically coupled (gear/pulley) with the static base.
 
You must log in or register to reply here.
Top