Voltage Regulator for a Small Motorbike, 6V AC Current

[alec_t]

to keep the volts up and lights on strong.

'Lights' = brake light only. So far your tests have used DC voltages. The important voltage across the bulb, when driven with AC, is the RMS voltage, which should be no more than ~6V if you want the bulb to have a long life.
Your DMM, on its AC range, hopefully will give a reasonably accurate indication of the RMS voltage, despite the bulb waveform being non-sinusoidal.

 

[sign216]

Umm, real world English please? If by DMM you mean my multi-meter, then it won't give a reading for the AC current. The reading jumps all over, and is entirely erratic.

Reading AC voltage on small bikes; It's been remarked on in moped forums, but no one has an answer. Those forums have good mechanical knowledge, but are weak when it comes to electrics.

 

[KeepItSimpleStupid]

DMM = DIgital Multi-Meter

Quick lesson:
What you want to be able to do is measure is the equlivelent AC voltage that has the same heating value as a DC voltage into a resistive load. So..If we had 120 VDC and a 1 ohm resistor, the resistor would dissipate 120 W. We would LIKE to have a similar meter when it looks at any AC waveform to report 120 VRMS if that waveform has the same heating value as that DC voltage. These meters, have limitations, but are usually call TRMS meters or True RMS meters.
RMS is a fancy word for the square root of average value squared. Literally. it's a Calculus concept.

Your DMM is not likely to be a TRMS meter. What it is likely is what's called, average responding, RMS reading. SO, it ASSUMES a sine wave below some frequency. Averages that and multiples by a "fudge factor".

You really only care about how much heat is applied to the filaments or 6 V RMS. This controller clips. The waveform will be all over the map.

Now, what you "could do" is have two 6V light bulbs, one across your regulator and one across a 6V power supply. Adjust your regulator to be about the same brightness as the 6V powered bulb. Better idea, grab a solar cell and measure the short circuit current which is proportional to intensity when mounted in the same place. if you had a 300 mA DC current scale and a solar cell that had a Isc < 300 mA, just plug the solar cell into the meter's current input and adjust for the same intensity.

6 VDC = some Isc of the solar cell when mounted in the same place and distance away.
Use bike and also adjust for the same Isc of the solar cell.

 

[alec_t]

You've lost me on the solar cell bit, KISS. How does that relate to trying to measure RMS voltage of a non-sinusoidal voltage? The OP already has DC measurements and shouldn't be measuring short-circuit current of anything.

 

[KeepItSimpleStupid]

Hey, I was in solar cell research. Intensity is proportional to current basically. The DMM has a low value resistor across the inputs. A shunt ammeter. So, a single solar cell is an intensity measuring device. The best bias point is short circuit current. The solar cell is flattest there. Since it's not actively biased at 0V, because of th resistor, it's technically not Isc, but it will be close enough. A 1 sqcm crystalline silicon device was about 25-30 mA at AM1.5 Global spectrum at 25 deg C.

Aside: I used them to calibrate every day. We shifted the spectrum of the cells with a filter for some materials. We had about 15 cells and about four had traceable calibrations. With those traceable ones, we created more. Each day we used the one we used the day before and one more.

So, just want the regulator to put out the same voltage that would make a 6V lamp on the motorcycle to be the same brightness as a 6V lamp operating on 6 VDC without eyeballing it. Intensity is also affected by distance - so they need to be in the same place relative to the lamp.

 

[sign216]

Here is my take on why the bike's voltage is erratic.

The two coils (head light, and ignition+brake light) are wound differently, spinning around magnets at varying proximity and speeds, and share the same ground. Even if the engine idles at a constant speed, the relative speed of a coil to a magnet is always changing.

A bike with a DC current has a regulator and a battery to smooth things out. But an AC bike will always have an unstable current.

That's my theory. In practice, people can't measure the voltage.


Edit: So KISS, essentially I could look at the light's intensity at different rpms, and then hook the light to known voltage sources to see how bright it gets?
That would give me a crude idea of how high the voltage reaches.

 

[KeepItSimpleStupid]

So KISS, essentially I could look at the light's intensity at different rpms, and then hook the light to known voltage sources to see how bright it gets?

Yes. If we're lucky, it won't change much. What you really care about is the lamps maximum intensity when mounted on the bike vs a 6 VDC driving the bulb directly.

You will be. sort of, measuring the voltage optically.

A solar cell used t easily findable at Radio Shack, but not any more. This **broken link removed** would be suitable. Just plug it in to the DMM using the DC current scale capable of measuring 50 mA or use your eye.

 

[sign216]

I finished constructing the regulator. Sorry it took so long, but work and wife often don't realize the important things, like amateur electronics.
After I test it on the bike, I'll seal the interior "board" with conformal acrylic (new to me).

Notes: I put a 3 amp fuse in the regulator's connecting wire (nice suggestion KISS). Chose 3 amp because it was the smallest I could get, and seemed a reasonable value.
Alec -T, thank you for the suggestion to buy extra semi-conductors when I was placing the order. Very helpful. And the advice to get transistor mount kits and "mica" insulators. I would have forgotten.

One of the two semi-conductors needed a ground, so for that one I didn't add an insulator sheet underneath, letting it ground to the metal box, and I'm attaching a wire (not pictured) from one of the box's screws to the bike's frame. The regulator's main wire (red) will be connected to the brake light wire.

There might be a delay in mounting to the bike and running it. It's 32 F (0 C) in Massachusetts, USA right now.

 

[KeepItSimpleStupid]

It looks really decent and small. Nice work.

 

[alec_t]

Neat. Will you be adding C1 internally or externally?
You might want to add some blobs of epoxy to support components physically, as an anti-vibration measure.
When it's operational, check the enclosure temperature to see if additional heatsinking is needed.

 

[gary350]

If the voltage is correct vibration is often why light bulbs like that burn out. Bolt a weight to the under side of the back fender to stop vibration. At a certain speeds it causes cause parts to vibrate at their naturat resonance frequency. Add some weight it changes the resonance frequency of the metal parts. Test it with 1/4 lb, 1/2 lb, 3/4 lb etc it wont take long to change resonance frequency to something that works.

 

[sign216]

Thanks for the kind words. Instead of epoxy, I'm thinking of using brush-on acrylic coating that's lighter.

I started the cycle up this afternoon, in 32F (0 C) and damn! Whenever I apply the rear brake, the engine dies. So I can't try the voltage regulator yet.

If you are familiar with moped style electrics, the rear brake light is powered by the ignition circuit. So if the rear brake is burnt out, the ignition loses it's ground and halts. But this brake light is good, strong and bright! I've got one suggestion to improve all the ground connections, especially for the rear light. I'll try that, and keep you all posted.

 

[sign216]

Alec_t, KeepItSimple, and Diver,

Sorry for the long delayed post. In short, the 6v AC voltage regulator works great!!!
In fact the bike runs so much better with it! The extra path to ground prevents a common problem with moped and small motorbikes; losing the ignition when the brake light is pressed or burns out.

Summary: Not only did this regulate the voltage and keep bulbs from burning, but it "cleaned up" the whole electrical system and the bike is now a reliable runner! It is transformed.

Cycle/moped vendors no longer carry 6v AC regulators for bikes, and the few on the internet are of dubious design and construction. Yours is superior.

Thank you,
Joe

 

[Pommie]

I like a success story, even if it takes two and a half years to come to a happy ending.

Mike.

 

[sign216]

Sorry for the delay in reporting. I built the unit during the winter of 2014-15, and rode the bike all summer of 2015 once it was sorted out. It was a father-son project, and I'm holding it now for my son, who considers it his bike.

The regulator really did make a huge difference. Changed it from a "vintage Italian of questionable reliability" to "steady solid performer." This blue bike is my profile picture (pic is before its limited restoration).

Making the regulator was the best electrical project I ever did. Its merits are;
-Original design from this fantasic board
-Construction from scratch by me
-It was terrifically successful in making the bike work
-No commercial option was available

A win for amateur electronics!

 

[KeepItSimpleStupid]

I am very impressed with everybody's work especially alec_t. He can really get involved sometimes. We (alec and a few others) had a thread going for at least a year on ETO for a pump controller for an aquarium. I provided some physical motor measurements. As of a few months ago, there is only live rock in the aquarium., but it was one elaborate setup where the tank was made and built in. The pumps needed a "flick" to chase the fish away so it would not turn on suddenly.

Your device is quite small and I'm not sure how it gets mounted, There are di-cast aluminum cases with an "o-ring". So, a few suggestions: 1) Let everyone know the part number of the case you used and the size, 2) post a pic of the finish mounted project and the model of the bike with a nice search string. 3) I really liked your testimonial and think you. 3) Electronic grade silicone or aquarium silicone might have been nice to support the components.

You could have used "heat shrink" around the leads of the FETS. I could send you a say 1.5 feet of Kapton Tape to use as an insulator under the FET pins if you want? It's a really expensive tape with a silicon based adhesive that's good for 200 deg. C temperatures. It's only about a mill thick and removes easily even after being exposed to 200 deg C. I have some that's about 1/2" wide. I've only used the silicon based adhesive. Not sure if this is the p/n I have: http://www.digikey.com/products/en?mpart=5419 GOLD 1/2IN X 36YD&v=19

This (Scotch 70) http://www.digikey.com/product-detail/en/3m/70HDT/3M12008-ND is not a bad tape for outdoor connections or even those in a car. You could have also considered head shrink with an adhesive.

FYI: Many of the sealed automotive connectors are available at Mouser.com. e.g. **broken link removed** (Delphi Weatherpak). With a lot of the connectors, you have to buy the parts separately. Generally they are give an IP rating or "Ingress Protection". Connectors require a crimp tool and they can get pricey. I've got one that generally works.

Corrosion of the connections could be a long-term problem if it isn't done right. The silicon tape, the adhesive heat shrink or the sealed automotive connectors should work well.

With wiring and things, one needs to include "drip legs" if appropriate. A very good illustration of a "drip leg" is power entry of an overhead power line.
The line from the pole "dips" before it connects. The water than runs along the wire falls off at the lowest point instead of following the wire into the meter/house etc.

Agreed, this may be 20/20 hindsight, but it may have been an overlooked detail.

Nonetheless congratulations are in order for sign216, the threadstarter, and alec_t.

 

[spec]

I enjoyed reading this thread- well done all those concerned.

spec

 

[alec_t]

Good to hear it's stood the test of time. Vibration, heat, voltage spikes and moisture are the enemies of automotive electronics.

 

[sign216]

KeepItSimple,
Thanks for the tape offer, but I'm going to coat everything in acrylic board "dope" and that should provide additional insulation and support. Also thanks for the construction tips. The installation onto the bike is nothing special; just found some space by the large brake light housing to stick it, wired it in w a fuse to the brake light, and added a dedicated grounding wire to the bike's frame.

Alec,
I'm hoping the acrylic dope takes care of any vibration and wetness issues. Your superior design resolved the voltage spikes, and as for heat...I hadn't thought of that. The metal die-cast housing may provide enough of a heat sink, but I'll feel the enclosure after a run on a hot day, and check it.

 

[Apothecary]

This thread is far from dead.....the OP isn't the only one out there needing a good 6v regulator like this