Voltage regulators

I am rebuilding my café racer and adding an Alton electric starter, since space is at a premium, I opted for a Ballistic lithium battery. These batteries require a slightly higher voltage than standard or AGM lead-acid batteries. I read that some got on quite well using the lithium batteries with a "standard" Podtronic or similar integarated rectifier/regulator and while the battery was not 100% charged, it was OK to get the bike going.

I ordered a Chinese replacement regulator, when I got it, I tried it on a 12 volt AC transformer to see how it would perform. I wanted to know if it was stable, would heat up... As soon as I connected it to my transformer and powered it up, my clip leads just melted, so I figured it was defective. I measured it and it was not shorted, yet just touching the input wires to the transformer just produced sparks :!:

What was going on with that thing? I borrowed a real Podtronic from my friend and did the same test... same thing :?: Very puzzling.

So I started to investigate and this is what I found:

This is the basic circuit for these regulators



The GREEN rectifiers (this is a 3 phase model) take the AC from the alternator and turn it to pulsating DC, the battery smooths it out. As the battery voltage rises, the control circuit turns on the RED SCRs to effectively short out the input or to put it another way, short out the alternator. I have worked all my life in computers and electronics and this does not make sense to me.

My test setup consists of a pair of transformers hooked up to give me 3 different AC voltages, about 13.5 V, 16.5V and 19.5V. Look at the waveforms at the transformer output with a Podtronic type of regulator. My transformers are 250 VA so more than a good 3 phase alternator.

This first picture is with the first tap (13.5 V)

AC current is 28A, AC volts is 9.6 and DC volts at the battery is 14.34 with no load



Second tap (16.5 V)

AC current is 25A, AC volts is 8.1 and DC volts at the battery is 14.4 with no load



Third tap (19.5 V)

AC current is 19A, AC volts is 7.2 and DC volts at the battery is 14.4 with no load



I added a load consisting a 6 ohm resistor in parallel with a 2 ohm resistor, this in theory would be a 8 amp load, but my wires are somewhat undersized. Now watch what happens...

First tap (13.5 V)

AC current is 5A, AC volts is 12.8 and DC volts at the battery is 11.8



Second tap (16.5 V)

AC current is 6.8A, AC volts is 14.1 and DC volts at the battery is 12.1



Third tap (19.5 V)

AC current is 7.3A, AC volts is 14.6 and DC volts at the battery is 12.2



As can be seen, as long as the voltage is below threshold, the rectifier acts as a basic bridge. Now watch what happens when I reduce the load to only a 6 ohm resistor or about 2 amps...



My conclusions with this type of regulator are:

NEVER use them without a battery
NEVER run them without a load

They work, but they are not kind to alternators and as far as using them with lithium batteries, this is asking for disaster, the voltage of lithium batteries being higher, they will short the alternator more than with a lead-acid battery, a shorted alternator will heat up and some day burn up. The less load there is, the bigger the hit on the alternator.

Next I will show what the old Zener diode does.

Stay tuned

Jean

Now the old Zener diode that many shun for obscure reasons... there are some reasons to use something else, the main one is the Zener is made for positive ground systems and the other is that a separate rectifier is required.

This is the alternator waveshape at 13.5 volts with only a rectifier and a capacitor, connected straight to the battery. I did not test it at higher voltages because the DC voltage of the battery would have gone too high.
On the first tap of the transformer (13.5V) my readings were:
AC current 0.6A AC volts 13.5 and DC volts at the battery 16.0 with no load



Now I connected the Zener and here are the readings

First tap 13.5 volts from the transformer:
AC current 1.7A AC volts 13.2 and DC volts at the battery 15.0 with no load



Second tap 16.5 volts from the transformer:
AC current 4.2A AC volts 15 and DC volts at the battery 15.9 with no load



Third tap 19.5 volts from the transformer:
AC current 5.2A AC volts 16.2 and DC volts at the battery 16.4 with no load



The old Zener is kinder to the alternator since it only clips the top of the AC waveshape, its regulation however is very approximate erring on the high side. There are ways to use it on negative ground systems by using stud mount rectifier mounting hardware which electricaly insulates the stud from the heat sink, but most people won't bother and will use a Podtronic or similar not knowing what I showed in my previous post. Out of the two so far, I would rather use a Zener than a Podtronic or other similar product.

On each test, the Zener is "wasting" more and more energy, even without a load, going from 23 watts to 63 watts to 85 watts. Obviously with a load, some of that wasted energy would go to the load and the rest out in heat.

Next a home brewed solution

Jean

With my nice lithium battery I wanted something a bit more precise than an old Zener and something less brutal than a Podtronic, this is what I came up with:



This is what is known as a linear regulator, effectively it is a variable resistor between the bridge rectifier and the load. It is not efficient like a switching regulator because as the engine revolutions go up, the voltage goes up and the excess from the set point is dissipated as heat. This type of regulator works best with a light load which is what I am shooting for with LED bulbs all around and only an electronic ignition as the main power draw.

I built it in the shell of an defective Japanese voltage regulator. Presently the components can be seen but when I am satisfied with the way it works, I will pour resin in it to keep the vibrations from breaking component leads. I don't know how well it will perform in the long run, but seeing what the others do, can't be any worse.









Now here are my test results:

First tap (13.5 volts)
AC current 0.2, AC volts 13.5 DC, volts at the battery with no load 14.4 V



Second tap (16.5 volts)
AC current 0.2, AC volts 16.5, DC volts at the battery with no load 14.4 V



Third tap (19.5 volts)
AC current 0.2, AC volts 19.5, DC volts at the battery with no load 14.4 V



I'm happy about the fact that the regulator takes almost no power and especially happy about the tight voltage regulation. As more current is taken from my regulator, the wave shape will flatten, a bit like it does with the Zener, this is normal, it will heat up too, but so do the others I have tested and nothing a bit of a breeze won't deal with.

The voltage is the correct one for lithium batteries, the alternator is not brutalised to regulate the output, the regulation is spot on, the only thing left to see is long term performance.

Jean

Another interesting bit of information:

With a battery sitting at 13 volts, here is the current draw with the regulator connected

Zener 0
Podtronic type 0.416 mA
Home built 0.080 mA

This means the Podtronic type will drain the battery 5 times faster than my home built linear regulator. The Zener has zero draw since that is below its design voltage.

Jean

Very nice presentation and very informative. A nice job of packaging.

Have you thought about designing a switching type regulator? That is, one that would open circuit the alternator, with a zero crossing detector to prevent reverse spiking.

Slick

Excellent report.

It would sure be nice if you could build a batch of 20 or so, my guess is they'd sell out overnight right here on AccessNorton

texasSlick said:

Very nice presentation and very informative. A nice job of packaging.

Have you thought about designing a switching type regulator? That is, one that would open circuit the alternator, with a zero crossing detector to prevent reverse spiking.

Slick

Click to expand...

A switching regulator would be better I'm sure but would go against the KISS principle as well as adding a transformer. If you look at PC power supplies, they manage from 250 to 500 watts and while the space required to do so is not that big, it is larger than what I have made. My output transistor is rated at 30A which is more than what the Alton's alternator is capable of.

Jean

grandpaul said:

Excellent report.

It would sure be nice if you could build a batch of 20 or so, my guess is they'd sell out overnight right here on AccessNorton

Click to expand...

I'm just a hacker, someone else can take the ball and run with it. Also, long term use has not been proven yet, I'm confident, but not enough to blindly make some and sell them. When I looked at the regulators, especially the Podtronic type, I thought there had to be a better way. It seems the regulator makers build them so they are bulletproof electricaly and physicaly, but they don't care one bit about the alternator.

Jean

I second GrandPaul's post. Think about it.

You could put a few out as "beta" sites to get a handle on reliability. I would buy one.... I do not quality as a beta tester .... Do not ride enough these days. Good luck if you venture out!

Slick

This is fastenating & scared stuff, wish I understood more but I think I got the concept, thanks Jean for posting this info. If I understand you correctly I am in danger of overheating my stator. I am running a 200 watt three phase Sparks rotor & stator with a Podtronics regulator / rectifier AND a Shorei lithium ion battery. The charging system has been this way for a while but I just recently started using the Li-ion battery. How bad of a risk do you really think this is?? Thanks again. Glenn T.S.

gtsun said:

This is fastenating & scared stuff, wish I understood more but I think I got the concept, thanks Jean for posting this info. If I understand you correctly I am in danger of overheating my stator. I am running a 200 watt three phase Sparks rotor & stator with a Podtronics regulator / rectifier AND a Shorei lithium ion battery. The charging system has been this way for a while but I just recently started using the Li-ion battery. How bad of a risk do you really think this is?? Thanks again. Glenn T.S.

Click to expand...

I really don't know if the alternator is at risk of burning up soon, after all there are thousands of these being used. I do know that a Podtronic type of regulator does short out the input as a method of regulating the output and I have never seen this done in any power supply for as long as I have been working and thinkering in electronics.

To say that they don't work is not a true statement, they do work but they also breakdown and the box full of defective units I tested certainly proves they can become defective. Sometimes when going to my friend's garage I would look briefly at electrical problems of his customer's bikes, more often than not, they would just replace the alternator and the regulator to get the bike going again, and this is on Japanese bikes with the same type of regulator and permanent magnet alternator. There is probably a fine balance between the alternator, the battery and the load, as long as there is no change, everything is fine but when something changes (battery aging, more or less load) the system is disrupted and a failure occurs. My friend has a Honda VTR1000, it had electrical problems on many occasions, regulators were replaced, the alternator was replaced, both many times. I haven't looked at it since I found out how these regulators work, but I'm sure there some imbalance somewhere.

I think the best way for you to see if there is a potential problem would be to take a current measurement on the input of the regulator, this is AC current, if at normal running RPMs and with a normal load, the current is very high, then there is a chance of the alternator or the regulator being stressed enough to break down over time. You could look at the waveform with an oscilloscope and if it looks like a chopped sinewave, it means the regulator is shorting the inputs, doing it's job of regulating the output the way it was designed, not pretty, not in finesse.

After seeing how these thing work, I would not put one on my bike that's for sure. Aside from building my own, I have not seen any other suitable replacement. I am waiting for an answer from Sachse http://www.elektronik-sachse.de/shopsys ... 0ceb7ae46b to see how they do things, either like I did or like the brute force method. I hope they use more finesse than the Podtronic type and even what I did.

Jean

I'm not convinced that shorting the coils in an alternator is comparable to shorting the output of a transformer. As you say, there are hundreds of thousands of alternators being regulated this way and the number of failures is very low. Most common part to fail on Japanese charging systems is the regulator/rectifier not the alternator. Some models have very few failures while others are almost guaranteed to fail. This points to design flaws in those that fail.

Ian

Nortoniggy said:

I'm not convinced that shorting the coils in an alternator is comparable to shorting the output of a transformer. As you say, there are hundreds of thousands of alternators being regulated this way and the number of failures is very low. Most common part to fail on Japanese charging systems is the regulator/rectifier not the alternator. Some models have very few failures while others are almost guaranteed to fail. This points to design flaws in those that fail.

Ian

Click to expand...

Power is power, if 200 watts is being shorted, that is a lot of current and it has to be dissipated as heat, in the transformer (alternator), the regulator or the wiring. If I look at my homemade regulator, the heat is dissipated only in the regulator and it is only the difference between what was produced and what is needed. As an example, if the alternator is giving out 25 volts and the set point is at 14 volts AND the draw is 5 amps, then 55 watts goes out as heat (25-14 X 5).

A 200 watts alternator with either a Podtronic type or a zener would "waste" 130 watts if the load is 70 watts (5 amps @ 14 volts) heat being the ennemy of electrical devices, this can't be good.

I know there are a LOT of these things out there and the failure rate is not dismal but it is not elegant design wise. Call me paranoid, but its not going into any of my bikes.

See http://www.electrosport.com/technical-r ... stem-works they say a lot of heat is produced, so cooling IS important.

While searching for motorcycle electrical system failures, I found this place http://cycleelectricinc.com/RECTIFIER.html they make a regulator for HDs that works EXACTLY like the one I made.

In the old days of video cassette recorder wars, the VHS came out as the winner, but it wasn't because it was technicaly superior, the same goes for regulators that short out the inputs (at least that's what I think)

Nice to see some discussion on this subject :wink:

Jean

Jean, I know you have nothing better to do, so why not test a Sparx regulator/rectifier if you can get one, and maybe a Tympanium?

I could send you one of each...

grandpaul said:

Jean, I know you have nothing better to do, so why not test a Sparx regulator/rectifier if you can get one, and maybe a Tympanium?

I could send you one of each...

Click to expand...

Nothing to do :roll: ha, I wish :!: I can find time to test them out, you still have my address?

Jean

Jeandr said:

grandpaul said:

Jean, I know you have nothing better to do......Nothing to do :roll: ha, I wish :!: Jean

Click to expand...

Click to expand...

Jean , thanks for posting

Jeandr said:

I can find time to test them out, you still have my address?

Click to expand...

I probably do, but it will be quicker if you just e-mail me when you can.

Interesting post.
A couple of things to comment on that explains how the zener system works on our Nortons and why lithium batteries may need special consideration

The Alternator (dynamo/magneto coil) is a current limited system due to the design and magnetic flux available.
These coils are designed to run constantly at the maximum design current and this is done in the basis zener regulator by the zener presenting a load to the coil/rectifier by shunting the coil to earth. The coil voltage then drops as the current is shunted as the coils have limited current avbl the voltage will drop until it goes below the zener conduction voltage and the zener stops conducting and the voltage rises, then the zener conducts again and loads the coil again. This results in a constant voltage determined by the zener. When a load is presented the current is shared by the zener and the load. But the total current is fairly constant. Zener regulator are normally designed so that the zener should still conduct about 10% of the total current available at the maximum load so as to keep it in conduction and thus regulates the voltage. In fact for most systems that use low resistance sources like a battery or large transformers they must have a series resistor to stop the zener drawing mega current and destroying itself. The resistor is calculated to keep the zener in conduction at max load hence the +10%.

The Norton system uses the coils ability to supply only a certain maximum current as the effective limiting resistance.

Now the coil is happy to be completely shorted to earth as it will only ever supply a limited current, not like shorting a battery as they have a very low internal resistance.

The coils will get no hotter then when running a normal load or completely shorted and in fact they will be cooler as the application of ohms law will show that power is current x voltage and if the voltage is 0v then current will be irrelevant as the the power will always be 0.

The difference as I see it with a zener V say a podtronics is that the zener does its regulation at a DC level after being rectified by the bridge where as the podtronics do it at an AC level in one leg of the bridge rectifier. The podtronics should run cooler in comparison.

The zener regulator or Podtronics are designed to charge a lead acid battery and the zener conduction point is designed to suit a lead acid battery. So I'm guessing using a lithium battery that has a higher terminal voltage will probably dump a high current through the zener or podtronics as they will try and conduct at the normal charging voltage and as the lithium battery is happy to deliver current at the higher voltage then it will most likely let the smoke out of the standard regulators or blow the fuse or both, or be very close to doing same.

That is how I see but I am quite happy to be proven wrong.

Cheers
Peter R

Hi Peter thanks for that !
so do you think I had better to let my Mithium battery on the shelf (pity , as I have two of them , waiting my next project!!)........your thoughts are welcome, for me (at least!). Cheers Pierre.

Yikes :!:
Remember the lithium battery fitted to the Boeing Dreamliner :?:

Boeing Dreamliner lithium battery catches fire;

“NTSB chairwoman Deborah Hersman told reporters on Thursday that tests showed a failure in a single cell of the battery spread to the rest of the battery in a way unanticipated by Boeing.”

http://www.theguardian.com/business/201 ... ttery-flaw


at least with a lead acid battery if one cell is flawed it usually goes dead :shock: