HT Motor + Shift Kit + 6V/12V generator

[Terminal Velocity]

switching regulator

hey steve. I was reading your thread because I've been looking at using a motor to generate some power for my headlight and ultimately other accessories. I've been doing some homework on the switching regulator but I can't seem to find one which outputs what I want. I need a minimum of 3 amps to run the headlight (that's no problem), but I'm always trying to think ahead and realize I may need more later...anyway, you said you built an 8 amp model. Would you be willing to share the schematic? If not maybe you could answer a couple of questions which I can't seem to locate on the net. For instance, can you run more than one regulator/controller in parallel to get a higher current output? What exactly is synchronized? my searches so far have revealed synchronization with an external clock...is this for the purpose of paralleling or is it for the sole purpose of changing the frequency of the controller? I don't have the money at the moment to build anything, but I want to get all my ducks in a row so when the time comes I can build it.

 

[AussieSteve]

I haven't drawn up a regulator for battery charging yet. Mine is a temporary circuit, (on a breadboard), to control a DC motor driving my generator for testing. The charging regulator comes later. (Still waiting for my optical tacho to arrive.)

It takes a 15-23VDC input and is adjustable from about 1.5V to 15V at 8A, or to 23V under lighter loads. Operating frequency is 20kHz.
There's no current-limiting for battery-charging, so an extra-low battery could overheat the MOSFET.
The MOSFET that I used is overkill, rated for about 80A at 25C, but I had it handy.
Any 30V+, 16A or better MOSFET would do the job, as long as it has a reasonably low Rds(on), or a heatsink would be needed.
You'll need to modify this circuit to at least add current-limiting, but much of the work is done.

It's possible to parallel linear regulators to increase current output but not a good idea with switching ones. Best to use the right sized regulator for the job.

Not sure what that synchronisation refers to. I'd need a bit more info. It might just relate to the fact that a switching reg, (DC-DC converter), runs at a fixed clocking rate and that the PWM synchs to that frequency.

Here's the schematic:-

N.B. This circuit can't tolerate more than an absolute maximum 30VDC peak input without changing components.
Also, of course, omit T1 and DB1 for generator input.
And, a bit of fine tuning of the timing components might be needed, to get things just right, to allow for resistor, capacitor and timer chip etc tolerances.
Further to that, the circuit could be made smaller by using a 556 dual timer instead of a pair of 555s, but I didn't have one sitting here.
Another important point is that D2 should be a fast-recovery diode with a reverse-recovery time of about 200nS or better. This diode limits voltage spikes from the motor on the 'drain' of the MOSFET to a safe level of about 1V more than the supply voltage. Without this protection diode, the MOSFET could quickly fry.

 

[AussieSteve]

Having said all that, there are a large number of dedicated DC-DC controller chips around. If you're building a permanent circuit, it's better to use one of those.
I just used the pair of timers because it was easy and they were sitting here.
The best circuits use an inductor for energy storage, output smoothing and higher efficiency.
Check out MC34063A, (my favourite for low-power circuits), and the TL494 and MAX787. There's tons more around, too.
I find that a search for 'TL494.pdf', for example, will yield the datasheets, but otherwise I have most older ones here that I could copy/scan and email if needed.
(Some sites charge for the datasheets, but if you look a little further down the results you'll usually find a direct link to a pdf file.)

 

[Terminal Velocity]

switching regulators

does "PSU" stand for power supply unit?

When I was looking over the schematic last night I had noticed there was no inductor. I was also surprised to see the 555's because I thought the controller chip wold have to be programmed to automatically adjust the duty cycle as the input voltage changed to keep the output voltage constant. Also, what is IC1 7812? It looks like you use it to power the 555's.

I am looking to build a dedicated circuit. At least for now, this is a proof of concept for me. I enjoy using my MB to commute to work but its a long ride (upwards of two hours). Needless to say, it is EARLY morning and very dark and often quite cool. Putting on more clothes just doesn't provide the warmth when you're immersed in the cold for 2 hours at nearly 30 MPH (55 miles). SO...I've investigated the possibility of using 12v heated clothing. This would require quite a bit of power and I don't even know if I would do it. One thing is certain, I WON'T do it without generating power on site (I don't want to imagine the battery requirement). Therefore, my first step is to manufacture a reliable power supply capable of running at least my headlight with a lot of room to spare. I suspect it would even be possible to use two separate power supplies tied to the same generator to power to separated circuits...which is fine, but they still each need to give me about 8 amps in a worst case scenario...a lot to ask of a small motor. Sorry, I digress...

I've looked at (or at least I am trying to understand) dedicated switching controllers. Few of the data sheets indicate the maximum power rating, yet they'll give an amp rating of say 5 amps. I've been to national semiconductors website since they have some webtool which is supposed to help you design a circuit based on their IC's. Basically, I can't seem to do what I want. Its as if the higher amp rating is only for lower voltages which I suppose is likely. So my quest becomes, how do I build a circuit for a high current application? I checked out the TL494 but the data sheet indicates Iout maximum is 250mA. The Max787 is no longer available but it was only capable of 5 amps at 5 volts output. I KNOW there are TONS of chips around, but I don't know how to find or at least assemble what it is I'm looking to do . I found a website using the LM2576 switching controller (there's a whole family of "simple switching" controllers). Indeed they look simple but...

In short, where might I be going wrong when looking for controllers? Is it possible I'm asking to do something for which there is no design? I'm looking to build a dedicated circuit but I would be happy to use your design (with the lower efficiency) if it means I actually have something that works. Thanks for your help with this so far. Believe it or not, you have been a big help.

 

[AussieSteve]

As I said, the circuit I posted is only a temporary motor speed controller, not a battery charger/regulator.

I mentioned current-limiting but forgot to mention closed-loop voltage control because this circuit wasn't designed for that. It was designed to set the RPM of a DC motor, in conjunction with an optical tacho. It does that beautifully. I added as many notes as I could think of.

If you don't know what a PSU is, I suspect that you might not have enough electronic experience for this project. It might possibly be better to buy ready-made stuff.
You got it right, though, PSU is Power Supply Unit.

All of the chips that you've looked at can be used merely as the driver, to drive a MOSFET or bi-polar power transistor and in turn either a flyback-style inductor or a transformer, depending on needs. The datasheets mention this and give circuit examples. For higher power circuits, you need to learn a little about half-bridge drivers and transformers for any decent efficiency.
An example of this is in my simple circuit - the 555 timer can only source about 200mA, so it drives the MOSFET to switch 8A+, 40 times the current.

For motor drivers or lighting-only applications, no inductors are needed - the inherent inductance of motors smooths the voltage and lights will flicker too fast to be noticeable at 20kHz.

Also, a bit of a look at 'switched capacitor' power supplies will show that inductors aren't necessarily needed in any application for a smooth output.
In fact, they will probably be the main switching PSU in the future, as switching frequencies increase. (The higher the frequency, the smaller the caps and coils need to be to do the same job.)
Even my circuit produces a pretty smooth output if an RC filter is used on the output, or even just a large capacitor. (Not needed for DC motor driving.)

All you need is a comparator sensing the output voltage and connected to the reset line of the second timer, to complete a closed-loop voltage regulator. (In the circuit I posted, the (LM311?) comparator would be set up to compare the output voltage to a pre-set reference voltage, then pull the second, (IC3), 555's reset low if the sampled voltage began to increase above the reference.)

Out of pure interest, what were you planning for a circuit board?

2 x 8A @ 12V, (14.2V if charging a lead-acid batt), (200-230W), is a big ask.
Have you decided on a generator and what battery did you have in mind?
I must point out that a 200-230W load will rob your engine of about 10% of it's power.

The 7812 is a linear regulator, 12VDC.
(The 78xx series is the most common series of regulators on the planet.)

I don't mean to sound rude, but I really do think that this might be beyond you.

I could design and test a circuit, but how much time should I put into MBc replies?
(That would take many hours.)
And how much time should I put into helping potential competitors, if my system works as I expect and I decide to start selling them?
Even if I did put in the time and post a good design, some p&$@# would copy it and use it commercially. (That's happened to me in the past.)
When I do design my regulator for the system I'm designing, it will be mine.
Sounds greedy, I know. That's life. (Wasn't meant to be too easy.)

I've tried to cover all of your points, but may have missed a couple.

If you're keen enough, PM me and we can talk. I'll help a bit if you really must design and build your own.

After all, this is a motorised bike site, not an electronics forum. I've already posted too much technical BS.

Sorry about all of the edits, Terminal_Velocity, I keep thinking of more to add and some to remove.

 

[Terminal Velocity]

generator

Sorry about that...I saw but didn't notice the "motor control" when I read your post and looked at the schematic. The circuit actually makes a lot more sense in the light of that.

Now see--I new it could be done. I must sound like an idiot at this point. I'm actually not stupid and I am fairly well educated (undergrad degree in physics). My problem is I don't have any education in electrical engineering and a lot of the terminology and methodology escapes me. I could imagine based on my internet research there is a way to relate so there wasn't so much power through the IC, but as stated earlier these things escape me. So I'm educable and I enjoy the challenges and so I set out to at least try things on paper. I certainly understand your concern with people copying your stuff for commercial purposes. Believe me, I have no grandiose ideas of manufacturing and selling anything and i would never be likely to buy one. I'm just a utilitarian who likes to make things myself when I can because I enjoy the challenge and then the reward.

Question: does this statement

"All you need is a comparator sensing the output voltage and connected to the reset line of the second timer, to complete a closed-loop voltage regulator."

refer to the schematic you posted or the other IC's i've been looking at? Your reference to the second timer leads me to believe it is your circuit but I can't quite tell.

my power requirements right now are pretty small...about 3 amps to power the headlight. I don't have a need yet for more power. But I do want the OPTION for more power. I figure it is easier to design for that from the outset so I don't waste money making smaller circuits to just throw them away later. In response to your specific questions:

Circuit board: haven't really thought that far ahead yet but would probably use a PCB to mount my components unless there is good reason not to. I've made a few home brew boards before and have some experience soldering smt components which I prefer to drilling holes. I like the space savings and I think they're easier to solder than through hole.

Generator: ultimately if I go to high power I was thinking about using RC brushless motors since they come in low kv and high amp ratings. Currently I plan on using an old heater blower motor from my truck (its still good) just as a proof of concept because it is free and I have it. I'm not going to spend money on a good motor until I know everything else works. Besides, my power requirement right now is only a couple amps (2.5 I think at 12 v). As far as driving the generator, I use a staton gearbox and so driving from my chain isn't even an option. It would have to be friction drive and I'm ok with that. I'm currently running everything from a 12 Ah SLA which gives me enough juice to get to work and with a little charge safely back home. I don't think the battery I plan on using is too critical because nearly the entire time I ride I am at cruising speed. There are only occasional stops when I get closer into work so I anticipate little "charging." I also figured I could run the generator at exactly 12 volts and never have to worry about overcharging...if need be I can charge the battery from the wall. I really will be running almost entirely from the generator almost all of the time...the high power requirement is only a winter thing IF I chose to get heated clothing. I may decide to never do that and ride in fair weather only...but I still want a generator which can power the headlight.

I've considered the tax on my engine...which is why this is all hypothetical for me. I want to prove the concept to myself in small steps. Step one, doable regulator. Step 2, mount the generator. Step 3: find a more appropriate generator when the need arises, etc., etc.

You've given me a lot of technical insight to think about for the project and I appreciate the time you've put into this discussion so far. We are all busy and I was trying to get some direction when I found your post. Potential competitor I'm not which is something you would have to take on faith if you choose to believe me.

I've enjoyed this exchange and I hope that I have not somehow offended you with my questions. I appreciate your honesty and I look forward to your response. Thanks again for your assistance.

 

[AussieSteve]

The comparator statement refers to the schematic that I posted.

Wow, I can't work with SMD components. My eyes aren't good enough any more, so I stick with through-hole for my boards. I only do it for a hobby now, stopped working a few years ago.
The generator I'm using is a brushed DC motor, for simplicity. It produces DC.
I assume that the truck blower motor is DC, so it should work. Getting the right gearing is important, too.

'Exactly 12 volts' isn't suitable for a lead-acid battery. (That's actually discharged, even standing they produce about 12.6V open-circuit.)
You need to charge to between 13.8V and 14.4V, to prolong battery life.
Cars typically charge to 14.2V or thereabouts.

Current-limiting is still important, too. Otherwise, if the battery voltage is a bit low,
too much current can flow through the MOSFET and burn it up.

 

[AussieSteve]

To update, my optical tacho arrived a few days ago and I realised during testing that my motor/generator was rated for 18,000 rpm and not 10,000 or less as I'd hoped, (I'd spun it up to about 23,000, it seems, thinking it was doing about 10,000, before I got the tacho).
Got another motor on the way, splashproof, twin sealed bearings and rated for 100W and 2500 rpm. Meantime, I sit on my hands again for a few days. Gives me a little time to consider what I'll use for a regulator - leaning toward buck-boost for overall efficiency, but it might be overkill.

 

[AussieSteve]

Got the new generator and set up a test rig.
In the first tests, I've had no trouble pulling 66W at 16V and it can produce 110W output at 12V, doing 3000rpm. (I'll never draw anything like that much, so higher voltages will be available.)
At 3000rpm, the short-circuit current of this generator is 21A and the open-circuit voltage is 23V - almost perfect for the application.
Now it's onto the regulator and the harder part, trying to find a suitable headlight.
I've ordered a cheap scooter regulator to see how it goes before re-inventing the wheel, but they have a built-in rectifier - not needed in this case because the generator produces DC. The inherent rectifier just 'steals' about 1.6V.
In the pic of the test setup below, the bulb is a 12V 21W indicator bulb. The generator is producing 13.9VDC at 2200rpm.
I was surprised at how fast the 'shutter speed' of my cheap little camera is, to freeze the drill chuck and coupling so well. The 'coupling' is a piece of 1/2" garden hose, slit to fit over the generator's timing belt pulley and to provide a 'universal joint' effect.
The drill is variable rpm, so I can use the optical tacho to set exact revs for each test.
It's all bolted firmly to a sheet of 5mm perspex. (There's a lot of torque involved in testing, especially high rpm short-circuit current measurements.)

(I couldn't reduce the pic's size without losing detail.)

 

[AussieSteve]

Got the headlight/battery/switch finished today.
Headlight is 2 x 18W/18W bulbs, battery is 12V 5Ah, switch is a motorcycle handlebar-mount type.
Works great.
First I'll add a horn, then back to the generator/regulator.
Some pics here:- Pics of mounted headlight