Dynamo Overvoltage Protection

[ConRAD]

I know that a specific chapter of German StVzo, I guess TA 24, specifically requires some overvoltage protection to be incorporated onto a dynamo; what isn't clear at all is why this protection is generally built-in on sidewall dynamos and not on hub dynamos ??

[andrew_s]

I think it's that there's a top limit to the permitted voltage out to the lights (i.e a 12 ohm resistive load, such as cycle light bulbs), regardless of speed, and that the design of hub generators is such that they don't need anything extra to keep the voltage within limits.

[ConRAD]

The only limit that I know for hub dynamos is the current and not the voltage. Indeed current is normally topped up to something around 500 ... 700 mA. Voltage converserly varies linearly with speed and can easily surge up to 100 V at 100 km/h!!
But what about sidewall dynamos, shouldn't they work more or less the same ?

[the snail]

My guess is that it's just difficult to fit the regulator into a hub. I think a lot of led lights have over-voltage protection built in anyway.

[Brucey]

there might be an implicit assumption that you will only be running a sidewall dynamo with the lights connected (which is not unreasonable....) and that -barring fault conditions- the voltage is thus limited by the lights themselves.

I have had no reason to try it, but I would be surprised if bottle dynamos don't also generate quite high voltages if driven 'open circuit' at high speed. Also, the few I've dismembered have had no regulation components within them, so if the voltage is limited, it may be done using the design of the magnetic circuit.

cheers

[andrew_s]

The STVZO regulations are written for old style incandescent lighting and sidewall generators, with amendments made for more recent technology.

The voltage regulations are shown on this diagram, in the LH chart (basically, it's no more than 7.5V, at least 3V at 5 km/h, and at least 5.7V at 15 km/h, into a 12 ohm resistive test load).
The regulations have nothing to say about what might happen with a different load, or no load.

A tyre driven generator has a linear relation between road speed and generator rpm, regardless of wheel size, but a hub dynamo spins faster at the same speed with a smaller wheel, so STVZO fixed their rules by specifying a range of allowable wheel sizes for an approved hub.

Modern dynohubs keep within the allowed range using the design of the magnetic circuit, but this may not be the case with other generators. These may therefore have simple back to back zener voltage limiters, or there could be quite a lot of electronics inside (Lightspin)

[ConRAD]

... modern dynohubs keep within the allowed range using the design of the magnetic circuit...

I don't think so.

- modern dynohubs loaded on a 12 Ohm resistor normally can give something around 6 V at 20 km/h on a 28" wheel

but also

- modern dynohubs loaded on a 40 Ohm resistor can easily reach something around 15 V at 20 km/h on a 28" wheel

so there's no magnetic circuit self-limitation design I guess !

[andrew_s]

... modern dynohubs keep within the allowed range using the design of the magnetic circuit...

I don't think so.

- modern dynohubs loaded on a 12 Ohm resistor normally can give something around 6 V at 20 km/h on a 28" wheel

but also

- modern dynohubs loaded on a 40 Ohm resistor can easily reach something around 15 V at 20 km/h on a 28" wheel

so there's no magnetic circuit self-limitation design I guess !

But STVZO don't test with a 40 ohm resistor.

Hub dynamos behave largely as a current source - the magnetic circuit design limits output to about half an amp or a bit more, if you are going fast enough, though whatever load you put across the terminals.

12 ohm & 6 volts is a half amp.
40 ohm and 15 volts is 0.375 amp - with a 40 ohm load and sufficient speed, you'll get up to 20 volts, easy.

Back in the day, before white LEDs had been invented, when using 2 standard filament bulb lights in series (24 ohm) "fast enough" was 13-14 mph, to give the 12 volts needed to power the two lights fully.

[ConRAD]

... STVZO regulations are written for old style incandescent lighting and sidewall generators...

Even worse I'd say. They are complex, of difficult interpretation and sometimes definitely obsolete. And on top of that they written in German
Yes, basic StVZo-TA24 is still based on incandescent lights, 20 km/h or something like that. They are based on sidewall dynamos and based also on a general assumption that the "normal" cyclist was supposed to go at an "urban" standard speed of 15-20 km/h.
The original purpose of StVZo, I guess, was to arrive to a sort of standardization according to which generators (sidewall type) were supposed to be designed to deliver 3W-6V on a 12 ohm at 20 km/h.
In the most common configuration two lights were supposed to be connected (IN PARALLEL):
- one front light rated 2.4W (i.e. 6V on 15 ohm), and one rear light rated 0.6W (i.e. 6V on 60 ohm).
- at that time sidewall dynamos were not overvoltage protected so StVZo was recommending to install a protection at least on rear light simply because in case of failure of front light the rear one was suffering a quick overvoltage therefore burning in turn.

Nowadays all this sophisticated electrical philosophy simply has no reason to exist anymore.

I agree on the fact that to have a sidewall dynamo overvoltage protected is still good, so you'll still have the protection "present" only when the dynamo is activated.

Different story for the hub dynamo, it would be a nonsense to have the protection built-in because in this case, lights-on or lights-off it wouldn't make any differerence, the hub is spinning all the time you know and as such the protection would uselessly overheat to dissipate the excess voltage.

[Brucey]

I wonder when this standard was written and also if it was ever adhered to in all respects...? I don't recall seeing an overvoltage device in any of the (many) basic German-standard dynamo light sets I have seen. It is quite normal for the rear light to blow shortly after the front fails.

Also I note that there is arguably a point to having an overvoltage device even with a hub generator; without it, when the front light is switched 'off' , at speed, there can be many tens of volts on the generator, and the wiring running to the switch on the headlight, all of which could be damp, too. This poses a much greater challenge to the insulation values of these parts vs any kind of normal operation with the lights on. If there is a fault, or the switch is operated when descending at speed, an electric shock is a genuine possibility, and if it occurs, a real hazard to the safety of the rider. [I note with interest that more recent shimano hub generators (even those with the same model and part numbers as previously) have moved to a design that has both terminals insulated from ground, which makes ground return wiring less practical, and presumably electric shocks less likely too.]

If there was a device that always limited the voltage to some kind of notionally 'safe' level, it might (depending on the voltage threshold chosen and the method of limiting) only create significant extra drag at speeds which would only be achieved by most riders when descending?

cheers