I have a project, where I want to get USB 5V out of a dynamo.
Its a standard 6v/3W dynamo.
I can make the switch mode regulator circuit, there are loads out there. And I can rectify the AC from the hub to DC. But, if there is no load on the circuit and I ride over 25mph, the dynamo can produce over 100v. The 5V regulators I’ve seen, has a Vin MAX of 60V.
I would like to limit the voltage after (or before) the rectifier, but keeping the circuit as efficient as possible, as its my spindly legs that’s powering it!
I have seen a few circuits with 2 zener diodes in parallel to Vin. Seems simple enough, but they don’t look all that efficient. The other reason I’m asking on here is that the posts are often over 6 years old. Has the tech caught up and is there a more efficient method? Preferably something that doesn’t get hot.
If not, can I get some help in selecting the correct zener for the job.
This isn’t website design. Six years is not very long
You can do a few things:
1 Keep a load on the circuit. Not unusual, many PC supplies require a minimum load to operate.
2 Clip the voltage with zeners. Nothing inefficient about that. The zeners won’t breakdown until above their threshold and the leakage current is minimal.
3 Design some wildly overcomplicated circuit to disconnect the output if the voltage goes over a certain maximum and reconnect it when it goes below that maximum.
Back in the days of Heathkits my motto was “Be Prepared”, no actually it was don’t build it if you can “buy it” cheaper and better. Their are plenty of things to build to learn from and I don’t think this one will be cost effective In the long run and you damage an expensive piece of equipment.
Yes, keeping a load on the circuit would be ideal, I don’t want to rely on that. During the day, the Dynamo will be charging a Li ion cache battery and whatever GPS is running. If both are full, there is no load. At night, a light will be on so all is good. Also having a constant load that isn’t contributing to something is a waste.
I have seen a wildly complicated circuit to automatically disconnect, think it’s in the second link above. Not for me!
So I suppose it leaves the zener. When you say breakdown, I just want to confirm it doesn’t actually destroy the zener? I’ll be on long multi day rides, don’t want to have to get the soldering iron!
So, can I get a hand in selecting the correct zeners please. Max voltage will be about 100v, I’d like to limit it to less that 50v. I’ve read that having two in series, with both in parallel with Vin is the way to go. Would each diode have to be able to limit the voltage down to 50v, or would they do it as a pair, ie. each one dropping 25v.
If the fail, do they short circuit or open.
Sorry for the questions, but the jargon that comes with diodes is new to me.
oldtemecula:
Back in the days of Heathkits my motto was “Be Prepared”, no actually it was don’t build it if you can “buy it” cheaper and better. Their are plenty of things to build to learn from and I don’t think this one will be cost effective In the long run and you damage an expensive piece of equipment.
Christopher[/quote]
Sorry but that option isn’t for me. This site is about learning.
Cheaper? That one that’s in the link, looks horrendous and the reviews prove that. You said Be Prepared, just like the Boy Scouts. That doesn’t mean taking an inferior piece of kit out on the road with no outside help available. It means knowing your kit inside out and knowing what to do if it goes wrong. I will be testing my circuit to failure so I understand it and what to do to put it right. And $70, I can make a perfectly good one for less that $10 I reckon. With my 3D printer, the housing can be bespoke to attach to a certain part of the bike, not a green lump like the Amazon one.
Luckily the components that go bang are very cheap, just the voltage regulator. Just a pain to replace though!
There is one, Sinewave Revolution, that gets great reviews but it’s £100. So until my circuit doesn’t work and I’ve exhausting all options, I won’t be buying the
Zener breakdown is the point at which the zener begins to conduct; it doesn’t mean failure.
However, by putting zeners on the output, you are still putting a load on the circuit. The difference is that now there is a threshold where the load switches on. That threshold is roughly the zener breakdown voltage.
It occurs to me that perhaps the simplest solution is a neon light across the output of the alternator. Something like [this might work. No guarantee (been years since I played with neon), but I think it will. The neon will clamp the output of the alternator to about 28V and that can be rectified and sent into your regulator.](6838 JKL Components Corp. | Optoelectronics | DigiKey)
I always think, “The only thing worse then investing a lot of time in a build that works poorly… is having a build work perfect out of the gate and the builder did not learn anything.”
I have never heard of the device you want to design until today as I am old and retired. What sparked my interest was your wanting to do it efficient and not generate wasted energy to heat. The dynamo you say can get up to 100v when unloaded and this will become wasted energy if you do not have an efficient method to transfer this extra energy into a battery for storage. Using a poorly designed circuit with a drop down resistor or zener diode will both dissipate the extra as radiated heat, similar to when you brake on your bike.
I enjoy exploring new ideas and am happy to see you undertake your exploration.
The more ideas you view the better the approach in end.
lyndon:
Zener breakdown is the point at which the zener begins to conduct; it doesn’t mean failure.
However, by putting zeners on the output, you are still putting a load on the circuit. The difference is that now there is a threshold where the load switches on. That threshold is roughly the zener breakdown voltage.
It occurs to me that perhaps the simplest solution is a neon light across the output of the alternator. Something like [this might work. No guarantee (been years since I played with neon), but I think it will. The neon will clamp the output of the alternator to about 28V and that can be rectified and sent into your regulator.[/quote]
That’s the first time its been explained in a manner I understand, thanks! Using the neon as an example helps me understand it too.
I’ll go with the zener. So if I choose a zener with a breakdown voltage of say 50v to protect the components after it I should be alright? I’ll buy a few and put a sacrificial load after them and see what goes bang first.
I always think, “The only thing worse then investing a lot of time in a build that works poorly… is having a build work perfect out of the gate and the builder did not learn anything.”
I like that statement! Very true.
I don’t mind some wasted energy when I’m hitting 25mph, but really want to avoid loosing too much at slower speeds. But then as a percentage of total energy, I doubt it’s much at all and I probably wouldn’t notice it. Especially as my bike will be laden with kit!
It’s a bit more complicated than that. The zener will load the circuit when it switches on, but it has almost zero resistance, so it will sink all the current the dynamo puts out and likely turn into a short lived LED!
I’m trying to think of a simple solution to this (short of a constant current load) and drawing a blank…
[edit]
One thing I’d suggest is that you put a series of small value resistors on the dynamo output and measure how much the voltage drops. If you get 6V with a 3W load, and 100V with no load, I’m guessing that a small load of a few mA will drop the voltage to a safe region.
Yes, I hope there will be next to no current if the dynamo manages to produce those high voltages. I’ve been chatting to a guy who makes and sells these sort of circuits. His have moved on and are now all integrated and digital. His old ones used an AC zener in parallel to the output of the dynamo. And in series with the dynamo, a thermal switch, which opens when the zener gets hot. It should hopefully only get hot when nothing is connected, so I shouldn’t notice it at all.
I’ll certainly be testing it A LOT before I plug in the other bits!
I think someone above mentioned using a current source setup and my thinking at the moment also favors this. Check out this webpage because I like what I see.
