Solar voltage limiter?

[star882]

"Otherwise, you can easily put a zener diode of the voltage you wish to limit to across the output (suitable power diode of course)."
You can also use a small zener and a power transistor.

 

[Burnt_Retinas]

Correct - no voltage loss. The zener is across the panel ie in parallel with the panel. You will only drop volts if a diode is in series with the panel. You WILL NOT loose volts using a zener in parallel with the panel. This is why the shunt regulator is so popular with low power panels. There is NO voltage loss. Higher currents will need different techniques as Inretech states, but this is not the case with your mod at 50mA.

Remember, the zener must be reversed biased to work, otherwise it's just a *normal* diode and you will short it out with the batteries!

Chris

 

[Entropy]

[ QUOTE ]
Doug Owen said:
[ QUOTE ]
Entropy said:
[What he may need is some sort of switching regulator. I believe NatSemi has some flyback controller chips that are designed to work in situations like this.

[/ QUOTE ]

Not really. If you check, the original problem was worry that the solar panel would go overvoltage and kill the cell phone. Such regulators don't work well at reducing voltages.

I still think the zener's the call.

Doug Owen

[/ QUOTE ]
There are a few switching regulator architectures that work both when Vout<Vin and when Vout>Vin. Some of the NS chips were in this category, and (I believe) dat2zip's Wizard boards are a current-regulated variant of this category.

 

[Doug Owen]

[ QUOTE ]
Entropy said:
[ QUOTE ]
Doug Owen said:
[ QUOTE ]
Entropy said:
[What he may need is some sort of switching regulator. I believe NatSemi has some flyback controller chips that are designed to work in situations like this.

[/ QUOTE ]

Not really. If you check, the original problem was worry that the solar panel would go overvoltage and kill the cell phone. Such regulators don't work well at reducing voltages.

I still think the zener's the call.

Doug Owen

[/ QUOTE ]
There are a few switching regulator architectures that work both when Vout<Vin and when Vout>Vin. Some of the NS chips were in this category, and (I believe) dat2zip's Wizard boards are a current-regulated variant of this category.

[/ QUOTE ]

This much is of course, true. They are not, however, *flyback* converters. Flyback converters depend on the 'inductive kick' from that comes from interrupting the current through an inductor in series with the supply. The induced voltage is then added to the 'battery' voltage. Such circuits don't therefore work as down converters since when the output is lower than supply the transistor never switches but the inductor 'shorts input to output'.

Again, even if we used the correct type converter for this use, it would be a lot more expensive and less reliable than the zener solution. Right?

Doug Owen

 

[star882]

"This much is of course, true. They are not, however, *flyback* converters. Flyback converters depend on the 'inductive kick' from that comes from interrupting the current through an inductor in series with the supply. The induced voltage is then added to the 'battery' voltage. Such circuits don't therefore work as down converters since when the output is lower than supply the transistor never switches but the inductor 'shorts input to output'."
That's a boost converter, flyback converters are isolated designs and can step either up or down(as an example, I have a PC power supply with a flyback converter that steps 170v down to 12v).

 

[Burnt_Retinas]

Mike, Inretech,

What is a "trace controller"? I've not heard of this and would like to know more. Any links to info would be nice.

Chris

 

[Brock]

Trace is a company, now Xantrex actually, anyway they make large charge controllers and voltage regulators. Typically in the 20 to 50 amp range, way overkill and over $ for what we are talking about here. But for large setups they are a lot more common.

 

[Entropy]

[ QUOTE ]
Doug Owen said:
[ QUOTE ]
Entropy said:
[ QUOTE ]
Doug Owen said:
[ QUOTE ]
Entropy said:
[What he may need is some sort of switching regulator. I believe NatSemi has some flyback controller chips that are designed to work in situations like this.

[/ QUOTE ]

Not really. If you check, the original problem was worry that the solar panel would go overvoltage and kill the cell phone. Such regulators don't work well at reducing voltages.

I still think the zener's the call.

Doug Owen

[/ QUOTE ]
There are a few switching regulator architectures that work both when Vout<Vin and when Vout>Vin. Some of the NS chips were in this category, and (I believe) dat2zip's Wizard boards are a current-regulated variant of this category.

[/ QUOTE ]

This much is of course, true. They are not, however, *flyback* converters. Flyback converters depend on the 'inductive kick' from that comes from interrupting the current through an inductor in series with the supply. The induced voltage is then added to the 'battery' voltage. Such circuits don't therefore work as down converters since when the output is lower than supply the transistor never switches but the inductor 'shorts input to output'.

Again, even if we used the correct type converter for this use, it would be a lot more expensive and less reliable than the zener solution. Right?

Doug Owen

[/ QUOTE ]
The circuit you describe is a boost converter (as used in the BadBoy and MadMax boards)

As another poster mentioned, a flyback regulator has an output that is isolated from the input, and can operate with the output voltage inside the input voltage range. (This is excellent for producing regulated +12V from a car BTW.)

I would consider an LM2585-based circuit to be more reliable than a zener - Depending on the solar cell, that zener may have to dissipate quite a bit of current. In a case where you're dropping from 8v to 6v, a flyback converter will also be more efficient.

 

[Burnt_Retinas]

I can hear what everyone is saying re switching regulators etc, and I agree a switching circuit for solar is good, as you can increase the current greater than that available from the panel when the panel voltage is above what you need, but also milk the panel for what it can offer when the voltage is below what you need, but for 50mA really? IMHO a switching circuit just isn't cost vs power effective for a 50mA set-up.

You can't beat the 100% efficiency of a shunt reg like the zener. Of course this only holds true when the power lost with a shunt reg, when it kicks in, exceeds the power to drive a switching circuit. Keep in mind that even below this level a switching reg will be burning some power!

Thanks for the "Trace" info too. Any idea what topology they use? Perhaps it's unique, or is it just that they happen to have a product that needs some peoples needs?

Chris

 

[zmoz]

Alright, I found a solar panel that is rated at 6v/200ma for the same price as the 50ma version. Will this zener diode still work fine at 200ma?

 

[Burnt_Retinas]

No. It'll cook the 1W zener when the batteries are charged, which is then the zener 'kicks in' and limits the voltage.

You'll have to use the other methods other people have suggested here.

Chris

 

[Doug Owen]

[ QUOTE ]
Burnt_Retinas said:
No. It'll cook the 1W zener when the batteries are charged, which is then the zener 'kicks in' and limits the voltage.

You'll have to use the other methods other people have suggested here.



[/ QUOTE ]

First part's true, with four times the current (and therefore four times the power) 1.2 Watts will 'cook' a 1 Watt part. But not a two Watt, or two ones in series.....

The idea still works just fine. You just need bigger parts for a bigger job.

Doug Owen

 

[Burnt_Retinas]

True Doug, thanks. I should have thought the same but didn't. I guess I was too focussed on answering the question, rather than providing an equally effective and cheap solution.

It may be worth noting, that a) if two in series are used, then they will each need to be half the required voltage ie 3V and, b) the power (read heat) still needs to be gotten rid of somehow regardless of the power capability of the part. One watt is not a lot however, so perhaps a large-ish PCB (PWB) piece of copper trace or similar heatsinking will do. Perhaps even the wires from the panel will suffice to take the heat away?.

Chris

 

[Squig]

True Doug, thanks. I should have thought the same but didn't. I guess I was too focussed on answering the question, rather than providing an equally effective and cheap solution.

It may be worth noting, that a) if two in series are used, then they will each need to be half the required voltage ie 3V and, b) the power (read heat) still needs to be gotten rid of somehow regardless of the power capability of the part. One watt is not a lot however, so perhaps a large-ish PCB (PWB) piece of copper trace or similar heatsinking will do. Perhaps even the wires from the panel will suffice to take the heat away?.

Chris

How would it differ for OP to use six 1v zeners in series across the 200ma pv output? Would increasing the number of zeners remove the need for heatsinking? I'm wanting to do this with a 12v 30-watt pv.

 

[magellan]

Great info, all, thanks!