12v regulated supplies

[Tusses]

ok - had a little think ... its been about 20 years since I tinkered with electronics so I will need to do my homework, so bare with me

 

1st thoughts are....

(figuers rounded up for ease)

input say 11v to 15v output would be 11v to 13v

 

the cuircuit would let the input take 2 routes to the output

 

Route 1 would be batt voltage available upto a set figuer when it would switch off say 12.5 v

Route 2 would be limited - ie a zener / power transister which would let the voltage rise to -say 13v- and no higher.

 

so both routes connected to the output unti the V goes high when the 1st one switches off (somehow ! ) shortly followed by the second one limiting the voltage.

 

got to do some reading up on SSR's and FET junctions etc to see which have the lowest drop accross the junction - or maybe use a relay in route 1 as when it switches in and out, the voltage difference between route 1 and route 2 may be close enough so as to not get a switching spike.

 

 

thats my 1st 'quick' thoughts on it - feel free to tare it apart !

[Ex- Member]

Yeah I guessed he meant alt reg too. This guy does actually munufacture the bulbs, as well as sell them so when he says going over 12 volts is damaging to the life of the bulb i see no reason to disbelieve him, so is it possible to generate a steady 12 volt supply for the bulbs if that indeed is the case? I would prefer not to shorten the life of bulbs this expensive

[chris w]

Yeah I guessed he meant alt reg too. This guy does actually munufacture the bulbs, as well as sell them so when he says going over 12 volts is damaging to the life of the bulb i see no reason to disbelieve him, so is it possible to generate a steady 12 volt supply for the bulbs if that indeed is the case? I would prefer not to shorten the life of bulbs this expensive

 

I installed some LEDs to illuminate the staircase into the boat at the stern and simply used a 12v, LM7812 regulator.

 

Chris

Edited April 1, 2008 by chris w

[chris w]

This is another one that comes up often. As is this often shown "solution".The above circuit will NOT work. The 78xx regulators need an input voltage at least 2.5 volts higher than the output. So a 12 volt one will not work until the input voltage is above 14.5 volts. With an input voltage of (say) 12 volts they simply do not work. They produce an output of a few volts.Gibbo

 

If one absolutely needs a constant 12v then you are correct and an SMPS is the solution.

 

However for stabilising LED voltage we are not really so concerned with undervoltage as ensuring there is no overvoltage. Afterall, if one has incandescent lamps in the boat (tungsten or halogen) the voltage on them is not constant either but we don't notice the subtle changes of brightness from moment to moment, as other equipment switches on or off and/or the batteries lose charge and the terminal voltage drops.

 

The point is that a 7812 type regulator WILL work at input voltages under the minimum specified and will still produce a healthy output voltage. So, for a typical 7812 regulator the minimum input is usually specified as being around 15v to ensure 12v out. However, the output will remain within a few tenths of a volt of 12v down to about 13.5v input. Thereafter, the output will drop very linearly against input voltage, with the output remaining at an almost constant 1.5v below the input voltage. This means that even at 12v INPUT, we will still get 10.5v OUTPUT which will still be perfectly bright for LEDs.

 

I use this system on my boat to run a string of LED lamps, viz: a simple 7812 regulator at about 60 pence each and a couple of caps for stability.

 

Chris

[Gibbo]

If one absolutely needs a constant 12v then you are correct and an SMPS is the solution.

 

However for stabilising LED voltage we are not really so concerned with undervoltage as ensuring there is no overvoltage. Afterall, if one has incandescent lamps in the boat (tungsten or halogen) the voltage on them is not constant either but we don't notice the subtle changes of brightness from moment to moment, as other equipment switches on or off and/or the batteries lose charge and the terminal voltage drops.

 

The point is that a 7812 type regulator WILL work at input voltages under the minimum specified and will still produce a healthy output voltage. So, for a typical 7812 regulator the minimum input is usually specified as being around 15v to ensure 12v out. However, the output will remain within a few tenths of a volt of 12v down to about 13.5v input. Thereafter, the output will drop very linearly against input voltage, with the output remaining at an almost constant 1.5v below the input voltage. This means that even at 12v INPUT, we will still get 10.5v OUTPUT which will still be perfectly bright for LEDs.

 

I use this system on my boat to run a string of LED lamps, viz: a simple 7812 regulator at about 60 pence each and a couple of caps for stability.

 

Chris

 

If you think sticking summat in there that will drop 1.5 to 2 volts is OK and good engineering practice then all I can say is "ok".

 

Considering the lengths we go to to save every last ounce of power and use huge cables to avoid excesive voltage drops it seems rather primitive to me.

 

Certainly not something I'd do. But there again, when it comes to electronics, I like to do things properly.

 

Gibbo

[chris w]

Well, perpetual motion is still on the drawing board so even an SMPS or an inverter or a DC-DC converter all have significant losses (around 10-15%), so no solution is "proper". As you well know, ANY use of a 78xx regulator will always drop 1.5-2v so I don't understand from where you are coming. You know full well that these devices are used in their millions. Every TV, radio, hifi, laptop etc used on a boat or at home will have 78XX regulators in them ALL dropping 1.5-2v.

 

This simple, very inexpensive solution solves the particular issue for which someone was seeking a solution whereby the output volts are limited to 12v maximum regardless of the input voltage (within the typical voltages found on boats, for the pedants).

 

 

 

Chris

Edited April 1, 2008 by chris w

[smileypete]

I've read this topic from the beginning, but you electrical boffins have lost me. I run all LED mr16 bulbs. The first lot i bought blew numerous led's in each bulb after a few months, bloody annoying at a fiver each

 

A couple of months ago I bought another 20 x 12V 3W HIGH POWER LED soft white single bulb led from e-bay been in the boat 6 months, but one blew last weekend, and i did have the engine running with lights on. Been on to another e-bayer who sells 30 led bulbs, the guy states that they can only run from a DC transformer, so I emaild him with regard to using them on the boat, the guy was very honest and gave this reply.

 

 

Hi Julian

Yes the bulbs are voltage sensitive they will cope with about 15 volts DC but anything above 12 volts DC will affect their life, the problem with boats ect. is if the regulator is not working properly you can have 18-24 volts while the engine is running, and if the lights are on during this time you will damage the bulbs, the only way to check this is to measure the voltage at the bulb when the engine is running, this is easy to check by using a volt meter across the supply at any bulb when the engine is running.

If you require any further assistance please contact us.

Best regrads

xxxx xxxxxxxx

Technical Director

 

Now if I cant regulate a steady 12 volts for the lights using the 12 volt system. Would it pay to use the inverter through a transformer to regulate a steady 12 volts to mollycoddle these 12 volt Led's Or is this wasteful. The lights i have now are great, give off ample light and look good also, the 30 single LED bulbs are cheap but a bit ugly. The energy saving from Led is such that sorting something out to run them efficiently would be brilliant.

 

Hi,

 

You could use one of these:

 

http://www.amperordirect.co.uk/products/12v_Stabiliser.asp

 

It will need some basic 12v wiring competence to wire it into the LED's lighting supply, maybe someone could help if need be.

 

If all your lighting is LEDs or less than 70 watts in total, it would make it easier as the stabiliser could supply the whole lighting circuit.

 

 

The only gotcha is that it will draw a small current even when no lights are on, so need switching off overnight or leaving the boat awhile.

 

One way round this would be add a couple of relays so it's only used when the engine ignition is on.

 

 

We're just arguing the merits of one system over another here

 

cheers,

Pete.

Edited April 1, 2008 by smileypete

[Tusses]

Hi,

 

You could use one of these:

 

http://www.amperordirect.co.uk/products/12v_Stabiliser.asp

 

It will need some basic 12v wiring competence to wire it into the LED's lighting supply, maybe someone could help if need be.

 

If all your lighting is LEDs or less than 70 watts in total, it would make it easier as the stabiliser could supply the whole lighting circuit.

 

 

The only gotcha is that it will draw a small current even when no lights are on, so need switching off overnight or leaving the boat awhile.

 

One way round this would be add a couple of relays so it's only used when the engine ignition is on.

 

 

We're just arguing the merits of one system over another here

 

cheers,

Pete.

 

that looks like it will give the correct output. Shame about the efficiency tho

[smileypete]

that looks like it will give the correct output. Shame about the efficiency tho

 

That's why I suggested adding relays so it's only in circuit when the engine is/will be running.

 

 

Needs a changeover relay to select between the stabiliser and normal unstabilised 12v.

 

Then another relay to switch the stabiliser on or off.

 

Both relays would only be energised when the engine ignition is on.

 

Would also avoid the problem of any small idle current drawn by the stabiliser while all the lights are off.

 

 

For non liveaboard use (24/7 off shoreline) I doubt the efficiency would be a big problem.

 

cheers,

Pete.

Edited April 1, 2008 by smileypete

[chris w]

Needs a changeover relay to select between the stabiliser and normal unstabilised 12v.

 

Then another relay to switch the stabiliser on or off.

 

 

cheers,

Pete.

 

You don't need two relays. One relay will do it.

 

Chris

[stuart]

This all seems a bit overkill to me with relays!

 

I've got one of the Amperor units for my TV (before I realised I didnt really need one!) and they consume about 80-100mA when powered on but no load.

 

Is this amount of power waste going to cause you a big problem?

[smileypete]

You don't need two relays. One relay will do it.

True, a 2 pole changeover will do it, I was looking at commonly available car relays.

 

 

This all seems a bit overkill to me with relays!

 

I've got one of the Amperor units for my TV (before I realised I didnt really need one!) and they consume about 80-100mA when powered on but no load.

 

Is this amount of power waste going to cause you a big problem?

Probably not, though it's more efficient to run 12V loads straight from the battery when suitable.

 

That said it may only be worth doing when living aboard 24/7 without a shoreline.

 

cheers,

Pete.

Edited April 1, 2008 by smileypete

[Gibbo]

Well, perpetual motion is still on the drawing board so even an SMPS or an inverter or a DC-DC converter all have significant losses (around 10-15%), so no solution is "proper". As you well know, ANY use of a 78xx regulator will always drop 1.5-2v so I don't understand from where you are coming. You know full well that these devices are used in their millions. Every TV, radio, hifi, laptop etc used on a boat or at home will have 78XX regulators in them ALL dropping 1.5-2v.

 

This simple, very inexpensive solution solves the particular issue for which someone was seeking a solution whereby the output volts are limited to 12v maximum regardless of the input voltage (within the typical voltages found on boats, for the pedants).

 

 

 

Chris

 

It isn't the losses that are the problem so your comment about perpetual motion and SMPS inefficiency is irrelevant.

 

It's because.....................

 

A 12 volt system has a voltage somewhere between 10.5 volts and 12.7 volts during discharge depending upon what is happening. Lights etc like 12 volts. You're going to put in a voltage reg that will lose about 2 volts. Your lights are now running on 8.5 volts to 10.7 volts.

 

Gibbo

[PaulG]

At risk of getting caught in the electronic crossfire between Chris W and Gibbo, I'd suggest using a "low dropout" regulator, LM2940 or similar. These should work ok down to about 12.5 volts, or when the battery is about 40% charged (or 60% discharged, depending on your outlook on life!). I think the original question was about protecting equipment from overvoltage, so I think this would do the trick.

[smileypete]

Hi,

 

I think part of the problem is when the engine is running the alternator can produce huge voltage transients, whether a linear regulator would block them enough I don't know.

 

SMPS have some filtering built in because they use an inductor. You could add filtering to a linear regulator though.

 

 

Where I got involved was to suggest a suitable car laptop adaptor may be modified to give a 12v output.

 

Gibbo then correctly pointed out that nearly all of them cannot give a lower output voltage than the input, by design

 

 

Sorry if any non-tekkie people reading this find it boring/totally confusing!

 

cheers,

Pete.

Edited April 1, 2008 by smileypete

[Gibbo]

At risk of getting caught in the electronic crossfire between Chris W and Gibbo, I'd suggest using a "low dropout" regulator, LM2940 or similar. These should work ok down to about 12.5 volts, or when the battery is about 40% charged (or 60% discharged, depending on your outlook on life!). I think the original question was about protecting equipment from overvoltage, so I think this would do the trick.

 

They would be ok but low drop regs all suffer from 2 problems which are relevant in this case.

 

They use a PNP pass transistor (I mentioned this earlier in the thread) which allows them to drastically reduce the voltage drop. The problem is that they are very susceptible to damage from high input voltage. That particular one has an absolute maximum input voltage of 26 volts. They also don't stop voltage spikes getting through.

 

One of those would last about 10 minutes in most 12 volt battery systems.

 

Gibbo

[Chris Pink]

They would be ok but low drop regs all suffer from 2 problems which are relevant in this case.

 

They use a PNP pass transistor (I mentioned this earlier in the thread) which allows them to drastically reduce the voltage drop. The problem is that they are very susceptible to damage from high input voltage. That particular one has an absolute maximum input voltage of 26 volts. They also don't stop voltage spikes getting through.

 

One of those would last about 10 minutes in most 12 volt battery systems.

 

Gibbo

 

so, observing your replies, it would seem that we are back to the zener diode solution; does nothing, so consumes no power, below voltage, drops power over voltage, when we can afford to lose power because presumably it is being generated at that time.

[Gibbo]

so, observing your replies, it would seem that we are back to the zener diode solution; does nothing, so consumes no power, below voltage, drops power over voltage, when we can afford to lose power because presumably it is being generated at that time.

 

How much power do you think it will dissipate?

 

Gibbo

[Chris Pink]

How much power do you think it will dissipate?

 

Gibbo

 

too much teecher

[Tusses]

zener as voltage reference to drive a power transistor.

[Gibbo]

too much teecher

 

 

Work it out. Then you'll realise the problem.

 

Gibbo

 

zener as voltage reference to drive a power transistor.

 

Too much voltage drop. Same problem as the LM7805 "solution".

 

Gibbo

[chris w]

At risk of getting caught in the electronic crossfire between Chris W and Gibbo, I'd suggest using a "low dropout" regulator, LM2940 or similar. These should work ok down to about 12.5 volts, or when the battery is about 40% charged (or 60% discharged, depending on your outlook on life!). I think the original question was about protecting equipment from overvoltage, so I think this would do the trick.

 

Paul

 

12.5v is the off-load battery voltage in your example. When running loads, there will be a voltage drop across the battery's internal resistance. So although the off-load voltage may be 12.5v the on-load voltage will be lower. If you are drawing 20A for example and the off-load voltage is 12.5v the on-load (20A drawn) voltage will drop to around 12.2v.

 

Gibbo is exaggerating the 78XX voltage drop to try to score points against me for some reason. If my battery bank ever went down to 10.5v on load, my fridge would have already cut-out and my inverter too. The lowest I ever see even under heavy load is about 11.5v. Normally they are running above 12v on a typical load.

 

So the lights could get down to 10.5v (the drop on a 78XX is 1.5v not 2v as Gibbo says). So they will be slightly dimmer, but this may be better than putting 15v through them. At the end of the day it's just another one of many possible solutions with some pros and some cons like all solutions.

 

I do agree with Gibbo that the zener solution is a non-starter for the reason he articulated. I don't think the people proposing zeners understand how they operate.

 

Chris

Edited April 1, 2008 by chris w

[smileypete]

They would be ok but low drop regs all suffer from 2 problems which are relevant in this case.

 

They use a PNP pass transistor (I mentioned this earlier in the thread) which allows them to drastically reduce the voltage drop. The problem is that they are very susceptible to damage from high input voltage. That particular one has an absolute maximum input voltage of 26 volts. They also don't stop voltage spikes getting through.

 

These any good for transient protection? :

 

http://www.rapidonline.com/productinfo.asp...=61786#techspec

 

Edit: I would look at putting lower value ones in series for better power handling if need be.

 

cheers,

Pete.

Edited April 1, 2008 by smileypete

[Tusses]

zener as voltage reference to drive a power transistor.

 

Too much voltage drop. Same problem as the LM7805 "solution".

 

Gibbo

 

Did you see my post #26 on this thread ?

[Gibbo]

Paul

 

12.5v is the off-load battery voltage in your example. When running loads, there will be a voltage drop across the battery's internal resistance. So although the off-load voltage may be 12.5v the on-load voltage will be lower. If you are drawing 20A for example and the off-load voltage is 12.5v the on-load (20A drawn) voltage will drop to around 12.2v.

 

If the batteries are at around 60% state of charge and a load is put on of around 25% the battery capacity the voltage will fall to WELL below 11 volts.

 

Gibbo is exaggerating the 78XX voltage drop to try to score points against me for some reason.

 

No he isn't. See below.

 

If my battery bank ever went down to 10.5v on load, my fridge would have already cut-out and my inverter too. The lowest I ever see even under heavy load is about 11.5v. Normally they are running above 12v on a typical load.

 

So the lights could get down to 10.5v (the drop on a 78XX is 1.5v not 2v as Gibbo says).

 

No it isn't. It's 2 volts. Recommended headroom to keep away from this is 2.5 volts.

 

http://www.fairchildsemi.com/ds/LM/LM7805.pdf

 

See page 8 - "Vdrop" spec.

 

I didn't write that. The people that make them wrote it. I think it's fair to assume it's correct.

 

Gibbo