Power holdup capacitor

[Smoke_Maker]

Here's the set-up, I have a circuit in some vehicles that needs to be powered for about 10 minutes after the ignition is shut off in order to refuel at the gas pump, the unit draws 15mA. I had the unit connected to the vehicle battery oringally and powered all the time but this proved to be too much of a draw for some of the vehicles with a small battery (Honda).

My first try was 4 (5.5 volt supercaps 1F) and a 1N4002 diode, huge inrush current, This was my knee jerk reaction try. For the second try I'm going to ask, am I on the right track or is there another solution I'm not seeing.

Thanks

[haklesup]

you're better off powering your circuit through a relay and putting a timer circuit of some sort on the relay coil. Simplest might be based on a 555 chip and transistor to interface it to the relay coil. Choose a relay with a low coil current or you'll make it worse. use the CMOS 555 and the power will be low, its output can also drive a larger load to vdd or ground.

[Chris Smith]

I would just use a common set of AA Rechargeable batteries for this charge.

Diode [s] included.

Nicads, NiMh, etc.

15 ma is nothing and any small battery pack will do the trick.

And if left unattended it will go dead all by it self and still recharge later.

[jollyrgr]

Try what Chris suggests.

Instead of powering the device directly, why not use a capacitor in an RC circuit to drive a CMOS analog switch/relay?

[philba]

the problem with a battery based solution is that the automotive environment can be pretty unfriendly to batteries. high heat could be the cause of premature failure. I'm not sure your customer want to have to change rechargable batteries in your device.

I would use a variant of hackelsup suggestion but connect to both the always on battery circuit and switched power. use a Pmosfet to turn off the unswitched draw after 15 minutes and the switched power to turn it back on. pretty simple and, depending on the mosfet, very little long term draw.

[Chris Smith]

I would try it for the cost first, 5 nicad packs are cheap?

2.4V 15MAH NIMH BATTERY PACK

CAT# NMH-15 [all electronics] 15 ma at one hour

$1.25 times 5 packs.

or

NI-CAD PACK, 3 2/3 AA CELLS, USED
CAT# NCB-23U 1.00 each [325 mil?]

or

AA NI-CAD BATTERY
CAT# NCB-AA $1.25 each [?750 mil?]

[Smoke_Maker]

Thanks for the ideas, I have tried these
http://www.chargeguard.com/what_is_charge.html
Charge Guards, but they draw 18mA which is worse. The cars parasitic draw is 7.8mA without my stuff on it. A perfect solution would be uA or less and I would like to find something already on the market if possible.

The batteries sounded like a less reliable solution for a 6 year vehicle cycle but hasn't been eliminated from a possible solution. For a KISS battery solution maybe 7 nicads packs for 16.8 volts and only charge them to 80% or there would have to be a charge supervisor circuit.

That's why I jumped to supercaps, has anybody seen a relay or relay circuit that will time out after a set time and draw no current when the relay is off. One of my first thoughts was to put the supercap in the relay coil circuit but that was just a added layer over what I need, I think the relay coil would draw more current than my unit.

[Smoke_Maker]

jollyrgr wrote

why not use a capacitor in an RC circuit to drive a CMOS analog switch/relay?

Tell me more about this circuit I'm not familiar with CMOS analog switch/relay, is it simple.

[dyarker]

Skip the relay, and just use the 555 and transistor. It is only a 15mA load!

Two diodes one from key ACC terminal, one from timer circuit to form an OR gate to load. Also, timer circuit should power itself sort-of, so at time-out it shuts itself off too.

[Chris Smith]

A CMOS 555 could run a timer, which switches a simple fet switch or a low power relay.

Low power and consumption.

[jollyrgr]

I read the comments to my suggestion and then re-read your question. I thought your current circuit was using 15mA constantly (24/7). And you wanted a circuit that would use less power to monitor the "shut down" time at the pump. Your circuit only draws 15mA and you only need it to be ON for 10 minutes. For some reason I thought you were using this circuit to drive a high current device such as a fuel pump.

A 555 timer will do what you want. But in answer to your question about the CMOS switch, here is an example of what I was thinking of:

http://www.aaroncake.net/circuits/relaytim2.htm

This uses a CMOS NAND as the active component, which would work. I was suggesting using an analog switch, such as the 4016. Here are some examples of circuits using the 4016.

http://www.doctronics.co.uk/4016.htm

From the first link pressing the button charges the cap C1. For your application the +9V would be an always on 12V circuit. Instead of the switch you would charge the cap from a "ignition on" source. When turned off the cap would energize the control for the analog switch (or inputs to the NAND gates as shown at the link). There are dozens of analog switches; the 4016 and 4066 being a couple. There are also single package devices but these are mostly low current devices.

But the 555 timer route is the likely way to go. Here is an example of such a circuit:

http://www.aaroncake.net/circuits/relaytim.htm

Replacing C1 with a 470uF 16V unit should give you ten minutes of on time. Replace the relay in the above circuit with your 15mA load/device.

[MrAl]

Hi there,

Since your circuit only has to run for 10 mins you could use the
relay idea and let it stay 'on' for 15 mins. As said, this requires
some sort of relay and a timer.
If you decide to use the battery idea, use NiCd's as they are
more robust in terms of the way they can be charged. A resistor
and diode and your cells are charging. You'll still have to rig
up some sort of relay that keeps the circuit off while the cells
are charging though.

[Smoke_Maker]

I see what you mean about the CMOS and the 555, I'm new at this so bear with me. I got the data sheet for the 555 and it shows the low state current as 10mA, am I reading this right and is this what it will draw in the off state?
http://www.semiconductors.philips.com/a ... 55_C_2.pdf
I think I can see how a 555 could be used, nope I was wrong, I drew out the circuit and I can't figure out how to stop the parasitic drain of the 555 it's self. I'm hoping that the 555 draws uA when off, I'm going to bread board it tomorrow and measure it.

Could I put a RC circuit in the base circuit of a transistor like the 2n3904 or a FET IRFZ44? I don't think I have ever seen this before.

This might work! low part count, source 100mA to be safe, bullet proof, potable with flying leads, and only 499 to make.

[MrAl]

Hi there,

There are CMOS versions of the 555 which draw less current.
You dont have to use a 555, you can use cap and comparator IC
like the LM339.

[Smoke_Maker]

I like Google, I kept typing in time delay relay but it worked much better with time delay circuit. This was what I was thinking of, running right off the ignition switch. Simple, low part count, and nA leakage draw.

I need something that is dam near bullet proof, now the question is which way to go 555 or transistor, following that decision is how do I protect it in the harsh automotive environment.