Gate IC - 6V power - are two button cells OK?

[Richard1234]

Hi,

I've just been designing a simple little circuit with some logic gate ICs in it powered off a 6v rail - I was planning on using some button cells for this. All logic gates in the circuit specify 2-6v Vcc.

However - fresh out of the pack, the button cells are 3.2V/cell! Will I risk damaging the gates before the button cells stabilise at 6v?

I could just add a simple diode in the supply rail to drop some voltage, buuuuut.... I already had the PCBs cut.

Thanks,
Richard

[Zero999]

What logic ICs are they? Do you have their part numbers?

If it's the 74HC series, the absolute maximum rating for the supply voltage is 7V, but it's not recommended to routinely power it at that voltage. I'd probably just add the diode anyway, as they're not expensive and it will provide reverse polarity protection.

[djacobow]

Nobody can really tell you. If you violate the max spec on a data sheet, you are taking a "risk". Is it a big or important risk? Depends on the part, the process, the version, the phase of the moon, etc. You just don't know. Theoretically, you could buy a reel of a thousand parts, violate the spec on all of them and have no problems whatsoever, but then on the next reel you buy, they could all fail under the same circumstance.

That said, what is the part?

The diode drop doesn't seem great to me because you're throwing away voltage all the time, which with button cells seems wasteful.

Maybe an LDO regulator? You'll want one with an enable that can shut down to near zero power.

[Richard1234]

Thanks guys!

The gates I was planning on using powered off 6V are:
2x SCHMITT INV = NC7S14M5X
AND GATE = TC7S08FU
NAND GATE = MC74HC1G00

Each of these gates has a specified quiescent current of <1uA, so I can't see the diode being too wasteful. Admittedly the diode will mean the gates will stop functioning at the diodes forward voltage more than what it did before. However, there are other items in the circuit that will stop functioning well before this point.

An LDO isn"t an option - I want the gates powered 24/7.

Bevan

[Audioguru]

You never told us the part number of the button cells. Instead of using two in series then simply use only one cell producing 3.2V.

[djacobow]

Well, depending on other current draw in the circuit, the diode approach might not work either. Those are all CMOS parts and though the datasheets promise static current draw of uA, I've seen such chips draw much, much less, on the order of nA. This if the outputs are all going somewhre high impedance (like another CMOS gate) or are currently low. This is generally a nice thing, but in your case it is something to consider because the voltage drop across a diode with "no" current in it is much less than the 0.6V or 0.2V you might be expecting.

It's easy enough to test, of course.

[NiHaoMike]

A diode connected transistor behaves much more like an ideal silicon diode at low current than actual diodes do.

[Richard1234]

Thanks dacobow and NiHaoMike,

Yep, I figured that was going to be the case. I was looking at diode datasheets, and noticed that the Vf voltage is always specified at various current flows, and drops as current tends towards zero.

Any other ideas other than upping the current pulled to try and get a diode in a current region predictable using the diodes datasheet?

As pointed out, the diode does make sense for reverse polarity protection.

I guess I could put a resistor to ground in an attempt to pull enough current to get the voltage drop I'm after. The question is, how much current would be required? I was hoping to get the dormant current low enough that the circuit could sit unused for several years without having to change the batteries.

I might just get the next boards cut with a place for a diode, and a dump resistor to pull some current.

Thanks for the tip about transistors behaving more like ideal silicon diodes at low current. Any suggestions on a SOT-23 package (or similar) transistor to look at?

Audioguru, I need 6v for other stuff in the circuit. Cells are CR1632.
 

Richard

[Audioguru]

The Energizer CR1632 battery has detailed spec's in their datasheet. It can have a current of 0.19mA then its voltage slowly drops to 2.6V in about 620 hours (26 days), then drops steeply after that. It is not called a "button" cell, it is a lithium coin cell.

[Richard1234]

Audioguru, I'm aware of the specs of the battery in terms of the current restrictions. It's 3V @ 0.130mAh.  I'm talking about pulling <10 microamps out of it. Not 190uA.

Just because the batteries discharge curve is spec'd at 190uA doesn't mean I have to go anywhere near that.

I'd like to pull the quiescent current of the ICs out of it without nuking some components which are 6v rated before the batteries start to deplete below 3v/piece, and in turn conserve battery life.

Richard

[Richard1234]

How about this?

https://docs-apac.rs-online.com/webdocs/1040/0900766b810408e9.pdf

Vf dataplot is down to 1uA. 

Look appropriate?

Richard

[Zero999]

Audioguru, I'm aware of the specs of the battery in terms of the current restrictions. It's 3V @ 0.130mAh.  I'm talking about pulling <10 microamps out of it. Not 190uA.

Just because the batteries discharge curve is spec'd at 190uA doesn't mean I have to go anywhere near that.

I'd like to pull the quiescent current of the ICs out of it without nuking some components which are 6v rated before the batteries start to deplete below 3v/piece, and in turn conserve battery life.

Richard

How many of these circuits are you going to make?

Is it commercial, automotive, defence, a one off and/or a hobby/college project? Will it be operated at high temperatures? Are other parameters being pushed to their limits, i.e. operating frequency, output current?

All of the ICs you've listed have an absolute maximum rating or 7V. They may not be guaranteed to operate properly at this voltage, over the full temperature range, but they should be able to withstand it for short periods of time, without damage. They're designed to operate properly up to 6V, at 125^oC, with no problems. It's pretty safe to say that operating at 125^oC, at 6V will be far more stressful to the IC, than 30^oC at 6.4V. Look at the table for the 74HC1G00, on the bottom left hand of page 2 the data sheet "DEVICE JUNCTION TEMPERATURE VERSUS TIME TO 0.1% BOND FAILURES". I'd be more worried about high temperatures, than slightly higher supply voltages, than recommended on the data sheet.

https://www.onsemi.com/pub/Collateral/MC74HC1G00-D.PDF

There's a myth going round that components automatically let out magic smoke, when their maximum operating specifications are exceeded. This is total nonsense. Whilst it's good design practise to not exceed maximum operating specifications and shouldn't be done lightly, it doesn't mean it shall never happen. The main risk is that reliability will be compromised, which can be a huge problem in a volume manufactured product, especially if it's operated in an adverse environment i.e. high operating temperature, but if you're only making a few, the components aren't being pushed to their other design limits, a slight over-voltage is no problem.

In short, if the components are run at room temperature, a low output current and you're only making a few, then don't worry about using a supply voltage of 6.4V, especially if it's just a hobby project. It's not worth trying to lower the voltage by 0.4V. All it will do is shorten the life of the batteries and add more components which could cause other problems.

[Richard1234]

Thanks Hero999,

It's a hobby project, but I might try and sell a few of them none the less.

Thanks for all the well made points. I'll try running it straight off the button cells. Next board run though, I'll add the diode for reverse polarity protection none the less.

Richard

[djacobow]

I should have said in my first post the same thing. If this is a hobby project for personal use, I would just gone with it without any voltage adjustment.