Hardware voltages not matching specs. Does it matter?

In the past, I bought some 5-volt / 3.3-volt level shifters, and was astonished how far off the actual voltages are, and yet they are considered "good", and they work.

Later, I bought some audio breakouts, and was again surprised that the 3.3-volt source they supply (from the 5-volt they run on) actually produces 4.1-volts instead. (In that case, I added an external 3.3-volt regulator to get things right.)

Now I'm using the following circuit to charge the internal clock battery, and once again, voltages just don't match what literature says they are going to be.

The source voltage (supplied by the laptop USB) is actually 5.17 volts, not 5.0. The output from the regulator is actually 3.41 volts, not 3.3. And the voltage between diode and resistor is currently 3.25, instead of 3.1, which will undoubtedly rise as current slows down from the battery reaching full charge.

I would think these differences too little to matter, if it were not for the literature on this specific battery stating:

(1) 3.1 volts is the accepted standard charging voltage. (2) The charging voltage may be between 2.8 and 3.2 volts. [3) Exceeding this limit will cause the battery resistance to rise, and may cause battery deterioration.

Actually, the clock came with an internal charger I could turn on, but the built-in charger is a resistor in series with a diode connected to the 5-volt source; clearly a "trickle-charger" by definition. So I didn't use it because the battery's literature states:

Under no circumstances should trickle charging, which is used for nickel-cadmium batteries, be used. Ignoring this precaution will cause the battery voltage to rise to about 5V, resulting in a deterioration of performance.

(What were the manufacturers of the clock breakout thinking?)

I feel like no one up there at these various manufacturing companies knows what's real, and they all just spout personal theories, claiming it to be the gospel truth when it isn't.

Should I take anything they say seriously? How can I be sure I'm not selling a product with batteries headed for "China Syndrome", due to my following their instructions?

(What were the manufacturers of the clock breakout thinking?)

Lots of boards at a few cents each, adds up to lots of dollars.

Weedpharma

The source voltage (supplied by the laptop USB) is actually 5.17 volts, not 5.0. The output from the regulator is actually 3.41 volts, not 3.3.

Those voltages are within typical specs.

If by "clock" you mean an RTC, most RTC chips were designed to be used with nonrechargable 3 V lithium backup batteries, which will last for several years.

However, the Chinese have become famous for selling cheap and nasty DS1307 and DS3231 modules with rechargeable batteries and a charging circuit of extremely bad design. One user even reported a small explosion.

Linear regulators need decoupling capacitors. What values are you using?

What are you using to measure voltage? If using a multi-meter and most all voltage readings are higher than expected, this could be a symptom of a weak battery.

weedpharma: Quote: "(What were the manufacturers of the clock breakout thinking?)"

Lots of boards at a few cents each, adds up to lots of dollars.

Weedpharma

Yeah, you may be very right; putting dollars ahead of the needs of those you serve.

jremington: most RTC chips were designed to be used with nonrechargable 3 V lithium backup batteries, which will last for several years.

The model I ordered came with a non-rechargeable battery, which lasted only 6 months.

jremington: the Chinese have become famous for selling cheap and nasty DS1307 and DS3231 modules with rechargeable batteries and a charging circuit of extremely bad design.

That's exactly what I got. Glad to know I'm not the only one who thinks the charger design is terrible.

Also, thanks for the link to some really good info on the subject.

[quote author=James C4S link=msg=2540194 date=1451361004] Linear regulators need decoupling capacitors. What values are you using? [/quote]

I wasn't using any at the regulator. But following your suggestion, I tried adding a 0.1 uf cap across the input and ground of the regulator, but didn't recognize any difference on the volt meter or oscilloscope. This voltage only going to the battery for charging, I doubt a little noise on the wire would matter.

dlloyd: What are you using to measure voltage? If using a multi-meter and most all voltage readings are higher than expected, this could be a symptom of a weak battery.

Good point. I just now tested that theory by connecting a second digital multimeter. But the difference between the two was only 0.03 volts. (3.24 vs 3.27). Batteries should both be good.