diodes in series to reduce voltage

[stevech]

To go from 5V to 3.3V, I could do so with two common and one Schottky diode in series to get about the right voltage.

For some current, I, flowing, would the diodes have to dissipate per Ohm's law per diode of heat?

so 0.7V times I amps for the standard diodes?

Is this a bad way to avoid putting in a 3.3V regulator chip/heat sink?

The load is about 180mA at 3.3V.

[dkinzer]

For some current, I, flowing, would the diodes have to dissipate per Ohm's law per diode of heat?

Yes. Every electronic component dissipates heat calculated as I*V where I is the current through it and V is the voltage drop across it. That said, as long as I is less than the specified maximum for the diode and you're complying with the manufacturer's recommendation, if any, for heat sinking, you're good to go.

Is this a bad way to avoid putting in a 3.3V regulator chip/heat sink?

Of course, "badness" is in the eye of the beholder. It is a "cheap and easy" way to drop voltage if the current is fairly consistent. To understand the issue, consider a "typical" I-V curve for a diode shown in the attached image. As the current through the diode varies, the voltage drop across it varies. The change depends on where you are on the curve.

There are 3.3V LDO regulators designed to be operated on 5V that are just about as simple to use as the diodes. Cost, however, is a separate issue.

[stevech]

thanks much.

The use of diodes in series is just an idea for an expedient for stuff on-hand. I should order some 5V and 3.3V regulators to have in the junk box.

[dkinzer]

The use of diodes in series is just an idea for an expedient for stuff on-hand.

If you're looking for a jury-rig, the diodes are a great idea. As with all such arrangements, you just need to be aware of the limitations.

[GTBecker]

If you're looking for a jury-rig, the diodes are a great idea.

Some examination of IC equivalent schematics will reveal that IC designers use diode drops all the time; hardly jury-rigs. I power 3.3v Bluetooth modules through diodes very happily.

[stevech]

I connected the 160mA, 3.3VDC WizNet810MJ to 5V using two diodes in series. Under load, the voltage is 3.4V. Seems to work fine. Diodes are just barely noticeably warm.

[victorf]

For an interesting and somewhat elegant solution to the 3.3 - 5 volt problem check out the 74lvcc3245. I tried to attach a data sheet but it exceeded the 512k limit. Sorry

Vic