Transistor as a relay vs mechanical relay

No moving parts.

Cheers! Rich

an

signal

etc?

Yup.

A basic question, indeed. Let's count the ways.

• Tranistors switch in microseconds, relays in milliseconds.

• Transistors switch directly on without contact bounce, relay contacts smack and bounce together for a millisecond or so, causing intermittent on/off for that time until the contacts settle in.

• Relays are mechanical devices with moving parts -- they will wear out. A typical relay is rated for somewhere between 10 thousand and 10 million operations before failure. And relay contacts usually wear out first, if they're swicthing at anywhere near rated load. Actually, they're supposed to.

• Relay contact bounce can cause EMI which can cause problems. Although you generally have to be more careful with transistors when switching inductive loads, there isn't a spark being created with switching like with relays. That EMI can cause the computer to spit up.

• A typical high current transistor will usually cost much less than a relay rated to switch equivalent current.

• A transistor will typically require much less power to operate than a relay coil, especially if you use a darlington transistor or a MOSFET.

I guess there are more reasons, but these are the ones that come to mind first.

Good luck Chris

It's none of the items mentioned. The signals from the PC may be on a separate ground than the other electronic circuitry that is receiving the signal. In this case, isolation is required by either by a mechanical relay or an opto coupler. Just a transistor is not adequate.

Harold

signal

etc?

Read the the intro, pages 1-6, for this book:

[snip]

A Darlington transistor requires a V drop of twice that of the single junction transistor. The C-E V drop of a single junction can be a tenth of a volt, but a darlington has to be at least 2 diode drops or about

1.2V to function. So the power wasted by a darlington is much greater than with a single junction transistor.

If you want to minimize this, use a regular power transistor, and drive it with another transistor connected common collector or emitter follower. This basically means do _not_ connect the collectors together in a darlington config.

That is not a useful statement of the facts.

True (assuming you mean a single BJT). Or it can be more, or less. A BJT in hard saturation can have 50 mV C-E drop.

Actually, the input BJT can and often does saturate to the kind of drop stated above. It keeps the output BJT out of saturation since the input C-E drop is in series with the base of the output BJT. This results in typical darlington C-E drops of 0.8 to 1.0 V. There is nothing about this situation that makes "2 diode drops" significant.

Assuming you mean a single bipolar junction transistor, the truth of your claim depends largely on what supply the base current is taken from. Where the input BJT forced Beta is X, then for a bias supply greater than X times the additional drop of darlington, your claim is strictly false, (meaning "much greater" is no greater).

That may be a good strategy if a low voltage bias supply is used and if the extra parts count is worth the power savings. But the OP should be aware that Darlington transistors have been used in many places by people familiar with the alternatives.

These days, stringing together BJTs as you suggest is rare. If the output drop is important, a single MOSFET is generally favored.

--
--Larry Brasfield
email: donotspam_larry_brasfield@hotmail.com
Above views may belong only to me.