I need a circuit to switch 9volt dc off on..cmos?

[domenico]

what can I use to make on off 4  9volt working circuits ? I don't need to make bypass , I need to stop dc 9volt flowing in the circuits
I've tried cmos cd4066 but I doesn't work
thanks
domenico

[brett]

Hi.
Maybe it's worth explaining why the CD4066 didn't work.
Not enough current capability, maybe?  If so, why not have each switch drive a mosfet (an MJE3055 costs $2 and switches 10 amps!).

For really bulletproof switching, optocouplers are good (but expensive).
At work they use computer logic output (5V) and 2N25s (from memory ??)
They are available in 6 pin DIL packages at about $5 each.
cheers

[domenico]

when I connect 9v to one of the switch of cd4066 it seems to burn because after that the switch stays always closed
I think I'll try with mosfet , does bs170 work for this application ?
thanks

[gez]

How much current is being drawn?  You're most likely going to exceed power ratings, CMOS switches aren't designed to handle 'large' currents (doesn't take much to fry them).

Mechanical switching/relays/power transistors if you really must use trannies...

[Paul Perry (Frostwave)]

A power mosfet will do fine. (??IR510?? it's a lomng time ago now!) I've controlled solenoids & motors by them.

[brett]

Oops!! That should have been an MTP3055  .
At AUD$2 and 10 amps of continuous current, they are very rugged.
(we switch 1/2 horsepower DC motors on and off with them at work)
cheers

[R.G.]

As you've discovered, switching power is different from switching signal. The big difference is that switching signal involves switching almost no current, but power switching always involves lots (relatively) of current.

So to do this successfully, you need to know whether you're switching microamps, milliamps, amps or car batteries. There are different answers in each of those areas.

An audio signal is usually well under a volt and milliamps at most; usually it's microamps, and it's always AC. To switch an audio signal, almost any switch element will work. The trick is to make it switch without distorting the audio signal and without introducing pops and noise. There is a whole discipline about switching audio.

To switch power to a pedal circuit, you have to switch at least milliamperes, and usually tens to hundreds of milliamperes. On top of that, it's DC, no AC involved. Audio-type switches will not do well for this because (a) they have to high a resistance and (b) their maximum current rating is usually too low. You haven't said, but I'm guessing that you're doing this kind of switching.

What you need is some device that can successfully switch up to 100ma, and have a low voltage drop when doing that. A CMOS switch like the 4016/4066/405x series can't do this because their internal resistance is perhaps 200 ohms and their current rating is milliamperes, not hundreds of milliamperes. I suggest reading the datasheet for the CD4066 and looking for the on resistance of the switching element and the maximum switch current rating. It's good education for you.

The simplest thing to switch things up to 100ma of DC is the ordinary, garden variety bipolar transistor. These do a GREAT job of switching this kind of load.

To switch DC you have to decide if you're switching the positive side, the negative side, or both at once. Usually you want to switch the side that's not ground, so I'll use that as an example. If you have an ordinary negative-ground pedal, you'd usually switch the positive side. You can do that with either an NPN or PNP transistor. To do it with an NPN, you hook up the collector of the NPN to your positive power supply (usually 9Vdc, so I'll use that as an example) and its emitter to the power connection of what you're powering. To switch this on, you cause the base to be pulled up to the collector, and to switch it off, you pull the base down to ground. This is a good setup and works well, but suffers from two defects: the voltage swing you have to apply to the base is 100% of the voltage you're switching, and the lowest voltage across the switch transistor is one diode drop.

We can do it differently. If we use a PNP transistor, hooked up with emitter to 9V and collector to the circuit being powered, we only have to move the base of the transistor down by one diode drop, and the transistor turns on - hard! We can pull the base even further down and cause the transistor to saturate to well below one diode drop. 50mv is entirely reasonable for a saturation voltage in this setup. You do have to limit the base current in this hookup, but a simple resistor works fine for that.

A 2N3906 transistor is rated for 100ma of collector current, and has a gain of well over 100. If you hook it up as I said above, you can use any resistance on the base that causes between one and 50ma to flow in the base and get the switching action you'd like. One 2N3906 per switched circuit works just fine.

It's a minor problem that you have to apply a low level to get the voltage to go high with a PNP switch. If that is a problem to you, you can use an NPN transistor to ground to switch the base current to the PNP, and then a positive voltage to a resistor in the NPN base to turn on the NPN. Now a positive voltage to the NPN turns on the PNP. You can also use a CMOS inverter driving the resistor to the PNP base to get positive input = current on.

I'll leave off the discussions of amperes and car batteries.

[domenico]

r.g. you are greeeeeeeeeeeeat !!! thanks that's what I need!!!
thanks thanks thanks!!!
domenico