Hi,

I need some help with a circuit that will reverse the polarity at it's two outputs when a switch closes and restores the polarity again when the switch opens.

The schematic I came with is the following:
https://picasaweb.google.com/lh/photo/QWAY9OXlsnVX2uvLHF5UDtMTjNZETYmyPJy0liipFm0?feat=directlink

Comments? Will it function as expected?

Dimitrios

 

First thing I can say that there are base resistors missing for the two transistors on the right.
When the switch is open the lower left PNP conducts, applying V+ to the base of the lower right NPN. BOOM !
Same thing for the upper two transistors when the switch is closed.

 

Oops...
Apart from the base transistors? Will the circuit perform as the table suggests when I add the base resistors?

Dimitrios

 

Not unless you like the smell of toasted transistors. Where did you get the idea for this arrangement?

 

Kermit,

Could you show me the short circuit if I add the base resistors? The idea is mine...

Dimitrios

 

Do you want two push-pull outputs or why did you connect the transistors at OUT1?
The two outputs signals are supposed to be inverted, right, i.e. if one is HIGH the other one is LOW?

 

Hmmm. All I want is a circuit that will reverse the polarity at the outputs when I open/close switch 1. I thought that with the switch open the lower transistors will conduct giving me V+ on output 1 and GND on output 2. With switch closed the upper transistors will conduct outputting GND on output 1 and V+ on output 2.
The base resistors are ommited in my schematic but I'll obviously have to add them in the circuit.

Dimitrios

 

I don't know if it helps you at all, but I recently needed a way to reverse the polarity of a circuit. I accomplished this using a 4017 counter, using the "1" and "2" outputs to power the circuit. On the first clock tick to the 4017, "1" goes high and "2" is ground. On the next tick, this reverses. Then the 4017 resets. Net effect, every time you send a pulse to the 4017, the output circuit polarity reverses.

 

Hmmm. All I want is a circuit that will reverse the polarity at the outputs when I open/close switch 1. I thought that with the switch open the lower transistors will conduct giving me V+ on output 1 and GND on output 2. With switch closed the upper transistors will conduct outputting GND on output 1 and V+ on output 2.
The base resistors are ommited in my schematic but I'll obviously have to add them in the circuit.

Dimitrios

No doesn't work like this.

Did you mean similar to this?

I didn't include anything to make absolutely sure that there will be no shotthrough under no circumstances.

 

Is there a specific reason you don't want to use a two pole double throw switch to do the deed? Would the switch bounce be a problem? There will be switch bounce with almost ANY mechanical switch even in the proposed scheme you first posted the output would cycle quickly between polarities for a short time immediately following the switch being 'thrown'

 

Praondevou,

isn't this the exact same schematic I drew on paper (sans the base resistors...)???

Kermit2
that's exactly what I want to get rid of with this solid state circuit - a DPDT mechanical relay.
The switch is not mechanical either. It will be either another transistor or a solid state NO relay (built in phototransistor?). Do you think switch bounce will still be an issue?

Dimitrios

 

Praondevou,

isn't this the exact same schematic I drew on paper (sans the base resistors...)???

Kermit2
that's exactly what I want to get rid of with this solid state circuit - a DPDT mechanical relay.
The switch is not mechanical either. It will be either another transistor or a solid state NO relay (built in phototransistor?). Do you think switch bounce will still be an issue?

Dimitrios

Yes , similar. the bases however should not be connected. For the leftmost transistors you also need two base resistors, one for each transistor.
You could tie the bases together if the transistors were inverted.

There is however still an error in what I drew. Just don't have the time right now to correct it.

It would be interesting to know what your load is. Do you really need two push pull output stages?

 

What the advantage - if any - of inverting the transistors and tieing the bases together?

The load is another solid state relay that will change add remove resistors in series, to change the total resistance, used as an input in a R/C servo driver circuit that will control and electronic speed controller that will power a brushless motor.

Yeah... it's one of those jobs that can be done much easier, faster and cheaper with a microcontroller, if only I could write a few lines code...

Dimitrios

 

If the transistor connected to V+ is a PNP and the one connected to GND is a NPN and you tie their bases together there will be a current flow from V+ through the two bases, really not desirable.

This is why they are usually connected the other way around when bases are tied together.

However, from what you describe you don't even need push-pull stages.

You get your signal from a transistor and will drive a solid state relay, that has an optocoupler inside I imagine.

Why don't you post the whole thing or at least the input / output requirements of this little circuit ?

From what I understand you have one input and want two outputs, one inverted to the other. When the input signal changes polarity you want both output signals to change polarity too. Is that it?
If so then we need to know:

- input voltage
- output current and/or voltage
- output sinks or sources current?

 

It will source 5 or a little more milliamps through R3 and maintain 4.5V(high) to 1.0V(low) output(with a 6 volt Vcc). The input and output caps are filters for switch bounce(drew this up before dinner and the subsequent additions to the thread since then, everything to the left of R4, plus C1 and C2, is just switch bounce simulation) You could add another NPN in a common emitter config and get the mirror image of the output across R3. Then a voltage follower could also be added, giving you the four transistor, two output circuit. Taking R3(load) much lower than 600 ohms reduces the voltage swing enough to be doubtful of its use as a switched digital input, should be enough current to drive some selected solid state relays that only need 5ma or 10ma to operate.

 

It will source 5 or a little more milliamps through R3 and maintain 4.5V(high) to 1.0V(low) output(with a 6 volt Vcc). The input and output caps are filters for switch bounce(drew this up before dinner and the subsequent additions to the thread since then, everything to the left of R4, plus C1 and C2, is just switch bounce simulation) You could add another NPN in a common emitter config and get the mirror image of the output across R3. Then a voltage follower could also be added, giving you the four transistor, two output circuit. Taking R3(load) much lower than 600 ohms reduces the voltage swing enough to be doubtful of its use as a switched digital input, should be enough current to drive some selected solid state relays that only need 5ma or 10ma to operate.

Kermit2,

to which schematic are you referring to?

Anyway, the circuit I have is this:
https://picasaweb.google.com/lh/photo/WZdxxVCGhYqYqZhsnQ4U29MTjNZETYmyPJy0liipFm0?feat=directlink

that is controlling a linear actuator that currently is using a DC motor and has internally wired two end switches.

I want to replace this brushed motor with another one that's brushless. The brushless motor will be driven by an ESC that needs a PWM signal compatible to the R/C servo standard. To produce that PWM signal I'll use either a commercial driver or I'll build a 555 based one (tons of schematics on the www).
The servo driver has a potensiometer as a user input control. As I only need three positions full throttle reverse-neutral off-full throttle forward, I plan to use two resistors in-series and a solid-state DPST relay that will "move the wiper" from the center - between the equal resistors - to the one side - to the other side.

That's about it!

Anybody willing to write a little code for me that does that and make my day? Hehehe

Until then I am trying to do it the analog way...

Dimitrios

 

Kermit2,

to which schematic are you referring to?

The one in my post and also in your quote directly above.
Hint: some people cannot see it if they are not logged into the forum

It is a moot point since, as I said, I made it long before the progression of the topic revealed what you truly needed.

 

I am logged in but still can't see it...

Dimitrios

 

The schematic is visible to me too, in posts #15 and #16. It could be a broken image while downloading. Try pressing Ctrl+F5 to reset the webpage cache and reload it.

 

Nope, no luck....

It doesn't really matter though. If the schematic on post #9 works that is... does it ?

Dimitrios