Ground triggered Latching Circuit using transistors

AnalogKid

Joined Aug 1, 2013
10,987
Late to the thread, this sounds like a simple toggle flipflop. Push once, on; push again, off; repeat. If so, there is a thing called an impulse relay that will give you DPDT 10 A contacts and the alternate action latching all in one part with zero electronics. Just sayin...

ak
 

Thread Starter

stillgrowingup

Joined Jul 15, 2015
210
Do you have a limit on how many transistors get used before you go to using an IC?
Not really. I've thought this circuit would contain 3 transistors, couple diodes, couple resistors and an LED. What ever is simple and reliable! .... last thing I want is a failing circuit cause of a poor design.

The attached schematic of the 4 relays is how I have done this in the past. But that's big and bulky. That's why I am looking to do this with smaller components, like transistors.

I found yet another transistor latching schematic. See attached. This schematic is for ground pulse on/off setup. Not sure what voltage the resistance, diodes and capacitor component are called out for. I would use BD139 NPN for the transistors. I have those. I like them, they seem to be very heavy duty.


Tony
 

Attachments

Last edited:

Thread Starter

stillgrowingup

Joined Jul 15, 2015
210
Latch/unlatch every push of the SW.
Max.

Hi Max,

I must apologize to you. When you said to reference the PDF. I thought you were talking about the pdf in post 13. :( ... I have looked at the correct PDF that you posted. I do see it being a latching set up using 3 ST4P relays. I have used these relays in the past at my work. I remember them being very noisy when they click. That's why I wanting to go transistor for less noise, plus a few transistors are smaller than a few relays. I will probably have 4 or 5 separate units of whatever circuit this topic produces.


To all,

Does the tim17 drawing in post 22 seem to be viable option? I found it here: http://www.electronics-tutorials.ws/waveforms/bistable.html

Your input is GREAT! :)

TONY
 
Last edited:

Thread Starter

stillgrowingup

Joined Jul 15, 2015
210
That is BAD... your transistor is rated for 1.5W only. You could place a resistor in series, but your lamp is going to glow much dimmer. Are you ok with that?

EDIT: Let me make sure I didn't say something stupid. Studying the datasheet right now.
Hmmmm ... what's an alternative to make to the lamp full?

TONY
 

cmartinez

Joined Jan 17, 2007
8,220
Yup... I said something stupid alright... not the first time that's happened either.

The transistor can handle up to 1.5A, and your lamp's only gonna draw 0.25A @ 12V ... so yeah, you should be fine.
I misread the transistor's specs at first. My bad.
I've attached the datasheet for you to see, if you're interested.
 

Attachments

Thread Starter

stillgrowingup

Joined Jul 15, 2015
210
Yup... I said something stupid alright... not the first time that's happened either.

The transistor can handle up to 1.5A, and your lamp's only gonna draw 0.25A @ 12V ... so yeah, you should be fine.
I misread the transistor's specs at first. My bad.
I've attached the datasheet for you to see, if you're interested.

LOL .. no worries. .... So no resistor or diode needed? I thought maybe a diode to protect the latching circuit from feedback through the lamp.

Does the rest of the circuit seem ok?

Thanks for the help.

TONY
 

dl324

Joined Mar 30, 2015
16,846
Please tell me what you think. Will it do the job and be reliable?
I thought it had already been established that that circuit wouldn't work. I certainly don't see how shorting the resistor makes the circuit change state. The inputs are usually the base terminals...
 

hp1729

Joined Nov 23, 2015
2,304
LOL .. no worries. .... So no resistor or diode needed? I thought maybe a diode to protect the latching circuit from feedback through the lamp.

Does the rest of the circuit seem ok?

Thanks for the help.

TONY
Your BD140 needs a resistor on the base to limit current. Maybe 1500 ohms. The emitter needs to go to +12 v and the lamp to ground. Other comments about the BD149 may also be important.
 

Alec_t

Joined Sep 17, 2013
14,280
Will it do the job and be reliable?
Simulation shows it won't toggle. It would be better to base your circuit on the tim17 bistable you linked to in post #22. That bistable will toggle ok with a ground-going pulse, provided the pulse source has a pull-up to >1V (e.g. to V+).
 

cmartinez

Joined Jan 17, 2007
8,220
:mad: darn! hp1729 bet me to it... gotta learn how to think fast here...
And yes, as he's just said, emitter and collector are inverted in your diagram. You need to correct that or your lamp's never going to turn on.
What do you think hp1729? Would a 1K resistor at the base be enough?
 

hp1729

Joined Nov 23, 2015
2,304
:mad: darn! hp1729 bet me to it... gotta learn how to think fast here...
And yes, as he's just said, emitter and collector are inverted in your diagram. You need to correct that or your lamp's never going to turn on.
What do you think hp1729? Would a 1K resistor at the base be enough?
The BD140 has a gain of only 40, so 1K to 1,500, yes. Somewhere in there. The BD140 can handle 500 mA. 12 V at 3 W is only 250 mA. It will work.
I don't know about his toggling latch. I have no talent with a simulator. I will have to build it and try it.

(edited to add ...)
I have to agree it doesn't toggle. And for LED D2 instead of 330 ohms, maybe 1K.
I have no suggestion for making it toggle without a lot more circuitry.
 
Last edited:

MrAl

Joined Jun 17, 2014
11,389
Hi,

I believe what you need would be equivalent to a JK master/slave flip flop.
Check into that and see if you can use that.

The idea goes like this...

First, a flip flop is either standard flip flop IC chip or two transistor arranged as a bistable multivibrator. That's the circuit with two NPN transistors where the one turns off when the base is grounded and the other turns on, then vice versa when the other base is grounded.

Ok so we have the basic flip flop desribed above, and we will need TWO of these, not one.

The first FF is the master, the second is the slave.

When the button is pressed, the master directs the signal to the appropriate transistor of the slave, and the slave changes state. That's the main job of the master.
When the button is released, only the master changes state, so the next time the button is pressed the opposite transistor base is grounded, and that changes the state of the slave back to what it was originally. Thus the slave is the output and the master just directs the signal.
This also means that BOTH states of the push button switch have significance: ON means the directed ground signal gets to the appropriate transistor base, and OFF means the master changes state.

Note also that the signal that gets to the slave may also be amplified with another transistor.
Also note that the switch must be debounced, which may require another transistor and capacitor type circuit. The master must not be allowed to get two pulses when it should only get one.

The final output of the slave would drive a high power MOSFET with low Rds to meet the high power bulb specification. This brings up a BTW, BTW, and that is that a transistor power rating does NOT have to equal the load rating, it only has to meet it's own voltage drop times current flow requirement (P=I*V in the transistor for the transistor power dissipation, and not P=Iload*Vload).

Some other notes:
A better debounce circuit could be built with a SPDT switch, maying the reliabilty almost unquestionable.
I have to wonder why a push on, push off switch could not be used here instead of circuity. That would be the only component requirement then right?

Some side notes:
Another idea that comes to mind is of course a low end microcontroller chip. These can be had for under a dollar USD and can do everything including debounce and state housekeeping.
A 555 timer might help in one of the circuits also.

Overall brief analysis:
There are four distinct states:
1. switch ON (activates one base of the slave, slave state 1)
2. switch OFF (flips the state of the master 1)
3. switch ON (activates the other base of the slave, slave state 0)
4. switch OFF (flips the state of the master to state 0)

So for states MS for master and slave, the states go as follows:
00 (initial states M and S)
01
11
10
00 (end, which takes us back to the initial state where both M and S are zero)

I almost forgot that the slave can be used to control the master which eliminates the debounce circuit. That's the main reason for using two flip flops. If you just want to use pulse steering diodes then you might get away with just one flip flop, but it would have to work every time.

I would also have to ask on what the limit on the number of transistors might be.
Also, what the max switch rate would be.
 
Last edited:
Top