1. Home
  2. Electronics Design
  3. Pulsing DC Signal Relay Control...

Pulsing DC Signal Relay Control...

Hello,

I am trying to construct a relay control circuit based on an input waveform. The waveform can take on two states - a 0 --> 14.4V DC square wave at 55Hz and 65% duty cycle OR a constant 14.4V DC:

formatting link

With the squarewave as input, the relay is to be in one state, with the constant DC, the relay is to be in the opposite state.

I have tried two approaches from an original post here:

formatting link

Including using a MOSFET with cap and 2 resistors and a circuit with a large cap, resistor and diode, but they seem to produce intermittent results.

The key factors are:

1) Simple circuit - as few parts as possible 2) Large caps (up to 10,000uF) and high wattage resistors are OK 3) Able to utilize relays with different characteristics (i.e. different coil resistances)

Someone on the original post mentioned a "monopulser", but I cannot find information on that anywhere. This sounds ideal though - something that keeps the input high (or low) when pulses exist and then flips states when the pulses disappear...

TIA

Reply to
afeudale
Loading thread data ...

formatting link

I recall that idiot question and told you at that time to go with a double pole switch. If you have a problem with that then tell us what it is....

Reply to
Fred Bloggs

On a sunny day (Tue, 10 Jun 2008 08:43:55 -0700 (PDT)) it happened snipped-for-privacy@hotmail.com wrote in :

So we only bring it up with AC:

  • - A B input ---------- C1-------- k diode a ---------------- | | | a | - | diode C2 relay k | + | | | | /// /// ///

In case of a 0 --> 14.4V DC square wave at the input: In A is a 14.4V square wave, from + 0.7 to - 13.7 V. In B is a - 13 V DC voltage.

In case of + DC on the input: There is undefined in A and nothing in B.

The electrolytic capacitors can be the same value, perhaps 25V volatge rating. The value depends on the relay current, so on the relay coil resistance. For 55 Hz, say less then 20mS and relay colil resistance of R, think RC time for discharge to a level before where the relais drops off.

Maybe 4700uF is enough....

Reply to
Jan Panteltje

formatting link

Simple: Two caps and two diodes. Make a voltage doubler, one cap C1 in series with the source connected to a forward diode, D1, anode. Second diode, D2 anode to ground, cathode to anode of D1. Cathode of D1 to cap, C2, and to relay coil. Ground other end of C2 and other end of relay coil.

State 1. AC wave form is rectified to peak value of waveform, 14.4 volts minus losses. This voltage pulls in the relay.

State 2. Straight DC won't go through C1, so relay remains unenergized.

Cap size depends on the relay current required. Relay voltage = 12 Volts.

Reply to
Bob Eld

I guess that would work real fine and dandy how ever, while your at it did you ever stop and think that maybe he's deriving that signal from a low current source ? In which case since you want to slop it out, a totem pole output trannies as a unity gain stage to drive this little radio shack of a kit idea circuit!. or Maybe, a driving transistor for the coil!

formatting link
"

Reply to
Jamie

Maybe something like this may be useful for that function.

D-Type Transparent Latch

formatting link

Reply to
Hammy

--- What you need is simply a retriggerable monostable multivibrator:

V+>------------------------------------+ | +-->RELAY COIL K| Vcc>--------------+-------+ [DIODE] | | | [Rt] | +-->RELAY COIL | +--+--+ | VIN---[R1]--+-----|----|T | D | | | OUT|--[R]--G N | +----|RC | S [R2] | +--+--+ | | [Ct] | | | | | | GND---------+-----+-------+------------+

And you adjust the values of Rt and Ct so that the output pulse from the one-shot: lasts for somewhat longer than the period of the 55Hz signal, like this:

|| ____________ ____________ ____________ VIN___| |_________| |_________|

___________________________ OUT___| |______________________________

In reality what will happen is that the one-shot's output will be extended for an additional RtCt time if it's high when VIN goes high and it'll stay high as long as VIN continues to toggle, like this:

____________ ____________ ____________ VIN___| |_________| |_________|

__________________________________________________________ OUT___|

However, when VIN _stays_ high, the one-shot will time out and OUT will go low, turning off the relay:

____________ ___________________________________ VIN___| |_________|

_________________________________________________ OUT___| |________

R1 and R2 are a voltage divider used to cut VIN down to something the one-shot wants to see, around 4V when VIN is high.

A good choice for the one shot would be either an HC123 or an HC4538.

JF

Reply to
John Fields 

--
One caveat: note that relay will close for one RC time when VIN goes
high and stays high:

       _________________________________________________________
VIN___|                                              

       ___________________________
OUT___|                           |______________________________

If that\'s going to be a problem for you it can be fixed by "ignoring"
the first edge, but at the cost of  an ON delay one RC wide.

I think...

 
JF
Reply to
John Fields

He can trigger on the negative going edge ....

Reply to
Fred Bloggs

In case of high impedance for the source, how about this:

input ----------------- R ------------------- 7,2V if AC, 14.4V if DC ----------- + | +10V - opamp --- R - driver transistor with diode. C | ///

R can be very high, and C maybe 1uF, no need for large caps.

Reply to
panteltje

On a sunny day (Wed, 11 Jun 2008 04:12:26 -0700 (PDT)) it happened snipped-for-privacy@yahoo.com wrote in :

Google REALLY screwed that drawing up, so here again:

In case of high impedance for the source, how about this:

-------------- + something | | | relay level comparator diode | |--------- input ---- R ------ 7,2V if AC, 14.4V if DC ----------- + c | opamp --- R - b C +10V - e | | /// ///

R can be very high, 1MOhm, and C maybe 1uF, no need for large caps.

Reply to
Jan Panteltje 

--
Yes, but then he loses the time from the first high-going edge to the
first low one if VIN is toggling, and he\'ll get a false actuation when
VIN goes from steady 14.4V to 0V.

JF
Reply to
John Fields

On a sunny day (Wed, 11 Jun 2008 09:17:14 -0500) it happened John Fields wrote in :

Use 4046 PLL, osc at 55Hz, use 'lock' output :-)

Reply to
Jan Panteltje 

--
Still, there\'s the time constant of the loop filter to deal with.


JF
Reply to
John Fields

On a sunny day (Wed, 11 Jun 2008 10:03:16 -0500) it happened John Fields wrote in :

Yes, if you want immediate action make double freq input with xor and gate delay:

delay input-------- gate ----- | XOR -------- 110 Hz small pulses ------- --------------

Then trigger a one shot with a timeout of about 10 mS. If the first pulse is absent it will flip over.

Could even be simply discharging a cap:

  • | R | -- k diode a --------------- FET ? | === | ///
Reply to
Jan Panteltje

--> wrote in

--
I don\'t see it.

There are only two valid input signals, either high all the time or a
55Hz, 65% duty cycle square wave.

If we say that high-all-the-time causes the relay to stay open, then
we can use the first low-going edge of the waveform to turn on the
relay and lose no time except the relay\'s own close delay time:

                                 -->|      ||  ||   |
Reply to
John Fields

On a sunny day (Wed, 11 Jun 2008 12:04:04 -0500) it happened John Fields wrote in :

In an XOR gate, the ouput is high as long as there is a difference between the inputs. If yo udelay one input by say 10nS (by adding a gate in the path), then there will be 10nS pos pulses on its output. There wil lonly be 10nS 110Hz pulses as lon gas the input is AC. In case of DC the output of the XOR is always zero.

The (invert them, now negative narrow pulses) discharge the capacitor via the diode. When these pulses stop, the cap charges, and at some level the relays is energised.

OK, what I was merely suggesting is that using the double frequency at 110Hz reacts faster.

Yes that is correct. I am not questioning your solution, merely added an other one. Mine is faster when it changes to DC I think (10ms)?

Reply to
Jan Panteltje

energised.

--
Dunno, let\'s see:

For a 55Hz 65% duty cycle square wave we have:


              6.4ms-->|  ||   |
Reply to
John Fields

Thank you everyone for the excellent replies. I have tried most of the circuits described (some successful, some not). I found that one of the missing pieces of information in my description was that the voltage source driving the relay (or logic, etc.) is the same source providing the signal (i.e. sometimes DC, sometimes fluctuating), thus making the use of logic circuits, opamps, etc more difficult (if not impossible since they relay on a stable power source). Here are specific replies:

Fred - there is no way to modify the switch (it is part of a vehicle) otherwise, yes I would make it a DPDT as you suggested Jan & Bob (similar ideas) - I tried the 2 diodes, 2 caps idea with a lot of optimism, but for some reason, the relay would still flicker, even with 10,000uF caps Hammy & John - the constant DC source required for the logic circuits is not available

Another piece of information is that the output does not have to change immediately - it can go for a few cycles before changing...

I am currently using a combination of everyone's ideas:

Resistor Voltage divider connected to a MOSFET with a capacitor on the gate and with a reversed diode to discharge the cap quickly when the input goes to zero. So far so good...but I'm open to any better ideas :-)

Thanks again for all the help!

Reply to
afeudale

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.