Buffer a negative input?

Question

I have a built a little relay circuit to control a light from the inputs of a motor controller board. When the motor runs I want the light to turn on, when it stops I want the light to stay on and when the motor reverses and stops in a "home" position I want the light to turn off. This is a 12V DC system and my motor enable and motor reverse are negative inputs.

I have a relay that latches (light on) when the motor is enabled and the reverse enable breaks the latch (light off). It works pretty well but about 1 in 20 times the latch will not break when the motor reverses. If I could extend the reverse enable for a short time this would fix the issue. If it was triggered by a positive input I would think that I could just add a capacitor as a buffer to extend the time. Not sure how to do this with a negative input. Any ideas?

Thanks.

John Fields · Accepted Answer

---1. When you say "negative input" do you mean the signal goes from   12V to 0V, or it actually goes negative, i.e. below ground?2. When you say "motor enable" is that the same as "motor run"?3. Is there a real "Home" signal or is the "Reverse Enable" used to   turn off the light before the motor finds its home position?4. Are the "Enable" and "Reverse Enable" signals momentary or   continuous?---JF

Jim Thompson · Answer

A posted schematic or URL link would be nice.  As Confucius say...		                                        ...Jim Thompson-- | James E.Thompson, CTO                            |    mens     || Analog Innovations, Inc.                         |     et      || Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    || Phoenix, Arizona  85048    Skype: Contacts Only  |             || Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  || E-mail Icon at  |    1962     |                    Obama, All Blow but NO Hard

Martin Riddle · Answer

"Jim Thompson"  wrote in message news:If it is a 'Negative input' eg: -10v.  One could use a Max232, Built in charge pump for +10 & -10 volts.Cheers

John · Answer

Sorry, let's see if I can clear it up.The motor controller board input is where I'm tapping into the system.The inputs are: Constant 12VDC, Constant ground, Switched ground tomake the motor turn on and run ccw (motor stops when turned off) andswitched ground to make the motor reverse (will make motor run cw whenactivated).Yes, the CPU closes a connection to ground, when the motor controllerboard sees this it turns on the motor.The CPU has feedback from the motor (optical switches) so the CPUknows when the motor is home and stops the motor.Continuous and CPU controlled.This might clear it up a bit. motor is turning the head and I have added the light to his face.Occasionally (about 1 in 25) the light will stay on after the headrotates back.

Joerg · Answer

Essentially the same way as if it were positive. RC with a diode-resistor series combination across the resistor so it delays in one direction of change but much less in the other. Assuming that's what you want.

As others have said, draw a schematic and post it. A photgraphed hand-sketch is fine. Coffee stains and flux splatters on there are ok, Jan does that all the time. So are beer mug circles, but avoid squished mosquitoes :-)

BTW, if you want to turn negative control signal into positive ones the easiest and cheapest method are usually opto-couplers. They come as duals, quads and so on.

whit3rd · Answer

Only easy if your signal sources all can drive a few milliamps. Not terribly cheap, either.

You can also connect a negative (voltage-level) signal to a pullup resistor, but not a pullup to GND, rather to the emitter of a grounded-base transistor. The collector then has a positive voltage range and corresponding current to the (-) signal you started with. A suitable pullup resistor on the collector gives you back a positive-excursion-only signal that tracks the negative-excursion input.

John Fields · Answer

---I'm still kinda confused, but the way I understand it, so far, is thatif the thing starts docked, in the HOME position, then to start acycle you simultaneously pull down the MOTOR ENABLE line and set thelamp latch, which starts the motor turning CCW and the lamp turningon.Then, after a while, you force the MOTOR ENABLE line high, which stopsthe motor with the monster head at the far CCW end of its travel.The light continues to stay on, however, and after a while you pullboth the MOTOR ENABLE and REVERSE ENABLE lines low, causing the motorto turn on and rotate CW until it actuates the HOME switch, at whichpoint you pull MOTOR ENABLE high in order to stop the motor, pullREVERSE ENABLE high so that the motor will turn in the right directionthe next time around, and reset the latch so the lamp will turn off.Right?If so, the next question I have is: "Which latching rrlay are youusing and how are you setting and resetting it?"Coincidentally, what's stopping the monster head at the end of...

John · Answer

Yes, I built a little board that goes in between the input to the motor controller board and the board itself. My board only controls the light.

It's Bigfoot :) and the CPU is controlling it, I'm just using the CPU's signals to run the light. The signal is are capable of reasonably high current as they come from a TIP102 transistor (normally they drive a solenoid)

Your close, the enable that the motor controller sees is just a path to ground. That's how it triggers the solenoids that are typically there. The designers just used these "switched grounds" as a convenient way to control the motor.

On my board the motor enable latches a relay that also drives the light, another relay breaks this latch when the motor reverses. It works most of the time but occasionally the light gets stuck on. My guess is that the CPU turns off both the motor run and reverse signals (grounds) at about the same time and occasional this will not break the relay latch.

There are opto position sensors that feed back to the CPU.

John Fields · Answer

---If the light stays on from the time the motor first starts until itgoes home, that can't be right.--- ---Sounds to me like transients from the motor might be causing themischief, especially if you're using a latch and the CPU to controlthe lamp relay.Unless there was some compelling reason not to, what I'd do would beto divorce the lamp circuitry from everything but the opto and driveit directly, like this:View in Courier.            +12>----+--------+------+        |        |      |     +--|--------|-+    |         |  |        C |    |     |[LED]->  B   |  [LAMP]     |  |        E |    |     +--|--------|-+    |        |  OPTO  |      D        |        +----G NCH        |        |      S       [R]      [R]     |        |        |      |GND>----+--------+------+That way, the lamp will always be ON when bigfoot isn't home, OFF whenhe is, and there'll be no timing or glitch problems.If you _had_ to use a relay, you could do it like this:+12>----+--------+------+-----------+        |  ...

John Fields · Answer

---But, silly me forgot that your opto might be wired differently. :-(Regardless, if you can post a schematic (preferred) or an accuratedescription (OK) of which opto it is and how it's wired, I'll modifymy designs to suit.---JF

Jamie · Answer

the unlatching relay can have a CAP across the coil to maintain a hold for a short time once triggered.. Use a diode in series to prevent back feed so that the coil in the relay can benefit from the charge in the cap.

Buffer a negative input?

Join the Discussion

Didn't find your answer?