Latch does not work with transistor loaded

[vajirkar]

I have a latch driving a transistor (indirectly through a few gates).
The transistor is used to drive a relay that drives a 110V linear actuator motor.
When the relay is not connected to the actuator motor, the latch works perfectly. A single button click is enough to set the latch, the reset button then resets it.
When the relay is connected to the actuator motor load, the latching is less reliable. Sometimes multiple button clicks are needed to set the latch.
Another point worth mentioning is that connection from the AND gate to the Limit Switches (LS1 and LS2) is a little far. About 2 feet of twisted wires connects the output of the AND gate to the Limit Switches.
I'm not sure if its the load that's causing this or the length of the wire.
BTW I have replaced the CMOS AND gate with LSTTL, but that did not help the latching issue.

View attachment 35684

 

[vajirkar]

I was thinking of adding an extra transistor driver between the AND gate output and the limit switch. Dunno if its going to help.

- Sid

View attachment 35685

 

[hotwaterwizard2]

Long wires tend to create a voltage loss. Try changing the resistor to a smaller size to compensate. Or try this circuit with the 2N2222 instead of the 2N3904.

View attachment 35686

 

[hotwaterwizard2]

Here is the MOD just try it.

View attachment 35687

 

[MP1]

vajirkar,
what is the current needed for the relay? Perhaps you have chosen the wrong resistors and it is drawing down the supply.
If you refer to the 10K as R1 and the 1M as R2, then:
R1 = Supply Voltage / ( Maximum Current Required / Minimum HFE * 1.3 )
R2 = 10 * R1.

MP

 

[GPG1]

Perhaps the wires connected to U4d.e are pcking up hash from the relay closing with a load and resetting the FF. Try putting a RC filter on U4d,e inputs.

 

[vajirkar]

I tried to narrow down the problem and it seems to be related to length of the wires to the limit switches. I simulated the limit switches by a pair of jumpers on the board and the latching is much more reliable. So its possible that in rare circumstances the relay-switching noise does affect the circuit but its probably not the root cause here.

With the same values of resistors (10K, 1M) I experimented using a darlington pair with hFEmin = 25000. The problem still persists. I'll try hotwaterwizard's mod when I get a chance.

- Sid

 

[audioguru2]

Hi Vajirkar,
So this is the rest of your "sparking relays" circuit.
I see the same problem as GPG did:
The inverters (which I hope are CMOS, not TTL) that are connected to the limit switches have a very high-value resistor of 1M on their inputs, and also unshielded "antennas" from the limit switches. A perfect way to pickup hash and your twisted wiring may make it worse. Change the resistors to 10K, and use shielded audio cables for the wires to the inverters' inputs. Connect the shield to ground only at the inverters' end. Try to keep the limit switches away from the relays, motor and mains wiring.
You have the most complicated "flip-flop" that I have ever seen.

 

[vajirkar]

Audioguru,
Yeah... that's the rest of the circuit.
Thanks for the techniques/precautions.

About the complex "flip-flop" , I'd love to hear comments and critical and otherwise. How can I improve this? I would have just loved to avoid the mess by programming the microcontroller myself. I just don't have all the tools for that. Maybe building a programmer may be a good investment in itself. In anycase comments are really welcome!

- Sid

 

[GPG1]

This is what I mean

View attachment 35697

 

[audioguru2]

Hi Sid,
I hope that you aren't mixing regular CMOS with TTL logic ICs, because they are not compatable without making changes. The "high" output voltage (2.4V minimum) from a TTL IC will not always be high enough to be recognised as a "high" input voltage (3.5V minimum) at the input of a CMOS IC. Usually a 2.2K pullup resistor is added to the the TTL's output to make them compatable. That's why 74HCXX CMOS ICs have the optional 74HCTXX family available.

Your schematic is not correct because it lists both CMOS and TTL ICs as substitutes. If you use a 7404 for U4-d and U4-e with their 1M input pull-down resistors, then your circuit certainly won't work. Even a 74LS04 (minimum input voltage is 0.8V, maximum input current is -0.4mA) with a 2.2K input pull-down resistor will not always recognise the input voltage as a logic "low".