Replacing a swtich and a relay by transistor, can it be done ?

[DBoulanger]

Hi everyone,

I'm actually experimenting with a Delay-Off and Delay-On circuits using NE555 timers.

I came across this circuit for the Delay-Off section and it works like a treat with a relay and a physical switch.  The thing here is that I'm not too sure how too proceed with the substitution of the relay and the switch.

Actually, I'm not sure if I should use a NPN or PNP transistor for either one of them (the S1 switch and the relay).  I'm not familiar enough with bipolar transitors to actually modify an existing circuit and be sure that I won't smell some "magic smoke".

The toy/project I'm trying to come up with is a somehow versatile Load Sensing switch.  I want to be able to adjust the Delay-On and Delay-Off, within 1 to 60 secs or so for each.  I kinda have an overall idea how to inter-relate the different blocks, but I need have the minimal amount of mechanical parts and use mainly discrete components with the absolute minimum of switches and relays.

So here is the Delay-Off circuit I'd like to customize with BJT transistors for the switching portion of the design.

Thanks in advance for you help.

[con-f-use]

It depends very much on what kind of load you want to switch. First your transistor needs to handle current and voltage of your load. Inductive loads require protection for your switching transistor. If your load is ground-referenced, you should switch the non-ground line, which might require you to use a PNP transistor.

[DBoulanger]

Hi,

There will be very minimal load to be controlled by this circuit.

In fact, this circuit will actually power up a Delay-On circuit which in turn will trigger a Triac to activate the final portion of the chain.  The load itself will be running at 120V AC.  The overall idea is to start an appliance, such as a vacuum cleaner, few seconds after the tool I'm using has been put in function and then turn off the vacuum cleaner few seconds after I turn off the tool, so that everything has been cleaned before the vacuum shuts down.

I was thinking of using a Triac but if a relay is more appropriate, I most likely will adjust accordingly.  Afterall, this little project is simply to get a better understanding of how the things work together and have some fun.

I could have considered to buy something already built, but that would have ruined the fun, right ?

There's nothing like something "Made at Home" and having a personnal touch to it, right ?

Dan.

[minime72706]

Not that this is helpful, but what is your beef with relays?

[con-f-use]

 - Clunky
 - Noisy
 - Slow
 - Current hungry
 - Moving Parts, Sparks -> Abrasion

So what voltage/current will you need to switch the triac on (datasheet)? In most cases you should be able to switch your triac directly from the 555 without a transistor.  Also I'd suggest to use a solid state relay that switches at zero-crossing instead of the triac.

[minime72706]

What will the delay circuit that you are trying to design be powered with? Full line voltage as well?

[DBoulanger]

Firstly, I don't have anything against relays, I can cope with them quite well. 

However, minimizing the physical dimensions of my finished product is one of my reasons to substitute switches/relays for transistors.

Also, knowing that the Delay-On and Delay-Off circuits will barely draw 40ma overall, that includes the leds current too, I have a hard time to justify using a relay that will draw as much current if not more, simply to drive very basic circuits.

If I'm sticking to the circuit samples I found on the net, I would end up with 3 relays and one of them will need to be able to handle loads of 10-20Amps under 120V.

So the final product compactness, minimal use of mechanical parts and learning different ways of doing things, pretty much sum up the reasons why I'm leading towards the "relayless" approach.

I know there are many ways to achieve similar/identical results, so overall this is more of a choice than a need.

Thanks for your input anyhow.

[minime72706]

Answer my recent question.

[DBoulanger]

- Clunky
 - Noisy
 - Slow
 - Current hungry
 - Moving Parts, Sparks -> Abrasion

You read my mind or there is a hidden camera/mic somewhere !

[DBoulanger]

What will the delay circuit that you are trying to design be powered with? Full line voltage as well?

I thought of different ways, but I haven't made a final decision yet.

Since the overall current consumption will be fairly low, and if I'm successful at converting the relays for transistors, an embedded transformerless power supply should do the trick.  A regular wall adapter alternative could be considered as a potential approach, connected externally, although it defeats my initial compactness idea.

If I can put my hands on a small transformer, I can integrate a small PSU on the board as well, but again, that would require precious space too.

At first glance, a transformerless power supply seems to be my best candidate, but nothing is casted in stone as of now.

[minime72706]

What will the delay circuit that you are trying to design be powered with? Full line voltage as well?

I thought of different ways, but I haven't made a final decision yet.

Since the overall current consumption will be fairly low, and if I'm successful at converting the relays for transistors, an embedded transformerless power supply should do the trick.  A regular wall adapter alternative could be considered as a potential approach, connected externally, although it defeats my initial compactness idea.

If I can put my hands on a small transformer, I can integrate a small PSU on the board as well, but again, that would require precious space too.

At first glance, a transformerless power supply seems to be my best candidate, but nothing is casted in stone as of now.

Capacitive/Resistive power supplies are inefficient and lack isolation, but they do work. The most annoying thing about them is that they draw the same amount of current regardless of whether or not they have a load attached. You end up burning the power off in the zener diode when you're not burning it off in a load.
So your goal is to delay the turn-on of a load which may be a small appliance that takes 120VAC? Is this correct?

[DBoulanger]

My goal is to turn on and turn off and appliance on 120VAC, with appropriate delays, when another appliance is turned on.

My approach is as follow, although some timing details need too be addressed :

) Appliance A is turned ON
) A current sense circuit is detecting activity, so it activates the Delay-Off circuit
) As long as there a current flow, the S1 switch in the Delay-Off circuit is held down, therefore the timer isn't started yet.
) Upon startup, the Delay-Off circuit activates the Delay-On one, the Delay-On circuit being the Delay-Off circuit's load.
) If the current stops flowing before the Delay-On circuit times out, the Delay-On circuit will be reset, or turned off (to be determined later), so that it doesn't trigger the Triac unnecessarly and is ready for the next event.
) When the Delay-On circuit times out, it will trigger a Triac to start appliance B
) When appliance A is turned off, the S1 switch of the Delay-Off circuit is released, therefore the shutdown countdown is started. If appliance A is restarted before the Delay-Off circuit times out, then the S1 switch will be re-engaged and therefore will nullify the countdown.
) Once the Delay-Off circuit times out, it will turn off the Delay-On circuit (un-triggering the Triac at the same time so that appliance B is turned off), and will turn itself off altogether.
) At this point, the only thing left on will be the current sensor, monitoring activities of appliance A.

I understand it might be a bit difficult to visualize the whole thing when reading it for the first time, but I think it is explaining my approach clearly enough.

Again, keep in mind that this little project is uniquely a learning platform for me, allowing me to learn new material, practicing and of course, having a good time at doing it.

Sharing ideas and concepts with other enthousiasts is a plus too, so it's a win-win situation !

[minime72706]

I'm interested in helping you, but I'm not functioning at a high-enough level at this moment in time. I'll figure something out in the next few hours. (I tend to, unfortunately, "wake up" in the mid-evening thanks to my lovely depressive brain chemistry)

BTW, I've seen issues arise with AC current sense circuits; however, I don't yet understand what you mean. Are you planning to use a current-sense transformer or some other type of circuit to detect when current is flowing in the hot wire leading to appliance A?

How much current will the appliances themselves be drawing? Would it be an arbitrary amount up to a number of amps or something more modest?

Using my current amount of understanding of your problem, I would replace your switch with a FET if what it will be driving is a similar delay circuit.



I'm designing something similar, but simpler, for my own practical uses. Winter is approaching and the air is getting dry, so my boyfriend set up the humidifier. The only problem is that it would need to be refilled multiple times a day. We had a pressure-rated water pump sitting around so we filled a large barrel with purified water and the required bacteriostatic agent. The pump brings the water from the basement to the first humidifier on the first floor and we intend to add another on the second floor. My boyfriend immediately complained that the power supply I gave him fails to start itself up with the pump attached. A home automation system switches a relay that provides power to the switching module and the immediate presence of the large pump load doesn't allow it to start up. I already made a relay-type delay circuit, but I decided this was a good excuse to use a SCR. My delay generator consists of a FET with a zener on the gate (something like 7-9V for a 12V system) and a RC circuit driving it. The transistor doesn't turn on until the voltage at the RC node is roughly the zener voltage plus the FET's threshold voltage. It works pretty well and the zener is a rather good idea in simple situations like this because you don't have much flexibility if your FET triggers at something like 1 or 2 volts.

[DBoulanger]

As a matter of fact, yes, I'm planning to use a current sense transformer or simply a split core transformer.  I was thinking about an opamp, a comparator or possibly a combination of some sort.  It doesn't have to be precise, but sensitive enough to detect relatively light loads.

Appliance A is expected to draw a current in the 0.4 to 8 Amps AC, depending of the tool being used.

For appliance B, I'm a bit unsure, but most of the time this will be an industrial vacuum cleaner, so 5-7 amps should pretty much cover it, but I will aim a bit higher, just to be on the safe side.

BTW, Appliance A is controlled manually, regarding the Power On and Power Off actions.

Overall this is some sort of "follow me" automated switch, with adjustable ON/OFF delays, no more no less, nothing really fancy, just a little personnal touch.

As of now, my first goal is to minimize the overall physical size as much as I can.  I think the relay -> transistor substitution is a good start to get the overall thing smaller.

The other elements that will compose the project are regular through-holes discrete components with potentially an exception, the Triac, which may require a heatsink, but I'm not considering it as a show stopper.

So for the moment being, I'll try to substitute the relay of a transistor, keeping the same fonctionality, otherwise I will have to modify the circuit and determine a viable alternative.

Back to the workbench !

[minime72706]

So basically you're trying to make your industrial vacuum cleaner (wall-mount?) turn on when you use something like a drill press or saw or what-not? i.e. anything that might generate dust.

I would check the consumption of your vacuum - residential uprights at least claim to use 10-12A.

A hall effect sensor is a good choice, though it probably has to be pretty sensitive. Remember that you have to split the conductors apart or they will cancel out each others' magnetic field.

If you use TRIACS for both these purposes, I would highly recommend the BTA24-600, which is rated for 25A, but in practical situations, I don't think designers let the current approach that level. I would DEFINITELY use heatsinks on both.

BTW relays do come in pretty small packages and those small packages can be shorter, at the very least, than a TO-220 semiconductor device. Just a thought.

-----

If I understand now, I figure that you want the switch in the example circuit to be replaced by a transistor that would be triggered by the detection of current passing through 'Appliance A', correct? This will start a timer which, if allowed to expire, will trigger both 'Appliance B' as well as a second timing circuit. This timer will then be used to shut off 'Appliance B' after some period of time? Is that correct? I'm not re-reading your previous posts because I'm lazy and I'm getting far enough that I am now unsure, so please correct me.

BTW since you expect your 'Appliance B' to be a specific vacuum cleaner, you should be a bad ass and actually custom-tailor a snubber circuit for that particular power factor. Rarely in applications I play with TRIACS do I actually have a specific load in mind. In that case you just throw something down, such as the suggestion in the datasheet, which is not ideal.

So I can understand avoiding relays, but given the environment this might be used in, why is limiting the size your main goal?

[DBoulanger]

Hi again,

If we omit potential false triggering for the moment being, the scenario would follow these lines :

) Current sensor starts the Delay-Off circuit
) Delay-Off circuit starts Delay-On circuit
) Delay-On circuit triggers the Triac on time out, vacuum starts.
) Current sensor "sees" that the tool is no longer used therefore it stops the Delay-Off circuit
) Delay-Off circuits times out, shut itself down along with the Delay-On circuit

The triac will be exclusively used to start the vacuum or any other controlled load.  The main tool is not controlled by any electronic component, it is strictly standalone.

The relay approach is not really out of question, but I'm considering it as my plan B, as I actually had in mind to make it as compact as possible.  In fact, I have no environment restriction as such, so the box could be the size of of shoe box and tha't would be just fine.  So the dimensions of the finished product is only a personnal goal or taste.

For the snubber circuit, I was aware of it when I did my researches on Triacs and other thingies.  I have some special resin and epoxy, normally used for garage floors, which is rock solid, so I figured to put the X-Type capacitor and the Resistor in a little mold, with 2 pins or connectors of some sort, so that the final snubber can be exchanged, replaced in a modular way.  So if I need to control a load which is way less or more inductive, or any other characteristic impacting the snubber specs, I could simply swap the snubber as easily as changing a car fuse.

Before starting this project, I knew I'd face a lot of new things, and that's the beauty of it.  I'm not in an absolute need for that "current load sensing master/slave" thingie, or whatever you want to call it, I simply decided to give it a try, exploring new avenues, learning more about electronics.

I could go the easy way, using relays to simplify my life and the overall construction, but that would defeat the purpose of the whole thing, the challenge !

If I opt for the relays, I'm pretty sure the whole thing would be completed rather quickly, and that's OK.  On the other hand, if I make it relayless, I'll have a better feeling about it.

Overall, this is strictly for a hobby, so time and efforts are not really a concern.  I don't give up easily, believe me !

Keep in mind that I'm not ignoring the relay approach, I simply consider it as my plan B.

Note: I've been successful at "transistorizing" the switch S1 on the Delay-Off circuit.  It might not be much for some of the readers, but for me, that's one step ahead.

Best regards.