Looking for electronic switching device ideas

Dr.Jeff

Dr.Jeff

True Classic
I'm building a piece of test equipment for a portion of my turbo X1/9 project, and need to come up with a "switching" or "pulsing" or "cycling" circuit. As everyone here should know by now, this is certainly not my area of expertise. So I'd appreciate some help. Hopefully there is some sort of device out there that will do this without a big investment (it may only get used a couple of times). Seems there are endless little electrical devices/circuits/modules listed on the Chinese sites for very little money, and I'm sure one of them is for this purpose - but I do not know what it would be called or which one to look for.

Basically I need to switch 12V on and off (does not matter if the positive or negative side is switched), again and again, rapidly cycling the circuit's power. The difficult part is the flash rate should be pretty high, around 20-30 times per second.

As for the load it needs to carry, approximately 40 amps normal (when the power is "on"). So a little cushion beyond that would be a good idea. If it is easier or beneficial to do so, maybe this can be split into four separate units, each with 10 amps capacity (other characteristics the same). Or maybe one switching controller with four relays splitting the load (if relays can operate that quickly)?

I'm sure there are other pieces of data needed but someone will have to tell me what they are. As can be seen, this isn't a precise or sensitive instrument. It will not be interfacing with any other equipment beyond the power source and the item being cycled, and will be performed on the work bench for testing only. I just need a way to allow power to be cycled on and off at a quick rate, continuously for up to 4 hours (which I'll control manually with a simple switch). I should be able to come up with an external 12v power source to drive this. So the switching portion is my main focus at this point.

Please tell me what else needs to be considered or suggestions for how to do this as simply and affordably as possible. Thanks
 
You might be able to find a suitable solid state relay to use as the power switching device. You could make a simple variable frequency pulse generator to drive it using a 555 timer IC.
 
The terminology you use (solid state relay and 555 timer) came up in my prior search on the subject. My problem is I'm really not able to build such a circuit, so I need to find some existing unit. Any references for something specific that does what you describe? Thanks
 
Edit - This sounds like you are making an injector controller. I'm no expert, but I think it would use an IGBT or MOSFET and a PWM driver. The typical eBay/Amazon PWM DC motor controllers run at several kHz so they would need modifying to lower the pulse rate.
 
Last edited:
A LM555, like previously suggested would do the trick in a monostable configuration (see section 7.4.1 of the attached datasheet). The output would need to drive a darlington pair of transistors to get the current driving capability you are looking for.

I built a very similar circuit to replace the Sepia time delay relay (for the door chimes in the X1/9), but it was in astable configuration.

For reference: https://xwebforums.com/forum/index.php?posts/282160/

Alternatively, a pulse generator or a PWM controller could be used easily - however you would still likely need to drive a darlington pair to achieve the current rating you want to switch.
 

Attachments

  • lm555.pdf
    1.5 MB · Views: 134
Great input, thanks to all.

Wish I could build some of the stuff described though. :oops:
I found several assembled controllers (timers) that use the LM555, but all of them say the fastest switching rate is .1 sec. While that could work, I'd prefer something faster. There is another one that says it can go to .01 sec; it really does more than I need but it's the only one I've found so far that has that cycling capability. Unfortunately ALL of the ready made units I've come across have a max rating of 10 amps (and frankly I suspect they are much lower than that in reality). I even thought about using four of them to split the load, but that's rather messy. So I looked into using one of them to drive a solid state relay capable of handling at least 40 amps. Unfortunately I haven't found such a component yet (at least not one that was affordable for this project). If someone could please explain more about either solid state relays (like some specific ones to look for) or what "a darlington pair of transistors" means I'd appreciate it. :)

And Rachael is almost right. This is going to be a part of a test rig for injection systems. With it only being something of an experimental or temporary item (likely to only use it once or twice ever), I really do not wish to invest a lot of funds in it. ;)
 
Jefco said:
https://en.wikipedia.org/wiki/Darlington_transistor
Click to expand...
Thanks, now I understand.

I believe I've found an arrangement that might work. It consists of a assembled timer/switching controller capable of .01 sec rate, with separately adjustable "on" and "off" times, but a max 10A load. Plus a solid state relay capable of a choice of loads (e.g. 40A, 60A, etc). From what I can tell the controller would trigger the relay, which can handle the total load for my project. The only thing I'm still not certain about is it recommends adding a diode across the load. That is something I can manage, but what diode? Sorry to know so little about this stuff - but that's why I'm asking.
Can anyone specify what sort of diode is appropriate to add across the load for protecting a 60A, 12V, DC-DC, SSR?

The neat thing is all of these components will be about $10 - even better than expected. Frankly I really do not know if it will work as I think it should. But for that cost it's worth a try.
 
The 555 can be made to operate over a fairly wide range of frequency by changing the values of the external components. The data sheet and app notes should have that information. If the power device is switching an AC signal, you could use a relay containing an SCR or Triac. If it is DC only, a relay containing a DMOS power FET would work or you might drive a DMOS device directly with the 555 depending on whether or not you want more of a turnkey solution or want to do some soldering. You could also use a bipolar Darlington device as mentioned above but you may need to pay a little more attention to the impedance at the base of the output transistor if you are going to crank the frequency up a bit. Usually, a few thousand ohms across the base-emitter junction of the output device should do the job. The other issue with using the bipolar device is that depending on how much the output transistor is overdriven, there can be significant storage delay when it goes into saturation. Probably not an issue at 100Hz.
 
I found this rather bizarre video (with thanks to ADP Training). The circuit shown may be of use to you and looks cheap to make. With this circuit you can vary both the frequency (RPM) and duty cycle (PWM). The circuit shown has no gate resistor on the MOSFET, which is a bit unkind to the LM393 but probably ok if driving only one MOSFET. Switching the MOSFET on or off is in effect momentarily shorting the LM393 output to ground via the gate capacitance, and adding a gate resistor will limit the gate charging current to a manageable maximum level. The LM393 output can supply 20 mA, so to drive 4 MOSFETs at 5 mA each, I think you would need a (14 V / .005 A) = 2.8 kohm gate resistor on each MOSFET. Hopefully someone with knowledge can chime in if I have it wrong!
 
Without knowing exactly what you're doing, microsquirt has an injector test mode that would allow you to experiment with injectors. You could set up a rig like this to measure output as well.
 
That's a nice simple little circuit for driving an injector. The LM393 is an open collector device so it is not turning on T1. Instead T1 is turned on by the 2.2k resistor. The LM393 is turning off T1 by shunting the gate to ground. Regardless, I don't think Jeff is going to build the circuit as it is not his expertise.

Jeff, what are the specs of the injectors you are using (resistance)? As for a diode, a "common as nails" 1N4001 should easily do the job. It needs to go across the injector in the reverse direction (cathode to positive).

Brian
 
tonyism7 said:
You could set up a rig like this to measure output as well.
Click to expand...
Tony, that rig is pretty much what I already have but using some lab grade hardware I had left over. Built it many years ago after I'd finished some lab research work.
I haven't viewed the video yet so I don't know if they addressed the method of activating the injectors. I'd like to simulate something similar to "real world" in having them pulsed rapidly rather than just steady stream (which is what I've been doing in the past). Ideally it will have a longer closed duration than open and I can set those times/rates.
The little timer circuit board thing I came across includes a display and some buttons so you can choose various functions and set various parameters. I think that will allow me to do this. But I do wish I could follow the suggestions given and build something to do exactly what I want. It is just a cheap Chinese item, I'll post the link below. Hopefully it is good enough for this temporary use.



beezee said:
Regardless, I don't think Jeff is going to build the circuit as it is not his expertise.
Click to expand...
You know me well Brian. When I was very young I thought I could do anything. Over the years I've come to realize (the hard way) that simply isn't true. I now accept that there are some things I do really well, a lot of things I can do but not particularly good or bad, and some things I cannot do very well at all. Electronics falls into the last category. Sad part is I once actually took a class in basic electronics and got an "A". But that only meant I could do the math on a multiple choice written test where a diagram was given and some specific values were left blank. I really never learned what was going on, and we never did any actual building of circuits, etc. Many times I've soldered wires to make repairs, but I really suck at even that. I've read things to understand something, looked up all the terminology, thought I got the general idea, only to forget it by the next occasion when another question came up. I guess it just isn't my thing. Which is why I highly admire all of you that do get it.

Brian, each injector has a peak of roughly 6 amps and average of about 3 amps. However I'd like to be able to run 8 injectors at the same time.
Which brings up a question about the diode. According to the pictures I found using the solid state relay, they installed one diode across the positive and negative leads before and after the load (injectors in this case). Since I'll be running all of the injectors together, it should be easier to have one SS relay capable of handling all of the injectors rather than 8 smaller ones. So do I need to install a separate diode over each injector, or one diode for the common leads that will power all 8 of them together?
And about the value of the diode, should it be enough amperage to handle the maximum load of the system, or does this only need to be a small value (same question for voltage). I actually have a couple spare 1N4001 diodes but did not think 1 amp was enough. As I understand, the purpose is to help protect the SSR from spikes.

The cheap controller I found:
https://www.aliexpress.com/item/4000089701886.html?spm=a2g0s.8937460.0.0.3a342e0eE7FTOw

Thanks for the help.
 
The injectors contain a coil wire which acts as an electromagnet and when energized it opens the injector. When the coil is denergized the magnetic field collapses which tries to keep the current flowing. WIth no path for the current the voltage can rise high enough to damage the control transistor. The diode provides a path for the current to flow thus preventing any damage. Since the diode only conducts when the injector turns off, it does not need to be rated for the injector current, only for the current generated when the magnetic field in the injector collapses. Now to determine that current is a little more complicated... Simple put, a 1N4001 should suffice.

In regards to the controller in your link, it uses a standard mechanical relay which will not switch fast enough for your needs.

Brian
 
beezee said:
Simple put, a 1N4001 should suffice
Click to expand...
Thanks Brian, that's great since I already have a couple of those. Do I need to put one diode across each injector, or will it work the same to have one diode for all of the injectors collectively. To explain, all of the injectors' electrical leads will be combined and driven together from one source. So possibly a single diode across the two leads before it splits to each injector? Might look a little like this:
282301206_L_f21448c8-aec1-4b22-ad3a-ad726e38379f.jpg


beezee said:
In regards to the controller in your link, it uses a standard mechanical relay which will not switch fast enough for your needs.
Click to expand...
I was afraid it was too inexpensive to work. The description says the timer can be set to .01 sec intervals. But I guess that means nothing if the relay can't keep up. I assumed it must have had some sort of solid state relay onboard to work with the timers potential speed. Back to the drawing board.

Would it be easy to remove the onboard relay and wire the rest of the circuit directly to the solid state relay I intended to use in order to handle the total load?
 
Perhaps you're overcomplicating the switching. If electronics aren't your thing, does it have to be electronically switched? Could you use a rotary switch for a light bulb & and a hand crank - geared 15:1 to turn the switch at 30Hz when you crank at 120 beats-per-minute, listening to the Circus March?

However you switch it, for 40 Amps, you will still need a relay.
Nomenclature for Solid State Relays goes something like SSR (Amperage Rating) (A or D for AC/DC control voltage) (A or D for AC/DC switched voltage).
So you're looking for an SSR 40DD, similar to this. I'd suggest getting one that includes the heat-sink, so you don't cook it and need a replacement in short order.
 
Dr.Jeff said:
Thanks Brian, that's great since I already have a couple of those. Do I need to put one diode across each injector, or will it work the same to have one diode for all of the injectors collectively.
Click to expand...

One diode should suffice.

Brian
 
Russe11 said:
So you're looking for an SSR 40DD, similar to this.
Click to expand...
Thanks. That's what I had found and intended to use with the controller I posted earlier. The nice part is only one relay of this type would be necessary to run all 8 injectors simultaneously (but with 60 amp capacity). But now I need to rethink the switching controller mechanism.
Honestly I will not be cranking anything by hand for several hours. o_O The simplest approach would be to use a 12V flasher to trigger the relay. The fastest I saw was "up to 200 cycles per minute", which isn't nearly as fast as I'd like. But if nothing better comes up I suppose it would be an option. Having the dual rate - "off" time around 20 times longer than the "on" time - would not be as critical with a slow flash rate (like the above idea).
 
You must log in or register to reply here.
Back
Top