How to relay a progressive controller

[Tipsy]

Forgive my not finding the answer via searches but my limited knowledge means I don't understand most of what I'm reading so am unable to make an informed judgement. Perhaps it'll be better for me to ask directly.

I have a car injection kit that incorporates a progressive controller powering two electrovalves in parallel. The car's electrical system is ~12V-16Vdc and the electrovalves can draw up to 20A. I am advised the controller uses Pulse Width Modulation in the 10-50hz range.

I would like to "relay" the controller's output but am told normal 12v auto relays are not up to the job. Someone suggested I need a solid state relay while someone else said I just need a MOSFET.

While trying to source either SSR's or FET's, I get confused over the differing types. Can anyone help me source (in the UK) what I need to make this work safely in the given environment?

Thank you in advance.

 

[Tipsy]

Aynone please?

For instance, would this SSR be ok to use as stated above for a Pulse Width Modulation control on a 12V electrovalve?
**broken link removed**

 

[Tipsy]

Anyone please?
Looking at SSR's I see one's spec'd to handle DC input but they all say output is for 24+AC. Do they not make them for DC loads?

 

[Nigel Goodwin]

Not normally, because they use TRIAC's - which are AC only devices.

 

[Tipsy]

Thank you kindly Nigel. Can you suggest an alternative to "relay" a 12vDC PWM of ~20A draw? Normal auto "coil" relays can't handle pulsing, I've been told.

 

[smanches]

You'll probably have to go the MOSFET route. There are hundreds that will work for that application.

 

[Tipsy]

I see. Any particular type as it will be working in a harsh environment? When I looked into MOSFET's I noticed they generally need at least one other supporting component - a diode. I also suspect at 20A (possibly surge higher) that it would need a heatsink. Are there any other components I'd need to buy at the same time?
Thanks smanches

 

[smanches]

I would google PWM motor controls or PWM relay control. Although you don't need the PWM part, you can just create the control signal yourself. You should still get a good idea on the switching components that are needed. There are usually dedicated parts for automotive. Just look for those when you're searching for specific parts. ST makes a lot of automotive stuff that I have used before.

At the very minimum (depending on how robust you want it) you will need:
Mosfet
protection diode for the mosfet (for backEMF)
anti-ringing cap for the mosfet (across drain/source)
gate drive resistor
gate discharge resistor
input filter cap (for entire system)
input transorb (for entire system)

Optional:
Dedicated mosfet driver instead of the gate resistors
15V zener to help protect the gate from transients (if the mosfet driver doesn't have one)

There are 1000s of these schematics out there. Just look at the common things around the mosfet itself and it should start becoming clear. Not all contain everything I mentioned above, but should have most.

 

[Tipsy]

Ok, bye for now and thanks! I'll have to get back to you about the supporting components (when I know what I'm talking about).

 

[Tipsy]

I have been researching PWM circuits and I can see me being enticed to build my own but time isn't a luxury I have for this project at the moment so....

Can anyone see if this SSR will be ok to use? Fleabay item number 250538409676
And would it be suitable for both solenoids (27A max) and motors (17A max) ?

 

[Sceadwian]

Those relays are for switching high voltage DC, not really useful in your application. You'd be better off getting some 20 amp power mosfets. You're only pulsing current through them and Mosfets have some pretty high pulse current limits well above their continuously rated current. The actual overall power you'll be switching isn't that high, you might not even need a heatsink. I'm not sure of the actual construction of an SSR relay like that but I'm pretty sure a Mosfet is going to have much better turn on and off characteristics than that type of relay.

 

[Tipsy]

So I may learn, can you please explain why you say the given Control and Load ranges of 3-32Vdc and 5-220Vdc are not suitable for 12-16Vdc?

Of course, I will continue learning about MOSFET's and PWM control as advised but I'd hate to pass up a quick solution if the SSR can provide it.

Much appreciated

 

[Sceadwian]

Tipsy you'd need a detailed datasheet of the SSR before I could say for absolute fact that that SSR wouldn't work for your application, only thing I'm worried about are the switching times. If that SSR is isolated it's probably via an internal optoisolator and those have relativly slow turn on and off times. I'm not sure why you think an SSR is a quicker solution than a FET. A FET actually has 1 less terminal than that SSR does and the switching speeeds are quiet fast. They're pretty easy to use as well, for a basic application all you need beside the FET is a gate resistor to ground to keep the gate from floating when the circuit was off.

 

[Tipsy]

Sceadwian, I did learn how switching times and delays etc need to be carefully calculated but the number of steps to reach the end result left my head spinning so I'm still none the wiser at this point.

I have asked for and got a datasheet from the SSR seller but it was just one page and not the array of info I'm used to seeing. It's also a bitmap so unsure how to copy it for you to peruse.

However, I seem to be going in circles with the differing advice I've been getting. The reason I assume an SSR would be more convient is because it's an all-in-one package with just four terminals. The reason I assumed MOSFET's were more complex was because I believe they need supportive circuitry in the way of protective components but I have since been shown an example where this is not case.

In which case, I'm all for going the MOSFET route as long as I can be sure I won't be damaging the controller it will be relaying or anything else attached to it and that the MOSFET will remain reliable.

 

[Sceadwian]

Even is supporting components are need they're not complex. Generally speakign all you need is a protection diode if the load is highly inductive and a gate resistor to avoid static zap from killing the mosfet gate when it's off, they're very easy to use. For an N channel FET you'd simply gruond the source lead and connect the loads ground terminal to the drain of the FET. a 10k resistor from the gate to ground and then any time you get over about 4 folds the FET will switch on, keep in mind most FETS require between 10 and 14 volts at the gate to turn fully on at it's rated current. The datasheet for a FET can be quiet intimidating, but they're really no that bad. One nice thing about FETS is they're usually capable of some pretty serious pulse currents, so you might be able to get away with a 5 or 10amp FET for your 20+amp pulses.

 

[Tipsy]

The load in this case is a pair (~10A & ~7A) of solenoids/electrovalves and I'm unsure if they are highly inductive. How can I tell?

When you say "an N channel Fet" what exactly should I be looking for? Something like IRF840/ IRF640 but with a higher amp rating and/or could I use two or more in parallel as they seem cheap enough?


My controller output is via the -ve output, where the load +ve is direct to the car battery's 12v+ve terminal. Does this make any difference to the Fet I should choose and the supportive component placement?

 

[Sceadwian]

Most common things that turn electricity into high degrees of motion/force are highly inductive.

 

[Tipsy]

Thanks for your help Sceadwian, I'll try and work out exactly which components I should go for and take it from there.

 

[Tipsy]

I've sourced a pair of BUZ11 MOSFET's to play with. Are these suitable for my intended application?

30A, 50V, 0.0040ohm, N-Channel Power Mosfet.
Nanosecond switching speeds.
Linear Transfer Characteristics
High Input Impedance

 

[smanches]

Those should work fine.