Solid state relay for 48 V 5 A load

[jelliott]

Is there such a thing as a solid state 'relay' (i.e. a PWM-compatible switching device) that actually behaves like an electromechanical relay, in terms of being switchable with ground OR positive voltage to the 'coil?' And can handle switching a 5 A load at 48 V, up 100% duty cycle, with only 12 V across the 'coil?'

Background: I've built a circuit (based on **broken link removed**) around a CD4047 monostable vibrator IC, hoping to modulate the power output of an electric heater based on the tach output of a brushless DC fan. This works well enough, but the whole idea for this circuit was an offshoot of my original idea for a low-speed cutoff (see this thread; the third post therein sent me off on this train of thought), and of course when the fan stops spinning, its open-collector tach (and my pull-up resistor) provide a perpetual 'high' input to the CD4047, defeating the original purpose.

It occurs to me that if I had a switching device that required both ground and +12 V to energize the load (like an electromechanical relay), I could switch the +12V with the CD4047 circuit and switch ground with the original LM2907 circuit (discussed at length in the aforementioned thread). But I'd also be open to suggestions that would allow use of the CD4047 circuit exclusively (e.g. how to adapt the fan's tach output to the CD4047 to go 'low' when it stops spinning).

 

[Reloadron]

Is there such a thing as a solid state 'relay' (i.e. a PWM-compatible switching device) that actually behaves like an electromechanical relay, in terms of being switchable with ground OR positive voltage to the 'coil?' And can handle switching a 5 A load at 48 V, up 100% duty cycle, with only 12 V across the 'coil?'

Yes, an SSR (Solid State Relay). However, is the heater an AC or DC element? Also, what is the heater voltage?

Beyond the CD4047 I have done similar circuits using a 74123, check out this link for an example. What you are building is a retriggerable one shot multivibrator. See page 11 of the link, Figure 12. Missing-Pulse Detector. Typical fans output 2 pulses per revolution I believe. So you set a R and C time to where if the incoming pulses from the fan drop below a certain RPM output or no output. Then use the Q or /Q output to do whatever you wish to do. You would likely use a CMOS version like CD74HC123 or HD74HC123A.

If the fan speed returns the chip will toggle. That can be changed by adding another chip depending on what you want to do. Anyway, I would likely start with a missing pulse detector approach.

Ron

 

[jelliott]

Yes, an SSR (Solid State Relay). However, is the heater an AC or DC element? Also, what is the heater voltage

The heater element is 48 V DC.

SSRs I'm familiar with stipulate that the 'coil' ground (Pin 85 if you speak DIN 72552) always be grounded, precluding me from manipulating both 86 and 85 as I could with a real relay.

Beyond the CD4047 I have done similar circuits using a 74123, check out this link for an example. What you are building is a retriggerable one shot multivibrator. See page 11 of the link, Figure 12. Missing-Pulse Detector. Typical fans output 2 pulses per revolution I believe. So you set a R and C time to where if the incoming pulses from the fan drop below a certain RPM output or no output. Then use the Q or /Q output to do whatever you wish to do. You would likely use a CMOS version like CD74HC123 or HD74HC123A.

If the fan speed returns the chip will toggle. That can be changed by adding another chip depending on what you want to do. Anyway, I would likely start with a missing pulse detector approach.

Looking at that Figure 12, it only shows an example with the pulsed input going low; what happens if it goes high (as is the case when my fan stops spinning)?

 

[Reloadron]

I'll have more for you tomorrow. There are DC SSRs just AC is more common.

Ron

 

[dr pepper]

If your electronically minded you could use a igbt module, there are dc i/p dc o/p ssr's sounds like ron knows of one for your application.
You can get ssrs these days that can accept just about anything on the i/p ac or dc, I've used a siemens one recently that takes 10v - 415v ac or dc on the 'coil', but as ron says ac such as this one are more common.

 

[alec_t]

how to adapt the fan's tach output to the CD4047 to go 'low' when it stops spinning

If you AC couple the tach signal (i.e. via a capacitor) to your circuit it won't matter whether the signal stays high or low when the fan stops.

 

[jelliott]

If you AC couple the tach signal (i.e. via a capacitor) to your circuit it won't matter whether the signal stays high or low when the fan stops.

I was thinking something like that might do the trick, but I'm not sure how to implement it. Is this just a matter of resistors and capacitors on the input (like your suggestion that saved the day in my other thread)?

 

[alec_t]

Here you go. The traces are for a 5V input pulse, but the same components would be used for any voltage (within reason!).

 

[Reloadron]

You can use a cap as alex suggest for coupling. Configuring a 74123 as a missing pulse detector it won't matter. The chip won't care if the input is high, it only cares about pulses.

I haven't uploaded a picture in a long time so hope I have it right. The only reason the 2N3904 Transistor is in there was to invert the pulses so they stop on a high level. When the pulses stop the Q out toggles low. The Q out would drive a DC SSR Like This One. Click the link for the product then that will get you to a data sheet. That or let the Q out drive a good power MOSFET. Note, this is merely one idea as to go about doing what I believe you want to do and that is shut down heaters if the fan fails. Working with large walk in ovens if any of the blowers failed the sensors were air flow switches which were a small paddle on a micro switch.

Ron

 

[jelliott]

Here you go. The traces are for a 5V input pulse, but the same components would be used for any voltage (within reason!).
View attachment 82638

Thank you again! I didn't have a 1 nF capacitor on hand, but even without that exact value this basic arrangement worked.

But what should I be using to switch the load? Ron's link is for an AC-controlled SSR, so that won't work, and I'm not confident that the TIP122 Darlington that I bought on a whim the other day is really up to the task (heater will be at 100% duty cycle when the fan is running full speed, and the TIP122 spec sheet indicates it can't quite handle its rated 5 A @ 100%). With your capacitive coupling solution I no longer need a device that can be switched on both sides of the 'coil,' but I'm hoping someone can point me towards a suitable MOSFET or SSR that's good for 5 A at 48 at 100% duty.

 

[alec_t]

Good to hear it works.
Using a parametric search on any of the major component suppliers' web-sites should throw up lots of power N-MOSFETs to do the job. Look for voltage rating >> 48V and an Rds(on) rating < 20 mOhm. If you're driving the FET from < 10V then it should be a 'logic level' type, i.e. with a low Vgs turn-on threshold (e.g. <3V).

Edit: A FET will need a reverse-biased 5A rated diode across the load to protect it from any back-emf spikes.

 

[jelliott]

A FET will need a reverse-biased 5A rated diode across the load to protect it from any back-emf spikes.

You mean like the flyback diode on the coil of an electromechanical relay? Is that really necessary for a resistive load?

(Thanks again, by the way--I don't know what I mis-clicked last time I tried such a search, but I found plenty of suitable MOSFET options from one of my usual suppliers this morning.)

 

[alec_t]

You mean like the flyback diode on the coil of an electromechanical relay? Is that really necessary for a resistive load?

Yes, as for a relay. Not essential for a purely resistive load, but connecting wires would have some inductance. Personally I'd use a diode for peace of mind.