Question:
Why not use solid state relays?
Mad Martha
2012-11-08 19:55:30 UTC
I was at a local hackerspace earlier this evening. I've only been a coupe of times before but people were discussing how to turn on and off a motor from a microcontroller for a robot they were building. I suggested using a solid state relay which was a comment which drew a lot of laughter from the group. I didn't want embarrass myself by asking why at the time so why would an solid state relay be a bad solution to the problem?
Seven answers:
Andrew S
2012-11-08 20:46:46 UTC
You'll find SSRs widely recommended here and in other places with a lowish skills base, but they are by and large ignored by more advanced engineers. They look attractive to the hobbyist - they have the ease of wiring and isolation of a relay but the speed of a semiconductor, and no need to calculate component values as you would for a transistor. However, in reality the downsides tend to be greater than the advantages so SSRs tend to be the mark of a noob.



Biggest of these is the sheer cost. I don't know exactly what kind of motor you were controlling but consider something fairly average for that kind of problem - 12V and 2A. Controlling that with a power transistor would cost around 60p in components here in the UK. Using a mechanical relay may cost £1.50. A suitable SSR is likely to cost well over £20.00.



Using a power transistor means a suitable type needs to be selected and component values calculated to suit. However, most electronic engineers have a mental listing of perhaps 10-20 different transistors that they are very familiar with and can simply select an appropriate one from those in the blink of an eye. You need to calculate a resistance for the transistor base which can be tricky for those without the required understanding but it is a matter of 10 or 15 seconds with a calculator for someone with experience.



Using a mechanical relay reduces the skills needed to make a working design but at the cost of speed of response. On the plus side it will remain a lot cooler. The losses in transistors tend not to be high as a proportion of the power handled but they can be enough to create cooling issues. Losses in a relay are orders of magnitude lower so the relay remains cool.



Using an SSR firstly you have to pay for it as has already been mentioned, then you design it in as for a conventional relay. However, you have the same heat issues as you have for transistors.



In practice I suspect I would probably opt for a combination of relay and transistor for this kind of project - use a relay to reverse the direction (it's easier than using four separate transistors) with a single transistor to turn the motor on and off. A transistor's switching speed allows for speed control even from a digital MCU output by means of pulse width modulation - essentially turning the motor on and off rapidly. You can't do that with relays.



On a more general note the kind of people that populate hackerspaces are in general a helpful bunch but you need to approach them with the right mindset. If you ask this kind of question people will be only too happy to explain but you will not endear yourself to them if you conduct yourself as if you have equivalent levels of expertise to people with greater skills in an area. As such it's best to ask this kind of thing there and then - you'll look far less foolish in the long run.



One final point about the voltages that the earlier respondent alluded to: there is certainly some truth to what he says although the maximum working voltage of any SSR I have ever seen is well below what you can do with a relay. In addition you need to be wary of considering motors or other inductive components purely as low voltage devices - a 12V motor may generate ten times that voltage internally and may output that voltage to the circuit as brief transients in certain conditions. This is why you need the suppression capacitors to control radio interference as sparks literally fly as a result of those voltages. Those voltages can certainly damage a transistor or even an SSR if they are above the rated voltage for the device. Mechanical relays are a lot more forgiving in that respect.
Ecko
2012-11-08 20:23:46 UTC
Solid state relays are normally intended for switching 120/240V AC loads like heaters. They are intended to switch around the zero crossing of the AC cycles. Usually the switching element is a triac device. This is great for heaters or lamps, not so great where inductance is present, like an AC motor. When switching a DC motor, it probably does not work at all, very likely staying "ON", but the truth is these guys were rude and uncouth, and should encourage others instead of trying to show how smart they are (or are not). The description I gave is how they are generally, but not all of them are this way I suspect.



The solution for switching a DC motor usually involves transistors, sometimes in a configuration that allows reversing, and generally the arrangement is so the transistors can be switched rapidly to vary the speed (pulse width modulation PWM). Look up H bridge for details. The link shows the basic idea, where there are 4 controlling signals from a logic source using the same supply voltages in this example. Each transistor is either on or off according to its control signal. Current can be passed through the motor in either direction. The second link shows a simple switch where the motor cannot be reversed (Figure 9).
Rick
2012-11-08 20:24:43 UTC
The problem with solid state relays is that most of them have zero voltage turn on control. Since the motor load is inductive, the current will not increase much when the relay turns on. This means the current throught the relay will not reach the holding current by the time the gate drive is removed. The relay will then turn off. Then he voltage goes above the turn on threshold and the gate drive is not reapplied. Sorry for the long and perhaps confusing explaination.
anonymous
2012-11-08 20:29:16 UTC
Solid state relays would be OK for applications requiring isolated control as when controlling an AC load with a low level DC voltage. Robots generally use DC at low voltages without the need for isolated control.



The laughter was uncalled for as a solid state relay could be used in some circumstances. There are some solid state relays designed for DC operation and would be perfectly suited to robots.
Lloyd J
2012-11-08 21:31:14 UTC
SSR's are usually used when a microprocessor needs to control a 1-2 Kw load driven directly from 120 or 240VAC. Robots usually use 12 or 24 volts DC and simple transistors can be used to eliminate the bulk and slow speed of mechanical relays. As you get into higher power, SSR's might become useful. Laughing at a newby question is very rude.
wahzai1981
2014-09-12 08:14:09 UTC
In my opinion there are a few reason why SSRs are not commonly used for motor control.



1.Price is typical 3~4 times for a simple emr.

2. Switching mode zero voltage switching excellent for resistive load e.g. lamps, ballast, heaters etc.

3. Short circuit if overloaded, worst case may exploded. Thus need additional fuse especially for inductive load because of this reason..

4. Only single phase or three phase...and single pole single throw NO. some users used to the interlocking control logic with emr.

5. Size for ssr may need big heatsink due to depends on loads.
Wile E.
2012-11-08 20:05:44 UTC
There are two advantages to using solid-state relays: 1) They can control heavily inductive loads without arcs discharging across metallic contacts. 2) They switch when the voltage is at zero, called zero-crossing switching. This, too, eliminates arcing across the contacts. In addition, the zero volts turn-on eliminates current shock in inductive loads, which increases power loss. The disadvantage is cost.



In low-voltage, low-inductance circuits, arcing is not a problem. Hobby robots are usually controlled by microcontrollers such as the Arduino, so inexpensive relays will suffice.

 


This content was originally posted on Y! Answers, a Q&A website that shut down in 2021.
Loading...