Crydom solid state relays

[buffalonymann]

181025-0821 EDT

buffalonymann:

I believe Russs57's point is that if you use your PLC output to control an ordinary electromagnetic relay, such as a P&B KUP, with mechanical contacts on its output, then you can obtain an output contact with very low leakage current compared typical SSRs.

The Crydom drawing shows one output switch terminal pair as A1-A2, another as B1-B2, and the third as C1-C2. I don't believe any one of these pairs is your L1-L2.

Gar, I'm calling them L1 and L2 out of habit, so L1 would be A2 and L2 would be B2. We're seeing 190VAC across A2(L1) and B2(L2) and C2(L3)

Considering only one SSR contact in the Crydom package, for example A1-A2, then is that contact the switch in a simple single phase series circuit consisting of a 240 V source, the A1-A2 contact, and a load resistor? I am concluding that is what you have said. In that series circuit do L1 and L2 refer to the load resistor terminals? If not, then where in the circuit are L1 and L2?

My bad for not being more precise - I overlooked that you have not seen a wiring diagram of our system.

We have 3 phase Delta 240VAC for supply power. Phase A to A1\Phase B to B1\Phase C to C1. Output is A2, B2, C2. Single Phase heaters are across A2 and B2, B2 and C2, A2 and C2

PLC output to control (120VAC) circuit

For a high power load on the output of an SSR the snubber leakage current will produce no significant voltage across that load resistance. If you see a noticeable voltage, but not full voltage, across the load, then the SSR is conducting during a portion of the AC cycle. It appears that your SSR is conducting over some portion of the cycle. In my OAC5 test the SSR had three distinct states --- no conduction, looking like a 1/2 wave rectifier, and full conduction, and very touchy input control voltage to get the 1/2 wave rectification. I can visualize in your case that at 18 V there is some partial cycle conduction, and not so touchy as in the case of my OAC5.

.

Ok, what you say about the conduction during part of the cycle is a feasible diagnosis, perhaps that is what we are seeing. My company didn't want to spend money to buy a scope for me, so I'm stuck with a $400 piece of hardware that allows me to use my laptop as a scope, problem is it's rated at 30 volts (rolls eyes) I'm dealing with a management group without any technical background whatsoever, and never have a clue about what I'm telling them.

 

[gar]

181025-0936 EDT

What is the brand of the device that is a scope simulator?

If this heater application is from a single phase source with a grounded neutral, then you can use your scope simulator. Suppose the scope simulator has a 1 megohm input impedance, then an external series 10 megohm resistor will give you 300 V capability. To see timing information, and waveshape you don't need an accurate voltage divider, and you don;t need to compensate for capacitance in the divider at 60 Hz to just get a rough view of the waveform.

In the single phase grounded neutral situation, and for a rough view you connect the scope simulator input ground terminal to your single phase EGC (may not be necessary if the scope simulator signal input common or ground is internally connected to EGC, this is a general problem with many new (meaning since about 1950) scopes because of the lack of isolation in the scope of signal ground and the EGC, and why many times I use a 2 prong to 3 prong adapter to isolate my scope).

If both L1 and L2 are floating off of ground, then you would need a differential amplifier input.

If no 10 meg resistor, then put 10 1 megs in series, or whatever is needed for the input impedance of the scope simulator.

.

 

[gar]

181025-1011 EDT

Simulator is really the wrong word, emulator is probably better.

.

 

[Russs57]

buffalonymann,

I suggested adding an interposing relay because it would be far more likely to drop out at reduced voltage.

Comes a time to experiment a bit. I'd like to know if the LED on the SSR is on when PLC commands heaters off. If so I might suggest a load be added such as a 10K 5 watt resistor across SSR control input (maybe use a pot so you can dial in required load?). You have already proved SSR will turn off if input voltage is low enough. It is the nature of solid state to have some leakage and the SSR input typically only needs a few ma's to turn on.


You say you have inherited a problem. Perhaps there is something wrong with initial design? They make a number of different SSR types. Might be worth talking to tech support at crydom.

FWIW, anything I have been involved with along that line would use a dedicated temperature controller and not the PLC. You will get far better set point control, especially if all 3 elements are heating the same medium. You can get them from simple proportional all the way up to "fuzzy logic". PID types are usually more than good enough. Best to single source a package from a known vendor in the field like chromalox.

 

[buffalonymann]

Link Instruments MSO-28

 

[buffalonymann]

181025-0936 EDT

What is the brand of the device that is a scope simulator?

If this heater application is from a single phase source with a grounded neutral, then you can use your scope simulator. Suppose the scope simulator has a 1 megohm input impedance, then an external series 10 megohm resistor will give you 300 V capability. To see timing information, and waveshape you don't need an accurate voltage divider, and you don;t need to compensate for capacitance in the divider at 60 Hz to just get a rough view of the waveform.

In the single phase grounded neutral situation, and for a rough view you connect the scope simulator input ground terminal to your single phase EGC (may not be necessary if the scope simulator signal input common or ground is internally connected to EGC, this is a general problem with many new (meaning since about 1950) scopes because of the lack of isolation in the scope of signal ground and the EGC, and why many times I use a 2 prong to 3 prong adapter to isolate my scope).

If both L1 and L2 are floating off of ground, then you would need a differential amplifier input.

If no 10 meg resistor, then put 10 1 megs in series, or whatever is needed for the input impedance of the scope simulator.

.

Sounds like you know your way around a scope; I have little experience, and need to read directions while using. I can rent a scope and would do so to get an idea whether our design is inherently flawed or we have excessive current leakage on the plc 120 vac output module. I would appreciate some direction on a proper scope to rent for examining these circuits

 

[Russs57]

They sell a 100X probe for it.

 

[buffalonymann]

buffalonymann,

I suggested adding an interposing relay because it would be far more likely to drop out at reduced voltage.

We could buy relay output cards, however we need to consider lifecycle. These heaters cycle on and off every second or two

Comes a time to experiment a bit. I'd like to know if the LED on the SSR is on when PLC commands heaters off. If so I might suggest a load be added such as a 10K 5 watt resistor across SSR control input (maybe use a pot so you can dial in required load?). You have already proved SSR will turn off if input voltage is low enough. It is the nature of solid state to have some leakage and the SSR input typically only needs a few ma's to turn on.


The LED is off when we command the SSR off, but we're still seeing substantial voltage on the SSR output nodes

You say you have inherited a problem. Perhaps there is something wrong with initial design? They make a number of different SSR types. Might be worth talking to tech support at crydom.

FWIW, anything I have been involved with along that line would use a dedicated temperature controller and not the PLC. You will get far better set point control, especially if all 3 elements are heating the same medium. You can get them from simple proportional all the way up to "fuzzy logic". PID types are usually more than good enough. Best to single source a package from a known vendor in the field like chromalox.

Done

 

[gar]

181025-1309 EDT

buffalonymann:

Your MSO-28 appears to be a very capable device. Since it has two analog inputs I assume that you can do differential measurements which means you can solve the ground reference problem. A big advantage of this device is that you can have a large screen display.

The description is quite good, but also very poor. As Russs57 suggested you could get two X100 probes to handle your voltage range. No need need to rent a scope.

You need to study the MSO-28, meaning read the manual and do a number of different experiments, basically play with the instrument. This scope can be of great value to you. Play with simple LR, RC, and LRC circuits in a single shot transient type of measurement. Look at a microphone output from striking some resonator, a tuning fork or piano key. Look at AC line voltage and see the waveform distortion, then do an FFT on that waveform. Many other things will give you signals to analyze.

.

 

[RumRunner]

Buffalomann,

If I may.

Since its introduction of SSRs some decades ago, a lot of wrangling between Electronics Engineers and Electrical Engineers had been going on.
I've been in several of these “duels” to no satisfactory conclusions.

Which is better...solid state or electro-mechanical relays (EMRs). The answer is : Neither

Each one has its good and bad points. Clear winners will emerge if and when the specific requirement fits the suitability of its application.

I will point out only those that are mostly encountered.

SSRs are good if you need something that don't create hazardous condition where arcing can cause explosion in a hazardous environment. They don't produce arc and they can get away with this nuisance when using electro-mechanical relays (EMRs)

SSRs don't make any noise in auditory terms.

SSRs are not really relays—they are merely electronic circuits. Unlike EMRs, they don't actually turn off—they just switch to null state.
When they (by design) transition to this null mode, they are still on but not enough current to activate the load it is connected to.

Some AHJ requires to have a physical disconnect like a knife switch to disable a motor when clearing jams for example. This is a must in the City of L.A.

Pushing the Emergency (red) button is not good enough for these guys.

One thing I might want to point out is: you indicated that you have a three-pole SSR. Individual poles of solid state relay are independent of each other which means one pole failure doesn't mean the others have failed—and they can still be conducting unlike those EMRs. Of course EMR contact get welded together.

Another thing you mentioned is the crossover from the control voltage to the output voltage (the usable power.)

This is highly unlikely because the opto-coupling could ensure that this doesn't happen internally.

OK, so you might say: what are you telling me?

What I'm saying is, the voltage you're getting even if the PLC command is given to stop-- is inherent to the design and characteristics of electronics components.
If the motor or resistive load appear to be off (nothing going on) then you can call it “off.”

Gar has good points to consider and so are the others.



Disclaimer:
This should not be taken as a means to perform your job safely. Suitable safety measure which you think is appropriate in your situation is preferable. It is always right to do things on the side of caution.
Have a good day.

 

[Jraef]

Another possibility to explore is in the power and control wiring arrangement. Someone not familiar with the “sensitivity” of the SSR’s control circuit may have run the control wires too close to the power wires so that they are picking up an induced voltage as soon as current flows in the power circuit. I’ve seen that happen many times because in an older elec-mech based control system it’s rarely an issue since the induced voltage is so low, so electricians not used to SS controls don’t know this can happen and do things the way they’ve always done it. But with solid state controls, all that need happen (in this case) is that it exceed 10V. So nothing happens until you turn it on, then a voltage is induced onto the controls, yet while your PLC is commanding it to be on, you can’t tell. But when the PLC goes low (or tries to), the induced voltage remains and is causing partial conduction through the SRR, which maintains the induced voltage. Once you remove the wire or drive the induced voltage below 10V, the problem is fixed.

If the power and control conductors are bundled together or in a common conduit, separate them physically by a couple of inches and see if the issue resolves itself. That’s the cheapest fix, unless it means running another conduit...

This is exactly what an inter posing relay might fix as well, but you have to be selective and not get one of the new ones with solid state coils! I’ve fixed this many times however with burden resistors across the input of the SSR control terminals. I’ve found that people who choose SSRs often do so precisely BECAUSE they want to eliminate elec-mech components that wear out, so adding an inter posing relay is anethema to that. But they don’t even notice the burden resistors.

 

[newtonwb]

Rockwell Automation has a nice little write-up for the load resistor mentioned a couple of times above. At our plant I've seen it done both ways.
1) Interposing relays
2) Load resistors

Adding load resistors doesn't require any additional wiring (typically) and drawings are easy to update...

You'll need a Rockwell Automation log-in to access this following...

https://rockwellautomation.custhelp.com/app/answers/detail/a_id/24498