Where to get schematics about SCRs full-wave triggering circuits ?

[eagle1109]

Just saw your post#18. Much better now.
But I miss some important informations:
* diode type
* SCR type
* transformer output voltage

Klaus

The diode and thyristor types are generic, I'm now trying to reach a level to run the simulation successfully then I would consider the ratings of voltage and current.

Because it would be basically for simulation, there's no intention to do a PCB of this project.


This is my last simulation, it's still not working !

The triggering signal is 250mV, I know it has to be more than that but does that mean I have to add another circuit; like, a transistor to deliver more voltage to the gate ?



- - - Updated - - -

Circuit #18 will probably work in a simplified simulation circuit, less likely with real components.

Power SCR need a certain amount of trigger current that can't be provided by an analog opto coupler. For the intended phase angle control, you want to generate a trigger pulse with variable timing. Circuit #18 isn't particularly suited for this purpose.

OK what should I do ?

A versatile method to provide isolated gate signals for multi phase SCR circuits are trigger transformers. Another popular means are opto couplers with triac output like MOC3020. They can be utilized for SCR triggering with an additional rectifier diode. There are also opto couplers with SCR output like Toshiba TLP748.

OK I would consider these drivers when I know how to trigger the thyristor.

Post #1 is asking for variable AC voltage, but most of the presented circuits have DC output. You understand the difference?

I'm sorry you're right, I'm actually working on DC controllers. I have to decide on a specific route now, then I would work on another route which is AC controllers to drive AC motors.

Phase angle control isn't well suited to operate AC motors will variable speed, except for brushed "universal" motors.

oh that's new aspect to me about phase angle control ! I know that AC motors control for speed would be like using traics, then I learned that there are more sophisticated stuff like VFDs.

The term "full-wave triggering" in the thread title is misleading. It's mainly used for controllers that switch full sine waves on and off in contrast to phase angle control which is cutting part of the wave.

I'm actually trying to work and save the simulation experiments for the circuits and applications in our course manual that I put the link in #3.

The manual is divided into 6 chapters excluding the first chapter which is about the development board that is in our laboratory so it would be 5 actual chapters:


CH1: Uncontrolled Rectifier
CH2: Controlled Rectifier
CH3: AC Inverter
CH4: DC Current Converter
CH5: Inverter Circuit

I really had to put this information in the main post to explain what I want to do. I'm actually want to do all the exps in the manual in simulation.

Considering CH1 is easy because it's diodes and should work perfectly. But starting with CH2 where difficulty starts. So I'm in CH2 so going to CH3 would be a long way to finish CH2 because I now have to get a grasp on triggering techniques.

If I did all these chapters in simulation, then I would consider that I have the basic idea about power electronics.

 

[KlausST]

Hi,

The diode and thyristor types are generic,

This makes no sense to me. How can you know the trigger level, hold current and so on?
For every design you need to know this.

You either learn to read datasheets, calculate with current, voltage, resistor...

...or you stay with the trial and error method. Never know whether it works or not, never know why it works amd never know whether your circuit runs close to low drive level (means function could fail with modified temperature)or overdrive level (means it could be killed).

it's still not working !

This is quite expectable for me. Without knowing the parameters to control the SCR...

Read an SCR datasheet. Read an SCR drive application note.
They don't tell to drive the gate with "voltage", they tell to drive the gate with "current". This is what you need to do.

Klaus

 

[eagle1109]

I tried the one with with triac output, but the output voltage went really crazy ! I don't know why !



- - - Updated - - -

I think I need to learn a lot of things in electronics !!

I'm trying to test the timing and output of a generic mosfet with 100k ohm pull up resistor and 1k ohm pull down resistor but I don't know why the signal on the drain is much slower than the signal on the source as there's no capacitor involved.



- - - Updated - - -

OK, I think I would start with something like this, it's very similar to the one on our laboratory dev board.



But also the problem is that they didn't mention the numbers of models, voltages and currents. It's just the conceptual circuit.

I don't know how to choose the models of the schottky diodes, pulse transformer and the other components.

They mentioned little things about the capacitor.

The fast diode D in the secondary circuit prevents negative gate current during the commutation swing of
the transformer secondary voltage and during dynamical VGK overshot. Purpose of the RC-element, RGK, CGK
is to filter unwanted glitches on the trigger line. A capacitance, CGK of 10 - 47nF is recommended to
achieve a discharge time constant τ = RGK · CGK ≈ 10 - 20μs (and RGK ≈ 220 - 2200Ω). The power loss PR of
RGK at maximum control angle via a half period of line frequency will be:

That means at VGK = 5V and extreme values (max. CGK = 47nF, min. RGK = 220Ω) and Τ ≈ 10μs, PR is
about 60mW. At τ ≈ 22μs with CGK = 10nF and RGK = 2200Ω, PR would be only about 6mW. The amount of
loss PR’ resulting from the discharge of CGK:

- - - Updated - - -

You're right, they mention the gate current is really high 1A @ 20mV .. wow this is a very high current with a very low voltage ! How to achieve that ?



- - - Updated - - -

The problem now is that there's no pulse transformer. I don't know if this one would work ?

 

[betwixt]

The who concept from the first post is rather odd and it gets worse when you start adding zero crossing optocouplers, triacs and MOSFETS.

As with all projects, you should start by defining your target. Any bundle of random components will do something but almost certainly not anything useful. Decide what you are trying to do first then 'home in' on the optimum solution to achieving it.

Go back to the first design and understand why it doesn't work before moving on to more complex problems. Ask yourself why the bridge rectifier?, how is the trigger voltage generated, can it provide enough trigger current. Most of all, what kind of trigger signal is actually needed, does it have to be rectified AC?

Before considering pulse transformers, which you probably don't need anyway, think of how you are proposing to generate the pulses. It requires a pulse generator circuit which adds a whole new level of complexity when all you are trying to do is turn an SCR on.

Brian.

 

[eagle1109]

I want to do the same exps in the manual I posted in #3, I don't know what hera used in their dev board they provided to our college. It's just closed boards for training.

So I want to do the simulations for all the experiments in that manual. The voltages and currents of the outputs are provided but the model numbers of the used components aren't provided, also the gate currents aren't mentioned either.

I don't know how to start.

First of all this is the first experiment in CH2 for controlled rectifiers.



Then they put this wiring diagram and that's all.



Everything is under the hoods I don't have their actual diagrams, so it's really difficult to know their design and values.

All I did is to expect the used components in relation to the supply voltage of the circuit which is 13V out of the main transformer.

 

[betwixt]

The schematics do make sense but as you point out, they are devoid of actual values. Treat them as conceptual, they show how it works without detailing the exact components used.

I think you need to grasp some concepts before getting to deeply involved. Do some research and see if you can answer these questions:
1. An SCR natively does not conduct at all - how do you start it conducting?
2. Once it is conducting, how do you stop it again?
3. What does 'phase control' mean and why do we use it?

I'll be back tomorrow to mark your homework :grin:

Brian.

 

[eagle1109]

The schematics do make sense but as you point out, they are devoid of actual values. Treat them as conceptual, they show how it works without detailing the exact components used.

1. An SCR natively does not conduct at all - how do you start it conducting?

Put a positive pulse at its gate, connecting the positive voltage of the voltage at the gate and the ground reference to the cathod

2. Once it is conducting, how do you stop it again?

I can't stop it until the voltage between the anode - cathode, gets reversed biased that the cathode voltage is more than the anode voltage

And that's why they invented the GTO, there are more sophisticated thyristors out there but of course would be more expensive than an ordinary SCR

3. What does 'phase control' mean and why do we use it?

Means that in case of DC control, you trigger the gate in phase of the rectified voltage that is coming out of a half wave that is every 360 degrees or full wave rectifier which is every 180 degrees.

I'll be back tomorrow to mark your homework :grin:
Brian.

Surely it's my honor that you mark anything I would answer or do.

- - - Updated - - -

For example, as I was going through this file for Philips, in their example diagrams they put the load after the cathode.



also in our manual provided from hera company, they also put the load after the cathode.



I think I learned something about that similar in the electrical circuits where in a basic electrical circuit, where there's a normal load; like, a lamb and a switch, some circuits put the load between the live wire and the switch and some circuits put the load after the switch to the neutral.

Like this one:

and this one:

But most other diagrams using SCRs actually put the load on the anode.

 

[KlausST]

Hi,

Brian will correct the answers tomorrow. Until then I give some comments on your answers.

Put a positive pulse at its gate, connecting the positive voltage of the voltage at the gate and the ground reference to the cathod

...and the gate current need to be higher than I_GT to safely trigger the SCR.
You need to verify that the trigger conditions of your curcuit meet the conditions in the datasheet. Like temperature range, anode voltage, anode current....

I can't stop it until the voltage between the anode - cathode, gets reversed biased that the cathode voltage is more than the anode voltage

No, it is not voltage related. It is current related. And no need to get reversed, it is sufficient that the anode(cathode) current goes below I_Hold.

Klaus

 

[eagle1109]

Here's my last simulation. It's a simple circuit but I don't know why the thyristor isn't fully activated ?

 

[KlausST]

Hi,

Why do you use R5? (It makes things more complicated than necessary)

What does tell you the SCR is not fully ON?

Klaus

 

[FvM]

The trigger voltage has to be applied between gate and cathode. R5 causes a negative gate voltage, in a real circuit it might cause damage of the SCR and trigger source.

 

[KlausST]

Hi,

and set your Scope input to "DC".

Klaus

 

[betwixt]

To reiterate: the trigger voltage is applied between the cathode and gate , that means if you lift the cathode voltage, the gate voltage has to be that much higher. Allowing the cathode to change is like trying to hit a moving target!

For simplicity, keep the cathode point your 0V reference, remove any components between it and 'ground'. Then all you have to do is provide the gate with enough voltage above ground to turn the SCR on. It is sensible to add a path to hold the gate voltage low when it isn't being triggered but your R3 value of 500 Ohms is unreasonably low, all it does is waste 24mA when the 12V is applied. It would work but a value of 10K would probably be more appropriate.

Brian.

 

[eagle1109]

Hi,

Why do you use R5? (It makes things more complicated than necessary)

I thought I only can measure output signal after the cathode.

I removed the resistor and put the oscilloscope on the anode and it worked.

What does tell you the SCR is not fully ON?

the voltage drop on the SCR would be very low; like 1V or less. That's similar to the transistor when it's fully on the voltage drop is very low.

The trigger voltage has to be applied between gate and cathode. R5 causes a negative gate voltage, in a real circuit it might cause damage of the SCR and trigger source.

I even don't have a practical experience with SCR circuits, the only practical experiences are microcontrollers, so I don't know how would a SCR should work .. but I'm trying to learn through the simulation. Getting error messages tell me that the circuit isn't balanced and I have to modify it.

To reiterate: the trigger voltage is applied between the cathode and gate , that means if you lift the cathode voltage, the gate voltage has to be that much higher. Allowing the cathode to change is like trying to hit a moving target!

For simplicity, keep the cathode point your 0V reference, remove any components between it and 'ground'. Then all you have to do is provide the gate with enough voltage above ground to turn the SCR on. It is sensible to add a path to hold the gate voltage low when it isn't being triggered but your R3 value of 500 Ohms is unreasonably low, all it does is waste 24mA when the 12V is applied. It would work but a value of 10K would probably be more appropriate.

Yep, I modified the circuit and finally worked without much problems.

I learned how to add the 150 resistor to lower the incoming current that was 80mA to approx. 50mA.



There're things that I still don't understand.

Why there's still a voltage drop on the SCR ? When it's fully on as I recall it should have a very low voltage drop like 1V or less.





Now I have to learn how to control the phase angle of SCR firing.

 

[betwixt]

Why there's still a voltage drop on the SCR ? When it's fully on as I recall it should have a very low voltage drop like 1V or less.

That is correct, and it is what the oscilloscope shows. The negative half cycles are when the SCR is NOT conducting, remember it only conducts in one direction, the tiny rise above zero on the positive cycles is the forward voltage drop, the rest of the half cycle is making the lamp light up.

Note that Vgt is the trigger voltage to start it conducting, not the voltage drop across anode and cathode when it is already conducting.

Brian.

 

[eagle1109]

That is correct, and it is what the oscilloscope shows. The negative half cycles are when the SCR is NOT conducting, remember it only conducts in one direction, the tiny rise above zero on the positive cycles is the forward voltage drop, the rest of the half cycle is making the lamp light up.

I really didn't understand anything ! What I understand and what I got in our actual laboratory oscilloscope is the same thing in the manual which is the positive half cycles. And what is mostly common about thyristor in this simple connection is to pass the positive cycles. So how it's showing the negative cycles on the oscilloscope ? This is the first time I get this.

Note that Vgt is the trigger voltage to start it conducting, not the voltage drop across anode and cathode when it is already conducting.

Yes I know that, the voltage for the gate is different that the voltage drop on the thyristor. But I still don't know why there's like +100V on the SCR when it's fully on ?

 

[schmitt trigger]

If your virtual scope's reference is where the ground symbol is, then the scope is showing exactly what it should be: The SCR's reverse blocking voltage

As simple as that.
You want the negative voltage to pass to the load also? Use a Triac instead.

 

[betwixt]

Think of the SCR as being a normal diode after being triggered. Substitute a real diode in the schematic and you will see what I mean. In conducting mode the real forward voltage depends a little on the current but will generally be around 1V.

The simulation voltmeters are ambiguous, to say "AC Volts" but have a '+' and '-' end is really meaningless and to display a '+' voltage is also meaningless when AC polarity is constantly reversing. However, the actual voltages are not far off what I would expect from DC meters in those positions. Consider that when you add a 'diode' (really the SCR) you are rectifying the AC. This has two consequences, first is the voltages are DC although pulsed and the second is that a voltmeter is calibrated for a steady voltage, not a pulsed one so the measurement will not be accurate.

Brian.

 

[KlausST]

Hi,

Eventually you dud whatvwe told you from the beginning (keep it most simple, remove R5)... and now it works.

It works like expected. Well done.

********
All is very basic.... and could be simulated very easily. Just use the simulation tool.
Let's focus on the conduction/non conduction thing.
Two parts are necessary for this:
* An AC power source
* a resistor as load.
...
And just for measurement: a scope.

* Connect the lower end of the power source with GND
* connect the upper end of the power source with ine end of the load
* connect the other end of the load with the scope.
(But don't connect the load with GND. This is like an OPEN switch, non conductive)

Now start the simulation.
--> the scope will show full AC voltage
But there is no current flow, no voltage across the load, no power at the load.
(You may add an AC voltmeter to measure the voltage across the load = one voltmeter connection to one load connection. The other voltmeter connection to the other load connection)

*******
Next step:
Same curcuit, but now connect the load with GND.
--> start simulation.
Now the scope shows zero voltage.
But current will flow. Full voltage now will be across the load.

*****
Next step:
Remove the GND wire. Add a switch at it's place and connect it with load and GND.
Now start the simulation with OPEN and CLOSED switch see what the scope shows..

****
Next step:
Replace the switch with a diode. Anode to load, cathode to GND.
Start simulation to see what happens.
(This is the same situation as when the SCR is conductive)

******
Now maybe something weird with "AC voltages".
(You may also recognize this in your post#34)
When the power source is set to an amplitude of 326V
* then the AC voltmeter across the power source shows 230V
* the AC voltmeter across the load shows about 163V
* the AC voltmeter across the diode shows about 163V
Usually (with DC supply) one would expect that: V_supply = V1 + V2.
But here with AC this is not the case.
163V + 163V = 326V .... and not 230V
(Now you may say, that 326V is the amplitude of the supply...but - although the same value - it is not related to 163V + 163V)
Please accept this weird AC voltage problem as it is ... the explanation is too complicated for now...

Klaus

 

[eagle1109]

I followed a method on this website, and it worked but I think there's something different about Proteus simulation.

https://www.electronics-tutorials.ws/power/thyristor-circuit.html

I works different with the SCR.


Not firing the SCR


Partial firing


Full firing


It works ok with diode to show only the +ve signal at load.




Different waveform without the ground.

So actually with a normal circuit of a half or full rectifier, what would be the ground anyway ? If I've done a fullwave rectifier for example, then I would take the 220V of the wall, to the full bridge then with a capacitor and a zener. I don't have a ground. I have only the full wave rectified output from the bridge circuit and referenced with the -ve side with the capacitor and that is how I would develop a 0V reference.

But I think it is different in a simulation world. That I have to put a ground.




I'm still not understanding why I get the negative signal. What I ever learned about rectifiers of halfwave that I should get the +ve output signal that has passed through the diode. Like this:

- - - Updated - - -

Why these people put the load at the cathode ?

https://learnabout-electronics.org/Semiconductors/thyristors_62.php