Using optotriac in 3 phase softstarter

[ahha]

I have a problem, would appreciate some help.
I am building a 3 phase softstarter using opto-triac moc3052 to drive 25A output triac. I am experiencing various motor rattling and doesn't matter how I change the snubber circuit, it behaves the same. What did I do wrong? Is optotriac not suitable? Do I have to use pulse transformer for the job! Pls help

 

[Marke]

hello ahha

Firstly, you need to ensure that the switching elements, particularly the opto triac are rated with a sufficiently high voltage. The normal rating for soft starters is 2.5 - 3 times the line voltage. For a 400 volt three phase supply, I would use 1200 volt SCRs and if I used an opto triac, it would also need to be rated at 1200 Volts. If you use a 600 volt rated opto triac to trigger your SCRs on a 400 volt supply, you will have voltage across the triac element that cause a breakdown in the device and trigger the SCR ON. This will cause the rattle you describe. The SCRs turning ON will protect the triac from damage in many cases. Does this help?
If not, more detail please.
Best regards,
Mark Empson

http://www.lmphotonics.com

 

[ahha]

Hi Marke,
Thanks for the prompt response.For your information I am using a MOC3052(peak off-voltage 600V) to drive a Teccor Q8025 25A Alernister Triac (Vdrm 800V) in 400V ac. Also I have tried the il4218(peak off-voltage 800V) optotriac and not much improvement. 800V Vdrm is the max. rating I can find for both the optotriac and the power Triac. Can you give me some suggested devices that are rated higher? Thanks
Paul

 

[cbarn24050]

Hi, several problems here, if I understand your post correctly. First you cannot use a triac in a 3 phase softstarter you need 2 SCRs per line. For a 3 phase 400v system you need 1000v devices. Same for the opto-triacs. Another thing you should be aware of is that the timing of the thyristor trigger pulses is not as simple as you may think, if your thinking of using a delay from the incoming supply waveform forget it, it dosen't work.

 

[Marke]

Hello ahha
If you are using 800 volt components on a 400 volt supply, you will certainly have problems. The 800 volt optotriac will begin to break down and drive current into the gate of the Alternistor and turn it on on voltage peaks.
Another problem that you can have is due to the dv/dt rating of the trigger device. It is important that the trigger device is able to withstand a very high dv/dt or it will provide gate current on transients and also cause the problems that you describe. You must use devices rated at 2.5 - 3 times the line voltage or you will have difficulty.

cbarn. You can use triacs provided that the voltage rating is high enough, but there are different families of triacs. The standard triac as used in low voltage light dimmers is not rugged enough, but there are varients such as the Alternistor that provide similar overload characteristics to SCRs, and provided that they are rated correctly, will perform perfectly well in three phase soft starters. I certainly agree, that there is a lot more to getting the firing pulses right than meets the eye and there is quite a science in that, but it is possible to derive the timing from the incoming phases. I have done it for many years and have a very large connected base of product using that technique. Once again, there are appropriate techniques for achieving this.
Best regards, Mark Empson

http://www.lmphotonics.com

 

[ahha]

Hi Mark and cbarn,
Thanks both of you for the valuable advices. It seems that using optotriac is not a feasible method in building a three phase softstarter, I will try to build using pulse transformers then. I was successful in using optotriacs to drive small 3-phase motors (<3kW,2poles), that is the reason I was trying to make it work for the larger machines and failed. As for the pulse transformer method is there any area I should be careful when designing with them? Pls advice
Regards
Paul

 

[Marke]

Hello ahha

There are soft starters on the market using optotriacs so they can be used, you just must be sure of the voltage rating dv/dt etc.

Pulse transformers. To control SCRs you need a fast rising edge on the output current to ensure you trigger all of the SCR, not just a portion. If you want reliable operation, you need a very low capacitance between windings. It is preferable to maintain gate drive for most of the conducting cycle of the SCR (or triac) to prevent premature commutation.

Best regards, Mark Empson

http://www.lmphotonics.com

 

[cbarn24050]

Hi Mark, do you have product in the field that uses triacs? If so perhaps you could give me the device part number. In the days before micros firing was derived from the incomming waveform but you have to extend them to a long train of puses, and produce 6, 1 for each scr. That method would not work for 3 triacs, which you can supply, for ahah.

 

[ahha]

Hi Mark, charn is right, most of the standard cirucuits that implying pulse transformers use 6 to drive back to back SCRs for the output, if I use just 3 triacs how will the drive circuit looks like? Also I am using microprocessor output to drive optotriac in the old design, do I need to buffer those output to create higher voltage and higher current before driving the pulse transformers? I recall +24V is the standard driving voltage people use for the pulse transformers, if I just want to use a single +5V supply for the entire softstarter, is it feasible?

 

[Marke]

Hello cbarn and ahha

There are many soft starters on the market using reverse paralel SCRs with only three pulse transformers. You can look at the Danfoss range for example. The pulse transformers have two secondaries, one for each SCR. AuCom soft starters are SCR based but use only three pulse transformers.

Triac based controllers are small controllers due to the availability of small devices bu not large devices at the required voltage rating.

The transformers can be driven from 24 volt or from 5 volt. The advantage of 24 volt is that the primary current is lower, while the 5 volt has the advantage of a smaller core for the same output. Both are used by different manufacturers.

Best regards, Mark Empson

http://www.lmphotonics.com

 

[ahha]

Hi Mark,
Thank you very much for your valuable advices, I think I have enough information to start working on my softstarter design, again thank you

 

[cbarn24050]

Hi ahha, best of luck, let us know how you get on.

 

[jbartos]

Suggestion: Reference:

http://www.semiconductors.philips.com/acrobat/applicationnotes/APPCHP6.pdf

CHAPTER 6 Power Control with Thyristors and Triacs

 

[ahha]

Hi Mark, May be you can answer my question. I don't understand why do we need to have triac devices 2.5~3times the rated line voltage. I have been doing a bit more readings during the past few days, also articles as suggestedby jbartos, it seems that dv/dt and di/dt during commutation is usually the culprit causing triac not to turn off. Of course we have to keep the rating of the triac above line voltave peaks, but why 3 times so high?

 

[Marke]

Hello ahha

There are always transient voltages on the supply. Some will occur with the commutation of the devices that you are using, but others will naturally be there from other switching devices on the equipment connected to the supply. It is a statistical decision based on the likelyhood of a transient being high enough to cause problems. At 800 volts, you will have problems just from the commutation of your own devices. At 1200 volts on a 400 volt supply, you should have years of good service unless you have a bad supply, typically resonant supplies can cause problems!
I have soft starters installed that have been operating faultlessly for over 20 years using 3 x line voltage rating for SCRs and Triacs.
Best regards, Mark Empson

http://www.lmphotonics.com

 

[geezer]

Check the turn-off characteristics of the opto-triac. These are quite slow and may conduct through line cross, causing the SCR or Triac to prematurely turn on during the next cycle.

 

[jbartos]

Suggestions:
1. The &quot;ringing transients&quot; tend to be filtered nowadays. There is an EMI concern and EMI industry standards. It appears that Marke did not worry about this with his voltage overrated approach in long lasting, smoothly working applications.
2. The higher voltage rated switching device has somewhat different V-I characteristics at low voltage levels than the lower appropriately rated voltage device.
3. The higher rated voltage device may be more expensive.

 

[Marke]

hello jbartos
1. Wouldn't that be nice!! In reality, the supply is very rich in transients etc and adequate filtering has a long way to go to have any real impact in most industrial applications, and yes, the equipment that I have been involved with for years has full EMC compliance. The overvoltage approach is the recommended levels from the semiconductor manufacturers, not my own theory.
2. Not so once you are dealing with SCRs. This is certainly the case with bipolar transistors FETs etc, but there is very little variation between a 600 Volt rated SCR and a 1200 volt rated SCR. When you get to the 3300 volt ratd devices, you can see some variation, typically a higher ON state voltage due primarily to a higher rt and Vto.
3. Very true but if you use a device that is too low in rating, it will tend of overhead trigger when you switch another phase and will fail prematurely.
Best regards, Mark Empson

http://www.lmphotonics.com