Does varistor make sense instead of RC Snubber?

[tyilgin]

Hi;

I am working on a design for three phase motor control in the range of 0.75 .. 2KW

As you know, when the triacs are not conducting, a few mA leakage occurs to the output via the snubber capacitor.

I'm thinking of using varistor instead of RC snubber to get rid of this leak.

May I know your thoughts on this change please?

Note:
1- I have come across many articles recommending R2 against leakage of MOC3052. I would appreciate it if you could write your comments on this subject.
2- I think BTA12-600 is at the limit. Even though I am using it in the prototype for now, I know that I will have to replace it with BTA16-800 in the future.

 

[KlausST]

Hi,

snubbers often are used for reducing the dV/dt problem that may cause false trigger.
A varistor can´t do this.

If you just want to limit the voltage (spikes) you may go for a varistor.
I successfully did this in some of my applications:
The benefit is:
* in my opinion they give better protection against high voltage on a high energy pulse
(in my case primary cut off while a transformer is involved)
* the are high ohmic in normal situations.

Klaus

 

[D.A.(Tony)Stewart]

The Snubberless™ versions (W suffix and T12xx) are especially recommended for use on inductive loads to prevent false triggering.

But to reduce leakage current getting to the load, you may use X caps to Neutral to bypass some of the leakage current and some resistance on Vgc to reduce input leakage.

 

[tyilgin]

Hi,

snubbers often are used for reducing the dV/dt problem that may cause false trigger.
A varistor can´t do this.

If you just want to limit the voltage (spikes) you may go for a varistor.
I successfully did this in some of my applications:
The benefit is:
* in my opinion they give better protection against high voltage on a high energy pulse
(in my case primary cut off while a transformer is involved)
* the are high ohmic in normal situations.

Klaus

Thank you very much for your interest

But to reduce leakage current getting to the load, you may use X caps to Neutral to bypass some of the leakage current and some resistance on Vgc to reduce input leakage.

I had completely removed it from my mind when I first heard it, thinking that snubberless versions could not replace the real snubber. Thank you very much for reminding me and for your interest.
By the way, I couldn't understand the paragraph I quoted. Can I ask you to explain it a little more?


So dear Klaus and Tony Stewart, are the Snubberless versions designed to minimize false triggering caused by the dV/dt problem or are they sufficient to prevent it completely?
If it is sufficient, it seems that I will be completely free from leakage current to the load by using varistor + BTB12-800CW.

 

[D.A.(Tony)Stewart]

Leakage and false triggering are design tradeoffs. You cannot eliminate shunt capacitance but you can embed it into the chip and still add more snubbing externally.

It all depends on your transient noise environment requirements with switched inductance spikes and line noise filtering to what is best suited for you application. False trigger consequences or leakage current consequences.

A TVS or MOV is a good solution current limited, with a Snubberless Triac. Line filters are also a good solution which also have Y cap leakage to ground.

 

[tyilgin]

Leakage and false triggering are design tradeoffs. You cannot eliminate shunt capacitance but you can embed it into the chip and still add more snubbing externally.

It all depends on your transient noise environment requirements with switched inductance spikes and line noise filtering to what is best suited for you application. False trigger consequences or leakage current consequences.

A TVS or MOV is a good solution current limited, with a Snubberless Triac. Line filters are also a good solution which also have Y cap leakage to ground.

I apologize for not being as knowledgeable about the subject as you are, and for not being able to understand anything from your conceptual explanations, considering the losses during translation.

Unfortunately I had to eliminate the TVS solution because this created additional costs of approximately over 5€ per engine.

 

[tyilgin]

I still don't have enough information about how successful the varistor + snubberless triac duo will be in terms of false triggering.

Also, if you would like to share your thoughts on whether a varistor of 14mm size and 4500A ripple current is at a level that can handle the sudden increases of a 2KW motor i would be grateful

 

[FvM]

False triggering of opto-triac or main triac can be caused by high voltage and dV/dt. Varistor doesn't reduce dV/dt.

If you decide for ZCD opto-triac, as you apparently do, high dV/dt at zero crossing can also cause skipping of half cycles when the voltage rises too fast to trigger ZCD. The effect is particularly dangerous for AC motor and coil loads.

 

[tyilgin]

False triggering of opto-triac or main triac can be caused by high voltage and dV/dt. Varistor doesn't reduce dV/dt.

If you decide for ZCD opto-triac, as you apparently do, high dV/dt at zero crossing can also cause skipping of half cycles when the voltage rises too fast to trigger ZCD. The effect is particularly dangerous for AC motor and coil loads.

I chose a triac with high dv/dt value, BTB12-800CW's dv/dt level is 500V/uS

MOC3052 is already Non-ZCD opto.
Based on this comment, do you recommend ZCD optu? Or am I ensuring that half cycles are not skipped by using Non-ZCD opto?

If there is still a danger, could you please share your advice on what I should change?

 

[FvM]

My fault, MOC3052 is non-ZCD. I would prefer it for 3-phase motor control.

 

[D.A.(Tony)Stewart]

If you have a high dV/dt solution, how much is your tolerance to leakage current? Bypassing the line current to ground Y-caps in a line filter is a possibility, yet you now have 5€ cost rise limits. So until these specs are all defined, I don't see a possible solution.

 

[tyilgin]

If you have a high dV/dt solution, how much is your tolerance to leakage current? Bypassing the line current to ground Y-caps in a line filter is a possibility, yet you now have 5€ cost rise limits. So until these specs are all defined, I don't see a possible solution.

This is the key point
I don't know how to calculate my tolerance for leakage current to compare with dv/dt. At this point I couldn't do more than choose a triac with a high dv/dt.

You mentioned your recommendation to bypass the Y-caps to ground and line filter in your message number 3, but I could not understand your recommendation.

Could you elaborate a little more please?

 

[D.A.(Tony)Stewart]

Bypass to Neutral , not ground. But ground leakage is permitted to say 3 mA. for residential and 30 mA commercial in some locations.

 

[tyilgin]

Bypass to Neutral , not ground. But ground leakage is permitted to say 3 mA. for residential and 30 mA commercial in some locations.

It is necessary to put approximately 9W resistor for 310V peak voltage and 30mA current. I will keep it in mind to give an idea about an alternative solution or solution.
Thank you

I am revising my system with 800V Snubberless triacs and 625V varistor. My biggest concern is that it returns from the field without causing any problems during the tests...

I'm not very knowledgeable about dv/dt calculation or observation, do you have any ideas on how I can examine it ?

--- Updated Nov 8, 2023 ---

Bypass to Neutral , not ground. But ground leakage is permitted to say 3 mA. for residential and 30 mA commercial in some locations.

It is necessary to put approximately 9W resistor for 310V peak voltage and 30mA current. I will keep it in mind to give an idea about an alternative solution or solution.
Thank you

I am revising my system with 800V Snubberless triacs and 625V varistor. My biggest concern is that it returns from the field without causing any problems during the tests...

I'm not very knowledgeable about dv/dt calculation or observation, do you have any ideas on how I can examine it ?

--- Updated Nov 8, 2023 ---

Bypass to Neutral , not ground. But ground leakage is permitted to say 3 mA. for residential and 30 mA commercial in some locations.

It is necessary to put approximately 9W resistor for 310V peak voltage and 30mA current. I will keep it in mind to give an idea about an alternative solution or solution.
Thank you

I am revising my system with 800V Snubberless triacs and 625V varistor. My biggest concern is that it returns from the field without causing any problems during the tests...

I'm not very knowledgeable about dv/dt calculation or observation, do you have any ideas on how I can examine it ?

--- Updated Nov 8, 2023 ---

Bypass to Neutral , not ground. But ground leakage is permitted to say 3 mA. for residential and 30 mA commercial in some locations.

It is necessary to put approximately 9W resistor for 310V peak voltage and 30mA current. I will keep it in mind to give an idea about an alternative solution or solution.
Thank you

I am revising my system with 800V Snubberless triacs and 625V varistor. My biggest concern is that it returns from the field without causing any problems during the tests...

I'm not very knowledgeable about dv/dt calculation or observation, do you have any ideas on how I can examine it ?

--- Updated Nov 8, 2023 ---

Bypass to Neutral , not ground. But ground leakage is permitted to say 3 mA. for residential and 30 mA commercial in some locations.

It is necessary to put approximately 9W resistor for 310V peak voltage and 30mA current. I will keep it in mind to give an idea about an alternative solution or solution.
Thank you

I am revising my system with 800V Snubberless triacs and 625V varistor. My biggest concern is that it returns from the field without causing any problems during the tests...

I'm not very knowledgeable about dv/dt calculation or observation, do you have any ideas on how I can examine it ?

--- Updated Nov 8, 2023 ---

Bypass to Neutral , not ground. But ground leakage is permitted to say 3 mA. for residential and 30 mA commercial in some locations.

It is necessary to put approximately 9W resistor for 310V peak voltage and 30mA current. I will keep it in mind to give an idea about an alternative solution or solution.
Thank you

I am revising my system with 800V Snubberless triacs and 625V varistor. My biggest concern is that it returns from the field without causing any problems during the tests...

I'm not very knowledgeable about dv/dt calculation or observation, do you have any ideas on how I can examine it ?

--- Updated Nov 8, 2023 ---

For a reason I don't know why, "edaboard.com" is working slower than you can imagine.

While trying to send my message, he kept repeating the content. After sending my message, I saw that the editing period had passed and I could not edit it.

I'm sorry for the confusion.

 

[D.A.(Tony)Stewart]

More likely it is your PC browser that is slow. Perhaps Windows is doing a backup, or update or something else. But latency does cause duplicates with edits.

>"It is necessary to put approximately 9W resistor for 310V peak voltage and 30mA current. "

NO.

It is common for LED Lamps to not turn off from Triac leakage, so in the past, an 8W tungsten lamp was used to make a dimmer work better. But Philips and others have solved that problem with integrated dimmable LED lamps with sufficient capacitance load.

 

[KlausST]

Hi,

edaboard currently is having some delay problems.
Sorry for this inconvenience.

You still have limited time to review and edit your post.

Klaus

 

[tyilgin]

More likely it is your PC browser that is slow. Perhaps Windows is doing a backup, or update or something else. But latency does cause duplicates with edits.

>"It is necessary to put approximately 9W resistor for 310V peak voltage and 30mA current. "

NO.

It is common for LED Lamps to not turn off from Triac leakage, so in the past, an 8W tungsten lamp was used to make a dimmer work better. But Philips and others have solved that problem with integrated dimmable LED lamps with sufficient capacitance load.

Only "edaboard.com" has been very slow in my browser and computer for a long time. I have no problems with other pages.

I can't connect the clues, probably because I'm busy right now. I will research it when I have time after my project, thank you.

--- Updated Nov 9, 2023 ---

Hi,

edaboard currently is having some delay problems.
Sorry for this inconvenience.

You still have limited time to review and edit your post.

Klaus

Even though I clicked the post reply button many times after writing my message, it didn't seem to do anything. After a while, thinking that it had crashed, I copied what I wrote and reloaded the page. When the page reloaded I noticed that I had sent my message with each of my previous clicks and the "edit" button at the bottom of the page was no longer visible.
It gets incredibly heavy, especially after pasting text with ctrl+v.
Thank you for your attention