Not only that, but there are too many times on this site where people recommend a straight capacitor only, without any damping resistor. Bad advice all around. This applies to switch contacts, diode snubbing, choke snubbing, etc.
This is the second article that stresses the need for the resistor (page 2-7, note 6). I suspect the advice will continue, under the claim of "I've done this many times and it works just fine".
You are dealing with the need to dissipate energy from a location said energy will do damage if not quenched. A capacitor all by its lonesome does not dissipate energy- it stores it, happily releasing it back into the location if circuit conditions allow (read- oscillation). The load itself is not sufficient, as X/R ratios are too high for effective dissipation.
This is the second article that stresses the need for the resistor (page 2-7, note 6). I suspect the advice will continue, under the claim of "I've done this many times and it works just fine".
You are dealing with the need to dissipate energy from a location said energy will do damage if not quenched. A capacitor all by its lonesome does not dissipate energy- it stores it, happily releasing it back into the location if circuit conditions allow (read- oscillation). The load itself is not sufficient, as X/R ratios are too high for effective dissipation.
They seem to think they can calculate R and C from circuit current, rather than circuit inductance. They are also trying to sell something. I would ignore them.CharlieLaub said:
There are sites on the web which give the correct way to calculate R and C for snubbers. This must include the circuit inductance, as this is the source of the problem. Fortunately it won't matter too much if the RC values are not quite right (we are stopping sparks, not designing filters), and in many cases the circuit will have sufficient resistance to damp the oscillation.
I've seen some of those formulas, but haven't been sure what to consider in the inductance term. For a linear power supply, is it essentially the primary winding inductance of the transformer?
-Charlie
The equations can only get you close. The basic idea is to get a decent approximation of the smallest value of C that performs the task. From there R is easy to find with a scope (biggest R that gets the job done). I don't bother with the equations at all; just go right for the scope and start with my box of CDE 940C caps and a 10K pot.
Well I can't just "go for the scope" because I don;t have one!
I will just throw on a 0.1uF X2 cap and a 47 ohms resistor and call it a g'day.
-Charlie
Probably, unless there is capacitance across the secondary too.CharlieLaub said:
I think I would use a smaller cap and a bigger resistor. You are getting a spark with a switch capacitance of at most a few pF, so just a few nF will make the voltage 30 times smaller. Too big a cap could introduce a switch-on surge.
Ooops, sorry I wasn't specific. Here are some additional details:
I am actually planning on using TWO snubbers. Based on comments in this thread it seems there are two areas where these are useful:
1. in parallel with the switch
2. in parallel with the transformer primary
If I understand correctly,
#1 will help suppress radiated RF "noise" resulting from sparking/arcing as the switch contacts are just opening or are closing.
#2 will help dissipate the "kick of current" generated by the inductance of the transformer windings that would otherwise have no where to go when the switch is opened, leading to a brief but high voltage spike in the primary winding.
The series RC snubbers I am planning to use for these are as follows:
#1 0.01uF (10nF) X2 rated cap and 180R-220R, 1W non-inductive, in parallel with the switch
#2 0.1uF (100nf) X2 rated cap and 47R, 1W, 1W non-inductive, in parallel with the transformer primary
The linear PS that I am using these with is typically rated around 150W, for a gainclone amp.
Let me know if I am way off in my approach.
Thanks,
-Charlie
P.S. it was cool for me to look in to the "inductive kick". I learned how and why the spark plug "ignition coil" in a car works!
It's common practice, although that is certainly not invalidating your point about safety.
I thought that these X2 "safety" caps were designed to fail to an open circuit. If so, that makes me feel a little better.
One thing that I didn't consider until recently was leakage through the snubber. Unless you remove the plug from the wall, the AC voltage will still be present, although current will be limited by the reactance of the cap.
Let's see... Xc is (2*PI*F*C)^-1 = 265257, R = 47, Z = (Xc^2+R^2)^0.5 = 265257 (effect of R is negligible). At 120VAC, the resulting current would be 0.00045 Amp or only 0.45mA.
Thoughts?
-Charlie
0.45mA is not enough to harm you, but it is enough to feel. It could, if unexpected, feel sufficiently unpleasant to make you (or someone else) drop something.
If you have a snubber across the transformer then there is little need for one across the switch as well, as the transformer will be by far the dominant inductance in the circuit.
If you have a snubber across the transformer then there is little need for one across the switch as well, as the transformer will be by far the dominant inductance in the circuit.
Most transformers I have had to snub, the values are closer to 0.01 uF and well over 1K. Which further proves the 'bad idea' of just using a capacitor.
You aren't looking to load the circuit at harmonic multiples of the fundamental; you want to snub high resonant frequencies of a tank circuit. These frequencies are typically in the many tens of kHz. Smaller cap, bigger resistance.
The values I quoted were from design tables for an AC snubber that I found in a technical doc, based on the load current and applied voltage. I have seen values similar to mind in linear PS schematics like the one I posted earlier in this thread. I've never seen a combination like the values you mention. Can you describe the transformer rating, etc. for that application?
-Charlie
Most transformers were around 100VA, and the purpose was to snub the ringing when the rectifiers turn off. This could be done on either the primary or secondary with good results. I tend to snub closest to the rectifier, which means the secondary. Regardless of voltage and current, values tend to fall in the same order of magnitude, which I would not have expected. A typical schematic is attached.
As far as across the switch contacts, you could go with 'typical' values from an application note, and no doubt they will work. They tend to be "one size fits all" values, as the C is increased and the R decreased. All this will accomplish is the same as scope-adjusted values, with additional dissipation in other lower frequency components that provide little benefit.
I often will just place a Mallory quencharc across the switch (and/or across the 120VAC line) for turnoff suppression. That's when I'm lazy and don't want to break out the scope, but they aren't exactly cheap. They are overkill IMO as they attempt to take into account all scenarios, but they do work.
http://www.mouser.com/catalog/catalogUSD/644/960.pdf
Across the switch contact, a quencharc will permit much greater leakage through it than a specifically tuned snubber. Food for thought.
As far as across the switch contacts, you could go with 'typical' values from an application note, and no doubt they will work. They tend to be "one size fits all" values, as the C is increased and the R decreased. All this will accomplish is the same as scope-adjusted values, with additional dissipation in other lower frequency components that provide little benefit.
I often will just place a Mallory quencharc across the switch (and/or across the 120VAC line) for turnoff suppression. That's when I'm lazy and don't want to break out the scope, but they aren't exactly cheap. They are overkill IMO as they attempt to take into account all scenarios, but they do work.
http://www.mouser.com/catalog/catalogUSD/644/960.pdf
Across the switch contact, a quencharc will permit much greater leakage through it than a specifically tuned snubber. Food for thought.
Hello,
I get a pop when switching off power. So I installed a 10n ceramic cap across the switch lugs, and voila, no popping. (I mean it was greatly reduced)
Having read some things here, I was thinking about upgrading it to a CRC network: 10n, 400V in series with 47R , 2W non inductive, and this series RC in parallel with a 100n, 400V capacitor. And this parallel combination across the switch lugs.
Any opinions?
I get a pop when switching off power. So I installed a 10n ceramic cap across the switch lugs, and voila, no popping. (I mean it was greatly reduced)
Having read some things here, I was thinking about upgrading it to a CRC network: 10n, 400V in series with 47R , 2W non inductive, and this series RC in parallel with a 100n, 400V capacitor. And this parallel combination across the switch lugs.
Any opinions?
I think you mean a Cornell-Dubilier Quencharc . Here is one of them at Mouser, including a link to the CD datasheet 104M06QC100 Cornell Dubilier | Mouser
Out of experimentation, to see whether it kills it completely. I still get a subtle transient, a little "click" and "bump", but I need to put my ear close. So yes, problem solved anyway, just trying another way of implementation.
Hi all,
I'm pretty late into this topic and had a chance to read it all. I had the same Pop issue when main switch is off. I'm only interested in putting snubber across switch.
Instead of putting RC ckt in parallel with the main switch as mentioned in this thread, can I just place a single MOV in parallel across the switch and call it a day ?
Let me know if this MOV works to put across the main switch?
This is the MOV that i'm currently looking at : https://www.amazon.com/gp/product/B00NQ9L8US/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1
Thanks,
Tom
I'm pretty late into this topic and had a chance to read it all. I had the same Pop issue when main switch is off. I'm only interested in putting snubber across switch.
Instead of putting RC ckt in parallel with the main switch as mentioned in this thread, can I just place a single MOV in parallel across the switch and call it a day ?
Let me know if this MOV works to put across the main switch?
This is the MOV that i'm currently looking at : https://www.amazon.com/gp/product/B00NQ9L8US/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1
Thanks,
Tom
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.