Hi everyone,
For a while I have been using a 0.022uF X2 capacitor in series with a 20 Ohm resistor to suppress the arc and prolong the useful life of power switches and relays in my class D amplifiers (24-36VDC, 5-15A). I usually connect the said capacitor/resistor across the load.
However, lately I have been reading more about this topic and I now feel that 0.022uF might not be enough.
Would a 1-10uF be a better value to use? If not, then what would you suggest and why?
For a while I have been using a 0.022uF X2 capacitor in series with a 20 Ohm resistor to suppress the arc and prolong the useful life of power switches and relays in my class D amplifiers (24-36VDC, 5-15A). I usually connect the said capacitor/resistor across the load.
However, lately I have been reading more about this topic and I now feel that 0.022uF might not be enough.
Would a 1-10uF be a better value to use? If not, then what would you suggest and why?
This is what I use to get to an adequate solution.
https://www.illinoiscapacitor.com/pdf/Papers/spark_suppression.pdf
https://www.illinoiscapacitor.com/pdf/Papers/spark_suppression.pdf
Thank you for the quick reply. I read that document before, but it is not very clear whether the formula in the document applies to AC, DC, or both. From what I have seen in vintage class A and AB amplifiers, very few designers (if any) used that formula. I typically see 0.047uF, 0.005uF, etc. ceramic capacitors across the power switch in older amplifiers.
While I was back in my country, I read a translated article from a European journal that claimed that one must use a 0.01-0.1uF capacitor across the DC power switch for arc suppression. Unfortunately, I cannot find that article anywhere now.
While I was back in my country, I read a translated article from a European journal that claimed that one must use a 0.01-0.1uF capacitor across the DC power switch for arc suppression. Unfortunately, I cannot find that article anywhere now.
Both work - a snubber limits the max voltage of disconnect transient for inductive loads by providing an alternate route for current - it doesn't matter where that current goes, just that its able to go and thus reduce the dV/dt seen by the inductor.
Across the load is more logical and practical as the snubber is tuned to the load characteristics, and it ensures galvanic isolation when the switch is open.
For DC use a free-wheel diode may be a better choice, the arc can be fully suppressed, whereas an RC snubber is always a compromize. Free-wheel diodes have to go across the load.
Mark Tillotson, thank you for suggesting to use a free-wheel diode. How do you feel about combining an RC snubber with a diode? An example can be seen here Relays Part 2 on Figure 4.2.
russc, thanks for the reply. No, I do not use batteries to power my amplifiers. I have a bunch of class D amps (dual TPA3116, TDA7498E, TPA3251, etc.) that I usually power with industrial SMPS power supplies.
russc, thanks for the reply. No, I do not use batteries to power my amplifiers. I have a bunch of class D amps (dual TPA3116, TDA7498E, TPA3251, etc.) that I usually power with industrial SMPS power supplies.
Thanks for the reply. Could you please explain to me why would connecting an R/C snubber across the load have no effect? From what I am reading, it actually seems to be the preferred way of arc suppression compared to connecting the snubber across the switch/relay. Here are two sources:
* https://stevenengineering.com/Tech_Support/PDFs/38EARCS.pdf
* https://www.idec.com/language/english/AppNotes/Relays/contact_circuit_protection.pdf
It is more convenient for me to switch the DC side
The diodes in the switching power supply won’t last a long time if the mains are switched frequently, so it makes sense to switch the dc side.
My class d setup is left powered on, establishing a Bluetooth connection allows for sound, disconnecting it effectively shuts it off. Is used a few times a week when I exercise.
My class d setup is left powered on, establishing a Bluetooth connection allows for sound, disconnecting it effectively shuts it off. Is used a few times a week when I exercise.
Thanks for sharing. 10nf-100nf either across the load or across the switch contacts is what I used to do for a long time. I might be overthinking it, but I now feel that a higher capacitor value would be a better choice for most class D amps.
Are there any disadvantages of placing a free-wheel diode across the load while, at the same time, placing an R/C snubber across the switch contacts?
I used to run a mobile disco with dual mains turntables.
I got an awful crack through the speaker when turning the tables on and off.
A 100nf 250VAC cap improved it no end.
Okay, thanks. I will try a Zener/diode combination in parallel with the load together with an RC snubber across the relay/switch contacts.
Thanks!