How to build a 21st century protection board

I've come up with a supply with rail disconnect mosfets to shut it down in case of a fault.
That's interesting (in a good way, of course!). It's exactly what I was thinking in post #461.

I looked up that mosfet driver thingie (this link sends a .pdf datasheet despite no filetype suffix):
Data Sheet - ASSR-V621 and ASSR-V622, Dual Channel Photovoltaic MOSFET Driver (6.5V/15μA) (95 KB)
and it says:
Turn Off Time TOFF 0.03 ms IF =10mA, CL=1nF
That's plenty fast enough, but I'm wondering what a "typical" gate capacitance would be, especially with the Miller effect. I can't wait to see what the switch-off will look like under real-world conditions.
 
That's interesting (in a good way, of course!). It's exactly what I was thinking in post #461.

I looked up that mosfet driver thingie (this link sends a .pdf datasheet despite no filetype suffix):
Data Sheet - ASSR-V621 and ASSR-V622, Dual Channel Photovoltaic MOSFET Driver (6.5V/15μA) (95 KB)
and it says:
Turn Off Time TOFF 0.03 ms IF =10mA, CL=1nF
That's plenty fast enough, but I'm wondering what a "typical" gate capacitance would be, especially with the Miller effect. I can't wait to see what the switch-off will look like under real-world conditions.

I haven't physically measured off time yet but they talk about onboard fast off circuitry. It's definitely faster than a fuse from accidental testing.
 
With opto iso and zero cross detection indication Arduino can handle it all. I am working on that too.

I have a few different opto/triac drivers to choose from. Some have the zero crossing circuitry built in. I've had no previous experience with triacs so this is all foreign to me. I'm still trying to figure out how many quadrants I need the triac to be active in. I think just two? I'm going to start out just using them as relays and use a resistor for soft start.
 
I just use 2 quadrants.
I have added a resistor across the output so the holding current is always met.
This means I can quickly pulse triac gate instead of holding gate voltage on for the whole cycle.
Pulsing the triac means I don't need to drop so much power in the mains dropper resistors.
An externally hosted image should be here but it was not working when we last tested it.
 
I have a few different opto/triac drivers to choose from. Some have the zero crossing circuitry built in. I've had no previous experience with triacs so this is all foreign to me. I'm still trying to figure out how many quadrants I need the triac to be active in. I think just two? I'm going to start out just using them as relays and use a resistor for soft start.

Switching a triac creates a lot of noise. With relay and resistor , you just have low
level radiated fields.
I have a triac dimmed work lamp , my amp can easily pick up it's noise. :(

I accidentally created a large DC offset while tinkering. This control board/
SS relay combo switched that speaker off in the 1ms range ... woofer
never moved more than a couple mm's.
I've done the same stupid error on my HK analog IC protection
(offset wise). It is FAR slower.
PS - all the tinkering I've done , this contraption "saved my butt" !! :)
OS
 
Switching a triac creates a lot of noise. With relay and resistor , you just have low
level radiated fields.
I have a triac dimmed work lamp , my amp can easily pick up it's noise. :(

I accidentally created a large DC offset while tinkering. This control board/
SS relay combo switched that speaker off in the 1ms range ... woofer
never moved more than a couple mm's.
I've done the same stupid error on my HK analog IC protection
(offset wise). It is FAR slower.
PS - all the tinkering I've done , this contraption "saved my butt" !! :)
OS

Are you talking about radiated noise or garbage on the mains feed? I don't mind bending up a little shield box to cover the triac board. It's safer that way anyway.
 
Are you talking about radiated noise or garbage on the mains feed? I don't mind bending up a little shield box to cover the triac board. It's safer that way anyway.

The lamps circuit seems to both radiate and feed noise into the line direct. :eek:
Could be a substandard implementation of your standard triac
lamp dimmer circuit.

As far as a box , modularize it ... put both the control board trafo and the
dirty AC stuff in one. My control board EI puts out more EMI than
my whole 1KVA main transformer.

Materials , I've tried regular steel - sucks ! "GOSS" - "grain orientated silicon steel"
is what they use on little e-waste audio EI's. Put this between EMI source and
electronics - goodbye noise !

OS
 
The lamps circuit seems to both radiate and feed noise into the line direct. :eek:
Could be a substandard implementation of your standard triac
lamp dimmer circuit.

As far as a box , modularize it ... put both the control board trafo and the
dirty AC stuff in one. My control board EI puts out more EMI than
my whole 1KVA main transformer.

Materials , I've tried regular steel - sucks ! "GOSS" - "grain orientated silicon steel"
is what they use on little e-waste audio EI's. Put this between EMI source and
electronics - goodbye noise !

OS

I use a little toroid for the control transformer but I'm all for modularizing everything. Less wire clutter! My specialty metal supplier stocks "soft iron" sheet for shielding. That may be GOSS? They also stock copper sheet.
 
Switching a triac creates a lot of noise. With relay and resistor , you just have low
level radiated fields.
I have a triac dimmed work lamp , my amp can easily pick up it's noise. :(

I accidentally created a large DC offset while tinkering. This control board/
SS relay combo switched that speaker off in the 1ms range ... woofer
never moved more than a couple mm's.
I've done the same stupid error on my HK analog IC protection
(offset wise). It is FAR slower.
PS - all the tinkering I've done , this contraption "saved my butt" !! :)
OS

I have recently shorted the OPS board output (with the speaker connected via the 21'st century board) to the negative rail with DMM lead :eek:

Only a light click in the speaker - also a couple of mm movement of the woofer - that's it. Lightning fast, despite of the fact it ends with a relay :)
Even the output transistors survived.
 
I have recently shorted the OPS board output (with the speaker connected via the 21'st century board) to the negative rail with DMM lead :eek:

Only a light click in the speaker - also a couple of mm movement of the woofer - that's it. Lightning fast, despite of the fact it ends with a relay :)
Even the output transistors survived.

My cat torture tested my CFA-CFPx2. He knocked both heat sinks flying while they were playing. The only sound I heard was heat sinks banging together. Everything was fine on restart. These do work very well!
 
Hi,
Just for your consideration. To prevent the noise from the triac what I do it is bypassed the triac with a relay once it is ON. This way you will eliminated the noise coming from the triac. I used the traic only to turn ON/OFF the amplifier at zero crossing once it is On there it is not need for the triac.