5v supply using PS_ON to enable 24v and 120v SSR



  • I've seen a few versions of people using the ATX PS_ON to signal a opto isolated relay, wanted to double check this is a sound setup for switching mains 120v on/off to both the 24v PSU and a 120v bed heat SSR when powering the Duet from a dedicated 5V PSU.

    Goal being primarily to not worry about the 120v SSR for bed heat being energized during a SSR fault, and secondarily just to feel better about leaving the printer "on" all the time since the 24v supply will also be switched off.

    Using these components and layout:

    Mean well RS-15-5 5v/15w/3A supply
    Mean well LRS-200 24v/200w/9A supply
    Crydom D2425 SSR
    Sainsmart 2 channel 5v opto relay (https://www.sainsmart.com/products/2-channel-5v-relay-module)

    layout

    The relay seems to be opto isolated and dioded, should I have any concern about the coil being energized off the same 5V supply powering the Duet though? Is there a failure mode that results in the 120v being switched hitting the 5V side?

    Since its a 2 channel relay I was probably going to parallel both relays rather than "waste" one of them.



  • Looks reasonable as long as the 10A limit on the relay is enough to power both the 24v PSU and the bed heater. You'll want to leave some safety room between the actual 120V current draw and the 10A limit.



  • Bed heater is 300x300 keenovo 750w so figuring 7A to be safe if the SSR were full open, but my PWM rate should be nowhere near that utilization. I'm running it on a 5/16 350x350 aluminum plate so it was chosen for size compatibility not wattage needed.

    The 24v supply's sheet shows 4A "typical" but I'm guessing thats fully loaded which I should be nowhere near that typically.

    Since its got 2 relays I was going to parallel them, or I guess maybe use 1 relay for the bed, 1 for the 24v psu to keep well below 10A per relay.



  • That's not really enough headroom in my mind.

    To keep things from getting too complicated and to keep safe, I'd use one for the bed and one for the PSU rather than paralleling them. In parallel, if one fails to energize, the total current will be going through the other one and you'd then be relying on software settings like the PWM frequency and duty cycle to ensure that the relay didn't get overloaded.

    In parallel, it's also easier to predict the current draw of each and maybe add fuses before the bed and PSU.



  • Good point on the failure mode for parallel, individual for the bed and the 24v psu it is then.

    The failure I'm most interested in is the 120v side of the SSR or heat pad energizing the frame or shorting in some way. My thoughts there are the power inlet fuse would be the protection against that along with the usual ground fault path tripping the breaker, but maybe ill do some fuses inline right after the relay though too.

    I looked around for a while and couldn't find a more legitimate relay module from an Omron or similar like this one, just components to assemble one. Would be nice to have a higher rated relay on it and known good parts...

    @gtj0 said in 5v supply using PS_ON to enable 24v and 120v SSR:

    That's not really enough headroom in my mind.

    To keep things from getting too complicated and to keep safe, I'd use one for the bed and one for the PSU rather than paralleling them. In parallel, if one fails to energize, the total current will be going through the other one and you'd then be relying on software settings like the PWM frequency and duty cycle to ensure that the relay didn't get overloaded.

    In parallel, it's also easier to predict the current draw of each and maybe add fuses before the bed and PSU.



  • The other thing you could consider is replacing the relay with another SSR (which is the arrangement I have). That way it would take 2 SSR "permanently closed" failures to energize the bed which would be extraordinarily rare especially if the SSRs were adequately sized and cooled. I have a 1600w bed heater so I used 2 40A SSRs and watched the entire assembly with a Flir camera to see if there were any hotspots.



  • That's a monster heater!

    Good point on the 2nd SSR, at least then I can get it from a known entity and they have plenty of headroom for my setup. I never even considered doing something like that.

    Gives some parts redundancy too so I could cobble together a working setup if one of the SSRs lets go.

    @gtj0 said in 5v supply using PS_ON to enable 24v and 120v SSR:

    The other thing you could consider is replacing the relay with another SSR (which is the arrangement I have). That way it would take 2 SSR "permanently closed" failures to energize the bed which would be extraordinarily rare especially if the SSRs were adequately sized and cooled. I have a 1600w bed heater so I used 2 40A SSRs and watched the entire assembly with a Flir camera to see if there were any hotspots.



  • Above makes sense -- as an observation. This may have changed in newer Duet iterations, but my original 1.02 version had odd behavior when powered from a 5V in -- the 5v power pins on the PanelDue and the BL Touch read 4.2-4.5 -- I had to ramp up on the input voltage to over 5.6v to get closer to 5v and still the panel due didn't like the lowered voltage. So I scrapped that idea, I didn't like the 7 screen being on all the time anyway, so I am using a more robust power system. A bit overkill, but here is my solution. I have a 4 channel relay module (common on amazon) -- I have a raspberry pi which is used for the camera streaming anyway, so I attached a small 3.5v cheap SPI screen to it and have a little webapp on that screen which currently displays 4 buttons, printer on/off, light on/off -- the light control is done through my home automation system, because it's there, and there is room on the LCD, so the webapp has the Pi toggle all 4 relays, and 2 relays bridge neutral and hot to the 24v PSU, and the the other 2 relays bridge neutral and hot for my heated bed (presently 1kw 120v, soon to be 1.2kw) -- this works really well as 2 relays per connection ensure that there is no way for the bed to be powered even with leakage current from the SSR (which even the best SSRs do leak when off)


Log in to reply