Backfeed concern with battery pack and AC/DC power supply

[the1snm]

About a year ago I bought one of the Hakko T12 clones like seen here on EEVblog and built a decent 3D printed enclosure to hold it and a cheap WX-DC2412 AC/DC power supply like this Amazon one. The clone accepts 16-24V3A and the power supply is "rated" 24VDC at 4A continuous and 8A peak and to it's credit has been working good for many soldering projects. Recently I've had several needs to solder at distances ~100ft from the nearest outlet and I decided to modify my printed case to accept a DeWalt 20V battery pack since I have a bunch of them. This isn't uncommon to do on designs on Thingiverse but they all do one or the other and I want to run both (not at the same time). I've already designed the new case to handle both and it's actually printing as I type this out.

I know just barely enough about electronics (I tinker quite a bit with low voltage stuff like Arduino's and such) to know that even if there is no AC line plugged into the AC/DC power supply dropping 20V in series at the output of the board before the soldering station board is one of the rookie mistakes you make and go duh afterwards because the battery pack could be backfeeding the power supply.

Idea 1 was to drop a Schottky diode in there (IN5408 or IN5822?) since I have a variety left over from a previous project but then I remembered I don't know enough about them to and while occam's razor says the simplest idea is usually the best one I don't think he had n00bs and electronics in mind at the time.

Idea 2 was me thinking well heck I know how a SPDT switch works inside and out and I have those laying around. While a pragmatic approach it's about as sexy as a dirty dump truck at a classic car show.

Idea 3 was to use a non-latching relay like (but not) an automotive 12V30A 5 pin model so when 20V battery power was supplied it would flipping the power across the 87/87a pins. But I don't own any 24VDC relays just 12V auto and 5V switching ones for Arduino.

Idea 4 was to ask my electrical engineer and enthusiast friends for some advice (hint here for those playing along at home).

Now I've not gone straight here expecting to have someone else solve this for me, I have done quite a bit of googling and seen wild answers all over the board not being very consistent. Responses ranging from don't worry about it, it'll be fine to not needing one because the AC power supply contains rectifiers and any backfeeding would not harm it to elaborate circuits that give me googly eyes first looking at them.

I've got decent enough equipment to check things out if need be RMS multimeters, Rigol DS1054 oscilloscope I'm still learning and other assorted goodies.

Any suggestions where to focus my attention or other idea's to look at or more info/pictures needed?

Steve

[pqass]

Consider a variation of Idea 3; non-latching relay with coil power saver engaging only when wall-power is available.

Use a series resistor matching the coil resistance so you can use a 12V relay in 24V service.
Also, save power by using another series resistor to keep the sustained coil current low but a parallel capacitor to give it the initial snap-on.  Adjust the 470R to taste.

See attached.   Notice the yellow current trace spiking on initial coil power-on but then leveling off.

Play with it on falstad.com here:
http://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgpABZsKBTAWjDACgBnEXG8QqniBQpeVKgDMAhgBsODNgGNwGQkIwplqtHihRYcCC0LRsxYSuKEshBDWL0WkNgDcQLFeBq9sGsF91UNFRoAXoIbABOIJZqGjHaujQYTmAYOjFg-Ny+WVQQqHqQ2DR4WJBg2IQoeGDEkEmFSGDw8IW+cGwA7tFa+JpCfU7dgplUGblsAEq6IlTuWkGhiyKhMOEA5m6pqn68TAm7AV09sf2YGk6b8eon1TpixxlmQou1F2wAJmd9swM6Gh8GFIAK7SAAux1+bxewUGnxh3EqCISAKBklBEOwyWyQjwOh8iJ2RyiBOwSN+ZNUeQ6ww0dxx2ygbAAHltmuAUIZcBzeLw-AMAGoAHQ4ACFJGCwQwIgBPFluX4oBDELbYCBEDR83giYUcADqMmkIoAygB7YERBTyVn7JCVATYfGEehakAAUWZYIikhFEQYHAAlhwwaaIiKQyKALYShQAC19DGkkhlIoUpoDRo4nQDYPjHB1IoN0kzZotVvl+1UWNUCFq3AwKtdDHW0ATgeDoZFpvEIo4kkjDF7AYAXoO+xxU+nM3722DJAA7K0izqxhjzkXAwPz9Yin0cMAoXVpjMigPr-M0XVyCLOAPl1k+HbECAIMngYi8zwgAAqq43HGBGQTQABWNb8RQAB1NTppV7bMwTjaBIm4epTkEHwxBAPBoBQNhNnQkJ0KRB4HxKFCNEdARICQV0ABlTUkD42CxKjtRWQRznAI4uGCRY8g8TCpFka0FSwAYHEwAZVFdY14LjcNTSHSNQQlMdpFNDgJw4btl0NSDoOlY5+KrXwBLYU1wCQQhRHAbAEEIPAXTEHCZks7g2CAA

[thm_w]

We discussed here: https://www.eevblog.com/forum/reviews/ksger-running-on-battery-power/

[the1snm]

Thanks for the links, it seems in the other thread up till the SMPS caps the Schottky diode sounded like the best option. I tested both the 1N5408 and the 1N5822 3A diodes I have and was a little surprised with the results. From most everything I keep reading the 5822 is generally preferred over the 5408 because of the lower forward voltage drop, faster recovery etc. When I connected the cathode of the 5408 to the + off my 20V battery I read about 125mv off the anode side. Repeating the same test using the 5822 I get about 5.22v off the anode side. I know I must be missing something in my naivete on the matter. Did I test this incorrectly or working off the wrong assumptions here?

[thm_w]

Measure the diode using diode function of your meter, with the power off, and see if it is ok. They both should be around 0.4 to 0.6V.

You say cathode to +20V battery, sounds like you've put the diode backwards?

[the1snm]

Sorry if I wasn't clear, I put the cathode side (with the stripe) end to the positive of my battery. I then connected the negative lead of my meter to the neg side of the battery and the positive lead to the anode side of the diode. My idea hear was to test if voltage was in fact passing through in the direction it shouldn't. Reversing the diode and testing resulted in near full voltage as expected. Testing the diodes on my meter resulted in .570v and .245v.

Thanks for taking a moment, I am trying to learn more of this and been doing more research on this and even though the Schottky reverse voltage is ~5.2v is it OK because the maximum reverse current rating is very low i.e. in the mA? I know this power supply isn't terribly expensive but I am trying to learn as much as I can before just slapping one of these diodes on the pos+ output then connecting a 20V 5Ah battery on there. I've let out the magic smoke many times in my life and just trying to hopefully not do it again.

[thm_w]

Ah understand now. Yes you will see some leakage through the diode, this is why 1N5822 read a higher voltage, as it has higher leakage current.

because the maximum reverse current rating is very low i.e. in the mA?

Correct, the spec is 2mA reverse at 25C, which should not be enough to cause any issues with a typical power supply.

[the1snm]

Awesome thanks so much, I suspected that was the case and it was surprisingly difficult to find confirmation of that online. I've give this a bench run here in the near future and see how it goes.

When rummaging through some unused items recently I found I have a pair of N mosfets (different manufacturers) on some unused 3D printer bed heating boards. I remember reading these can be better than the Schottky's when done right. Might be overkill though for what I need since I'm not really concerned about the forward voltage drop of the Schottky and it's only 3A max draw.

[thm_w]

A good mosfet will have a lower voltage drop than the schottky yeah.

With those heated bed modules, you'd have to have some logic to switch between inputs. You wouldn't want both mosfet modules on at the same time, they would short out.
This is where diodes are safer, they can perform the OR logic for you.