DPS5020 dual power supply build

[eccerr0r]

Re: DPS5020 dual power supply build

I'd imagine Canada have much similar rules as USA which is described in the NEC. However even https://en.wikipedia.org/wiki/Electr..._North_America has some wording implying that neutral switching is taboo in most circumstances:

North American standards state that the neutral is neither switched nor fused except in very narrowly defined circumstances.

Agree that it should not be fused, but should be okay to be switched IFF hot is switched off first, and hot is on last.

Need to find the explicit rule, and then again whether this is only for building wiring or does it apply to end run pluggable equipment as well...

[megaraider]

Re: DPS5020 dual power supply build

That would agree with what I see so far - meaning a DPST is okay since neutral is not ground, but I still don't quite understand the statement in post 105...

I'm not aware of the electric regulation on Canada.

Nevertheless, it's a common mistake wrong labeling and interpretation about Earth connection.
While the Line is sometimes referred as Hot without raising any problems,
the Neutral is often mislabeled or referred to as Ground leading to wrong assumption and confusion with Earth .

Therefore, with exception for the electric house panel where Neutral, and Earth (off course), must always remain connected (only the Line can be switched),
it's very common to switch both Line and Neutral.
In contrast switching Earth is totally forbidden.
.

[eccerr0r]

Re: DPS5020 dual power supply build

That would agree with what I see so far - meaning a DPST is okay since neutral is not ground, but I still don't quite understand the statement in post 105...

[stj]

Re: DPS5020 dual power supply build

neutral is not considered grounded in the device or even a building - it would be a dangerous asumption.
any bonding is at the power company's side - either at the box or the transformer.

[eccerr0r]

Re: DPS5020 dual power supply build

That is right that dual contact switches one will make/break before the other and that one will die first, but I was wondering about that electrical code requirement for switch both line and neutral at the same time... Switching protective ground is clearly very bad, but would think that neutral would not be so bad...

I've seen a lot of computer equipment that has both line and neutral switched, and seen/have some DPST wall toggle switches. Is this really against code? The most I saw in the electrical code rules is that if neutral gets disconnected, line MUST be disconnected at the same time, which implies DPST are OK.

On the other hand, I also saw that "grounded" conductors must not be switched, and whether neutral, which is "ground" also counts as "ground" that must not be switched. Now that remains a curiosity for me...

[Dannyx]

Re: DPS5020 dual power supply build

It turns out dual-switch contacts are never identical, so one will close last (power-on inrush current) and another will open last (power-off arc). I had a boss yell at me that it would not last any longer than using one switch contact, when I tried them in parallel.

Ah yes, now that you mention it, I remember hearing about this before (including the boss part ), so you probably answered somewhere in the earlier posts.

Electrical codes here in NA forbid switching both line and neutral in a device. You can only switch line or hot. But I see your breaker panels and motor contactors seem to disconnect both.

I think this is because EU plugs are not polarized, so either conductor can become hot at one time....just a thought. I, as an ordinary do-it-yourselfer, always choose to cut both wires whenever possible "around the house", like if I wire up a simple power strip for instance...

[redwire]

Re: DPS5020 dual power supply build

It turns out dual-switch contacts are never identical, so one will close last (power-on inrush current) and another will open last (power-off arc). I had a boss yell at me that it would not last any longer than using one switch contact, when I tried them in parallel.

Electrical codes here in NA forbid switching both line and neutral in a device. You can only switch line or hot. But I see your breaker panels and motor contactors seem to disconnect both.

I would use two snubbers, one across each switch contact if you want to open both line and neutral. That is the best way to have no spikes when you switch off. I tried a MOV across the transformer primary but still got a spike.

[Dannyx]

Re: DPS5020 dual power supply build

This is great info. Sorry for replying so late, but it's only recently that I started working on the project again

If I use a double-gang switch to cut both the L and N AC lines, should BOTH sides of the switch get a snubber across them ? Silly question perhaps.....

[redwire]

Re: DPS5020 dual power supply build

To suppress back EMF from an inductive load, an RC snubber is commonly used. A DC inductive load usually has a diode across it to clamp the back EMF.

You can guess a snubbers value based on the transformer's size.
I would use 33-47nF and 22-100 ohms 1-3W as a ballpark for a 1,500VA transformer. The film cap needs to be HV rated, a junkbox X-cap is OK but if the cap fails short, the resistor burns up.

Leakage current from the cap doesn't matter, the transformer might have 1mA flowing all the time but it will stay asleep. Although some loads will still have weak ghost power from that.

[Dannyx]

Re: DPS5020 dual power supply build

You are switching AC to the AC load so a NO NO to put diode across the load, you will see nice fire works when that diode conducts.

Yes, I know that That's why I mentioned DC as a separate topic.

[budm]

Re: DPS5020 dual power supply build

You are switching AC to the AC load so a NO NO to put diode across the load, you will see nice fire works when that diode conducts. You will put diode across the RELAY DC winding with proper DC polarity to the relay winding.
RC snubber of for the switch contact.

[Dannyx]

Re: DPS5020 dual power supply build

that question depends on the inrush of the load.

the idea is to supress any contact arcing while keeping leakage across open contacts to a minimum.

Yeah, that's tough to determine.....say a beefy transformer in my case, though that's not really helping much.

[stj]

Re: DPS5020 dual power supply build

that question depends on the inrush of the load.

the idea is to supress any contact arcing while keeping leakage across open contacts to a minimum.

[Dannyx]

Re: DPS5020 dual power supply build

Question then is what value should the cap/resistor have and what should the cap be ? One of those yellow "X" caps ? I reckon it's dependent on the voltage/current you're expecting to see there, though I think a higher voltage rating for the cap would work on a 12v system too.

[megaraider]

Re: DPS5020 dual power supply build

C) both ?

Yes, both.

[stj]

Re: DPS5020 dual power supply build

it's good to put a film cap across the contacts, but you want a resistor in series to limit the current.
you can buy caps with the resistor built in for use as snubbers.

[Dannyx]

Re: DPS5020 dual power supply build

Still on the topic of PSUs in general, since I didn't want to open a whole new thread just for this, I thought I might as well ask here: when you've got a mechanical switch/relay and a load, particularly an inductive one, should you place:
A) a capacitor across the switch
B) a diode across the load
C) both ?

Obviously in AC circuits you can't put a diode across the load, but what about DC ? This PSU will use a couple of relays and I want to make sure it's as industry-standard as possible.

[Dannyx]

Re: DPS5020 dual power supply build

Yes, you are right: the metal shell of the micro USB connector does not contact the GND of that output header which plugs into the PSU module itself. I checked it today.

[Dannyx]

Re: DPS5020 dual power supply build

I just remembered I faced another intricacy of transformers when I built my first supply: the transformer got rather warm (to hot) when unloaded and seemed to stay cool when actually LOADED, ironically....I suppose it's got something to do with induction, but again: transformers to me are one of the most complex electronics parts, with all their particularities and physics they involve, so I'm not really sure why. I left it plugged into the mains on the workbench without anything attached to the secondary (I think I simply forgot about it) and when I came back after an hour or so, it was too hot to touch...Yeouch !....so that got me working on designing some sort of cooling system, especially since the first case is wood...not the best design idea, but it was the best we could build from absolute scratch. This one is metal and has plenty of space inside, plus ventilation holes, so it should be OK to go fanless.

For my first one, so the fan wouldn't be on all the time, I modified a car mobile phone charger that normally plugs into the lighter socket to increase/decrease its output based on heat. I replaced one of the feedback resistors with an NTC and tacked that down to the top of the transformer, so as it got hotter, the NTC's resistance dropped and increased the output voltage, making the fan spin faster...don't think I'll go through that trouble with this one though, at least not yet.

[redwire]

Re: DPS5020 dual power supply build

Second side done

Looks really good. I add a PE GND green binding post in case I need it.

The DPS5020-USB board has two opto-isolators, so the USB should be OK. The shell can touch chassis PE ground or float unconnected.