Underwurldes - Diy Auto Top Off Project

[SkiFletch]

Thought so Andy. They look like they can support up to 18V in... So would an 18V transformer work? Here's the datasheet for the specific lin reg I was planning on using (L7812ABV is the specific model since there are many in the datasheet)

 

[Underwurlde]

35V in max. 18VAC absolutely fine (you'll see much higher than this voltage once rectified and smoothed, say about 25VDC @ Vin. Might have to keep an eye on tab temp. Plan in the need for a Heatsink of some kind & test with relay closed & LEDs on etc.

Andy

 

[SkiFletch]

You mean a heatsink on the tab of the rectifier? I can do that The rectifier tab also must be electrically isolated from other parts of the circuit yes?

 

[Underwurlde]

I'm sure you mean regualtor. Yes, heatsink the tab. Bolt it to a piece of metal. The tab as you suspect is not electrically isolated... it is common to the middle pin (GND) - check that, it's from my poor memory.

Andy

 

[SkiFletch]

Yeah, I always do that, mix up rectifier and regulator, my bad. I too was under the impression that the tab was common to the middle pin of the regulator, so I'll just make sure to keep it isolated

 

[sophos9]

FINALLY, the pump turned up today which was great!

Changed connectors and set everything up to test. Checked the float levels, plugged in and it started topping up and topping up and topping up and topping up and topping up some more...

Too much top up, the unit would not switch off. Now I've been testing and the problem seems clear (not that I have a clue about the solution...).

These are the conditions:

• When the pump is NOT plugged in, the unit works fine.
• When the pump IS plugged in, it will not shut off - just keeps on pumping

When the pump is plugged in and the sensors have activated it, after the time delay you can hear the relay click and for a split second the red light comes on before going straight back to amber and the pump continues pumping

Andy, any idea?

 

[sophos9]

Been playing around bit more and what seems for no reason, the unit now operates between going off first time or taking up to 6 attempts, the norm is 3? Could this be down to interference of some sort?

 

[Underwurlde]

Sorry Sophos - I saw your post last night and then had to go out.

It is not unusual for any electronic design to have teething problems - par for the course really as so many things are happening - also I am designing this thing long distance!

I thing this is down to a similar problem like we saw on that other temperature device thing: electrical noise is getting in on the float input line and is causing multiple triggers and hence it is on all the time.

There are a number of ways around this - decrease the line inpedance (thus decrease 1K resistor R7) and hence a larger current in otherwords more energy in the source of the noise is reqired to trigger the input or we filter the noise out (like we did on the temperature guage thing).

So, like we did before, add a small capacitor across the input to the float swtich.
To impove it, at the same time you could also make R7 a lower value, say 100ohms or something like that.

Don't worry about this, quite normal and I think now looking at what I have in my design I should have added a filter to the float switch input of first stage 555 monostable.

Andy

 

[sophos9]

Hi Andy, hope everything is going well!

Sure, totally appreciate it and thanks again (as always) for the help! Will get the bits and drop em in, sounds like it will sort the problem, I've noticed the T5's kickstart the filling cycle so this will be great!

Once its sorted I will get the photos up

 

[SkiFletch]

*waits with baited breath and parts list*

 

[SkiFletch]

Ok Andy, here's the Parts List I have preliminarily (pending Sophos' findings). Damn I love google docs

All items are from Newark.com. I'm pretty confident about most of the components, those I'd like you to look over are the transformer, rectifier, relay, and regulator. Really wanna make sure those will do the job. The remaining components seem very normal and hard to screw up . Just plug the part number into any Newark search field and you'll get the description, datasheets are found on the left hand side near the acrobat link if you need them.

Thanks again for perusing this Andy, you've been a tremendous help.

 

[sophos9]

Should be getting the parts today so will post up ASAP

Thanks again for perusing this Andy, you've been a tremendous help.

Second that

 

[sophos9]

Changed R7 to 100ohms and it works perfect everytime now! Thanks Andy...

Will sort out and follow up with some photos

 

[Underwurlde]

Ha! Way to go.

TBH, i'm not enitrely happy with just bunging in 100ohms. It's not an elegant solution.... akin to using a sledgehammer to crack a walnut.

But, then again, I hear my golden rule ringing in my ears: If it ain't bust, don't fix it.

However, just to go back to filtering ALL noise out, I would still be tempted to put something like 100nF capacitor at the input pin IC4, right at the IC, in otherwords, put it from pin 1 to pin 2 on IC4.

Andy

PS, Ski, I will look at that parts list for you in about an hour's time after the sprogs have gone to bed....

 

[sophos9]

It's not an elegant solution.... akin to using a sledgehammer to crack a walnut.

Why does that sentance amuse me... (starts thinking of my first board attempts....)

However, just to go back to filtering ALL noise out, I would still be tempted to put something like 100nF capacitor at the input pin IC4, right at the IC, in otherwords, put it from pin 1 to pin 2 on IC4.

Whats the negative of using a 100ohm?

Thanks again, as always...


PS... What do you know about a failed microwave