HOWTO - Make a BrewPi Fermentation Controller For Cheap

[day_trippr]

You're good. A series-pass regulator dropping from 12VDC to 5VDC, the big honking caps won't hurt if the wall wart has some ripple, and the sink looks adequate, especially given the fan in close proximity...

Cheers!

 

[100amps]

Or just use a 120v panel mount indicator and install it on the high voltage side

That would definitely be the best place to monitor, and I suppose there might be decent looking 120v panel lamps available, but it invokes images in my mind of the jewel from a 1965 Fender Twin. Lol. And I have lots of LEDs kicking around, which got me thinking (be afraid), so I came up with a way to use LEDs on the power side without generating too much resistor heat.

..Todd

 

[wbarber69]

Not bad. Not at all. I love how this thread has sparked so many ideas for me to steal

 

[day_trippr]

I know exactly which site that schematic came from.

I wouldn't go that route, but it should work...for awhile, at least, until an LED breaks down from the reverse voltage...

Cheers!

 

[thekraken]

I know exactly which site that schematic came from.

I wouldn't go that route, but it should work...for awhile, at least, until an LED breaks down from the reverse voltage...

Cheers!

Could he just put a fet on the relay pins from the arduino that opens up vcc to the leds?

 

[wbarber69]

And incur the voltage drop

 

[thekraken]

And incur the voltage drop

I thought he was more worried about exceeding the current limit on the arduino's io pins.

 

[day_trippr]

A cheap n-channel FET with a TTL compatible gate (eg: 2N7002) and wired from 5V through say a 220 ohm resistor to an LED would be a safe, sane and long-lived solution. Even a cheap FET such as the 2N7002 has RDS(on) spec in the single-digit ohm range - you're not going to notice the drop at the LED

Cheers!

 

[100amps]

I wouldn't go that route, but it should work...for awhile, at least, until an LED breaks down from the reverse voltage...

Cheers!

Wouldn't the opposing polarity of the diodes alleviate the reverse voltage across the currently inactive one?

And yeah, I am trying to avoid drawing more VCC current from the Arduino pins.

..Todd

 

[100amps]

A cheap n-channel FET with a TTL compatible gate (eg: 2N7002) and wired from 5V through say a 220 ohm resistor to an LED would be a safe, sane and long-lived solution. Even a cheap FET such as the 2N7002 has RDS(on) spec in the single-digit ohm range - you're not going to notice the drop at the LED

Cheers!

Not sure I'm following that.

Got a napkin sketch for me? I'm out of my prehistoric audio circuit comfort zone here.

What's driving the FET 's gate? And what's powering the LED?

..Todd

 

[gezzanet]

love this controller. beer still fermenting at 18 degC with outside temp 43, cool.

 

[BrewCityBaller]

love this controller. beer still fermenting at 18 degC with outside temp 43, cool.

43*?! That's a scorcher. It's -5*C here in the Northern Hemisphere...

 

[thekraken]

Not sure I'm following that.

Got a napkin sketch for me? I'm out of my prehistoric audio circuit comfort zone here.

What's driving the FET 's gate? And what's powering the LED?

..Todd

Attached drawing is what I was thinking.

DISCLAIMER: I haven't had a lick of electronics classes, instruction, or training.
WARNING: Follow my schematics at your own risk!

I only post this drawing because I'm thinking of doing the same thing my self.

day_trippr, I drew the schem with npn transistors as you suggested, but if we're using the sainsmart relay where the input pins are inverted wouldn't we need to use pnp transistors?

 

[100amps]

DISCLAIMER: I haven't had a lick of electronics classes, instruction, or training.

You too huh? Glad I'm not the only one. But I see what you meant now.

but if we're using the sainsmart relay where the input pins are inverted wouldn't we need to use pnp transistors?

The inversion occurs after the relay inputs though, doesn't it. Should be normal at this point still.

Also thinking each LED node should probably get it own limiting resistor instead of putting it on the relay board's supply. Maybe it's not a big deal as long as you calculate the current values accurately.

It'd be more useful to have the indicator on the relay outputs though, so you can see if the relay board's output is alive.

..todd

 

[thekraken]

The inversion occurs after the relay inputs though, doesn't it.

I believe that the relay pins are high in the off state and low in the on state, so it's when the current stops on pins 5 or 6 that is when the relay kicks on.

It'd be more useful to have the indicator on the relay outputs though, so you can see if the relay board's output is alive.

Good point, agreed. The fiber optic filament idea that you mentioned a while back seems worth looking into (it'd at least be the safest option). [Edit] Though, I guess then you'r still not indicating off the actual 120v output itself...

 

[wbarber69]

Plus if you use a 120v indicator light on the output side of the relay you don't need any fet's or resistors or anything. Just a simple indicator light.

 

[thekraken]

Hey, you could use a pair of these old school looking light bulbs as your indicators, and maybe some nixie tubes as your temperature display. Don't ask me how to wire it up.

 

[100amps]

I'll have to look for a reasonably small 117AC indicator light. I've never seen one that wasn't huge. NOt enough room for those unless I find something LED size.

But any concern about sanity and safety with those LEDs is laughable IMO. I work with tube guitar amps on the bench normally, and some of the commercial products that sell by the millions would make you scramble for your rubber suit if you saw some of the 500-600V 'techniques' being used.

..todd

 

[100amps]

I believe that the relay pins are high in the off state and low in the on state, so it's when the current stops on pins 5 or 6 that is when the relay kicks on.

But it's the logic state that's inverted, not the polarity, right? The voltage is still +5v. (I'm referring to the question about using PNP devices.)

I do have some nixie tubes in my junk box. Hmm... Nah, retro isn't going to fit in my box.

..Todd

 

[wbarber69]

Check out RadioShack for the 110 indicators. They have a small diverse selection.

Or check an automotive store they usually have those little blue and green LEDs for dumb kids to stick all over their rice-burners.

 

[100amps]

RadioShack left Canada about 10 years ago. They sold the whole chain to Circuit City, renamed it the Source. When CC failed, the Source was acquired by Bell Canada (one of Canada's big 3 phone/cable companies). They try to imitate RS but mostly sell dollar store quality Made in China disposable crap.

Automotive store for 120v stuff? I'll check Digi-key, see what they have. But since I have boxes of LEDs, caps and resistors, I'll probably use what I drew.

Now if I could connect the gate of a FET to the 120VAC to turn on a 5v LED supply, I'd be happy.

..Todd

 

[thekraken]

But it's the logic state that's inverted, not the polarity, right? The voltage is still +5v. (I'm referring to the question about using PNP devices.)

I do have some nixie tubes in my junk box. Hmm... Nah, retro isn't going to fit in my box.

..Todd

The problem with the schematic I drew, using the sainsmart, when the relay's are off pins 5 and 6 are high, so the npn transistors gate would be open and led's would be lighting up. The LED's would work opposite as they are supposed to, they would be on when the relays are off and vice versa.

BUT it's moot since we want indicators for if the actual 120 plugs are hot or not.

 

[gezzanet]

I used Indicators on the 240v side Jaycar sl2632 240v bezel $2.75aud. Xxpost 1556. You should be able to get something similar for 110. Small and simple.


Sent from my iPhone using Home Brew

 

[FuzzeWuzze]

I wonder how large they will let this thread go before they lock it to make a new one We gotta be getting up there by now...

 

[wbarber69]

The soldering stainless thread was up in the 200 page range when I finished my pots

 

[gezzanet]

I wonder how large they will let this thread go before they lock it to make a new one We gotta be getting up there by now...

You've created a monster.


Sent from my iPhone using Home Brew

 

[wbarber69]

But I see that someone has gone in and deleted all the unrelated posts in there. It's only 75 pages now.

 

[balrog]

But I see that someone has gone in and deleted all the unrelated posts in there. It's only 75 pages now.

..not seeing only 75 pages. I see 239. And...I'm not drinking...so....?

Ah. You speak of soldering SS thread with the deletion. My apologies.

 

[wbarber69]

They might have split it up

I see 74 pages and only 1400+ posts

 

[day_trippr]

Attached drawing is what I was thinking.

DISCLAIMER: I haven't had a lick of electronics classes, instruction, or training.
WARNING: Follow my schematics at your own risk!

I only post this drawing because I'm thinking of doing the same thing my self.

day_trippr, I drew the schem with npn transistors as you suggested, but if we're using the sainsmart relay where the input pins are inverted wouldn't we need to use pnp transistors?

I should have remembered that, as when I coded up a fan driver using a spare channel through a dual Sainsmart relay board the inversion caught me. Coding negative logic when you're three-quarters in the bag is fun!

Yes, a p-channel fet solves that nicely, they'd swap right into your schematic. btw, wrt to that drawing, your would preferably use a separate current limiting resistor for each of the two LED circuits. Sharing will cause brightness changes when both sets of LEDs are on versus only one.

Anyway, apparently the OP wants something to glow when the actual 120vac lead is hot. This "inferential indicator" wouldn't solve that...

Cheers!

 

[day_trippr]

Wouldn't the opposing polarity of the diodes alleviate the reverse voltage across the currently inactive one?

And yeah, I am trying to avoid drawing more VCC current from the Arduino pins.

LEDs are not actually diodes in the sense that they are not designed to tolerate inverse voltage like actual diodes are. While a diode might have a PIV tolerance in the many hundreds of volts, in fact many LED specs don't even state a PIV tolerance, and the ones that do almost always use the forward voltage with a note limiting the available current to a few microamps.

In the proposed application there will be periods where neither diode is fully conducting, and in those periods one of the LEDs will be seeing inverse voltage with the potential current well in excess of microamps.

Anyway, nobody's gonna die going that route, but I'd recommend using LED sockets so they're easy to replace...

Cheers!

 

[100amps]

LEDs are not actually diodes in the sense that they are not designed to tolerate inverse voltage like actual diodes are. While a diode might have a PIV tolerance in the many hundreds of volts, in fact many LED specs don't even state a PIV tolerance, and the ones that do almost always use the forward voltage with a note limiting the available current to a few microamps.

In the proposed application there will be periods where neither diode is fully conducting, and in those periods one of the LEDs will be seeing inverse voltage with the potential current well in excess of microamps.

Cheers!

Understood. Thanks.

 

[100amps]

I wonder how large they will let this thread go before they lock it to make a new one We gotta be getting up there by now...

This topic is still young and fresh compared to some of the amplifier design topics over on DIYAudio.com.

One day someone's going to invent a killer forum app for DIY projects so folks can find the relevant good bits without the OP having to research and stuff everything into Post #1. One day...

..Todd

 

[aubiecat]

The day I started brewing beer on my kitchen stove a few years ago I had no clue it was going to end up earning me an electrical engineering degree.

 

[wbarber69]

+1

 

[atoughram]

microamps.

Milliamps? I dont think microamps will light a standard LED but I could be wrong.

 

[greenebo]

Hello all. I've been reading this post from the start and had it convince me that I needed a RPi and all the parts to make my own BrewPi controller. Now that I have fumbled through the setup to get my temp probes to talk and my beer to ferment at the temperature I set, I want to watch it through streaming video along with my brewing friends when we're not in my home network. I've read until my eyes bleed about setting up static IPs and all that for the website, but I'm as lost as I can be to set up streaming video for my BrewPi webpage. I'm a previous AT (in the Navy that means Aviation Electronics Technician), but I'm the kind that drop checked boxes and reinstalled them and not the ones that knew how to change resistors and knew what a capacitor looked like. Is there a complete A-Z post that a simpleton could read to go from having a webcam to having a webcam that can display live video - and then live video on the BrewPi application so I can show friends how the swimming yeast make alcohol?

If not, I still love the HBT community, I've successfully gotten two associates back into brewing and a working on getting around 5 untrained associates hooked on home brewing so it's win-win either way. I would prefer someone tell me Barney style, but if the worst that comes out of this is about 5 new sources of homebrew, I'm cool with it.

Thanks for reading this!

Chris

 

[100amps]

In the proposed application there will be periods where neither diode is fully conducting, and in those periods one of the LEDs will be seeing inverse voltage with the potential current well in excess of microamps.
Cheers!

Hey day_trippr, I'm loving your electronics expertise here. Do you see any room for subtle improvement in the basic schema I presented for running LEDs on 110VAC, to extend the life of the LEDs? Maybe a Schottky or FRED in parallel with the LED to speed up that near-cut-off reverse voltage situation? (I'm totally guessing, no training here at all.) Or anything else you can think of that would improve that particular situation?

..Todd

 

[100amps]

Milliamps? I dont think microamps will light a standard LED but I could be wrong.

day_trippr was referring to reverse current handling capability -- when the LED is *not* lit because the polarity of the current is backwards. Apparently LEDS have a fairly wimpy back door that's easily broken down by [microamp] current.

..Todd

 

[day_trippr]

Hey day_trippr, I'm loving your electronics expertise here. Do you see any room for subtle improvement in the basic schema I presented for running LEDs on 110VAC, to extend the life of the LEDs? Maybe a Schottky or FRED in parallel with the LED to speed up that near-cut-off reverse voltage situation? (I'm totally guessing, no training here at all.) Or anything else you can think of that would improve that particular situation?

..Todd

Have to think about it. First thought is to add an actual forward-biased diode in series with each of the two LEDs, but I'm not sure that actually alleviates the PIV problem...

Cheers!

[edit] After mulling this over in the shower, and then checking the reverse leakage current vs voltage and temperature specs for a random pair of generic diodes, I believe adding a series diode to each LED will indeed protect the LEDs nicely.