Homeroasters - Home Roasting Coffee Community » DATALOGGERS - CONTROLLERS - RATE OF RISE METERS » Dataloggers/Controllers/Rate of Rise Meters
Who is here? 1 guest(s)
Page 1 of 2: 12
 Print Thread
DIY SSR with Picmicro cpu
bvwelch
Top  #21 Print Post
Posted on 08/29/2008 7:03 PM

Quote

Island Addict wrote:
For my prototyping, I've been using an SOIC to DIP breakout board (http://www.sparkf...cts_id=494), with the TC leads soldered into the DIP holes.

The critical thing is to make sure the cold junction is at the same temperature as the on-chip temperature sensing diode. Avoiding sudden changes in ambient temperature is important, as is, I guess, making sure the thermocouple cable itself doesn't conduct heat down to the cold junction. In my final setup I think I'll put the MAX chip inside a separate metal box to help keep the temps stable around the chip.


I'm not sure what TC you are using, but the cheap ones I use, wear out fairly often, so I'll need to come up with some connector or other.

Yes, I found out the hard way that you need to keep your TC amp/circuit far, far away from the roasting chamber's 400+ F temp. On one of my prototypes, I did the basic TC conversion with a little 8-pin picmicro, and then sent the results via opto-isolator to the main PID controller.

-bill
 
bvwelch
Top  #22 Print Post
Posted on 08/29/2008 7:23 PM
OK, there are a couple of different types of circuits we could use for controlling the heat. One is popular for "dimmers". The idea is to dim the light without flickering. So, they do this by sending "something" to the light bulb at 60 Hz. What you might call a "modified" or "partial" sine wave.

Another approach, which is the one I tend to use, is to turn on/off the voltage at the "zero crossing" of the sine wave. This is nice, in that it doesn't send noise spikes all over your house and possibly mess up your X-10 or other gizmos on your AC line.

The only down-side, is that you basically have 8mS or so granularity on your on/off times. Not a problem for a nichrome wire at all.

I need to double-check this schematic against my notebook, but here is an early sketch.

Note the little "driver" chip. That one has built-in zero crossing detection, so you don't have to worry in your cpu code-- just turn it on or off for multiples of about 8mS, and it will do the heavy lifting for you.

Your cpu would connect to the left side of the chip. I've drawn a "ground" symbol, but what I mean, is your cpu/system ground. Inside the chip is a LED which is isolated from the rest of the driver chip, so all you need to do is adjust the resistor to get the correct current/drive into the internal LED.

bvwelch.com/roast/cpu_heatcontrol.jpg


Here is a pin-out of the triac I like for heat control:

bvwelch.com/roast/bta20.jpg


More later,

-bill
Edited by bvwelch on 08/29/2008 7:29 PM
 
Island Addict
Top  #23 Print Post
Posted on 08/30/2008 10:15 AM
Okay, I think I get it. I like the optoisolator and the low noise zero-crossing detection.

Does the triac need a heat sink?
Rob
 
bvwelch
Top  #24 Print Post
Posted on 08/30/2008 2:59 PM
Rob,

Yes, the triac needs a heat sink. Since it has an isolated tab, I fastened to the large metal bottom plate of the P1. In the photo (see other thread), you can't really tell it, but the triac is mounted beneath one of the circuit boards. The cool air from outside comes in all around that area, which helps.

-bill
 
seedlings
Top  #25 Print Post
Posted on 09/02/2008 12:39 PM
Just for a laugh, is there another easy, cheaper way to get the same results as -->THIS AUBER CONTROLLER<---?

As few as 4 segments would be just fine with me!

CHAD
Edited by seedlings on 09/02/2008 12:59 PM
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
 
bvwelch
Top  #26 Print Post
Posted on 09/03/2008 9:47 PM

Quote

seedlings wrote:
Just for a laugh, is there another easy, cheaper way to get the same results as -->THIS AUBER CONTROLLER<---?

As few as 4 segments would be just fine with me!

CHAD


Honestly, that unit may be worth the asking price.

We can do the PID function very cheaply-- I've written here about it before. But there is more to it than that-- for example, we need an interface to the TC. A likely choice, that was mentioned previously, the MAX6675, may cost as much as $12 to $14 in single unit quantities.

Another thing to consider would be some sort of front-panel LCD and some buttons, etc.

I'd prefer my own PID, simply because I enjoy programming.

Stay tuned to the other thread for news. :-)

-bill
 
scotthal
Top  #27 Print Post
Posted on 09/04/2008 10:21 AM
I've been using an Atmel Butterfly evaluation kit (<$20) for prototyping a NTC thermistor based data logger; unit has a built in LCD readout and simple joystick. Board has headers for the PWM outputs, & an SPI compatible interface.
Edited by scotthal on 09/04/2008 10:23 AM
Food for thought; coffee for concentration
 
bvwelch
Top  #28 Print Post
Posted on 09/04/2008 10:57 PM

Quote

scotthal wrote:
I've been using an Atmel Butterfly evaluation kit (<$20) for prototyping a NTC thermistor based data logger; unit has a built in LCD readout and simple joystick. Board has headers for the PWM outputs, & an SPI compatible interface.


Very cool! Let us know how it turns out.

I tinkered with some thermistors from US Sensor that are rated for high enough temperature, but I lacked the patience to work out the math and/or characterize their reading versus a K-type TC. However, if we could do that, it sure would make the expense of getting temp readings very very cheap!


-bill
 
Page 1 of 2: 12
Jump to Forum: