Who is here? 1 guest(s)
 Print Thread
4-channel TC meter and datalogger project
JimG
Just an FYI: building on Rama's work with performing RoR averaging in pBourbon.pde, I made a few very minor tweaks that make it simple to turn this feature on or off.

A value of SAMPLESIZE = 1 will turn this feature off, while a value of SAMPLESIZE = 20 will maintain it.

Also, if the line in the config file containing the averaging information is missing, the program will now default to SAMPLESIZE = 1. This will maintain compatibility with older config files.

I like being able to alter this behavior using only simple text file edits to the config file.

Jim
 
JimG
I captured some data from a roast this morning.

Using Excel and the captured CSV file, I computed an "exact" RoR profile. I did it by examining the actual BT at points 5 seconds prior, and 5 seconds after the time of interest, then dividing the difference by 10 seconds. This gives a fairly smooth RoR trace without lag.

On the same graph, I plotted the realtime RoR values that came directly from the aBourbon application using a filtering level of 85.

The two plots (blue = realtime from aBourbon; red = post-processed exact value) are shown below.

Although you cannot see this well in the plot, looking closely at the results tells me that the lag time is around 5 seconds for a filtering level of 85. This is acceptable to me. With my roaster (Hottop) and BT probe, this level of filtering also gave a "smooth enough" trace on my PC during the roast.

Conclusion: I'm going to stick with RISE_FILTER = 85 for my setup, and live with the 5 second lag.

Jim
JimG attached the following image:
roast201010240958ror.jpg
 
farmroast

Quote

JimG wrote:
I captured some data from a roast this morning.....

Conclusion: I'm going to stick with RISE_FILTER = 85 for my setup, and live with the 5 second lag.

Jim
That's plenty good.
what did you roast? I'm curious about you readings. I'm assuming the bottom readings are in seconds? At about what time did you hit 300f BT and start of first?
Ed B.
DreamRoast 1kg roaster, Levers, Hand Mills http://coffee-roa...gspot.com/
 
JimG
This was a test roast (Indian Mysores Nuggets). I wanted to log the built-in profile on the Hottop D, I wanted to test the RoR filtering, and I wanted to finish up the Mysores since it was getting old. So I did a test roast to kill three birds with one stone.

To answer your questions (also see profile graph below), it hit 300F at around 9 minutes, it hit 1C (380F) a little past 15 minutes. I ejected at a BT of 430F (around 16:30).

This is much different than the profiles I normally use, BTW.

Jim
JimG attached the following image:
roast201010240958btet.jpg
 
farmroast

Quote

JimG wrote:
To answer your questions (also see profile graph below), it hit 300F at around 9 minutes, it hit 1C (380F) a little past 15 minutes. I ejected at a BT of 430F (around 16:30).

This is much different than the profiles I normally use, BTW.

Jim

The RoR graph looks really good!!!!! I was guessing it was not a typical profile.
BTW Happy anniversary! It was Oct.12,09 when I first posted the RoR meter query.
Ed B.
DreamRoast 1kg roaster, Levers, Hand Mills http://coffee-roa...gspot.com/
 
rama

Quote

bvwelch wrote:
Thanks for the update Rama. That sure looks better to me. Did the change in the Arduino filter have any adverse effect to your 'average RoR' ?


Sorry for delay in response...
I didn't notice any adverse affect to the reported 'average RoR', although now its an average of an average and I suspect even more prone to lag.

Given the Hottop is not exactly quick to change direction, this seems right inline with its behavior. ;)
 
rama

Quote

JimG wrote:
Just an FYI: building on Rama's work with performing RoR averaging in pBourbon.pde, I made a few very minor tweaks that make it simple to turn this feature on or off.

A value of SAMPLESIZE = 1 will turn this feature off, while a value of SAMPLESIZE = 20 will maintain it.

Also, if the line in the config file containing the averaging information is missing, the program will now default to SAMPLESIZE = 1. This will maintain compatibility with older config files.

I like being able to alter this behavior using only simple text file edits to the config file.

Jim


Clever, thanks for the mods. I didn't think of the backwards compatibility problem.

BTW, the Excel spreadsheet was also very helpful. Seems a rise filter of 85% is the magic balancing point!
 
JimG
Patrick (patf11) has joined the software development effort in a big way! I have only had time for a quick glance at his new code, but it looks promising.

If you want to browse the source code on the project googlecode site, you will see new branches for aCatuai_RC1, pBourbon_RC1, and a new library to support the MCP23008 port expander.
http://code.googl...owse/trunk

The "RC1" suffix stands for "release candidate 1", BTW.

Thanks, Patrick. I am looking forward to trying out your new code.

Jim
 
JimG
I've made a few TC4 version 3 boards with all of the surface mount chips pre-soldered. This includes the MCP3424 ADC chip, the MCP9800 temperature sensor, the 24LC512 EEPROM chip, plus the surface mount capacitors, resistors, and diode.

If you want a partially assembled board like this (at cost), send me a PM.

(Note: you would still have to buy and solder the through-hole components).

Jim
 
gmohr
Just ordered a couple of the boards and then I see this post...

I just want to take the time to say "Thanks" for all the hard work you
all have put into this project. Not sure when I will get mine built since
I just bought a new B model Hottop and need to learn to use it. My
trusty Version 1 Gene Cafe finally gave up the ghost. but I may find a
use for one of the board as a PID using the old Gene Cafe as a project...

Gene
 
stoneguard
Arg. I have the unassembled board and definitely would have paid for the partially assembled one, as I am muddling along at the moment trying to figure out how to proceed Grin. Circuit boards are not an area I have touched on before (some anxiety issues keep me from just jumping into things at times). But I do agree, thanks for all the hard work you guys have done. I will hopefully have a completed board before the end of the year.
 
JimG
Gene -

Your boards haven't been mailed. If you want them partially populated, I can take care of this before mailing. The cost of the SMD's is close to $12 per board (MCP3424 $5, MCP9800 $2, 24LC512 $2, misc. caps and resistors approx. $3). If you want the SMD's, let me know and we'll work out payment later.

Phil -

Send the boards back and I'll add the SMD's for you if you'd like. Or I can send out 2 more and you can return the bare ones when you have a chance.

Jim
 
gmohr
JimG,

Message sent. What do you think of using the TC meter/datalogger as
a PID and controlling a Gene Cafe? What I am thinking of is controlling the fan and heater with this board and let the original board take care of the time and the motor...


Gene
 
JimG
Gene -

That is exactly the type of application contemplated for this project.

A great place to start would be to read about Randy's "Kona" application:
http://forum.home...post_25021

I'll also let everyone know that I've placed an order for ~15 LCD adapter PCB's (which I've not-so-cleverly named LCDapter). This new little gizmo will connect to the TC4 using a 4-conductor telephone cord and provides a 4-button keypad, a pin header to solder-connect a 16 x 2 LCD with backlight, and 3 LED's that can be used as indicators.

The bare LCDapter boards should be available in around 2 weeks and will probably cost $7 each.

Jim
 
gmohr
Jim,

That sounds awesome! They will be perfect for the project with the Gene Cafe!

Gene
 
randytsuch

Quote

gmohr wrote:
JimG,

Message sent. What do you think of using the TC meter/datalogger as
a PID and controlling a Gene Cafe? What I am thinking of is controlling the fan and heater with this board and let the original board take care of the time and the motor...


Gene


Do you know what voltages the fan and motor run on?

I have an older alpenrost I am using now, I control heater and fan on 115V, and the drum motor and another motor which open/closes the flaps for cooling, on 12VDC.

The 12VDC motors are pretty easy to control, although it takes a few extra parts to do it.


Randy
 
gmohr
Randy,

From what I can see so far the heater is 115v, the fan and motor are 12vDC.
 
JimG
FWIW, I/O3 on the TC4 board is intended for use with DC motors/fans. Its PWM frequency is 490Hz.

The TC4 board cannot provide the drive current for the motor or fan, but can be connected through a resistor to the base of a TIP120 transistor that can probably switch the full current. Of course, there are other options.

Jim
 
gmohr
Okay, I will bite.... What kind of options? My idea revolves around inserting the tc board somewhere in the control path of the roaster. I first have to get a good idea of what currently controls what on the Gene Cafe control board. Iam all ears for options.

Gene
 
SteveN

Quote

JimG wrote:

I'll also let everyone know that I've placed an order for ~15 LCD adapter PCB's (which I've not-so-cleverly named LCDapter). This new little gizmo will connect to the TC4 using a 4-conductor telephone cord and provides a 4-button keypad, a pin header to solder-connect a 16 x 2 LCD with backlight, and 3 LED's that can be used as indicators.

The bare LCDapter boards should be available in around 2 weeks and will probably cost $7 each.

Jim


Will a 20x4 LCD will fit ok and still leave room for the buttons?
 
JimG

Quote

gmohr wrote:
Okay, I will bite.... What kind of options?

I had in mind that other transistors, besides the TIP120, might also be used.

Jim
 
JimG

Quote

SteveN wrote:
Will a 20x4 LCD will fit ok and still leave room for the buttons?

Sorry, no. Electrically, using a 20 x 4 works fine. But the buttons and LED's would be covered up.

Jim
 
gmohr

Quote

JimG wrote:

Quote

gmohr wrote:
Okay, I will bite.... What kind of options?

I had in mind that other transistors, besides the TIP120, might also be used.

Jim


Ahhhh I think I see where you are going with this... I also have some
TIP32s and a couple of others in my Ham Radio parts.... time to break out the design books and see what I can create.... I may have to rebuilt the
control board for the Gene from scratch and possibly make a few mods to
the case... This is going to be fun!

Gene
Official Office Coffee Snob
Gene Cafe, Hottop B, Mazer Mini, Expobar Brewtus III-R.... Oh.. pilot, sailor and photographer.
 
stoneguard
Hey Jim,

I only actually bought one board. I think I have all the parts in, I just need to attempt to actually put it all together. If you want to PM me with any advice, I would be appreciative. I am not exactly sure when I will get to it, but hoping to have some time today or tomorrow to at least look at it all again.
 
JimG
For soldering the SMD's with an iron, there are a lot of videos and how-to pages online (use da Google).

My personal list of top 4 helpful things:
1. Put plenty of flux on the pads first
2. "Push" the solder into the pin/pad interface, starting at the edge of the pad.
3. Clean the iron tip (I use one of those brass brillo pads) after soldering each pin
4. Use fine solder wick to clean up solder bridges

The MCP9800 is a little tricky because it is so small. Do this one last.

Jim
 
Jump to Forum: