2nd Roast - feedback appreciated

[brew]

Thanks for any feedback.

[SusanJoM]

how many grams?

and how did you decide on that Charge temperature?

Susan

[brew]

350grams.  I actually charged at 410.  I forgot to click the button.

[rodraguirre]

It would be interesting if you also include your events information (gas & fan), even if you didn't capture it in artisan. Did you take notes of it?

It will be also very good if you enable the RoR (it's called DeltaBT in artisan). You can do it even after your roast is finished and save your graph again with the new curve.

http://coffeetroupe.com/artisandocs/configextras-sensitivity-and-filter/

This will provide valuable information for you and for others. When you do so, I recomend you to configure the max setting for the DeltaBT axis at 100, so the curves don't stay on top of each other.

[brew]

I've added the RoR like you suggested.


I was attempting to follow a roast plan that SusanJoM had posted except I am not using a variac and am guessing at my fan speeds.  I know min and max speeds but for now that is all I do know.  The following fan rates are guess on my part to mimic the speed suggested:


fan 13
2.5kPa at Charge
fan 26 for 30 seconds at 300
365 gas down to .5 kPa
390 fan 26

[SusanJoM]

fan 13
2.5kPa at Charge
fan 26 for 30 seconds at 300
365 gas down to .5 kPa
390 fan 26

As soon as I charge, though I crank the gas to 3.75 kPa.  In fact as of a day or so ago, I do that just before I charge.
And I am charging much hotter than you are.

[rodraguirre]

ok, now seeing your RoR curve: See that enormous flick in the end? That's unrealistic. What happens is that you smoothed your RoR curve so much that it ends up considering your post-drop BT readings to generate the average RoR.

so, basically, everything you see after ~7:05 in your RoR line is really a misrepresentation. Try to be less aggressive in your smoothing settings. Although the line will be a bit uglier, you'll be able to notice a much better trend by looking at it. and get better insights out of it.

Other important question is: how do the beans look? What about the aroma? did you like it?

And definitely one of the most important factors: did you measure your weight loss?

[brew]

Awesome...thanks guys.  You've given me a lot of info to consider.


rodraguirre - i am comparing this output to what I was doing on a Behmor.  I know that some people are putting out exceptional coffee with them but I never really could and got lucky every once in awhile.  The bean tastes very good and I believe it will be acceptable knowing that I can do a better job.


Finish weight is 307g

[SusanJoM]

ok, now seeing your RoR curve: See that enormous flick in the end? That's unrealistic. What happens is that you smoothed your RoR curve so much that it ends up considering your post-drop BT readings to generate the average RoR.

so, basically, everything you see after ~7:05 in your RoR line is really a misrepresentation.

Are you sure?  What IS his smoothing factor?  I've never seen one that would account for that steady and absolutely real increase in his dBT.

And the rise in his dBT is completely consistent with the speed of the second two thirds of his roast.  Edit: Okay, maybe I have this  sentence wrong.  I looked again after your post a ways down the page.

So, Brew, what is your Sampling Interval, and what are your Smoothing Settings.

At the bottom you will see the results of a profile using a 3 second sampling interval and a smoothing factor of 10.

[rodraguirre]

Yes Susan, I'm completely sure.

Mathematically, the RoR is a derivative of the BT. Look at the OP's BT (not RoR) behaviour after 7 min and before FC. See how the line starts deceleration?

That is completely contradictory with the enormous flick we see in RoR within the same time frame.

Anyway, I took the time to subscribe to the Artisan users mail list and wrote about this (thank you for pointing me that!). I hope Marko will step in and either give his reasons or come up with a solution.

[rodraguirre]

Adding to that, the only reason why we have weird looking RoR curves if we don't use any smoothing setting is largely based on the fact that the BT readings are not continuous. If we had a way of getting continuous, uninterrupted temperature readings (sampling interval tending to zero), then the RoR would be way closer to reality without any need for smoothing.

The only way of getting a better representation for RoR in our case, with 3+ sampling intervals, is by getting an average of the surrounding points for each specific point in time. The larger the surrounding area, the smoother the line, but the downside is obviously that when you get closer to the drop, your RoR will include post-drop BT readings to calculate that average.

[rodraguirre]

I love it 

And I'll prove it. Already think in a visual way of doing so. Will execute it once I get home

[SusanJoM]

Please don't do it on my behalf.
I don't need convincing.
I thought you had something wrong and I didn't like the idea that it would go unchallenged.

You may have it all just right, so....carry on.
The subject matter has moved out of range of my interest....

[brew]

SusanJoM - the sampling interval is 3s.  This is after I changed the settings to match the settings you posted.

[SusanJoM]

Well, he's got that right.  That is very aggressive smoothing.
I'd probably better go and cook up a big pot of crow .....