Delta BT Flick

[nickthorpie]

Hey all,
As per Rao's recommendations, I am able to maintain a constantly decreasing delta BT, up until the seconds leading up to 1C. As you have all seen, the curve pretty much does everything you don't want it to; It drops heavily, and then increases just as quickly as it started.

After some quick reading, it appears that what appears on the screen might be a bit misleading.. The general consensus is that the steam escaping from the beans cools the probe, and when the steam dissipates, the temperature reading goes back up(flicks) to a more realistic level.

What do I do at that first crack time? Do I crank up the power, then shut it off to produce that smooth ROR, or do I just let the end of the roast just do its thing? Has anyone Cupped the difference? I feel like I'm doing everything right and then butchering it at 1C

[folkery]

I wouldnt increase the heat to improve the flick. Scott Rao mentions not to do that. In his blog he talks about it often. Increasing the heat only produces more unwanted toasty flavors. He suggest finding a way to get an ever declining RoR by other means. I'm thinking he means airflow. Airflow can increase convective heat transfer to the beans. Just a little may fix the flick. Also try to anticipate the flick, maybe start more airflow just before FC begins.

[folkery]

Also from you graph it does look like the flick starts before FC. What steps did you perform before that? I'm curious.

[folkery]

I also have a hunch that your roast is progressing too fast as I can you have 5 kpa mid roast. Also from your BT line it is so steep it doesn't bend forward tapering off at the end of the roast.

[Gregr]

Scott Rao doesn't use a Huky    If he did then he would know that we can mitigate that big dip without imparting any roast notes at all. It is tricky though and takes a little practice but it absolutely can be done so stretching the final phase out a bit is possible if that is something you want to do. I suspect much of Rao's comments are aimed at roasters that don't have the agility of the Huky.
If you're a fan of the steadily declining bean temp delta ignore the rest of the roast and note that once first crack starts I raise the heat from 1kPa to 1.5 or so for about 30 seconds, then lower it back to 1.0 til first crack is over, then .5kPa for maybe 30 more seconds, then the heat goes off for the rest of the stretch, if any. The fan remains at 25% until first crack is done, then 75% until the heat goes off- then full fan. The numbers vary a little bit depending on what the beans are doing but the overall technique works very well for keeping enough momentum going without any roast notes appearing. That roast was for espresso, by the way, and the shots are very rich and creamy smooth.

[nickthorpie]

Also from you graph it does look like the flick starts before FC. What steps did you perform before that? I'm curious.

So I didnt track power input in these roasts, but I typically am around 4-5 kPa up until the 6 min mark where it tends to stall, and I drop it to 4kPa. Fan is around 50% for these, and I'm doing 1Lb roasts, which might be part of the reason I'm using so much heat. (Id like to do less but I sell it at my pop-up coffee shop that I'm using to fuel the money side of learnig to roast)

I also have a hunch that your roast is progressing too fast as I can you have 5 kpa mid roast. Also from your BT line it is so steep it doesn't bend forward tapering off at the end of the roast.

Yea so I definitely am still learning a lot and need as much guidance as I can get. Its hard to find claims that are backed up by some sort of proof.
With that in mind I want to ask:
Why should it progress slower? Shouldnt I be hitting 1C around the 8:30-9:00 mark?

In what ways does the steepness of BT curve affect the crazy stuff happening at the end?
       -does going into 1c with a lower dBT lower the dip and flick?

[nickthorpie]

.
If you're a fan of the steadily declining bean temp delta ignore the rest of the roast and note that once first crack starts I raise the...

First, let me say that your graphs over on artisan alliance have been extraordinarily helpful.

Your first line exactly describes my problem: I dont know if I'm a fan of that steady dBT. I'm trying to get to the point where I can repeat a roast profile back to back. Once I'm there I'll let the cupping table tell me what I'm a fan of.

[Gregr]

Thanks!
I know just what you mean but I think you'll be able to control your roaster with good predictability fairly fast and then move on to the fun stuff (cupping). Having Artisan to refer to is invaluable and then there's the fact that the Huky is very consistent.
As far as momentum into first crack goes- my approach at that point will be different than the Rao style but what I see [rough numbers] going into first crack is the BT delta at around 18F/min and, using that little dance I described earlier, roughly 12F/min coming out. That's enough momentum leftover for a nice long stretch if I want.
There's a guide to roasting in the Rao style with a perforated drum in the Resources section. I took a quick look at it and his heat settings were significantly lower than yours so maybe you're using too much heat and then tapering off too much or too quickly? I dunno- hard for me to say since I've had a solid drum since day one.

[folkery]

From what I can see your roasting really well. What counts is how it tastes when you brew it. Maybe the coffee prefers the flick.

On another note a steep BT profile is a FAST roast; all this heat is building up pressure within the bean and then releasing it at 1C. The first half of a roast is endothermic (absorbing heat) and the last half is exothermic (generating heat). The second half of the roast begins around 320-340 degrees Fahrenheit. I think from your roast your building up so much pressure, 1C is cooling it down, flicking the ROR downwards.

I do 1lb roasts. I use 2.5 kpa around 340F. lm reaching first crack around 8-9 minutes. Before that at around 370F I bring my kpa dialed down to .75. Typically you don't want to blaze through 1C too fast. You want the bean to have absorbed enough heat in the beginning to produce it's own heat to bring it through the end of 1C. 

[folkery]

http://scottrao.com/blog/development-time-ratio/

Note that the first roast graph on this page has some similarities to yours. The ROR stretches out horizontal for a longer period of time, however doesn't decsend lower than a varying 10-20 degrees.

Further on down the page are two successful roasts with a gradually declining RoR that descend below 5 degrees.

These are examples of using less heat mid roast bowing down the BT curve for a descending RoR.

[slickrock]

...
After some quick reading, it appears that what appears on the screen might be a bit misleading.. The general consensus is that the steam escaping from the beans cools the probe, and when the steam dissipates, the temperature reading goes back up(flicks) to a more realistic level.

What do I do at that first crack time? Do I crank up the power, then shut it off to produce that smooth ROR, or do I just let the end of the roast just do its thing? Has anyone Cupped the difference? I feel like I'm doing everything right and then butchering it at 1C

This flick mythology stuff has been something that has been bugging me for a long time, because most of the time its basically a mirage. Just like you implied, it not as much a "flick", but rather a temperature "dip" that shows up in the instrumentation falsely due to off-steam cooling.  What's important is that your roast should return to the temperature level at the end of the roast cycle as if dip never occurred in the first place (which is most likely the case if you don't try to compensate for it). If you hit that level then your roast should be fine.

[hankua]

I don't use a perforated drum with the Huky so can't comment on your gas/air settings; however I recently tried Greg's 1C boost with my other 500g machine and there were no adverse defects on the beans. It's worth a try especially for a full city or beyond roast.

Rao's market is artisan small batch roasters; the Huky does not fit into that category. Unfortunately or Fortunately we in Hukydom have to plow our own fields.

One thing I noticed was one roast went into 2C and the other finished Full City just before the onset of 2C it appeared. Several things I think are important are: the total roast time, RD time between 1C-Drop, finished bean temperature, and weight loss between green and roasted.

Couple of suggestions if your not doing already: Erase "Roaster Scope" and insert the green bean origin and relevant data. In the box marked "beans" you can add more relevant information that will show up on the bottom info line. Lastly fill in the boxes with charge weight and roasted weight and don't worry about some stuck beans getting lost. If it's consistent between batches (lost beans), good enough for comparisons. That will also show up on the bottom line.

If you really want to get geeky, start a spread sheet and track the major metrics during the roast, some of which I mention above.

[SusanJoM]

If you really want to get geeky, start a spread sheet and track the major metrics during the roast, some of which I mention above.

In fact Artisan will do it for you.
File > Report > Ranking
will let you select any of your .alog files for inclusion.
It's pretty cool....