Optimal airflow for North 1K
-
EddyQ
- Posts: 1051
- Joined: 8 years ago
I've owned my used North TJ-67(1K) for a little over two months. I've managed to make some pretty good coffee with gas and airflow settings I came up with based on other HBer posted profiles, previous owner profile and some empty roaster testing. It is time to determine for myself where the optimal airflow should be (if one exists).
With all the recent thermal discussions and Rob Hoos article in the recent Roast Magazine, I felt is time to run some more testing with hopes to find or verify the optimal airflow and its effect it has on the roast.
https://www.roastmagazine.com/resources ... achine.pdf
This is a really great article and worth reading a few times. However, Figure 3 was interesting and confusing. It shows there is an optimal point for airflow where too little or too high would hinder heat transfer. But dotted lines of "perceived change" I simply did not understand. So, off I go to experience a bit of this.
My simple experiment is to find the effects of changing only airflow quite drastically on two otherwise identical roasts. Would RoR change significantly? Would one be flat and lacking flavor? Would one taste more roasty due to roast defects such as scorching? Time to find out.
My North 1K is a older model that does not have the stable VFD type fan as the newer models. So I added a Magehelic gauge to aid controlling airflow. The gauge taps into the bean chute at the analog temp fitting at the air exit point of the drum. The gauge range is 0 to 0.25" water column and with max fan will just hit the 0.25"wc. With all the previous roasts that I have done, I have determined that I would not likely ever need airflow higher than 0.1"wc on gauge. In fact, a little above 0.1"wc beans begin to get sucked into bean chute. So, I set up artisan where 100% = 0.1"wc. Zero results in 0.01-0.02"wc or 20%.
I roast with propane which has a lot of BTUs. Gas gauge max is roughly 5-5.5Kpa and so I log artisan power with 5.5Kpa = 55% and so forth for lower settings.
My drum speed is fixed at 70rpm. While this may not be optimal, I do not plan of changing it for this experiment.
Each batch was 1 pound of Honduras beans (with a decent moisture content of 11%). I charged with a stable temp of 320F. At charge, gas was minimal and airflow was minimal. 2min into roast, I increased airflow to 0.4"wc for slow airflow roast and 0.8"wc for fast airflow roast. With this airflow update, I adjusted gas accordingly with a max MET target of 500F. My normal airflow to date has been somewhere in the middle of these two extremes in the are of roast where heat transfer is most needed. Throughout the roast I left airflow constant and adjusted only gas. I tried my best to keep the two roast MET curves the same so that the beans experience only airflow, not air temp, differences. Now, I am expecting differences going on with drum thermal path to beans, but I have no idea how to be certain what is happening there. What I expected to see is differences in RoR with these. A third roast in the middle may be needed to be certain, but for now I thought I would fester with results before throwing in another pound of greens.
The slow airflow roast progresses better than I expected. I really thought I would ruin the coffee with this test. There is a very noticeable crash. I dropped at a rather high BT and should have dropped just after FCs. But the moisture loss turned out to be only 15% and no oil (yet). I hope the combination of these faults don't hinder the taste and fog the experiment results.
There was a few challenges with the fast airflow roast. First was keeping the MET curve matching the first roast. At the 2min mark, where I changed gas/air, it was difficult knowing exactly where to set gas. But I think the roast came out reasonably close. Second challenge was holding the 500F MET. I think I know what Rob Hoos's term "Perceived Change" means. It means tiny changes of gas make a bigger difference when airflow is higher. My MET was noticeably less stable with fast airflow roast. I ended the roast at nearly exactly the same BT as the slow roast, but did manage to reduce the flick a bit better.
So the gas settings on the slow roast peaked at 2.8Kpa while the fast roast peaked at 4.5Kpa. That is a rather huge change of BTU input, yet when mixed with cool air such that temps are similar, it seems the RoR are similar. Taste is king.
Tomorrow I plan to cup and pull a few espressos with both roasts. I will repeat in 5 days. I'm pretty excited and hope to report back some interesting news. Stay tune ...
With all the recent thermal discussions and Rob Hoos article in the recent Roast Magazine, I felt is time to run some more testing with hopes to find or verify the optimal airflow and its effect it has on the roast.
https://www.roastmagazine.com/resources ... achine.pdf
This is a really great article and worth reading a few times. However, Figure 3 was interesting and confusing. It shows there is an optimal point for airflow where too little or too high would hinder heat transfer. But dotted lines of "perceived change" I simply did not understand. So, off I go to experience a bit of this.
My simple experiment is to find the effects of changing only airflow quite drastically on two otherwise identical roasts. Would RoR change significantly? Would one be flat and lacking flavor? Would one taste more roasty due to roast defects such as scorching? Time to find out.
My North 1K is a older model that does not have the stable VFD type fan as the newer models. So I added a Magehelic gauge to aid controlling airflow. The gauge taps into the bean chute at the analog temp fitting at the air exit point of the drum. The gauge range is 0 to 0.25" water column and with max fan will just hit the 0.25"wc. With all the previous roasts that I have done, I have determined that I would not likely ever need airflow higher than 0.1"wc on gauge. In fact, a little above 0.1"wc beans begin to get sucked into bean chute. So, I set up artisan where 100% = 0.1"wc. Zero results in 0.01-0.02"wc or 20%.
I roast with propane which has a lot of BTUs. Gas gauge max is roughly 5-5.5Kpa and so I log artisan power with 5.5Kpa = 55% and so forth for lower settings.
My drum speed is fixed at 70rpm. While this may not be optimal, I do not plan of changing it for this experiment.
Each batch was 1 pound of Honduras beans (with a decent moisture content of 11%). I charged with a stable temp of 320F. At charge, gas was minimal and airflow was minimal. 2min into roast, I increased airflow to 0.4"wc for slow airflow roast and 0.8"wc for fast airflow roast. With this airflow update, I adjusted gas accordingly with a max MET target of 500F. My normal airflow to date has been somewhere in the middle of these two extremes in the are of roast where heat transfer is most needed. Throughout the roast I left airflow constant and adjusted only gas. I tried my best to keep the two roast MET curves the same so that the beans experience only airflow, not air temp, differences. Now, I am expecting differences going on with drum thermal path to beans, but I have no idea how to be certain what is happening there. What I expected to see is differences in RoR with these. A third roast in the middle may be needed to be certain, but for now I thought I would fester with results before throwing in another pound of greens.
The slow airflow roast progresses better than I expected. I really thought I would ruin the coffee with this test. There is a very noticeable crash. I dropped at a rather high BT and should have dropped just after FCs. But the moisture loss turned out to be only 15% and no oil (yet). I hope the combination of these faults don't hinder the taste and fog the experiment results.
There was a few challenges with the fast airflow roast. First was keeping the MET curve matching the first roast. At the 2min mark, where I changed gas/air, it was difficult knowing exactly where to set gas. But I think the roast came out reasonably close. Second challenge was holding the 500F MET. I think I know what Rob Hoos's term "Perceived Change" means. It means tiny changes of gas make a bigger difference when airflow is higher. My MET was noticeably less stable with fast airflow roast. I ended the roast at nearly exactly the same BT as the slow roast, but did manage to reduce the flick a bit better.
So the gas settings on the slow roast peaked at 2.8Kpa while the fast roast peaked at 4.5Kpa. That is a rather huge change of BTU input, yet when mixed with cool air such that temps are similar, it seems the RoR are similar. Taste is king.
Tomorrow I plan to cup and pull a few espressos with both roasts. I will repeat in 5 days. I'm pretty excited and hope to report back some interesting news. Stay tune ...
LMWDP #671
-
chris_n
- Posts: 389
- Joined: 11 years ago
thanks for the insightful post. very interested in following this progress! would love to see more pics and details of your magnahelic gauge mod!
-
drgary
- Team HB
- Posts: 14394
- Joined: 14 years ago
Hi Ed,
You participated in my other thread where I wrote about how constant high airflow stripped all of the flavor from my coffee -- and for other readers, we have different roasters because mine has a perforated drum. Batches after that thread with low airflow were much better.
Constant high airflow doesn't seem to work with my roaster
I wonder whether keeping airflow fairly constant and relatively low will work with your roaster, while adjusting gas. I took a tip from keno and tried yesterday to approach 1C with some momentum, reduce power on the way to 1C to avoid the crash and reduce it some way into 1C to avoid the flick. That coffee is smelling very good today. I kept airflow constant and at the lowest setting to keep air moving. I'll be watching this thread to see whether there's an airflow speed where flavor is affected. Here's yesterday's roast. Fan and gas were actually off at the start, and I did a long soak, starting gas at 1:45 and starting air near end of dry, although I usually start the fan a bit earlier, when the beans hit 270°F. I'm still working on the device configuration to get the readout less jumpy.
You participated in my other thread where I wrote about how constant high airflow stripped all of the flavor from my coffee -- and for other readers, we have different roasters because mine has a perforated drum. Batches after that thread with low airflow were much better.
Constant high airflow doesn't seem to work with my roaster
I wonder whether keeping airflow fairly constant and relatively low will work with your roaster, while adjusting gas. I took a tip from keno and tried yesterday to approach 1C with some momentum, reduce power on the way to 1C to avoid the crash and reduce it some way into 1C to avoid the flick. That coffee is smelling very good today. I kept airflow constant and at the lowest setting to keep air moving. I'll be watching this thread to see whether there's an airflow speed where flavor is affected. Here's yesterday's roast. Fan and gas were actually off at the start, and I did a long soak, starting gas at 1:45 and starting air near end of dry, although I usually start the fan a bit earlier, when the beans hit 270°F. I'm still working on the device configuration to get the readout less jumpy.
Gary
LMWDP#308
What I WOULD do for a good cup of coffee!
LMWDP#308
What I WOULD do for a good cup of coffee!
-
EddyQ (original poster)
- Posts: 1051
- Joined: 8 years ago
Hi Gary,
Yes, I was thinking of that thread but decided to start a new one since my roaster has a solid drum. This North is quite different than my Quest with regards to airflow. Airflow can easily be too much which never happened with the Quest. Since it is still new to me, I feel I need to quickly get a good understanding of airflow and impacts now.
I'm coming to the conclusion that Rob Hoos's figure 3 showing a peak of heat transfer wasn't quite the same conditions as mine. With my Quest, I could detect less heat transfer when airflow is very low. And you can see at 4:30min into your roast when the higher airflow carried in more heat. But the drop of thermal transfer at high airflow, I did not get with my roast. Maybe I needed to max my gas and control MET with air to find the true maximum? But I really think what Rob has plotted is a drop in heat due to lack of BTUs to heat the cool incoming air. That is a different problem than what I have experimented with.
Ok, so I really believe I needed at least one more roast with airflow between the two extremes to draw a more concrete conclusion. And definitely less development would have helped as well. I pulled two side-by-side espressos on Monday morning (1 day after roast) and again this morning. This mornings espressos were rather similar to Monday, but a bit more mild and less astringent. Both have a rather strong bold astringency that lingers for a long while. Aroma of both was more roastiness than anything floral from the beans origin. However, there was some nice sweetness that came through with both. Unlike yours Gary, my fast airflow roast had noticeably more molasses sweetness that blended nicely with the bold chocolate aftertaste. The slow roast was more flat, earthy, tobacco that blended into a less interesting bold chocolate aftertaste. Now, I am thinking the differences would have been greater had I dropped earlier in the development.
Which brings up a somewhat separate concern. My drop temperature was a whopping 438F and I did not detect 2C. Furthermore, only a few tiny spots of oil showed up on the slow airflow roast and none on fast airflow roast. Had this been my Quest, the beans would be charcoal. So, I am concerned my BT probe isn't in the greatest location or for some reason getting impacted by high METs or something else. The airflow didn't seem to matter. Any thoughts?
So, I feel I have learned a bit with this experiment. But I feel I really need to do another run. I may have enough beans for 3 roasts, but if not I may try staggering the airflow to the higher end. Bring the low airflow case a bit higher and high airflow case to max (max gas). Also run them to lower development (drop just after FC). Any thoughts?
Yes, I was thinking of that thread but decided to start a new one since my roaster has a solid drum. This North is quite different than my Quest with regards to airflow. Airflow can easily be too much which never happened with the Quest. Since it is still new to me, I feel I need to quickly get a good understanding of airflow and impacts now.
I'm coming to the conclusion that Rob Hoos's figure 3 showing a peak of heat transfer wasn't quite the same conditions as mine. With my Quest, I could detect less heat transfer when airflow is very low. And you can see at 4:30min into your roast when the higher airflow carried in more heat. But the drop of thermal transfer at high airflow, I did not get with my roast. Maybe I needed to max my gas and control MET with air to find the true maximum? But I really think what Rob has plotted is a drop in heat due to lack of BTUs to heat the cool incoming air. That is a different problem than what I have experimented with.
Ok, so I really believe I needed at least one more roast with airflow between the two extremes to draw a more concrete conclusion. And definitely less development would have helped as well. I pulled two side-by-side espressos on Monday morning (1 day after roast) and again this morning. This mornings espressos were rather similar to Monday, but a bit more mild and less astringent. Both have a rather strong bold astringency that lingers for a long while. Aroma of both was more roastiness than anything floral from the beans origin. However, there was some nice sweetness that came through with both. Unlike yours Gary, my fast airflow roast had noticeably more molasses sweetness that blended nicely with the bold chocolate aftertaste. The slow roast was more flat, earthy, tobacco that blended into a less interesting bold chocolate aftertaste. Now, I am thinking the differences would have been greater had I dropped earlier in the development.
Which brings up a somewhat separate concern. My drop temperature was a whopping 438F and I did not detect 2C. Furthermore, only a few tiny spots of oil showed up on the slow airflow roast and none on fast airflow roast. Had this been my Quest, the beans would be charcoal. So, I am concerned my BT probe isn't in the greatest location or for some reason getting impacted by high METs or something else. The airflow didn't seem to matter. Any thoughts?
So, I feel I have learned a bit with this experiment. But I feel I really need to do another run. I may have enough beans for 3 roasts, but if not I may try staggering the airflow to the higher end. Bring the low airflow case a bit higher and high airflow case to max (max gas). Also run them to lower development (drop just after FC). Any thoughts?
LMWDP #671
-
EddyQ (original poster)
- Posts: 1051
- Joined: 8 years ago
Chris, my magnehelic mod isn't that complicated. I removed the analog temp gauge and installed pipe adapter to 1/4" hose several feet down to the gauge which is vertically mounted near my gas gauge. I did install a in-line filter, since there is "stuff" in this exhaust gas that I did not want inside my sensitive magnehelic. So far, I love it.chris_n wrote:thanks for the insightful post. very interested in following this progress! would love to see more pics and details of your magnahelic gauge mod!
LMWDP #671
-
Tonefish
- Posts: 1401
- Joined: 7 years ago
Ed, the magnehelic is actually sucking air from around your roaster into the gauge and then out into your roaster exhaust, so if you want to protect your gauge you should put a filter in the air input of the magnehelic itself.EddyQ wrote:Chris, my magnehelic mod isn't that complicated. I removed the analog temp gauge and installed pipe adapter to 1/4" hose several feet down to the gauge which is vertically mounted near my gas gauge. I did install a in-line filter, since there is "stuff" in this exhaust gas that I did not want inside my sensitive magnehelic. So far, I love it.
LMWDP #581 .......... May your roasts, grinds, and pulls be the best!
-
EddyQ (original poster)
- Posts: 1051
- Joined: 8 years ago
Thanks Jim, I have a filter on that end as well! But air does not flow through the magnehelic due to a silicone diaphragm. So no outside air will make it into the roaster through the gauge. Only a tiny amount of air would flow in or out (either port) depending on if pressure difference is increasing or decreasing.
http://blog.dwyer-inst.com/2017/03/09/t ... eQ1Ce.dpbs
http://blog.dwyer-inst.com/2017/03/09/t ... eQ1Ce.dpbs
LMWDP #671
-
Tonefish
- Posts: 1401
- Joined: 7 years ago
Ah, yes, thanks for posting the link. I never looked into the design to see that while the roaster exhaust sucks air from the gauge and the gauge inlet has air going into the gauge from the environment around it, that the air doesn't actually flow all the way through the gauge, just into and out of it.EddyQ wrote:Thanks Jim, I have a filter on that end as well! But air does not flow through the magnehelic due to a silicone diaphragm. So no outside air will make it into the roaster through the gauge. Only a tiny amount of air would flow in or out (either port) depending on if pressure difference is increasing or decreasing.
http://blog.dwyer-inst.com/2017/03/09/t ... eQ1Ce.dpbs
You wouldn't want your clean air polluting your roaster exhaust. 
Good thing you have the filter on the important end though. Happy Roasts! ... and if you're not already loving that North, I'm sure you will sooner than later. 
LMWDP #581 .......... May your roasts, grinds, and pulls be the best!
-
drgary
- Team HB
- Posts: 14394
- Joined: 14 years ago
Ed,
Yes, I think it's a good idea to start your own thread for this because our roasters are different. For BT probe, mine is inserted in the exit door and bent so it goes very close to the front bottom of the drum and doesn't hit the fins. I'm interested to read that higher airflow didn't strip flavors from your roast, which shows how different our roasters are with a perforated versus solid drum.
It can be difficult to hear 2C in any case. My typical temperature for that is 430°F.
Like you I'm practicing a lot with a single bean to see what works best. Roasting is so finicky that I'm now experimenting with lower charge temps and longer dry for increased complexity, longer soak to get there, then more gas for the ramp, tapering that off about 20°F before 1C start, backing off again into 1C (still figuring out when to anticipate), and then seeing if there's enough momentum through 1C to ensure enough development without losing origin characteristics. Doing this with the smallest practical charge for our roasters is a challenge, and it will all differ with larger batches. This kind of variability keeps a hobby interesting!
The roast shown above cups well in the second day post-roast, although I think I need a faster ramp to preserve sugar and make it more juicy.
Yes, I think it's a good idea to start your own thread for this because our roasters are different. For BT probe, mine is inserted in the exit door and bent so it goes very close to the front bottom of the drum and doesn't hit the fins. I'm interested to read that higher airflow didn't strip flavors from your roast, which shows how different our roasters are with a perforated versus solid drum.
It can be difficult to hear 2C in any case. My typical temperature for that is 430°F.
Like you I'm practicing a lot with a single bean to see what works best. Roasting is so finicky that I'm now experimenting with lower charge temps and longer dry for increased complexity, longer soak to get there, then more gas for the ramp, tapering that off about 20°F before 1C start, backing off again into 1C (still figuring out when to anticipate), and then seeing if there's enough momentum through 1C to ensure enough development without losing origin characteristics. Doing this with the smallest practical charge for our roasters is a challenge, and it will all differ with larger batches. This kind of variability keeps a hobby interesting!
The roast shown above cups well in the second day post-roast, although I think I need a faster ramp to preserve sugar and make it more juicy.
Gary
LMWDP#308
What I WOULD do for a good cup of coffee!
LMWDP#308
What I WOULD do for a good cup of coffee!
-
EddyQ (original poster)
- Posts: 1051
- Joined: 8 years ago
Your post actually got me thinking I could actually keep air in my filter (roaster end) cleaner if air was flowing from the gauge to the roaster exhaust. Although this would be at the expense of losing precious pressure. However, if I tapped into the exhaust duct closer to the fan, I likely would have more pressure and could "leak" clean air through the filter resulting in it staying clean much longer. If I have issues with dirty filters, I know exactly what to do!Tonefish wrote: You wouldn't want your clean air polluting your roaster exhaust.
LMWDP #671
