TJ-067 Gas Roaster Guidelines for Different Load Sizes

#1: Post by **drgary** » November 19th, 2014, 6:05 pm

Hello All:

Now that I've got my roaster giving me meaningful temperature readouts I asked H-B member Dregs (Barry) offline about his recommendation for using the TJ-067 gas roaster (propane or natural gas) with different charge sizes. This was after osanco (Steve) mentioned Barry had been experimenting with small loads. Barry's response was so helpful that I got his okay to post it here, verbatim.

Dregs wrote:While most of my batches are sized for the "sweet spot" on this roaster, 20 to 28 oz, I have done batches varying in size from 4 to 44 oz. The roaster is still easy to control and very responsive in that broad range. In fact, I recently picked up some Panama Perci N2 Level 39 over at GCBC ($58/lb, just crazy) and did 3 batches of 4 oz each before roasting a larger batch. The roaster performed flawlessly. My taste buds, on the other hand, lack the sophistication to appreciate that level of bean. Anyway, I have the confidence in the roaster to do a 4 oz batch, even with that bean.

My general approach for any size batch (after letting the roaster heat to about 450F and cool to charge temp) is:
-Preheat to a specific temp on the BT thermocouple. This temp varies from 255F for 4 oz batches to 325F for 44 oz batches. I determined that temperature range based on results of many batches. I keep the exhaust fan low (my fan stops at 10, so 12 is my minimum during a roast).
-Gas is off for the 1st minute after charge with the fan on 12. I usually hit the turning point at around 1 minute. At 1 minute, I turn the gas on. The level is about 40% of max for 4 oz batches, up to 80% of max for 44 oz batches. I also turn up the fan a bit, to about 20. I adjust mostly with the fan to hit my target End of Dry (300F) time.
-At EoD, I raise fan speed a few points, but don't change the gas level for 1:30 to 2 minutes (although large batches may need max gas at this point). Then I continuously lower gas and raise fan speed throughout the ramp to hit my target time for 1Cs. I use the projection lines and HUD readout in Artisan to help do that. I try to get to a gas pressure of 30 - 60% of max and a fan speed of about 40 at least 1 minute before I expect to hit 1Cs. Obviously, larger batch sizes require more gas than smaller.
-If I have gotten into 1Cs at an appropriate gas and fan level, I make no adjustment for about a minute. I've been trying to do roasts as Scott Rao recommends. Since "exothermic flash" will cause the ROR to drop as the beans enter 1Cs, you want to reduce heat input before, not at that point. As 1C is ending, I drop gas pressure and increase fan further. I am usually at 20-30% of max gas and fan setting of 40 - 70 to coast to drop. I rarely roast darker than Full City.

The one problem with smaller batch sizes is that the BT thermocouple does not read accurately. With anything less than about 8 oz, it reads low. With a 4 oz batch, first crack will likely read 10 - 12F lower than a 16 oz batch of the same bean. I really don't find that to be a problem. I know it will happen, so I just compensate in setting my time/temp targets. I do reduce the drum speed for smaller batch sizes:
<16 oz 40RPM
16-28 oz 60RMP
>28 oz 80RPM
The slower drum speed helps to keep the BT thermocouple in the bean mass.

#2: Post by **keno** » November 19th, 2014, 6:20 pm

Gary, glad you got it working. Looking forward to hearing more about your experiences with it!

#3: Post by **LDT** » November 19th, 2014, 7:00 pm

Gary, This is the kind of information I was talking about. It will be extremely helpful as you roast more. As I said the other day, I found it helped early on to do more, smaller roasts than fewer larger roasts just to get more experience at the controls. I was impressed with how responsive your roaster is and I know you'll really start loving it more and more.

November 19th, 2014, 7:59 pm

Laron and I roasted some Brazilian Candado peaberry on the fly and he noticed 2c starting. We dropped immediately. I'd been intending to brown some beans for temperature studies I'm doing while reviewing the Elektra Microcasa a Leva and the Ponte Vecchio Export. It came out too good, a really nice chocolaty base for capps. So I took no chances today and browned 1kg of mummified 2007 Ethiopian greens gifted to me by Rawman to help break in the roaster.

So yes, it was Laron's suggestion that got me searching for minimum load capacity to try many profiles.

November 19th, 2014, 10:55 pm

Great post, Gary! We're doing much more small batch sample roasting lately and it's comforting to hear we're doing much of the same.

Thanks to both you and the illustrious Dregs for sharing.

November 19th, 2014, 11:25 pm

Hi Steve,

I'm now starting to get a sense of how good this roaster really is.

#7: Post by **achipman** » November 21st, 2014, 2:29 pm

Great post! It helps add some verified context to what I am experiencing in my own batch size experiments.

This has been all the more challenging with the recent weather in the mid-west. The preheating is very important here.

#8: Post by **Dregs** » November 21st, 2014, 9:04 pm

I would very much like to hear how the approach to roasting taken by other TJ-067 users, and users of other roasters, differs from the outline in the first post above. I realize that comparing how we roast, even between identical roasters, is difficult. We use different thermocouples, run on different power sources (propane with or without Steve's supercharged regulator, or natural gas) and use different logging hardware/software. Still, I wonder if it is possible to cut through some of the equipment-based differences to get at real differences in roast approach.

For example, I can reach a target time for End of Dry by either charging at a higher temperature with low gas pressure, or charging at a lower temperature with higher gas pressure until hitting the target. I haven't experimented enough to know whether/how that might affect taste in the cup. Which approach are you using?

As another example, I ended up with a really good-tasting batch of beans when I accidentally slowed the roast enough so that the ET line fell below the BT plot toward the end of a roast (see graph below). I had seen this occurrence on some graphs posted by experienced roasters (BDL's, in particular). I always thought I would stall the roast if I let that happen. Surprise, no stall! It took me far too long to realize that wasn't just a defect in thermometry. Then again, I've turned some beans into cardboard trying to do that. Have you experience this inversion with your roaster or do your plot lines always stay "where they belong"?

So, please chime in with any differences in the approach you take to roasting compared to what I do. I'm still a novice at roasting - there have to be better ways to do things than what I found. Let's help each other learn!

Dregs

November 21st, 2014, 11:17 pm

Barry,

Thank you for your contributions here. To follow one of your questions, does ET falling below BT matter when the beans are generating their own heat (exothermic)? As long as they're continuing to generate enough, especially at that phase where they're mellowing out acidity, why would it be a problem? You might offer a hint by telling us what was special about that particular roast. What flavors were enhanced?

#10: Post by JK » November 22nd, 2014, 12:13 am

I have been experimenting with leaving the gas off longer.. I charging at 360* for 24oz roasts and leave gas off till a 1:30 to hit my drying times for roasts I want 6min drying time.. This way there is more stored heat.. I always try for a 175* turn.. I read that several places on the Roasts Guild site and Scott Rao also shows 175* turn in his graphs..

I'm not sure if 175* is a good TP temp but its what I use..
Maybe someone knows if it matters and can comment