It will warm up faster in the heat.

So do I.

3 roasts in.

Had to remake the gas lines to the pilot light and burner valve after finding leaks during the first roast.

First roast was a get to know you session, took some single source Ethiopian Natural Process beans to end of second crack to check temps on the new RTD probes.


Second batch same beans, had exhaust fan on full the whole roast started with 5lbs of pressure on gauge at burner. Moved to 6lb on gauge at the end of dry.






Third batch this AM. Kept the burner at 6lbs pressure. Started exhaust fan on 25% then moved up to 75% at the end of dry.




BT temps on these roast seem to be reading a little low based on FC times, but won't adjust in Artisan until i get a few more roasts in to make sure it's not just an anomaly.

Your phase percentages look good on the second batch* for an Ethiopian natural. Charge temp looks good taking into account that your RTDs are probably reading 20F lower than "normal."

Carefreebuzzbuzz may be able to chime in for some tips on RTDs and artisan. The RoR has noise in it that could likely be smoothed to provide more insight.

*EDIT: Meant to say third batch...

Milligan wrote:Your phase percentages look good on the second batch for an Ethiopian natural. Charge temp looks good taking into account that your RTDs are probably reading 20F lower than "normal."

Carefreebuzzbuzz may be able to chime in for some tips on RTDs and artisan. The RoR has noise in it that could likely be smoothed to provide more insight.

Thanks, the 3rd time truly was the charm on temp control and timing.

I am sure the RTDs are sampling too frequently so they should smooth out.

I may back the BT probe up about .5 inches and see if that helps before compensating on the symbolic reading in artisan.

I wouldn't be too apprehensive of smoothing. Raw data from probes or any instrument typically needs some noise reduction or filtering applied to be more meaningful. RoR is a derived measurement. The derivative of a raw (and noisy) signal exacerbates the noisy parts (that's where the big/quick changes are.) We are concerned with what is actually happening in the roasting chamber. On a traditional gas drum roaster there is no way it is acting like what an unfiltered RoR graph represents so we use filters and smoothing to reduce the measurement noise from the probe to better represent the actual value. IIRC RTDs are even more sensitive than the typical 3mm thermocouples so signal processing is even more important. There is a balance between too much where you lose insights and too little where it is hard to discern trends. Artisan provides a few ways to help.

I'm enjoying seeing your progress. I'd like to have a USRC machine like that right next to my 5kilo but I've got something else on order. Hard to find those 1lb USRC machines out in the wild. You've got a good find there!

Milligan wrote:I wouldn't be too apprehensive of smoothing. Raw data from probes or any instrument typically needs some noise reduction or filtering applied to be more meaningful. RoR is a derived measurement. The derivative of a raw (and noisy) signal exacerbates the noisy parts (that's where the big/quick changes are.) We are concerned with what is actually happening in the roasting chamber. On a traditional gas drum roaster there is no way it is acting like what an unfiltered RoR graph represents so we use filters and smoothing to reduce the measurement noise from the probe to better represent the actual value. IIRC RTDs are even more sensitive than the typical 3mm thermocouples so signal processing is even more important. There is a balance between too much where you lose insights and too little where it is hard to discern trends. Artisan provides a few ways to help.

Thanks, definitely not hyper fixating, but want to start from an accurate baseline to avoid too many measurement changes over time to see variances due to controls and environmental conditions.

Milligan wrote:I'm enjoying seeing your progress. I'd like to have a USRC machine like that right next to my 5kilo but I've got something else on order. Hard to find those 1lb USRC machines out in the wild. You've got a good find there!

I found it on facebook marketplace of all places about 2 hours from me. Was on the road the next day to pick it up. It really was a find as well as a steal, though there was a lot of deferred maintenance I've had to tackle. There was so much chaff and soot inside the top of the collector that the exhaust was barely blowing. That cleaned up well thankfully. The propane leaks were an exciting discovery, but that was a relatively easy fix. I still have a few snapped fasteners I need to drill out and replace, but that's more appearance than functional.

Just glad to be roasting at this point.

Awesome roaster and great timing! I just picked up the same USRC sample roaster in racing green last week. Got an amazing deal as well. I happen to live in Oklahoma City so I have USRC 15 min away for support which is always comforting.

Just got it hooked up to artisan via phidget (temperature sensor) and roasted my first few batches tonight. Currently I have only rerouted the BT from the two pronged port into the phidget but would like to add a ET. Very apprehensive about drilling into this beast... how did you do it? Did you need to create and tap a thread or just drill a hole? If just a hole how big did you drill?

Also, I was wondering if you're using propane or natural gas. I'm on propane and have barely increased gas past "1" on the side gauge and my roasts were finishing under 8 min with appropriate curves. Wondering if I need to decrease the propane pressure regulator valve before the roaster(the Honeywell section). Any experience there?

Thanks!
Matt


P.s. Maybe we can keep this thread going as we both test out these new (to us) beasts!