on a mac I load individual JSON files from from roasTime via the following location:
/Users/USERNAME/Library/Application Support/roast-time/roasts
if you are handy with Python you can probably make use of my code here:
I’ve been on and off working with it for quite a while and close to digging into an analysis of phase times and average RORs in phases. But I’ve been a bit to busy to get it done. I don’t know if my last work was pushed to that repo but we should talk more if you are trying to do more analysis in python.
You can see below that I have several derived or engineered features beyond what is recorded in RoasTime.
This is super helpful, So, the json files are manually downloaded from RT? I did something similar with Python, albeit less sophisticated, but I’m thinking that unless I get the taste data it’s going to be difficult to predict /fine tune a profile. Someone suggested the bean colouring in lieu of the cupping which might work if I can get a process to accurately determine the exact bean colour.
There is a port for “AIR” on the circuit board in the control panel. You can plug a probe into it and use that.
One thing to note, my experience with the Bullet may differ from others, as my Bullet seems to have issues with the airflow. Replacing the chaff collector seemed to improve things for a little while but then it warped again and airflow issues resumed. So something is the matter. I notice that the rubber seal between the top air channel and the chaff box opens up, and hot air escapes upwards instead of into the chaff box and eventually that warps the chaff box and causes additional airflow leakage. I’m not sure if this is a common issue or if it just happens on mine. And there is some conspicuous spacing on my front plate, so it could be allowing cooler air in that way too. Or maybe this is just how the Bullet roasts? In any case, I notice it suffers from either too little airflow or too much, and the air temps are never as hot as I’d prefer.
The airflow temperature at the end of the roast has a high correlation to how thorough the heat application was during the roast.
Let’s say I get to 440F IBTS at drop, and my airflow might read 400F or 391F, despite the IBTS being consistent, the roast development flavors are not.
That all being said, I am adapting to a new induction module that is less powerful than my previous one, so I am still trouble shooting a new profile for 800gram batches. But during that time the airflow began having issues progressively, so many of my roasts don’t work after a couple weeks.
There is a fairly standard 2pin option for airflow inside the control panel on the front of the machine, and you can plug a 2 wire temperature probe into it. The standard one from Aillio was made to fit into the front plate if you drill a hole into it, but that seems a bit extreme.
What I did was insert a longer heat probe through the rubber chute plug. But be careful, you can cause air leaks in the chute plug. That is a crucial contact point for airflow. I put a small weight to hold down the sides of the rubber so they cannot allow external air to disrupt the fan draw.
Yes, but technically I did not discover it, but it is already a part of the Bullet’s early design. I think it was not done correctly early on, and the data was labelled as not useful. Mainly because it is so different from other roasting machines that might lead with airflow, whereas the airflow temp on the Bullet lags behind the bean and drum heat. Or at least on my machine it does.
Airflow heat is probably the most essential to getting an even roast. As drum heat is somewhat inherently unevenly applied to the beans. And that is not my opinion, but stated by people with a lot more knowledge and science than I have. The Loring machines are a testament to this thinking of airflow being the single most important way to roast.
None of this should detract from the many successful roasting stories here. I’m not trying to say that the Aillio bullet can’t produce a nice roast profile. Think of it like the difference between technologies and pros and cons and strategies all differ. I mean, can’t we get good espresso from an D61 heat exchange espresso machine? Probably.
When I copy other users’ roast profiles I don’t get good results on my machine. When I first implemented the new chaff box it did improve things at first, but then the issue with airflow seems to have developed again.
As for the new induction board I have, it does not product the same amount of power as my last one, so I get more sour batches using the same profiles.
I notice that induction boards can vary by 100watts (120v) depending on the temperature of the IGBT (not IBTS). So there have been days where if I have sun coming through the window I get 1375watts at P9, and if I am roasting on a cloudy day, I can get as much as 1505watts at P9. However, if my room temperature is 80F or below, the power draw of the induction board is more stable. 1469watts to 1495wats typically. That still has an effect on the roast, but is manageable. If I see a big drop in the pull of power, I can expect a sour or herbal roast, but I can sometimes compensate with not pulling back on power as early. It helps but there is no way to go back and charge at a higher temp obviously.
@jimmybulletroaster
Not sure I follow your description of the ‘airflow’ facility in the Bullet. Are you saying that there is a connector for a thermocouple inside the control panel? If so, is there also software already in the unit to read and record/display this somewhere?
Sometime back I made a request for Aillio to add an interface to RT to connect a recording thermometer via a computer usb port and display the input along with the other traces in RT. Artisan can do this.
I would use it to add an exhaust air temperature trace much the same as other roasters show environmental temperature. It might not prove particularly useful, but air convection is an important factor in roasting and I like the idea of seeing for myself what is going on with that during a roast. My manual method is way too labor intensive but I did learn useful info about how a Bullet works. Having an integrated trace might make it possible to study the effects of drum speed, ambient air temperature, fan speed, etc.
The way I see it, getting good roasts from a machine depends a lot on understanding what controls that machine offers and how those factors interact. You have to understand what roasting environment a machine creates and then you can best manage it to get the best results it offers.
As far as adding a thermal probe, it is pretty easy. The software already is designed to do the line trace. The issue is that it doesn’t record temperature. However, I have a secondary probe that I use to see the temperature and then I take notes. But the line helps in seeing the trajectory.
So it is still a little annoying to use, but I am used to taking notes way back when roasting on a Probat, so I don’t mind. At least it is all graphed out.
Open your control panel and look at where the temperature probes are connected and then look for the open connection point titled “AIR”. Once you plug in a probe it will record the line graph for the probe, no need to do anything for that. You just need a secondary probe for seeing the temperature. RT4 does not show the temperature, it just records a line. It may have had temperature in an early version of RT or something, but I think the improvements made to RT overall make me hesitant to go backwards, so I just run with 2 probes to record the airflow data.
For adding a thermal probe, you might be able to order one from Aillio. I can’t speak for them on what they can or cannot offer in that regard.
The connector looks like a very common connector, so I hope it is easy to reverse look up. Idk if you have calipers to measure it. Pretty standard looking stuff. I could be wrong.
Aillio originally designed it to drill into the faceplate, but I didn’t want to do that, and my readings in the bean chute seem accurate enough to track consistency. Perhaps readings are a little cooler but being in the bean shoot. Also a lot of possibility of the bean chute cap leaking cool air in, so that can throw off a reading if not careful.
@jimmybulletroaster Thanks for the reply !
Is this the port you are talking about ?
Do you know the spec of the thermistor I should buy ? I am a bit lost ^^’
I think I just used the same thermal probe as the bean probe part. You can order that from Aillio.
However, just a reminder, it is just going to draw a line, with NO number value for the temperature. I have to read the actual temperature with another probe.
I don’t have my machine online, primarily to avoid issues and Bugs
But also, aillio doesn’t have an intuitive way to share or upload profile files. They kind of require an all or nothing approach to sharing. I’m trying to keep dialogue with them about this, but they have a small team and focus on the urgent functions and customer support first.
