On Idle Noise - Page 6

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
Supporter ★

#51: Post by Almico »

MaKoMo wrote:Note that this is all external to Artisan and not connected to Idle Noise:

Your TMP1101 is obviously not attaching to the Phidget driver. It should popup up in a similar way as the TMP1200 and listed by name. Either the module is dead, the patch cable connecting it to the HUB is dead or the HUB port you are using is dead.

For a red marked module the driver holds a firmware update. You should click that red text and follow the instructions to update the firmware of the module. Note that the Phidgets driver and the firmware of the modules is regular updated. Always install the latest Phidget driver and all available firmware updates.

Your first "RTD"-roast looks good. What was the sampling rate on that one?

Thanks Marko!!


#52: Post by mbenedet »

Alan, any luck with connecting the RTDs and Aritsan? Super excited to see your results

User avatar
MaKoMo (original poster)

#53: Post by MaKoMo (original poster) »

AlexGS wrote:If I count right it should be Y5, right?
No. The first channel of the first extra device I defined here to gather the ET is bound to the variable Y3. Y5 refers to the first channel of the second extra device (which is here not configured!). Y1 is bound to ET and Y2 to BT.

User avatar
Supporter ★

#54: Post by Almico »

mbenedet wrote:Alan, any luck with connecting the RTDs and Aritsan? Super excited to see your results
Me too.

The RTD seems to hold great promise. Idle noise was practically nil, yet the probe was very fast to respond when a change in probe temp occurred. Before, with my old Artisan settings, the RoR would jump in one direction and them seem to oscillate with very steep peaks and valleys back and forth until it stabilized. With the RTD, it jumped and then settled right back down. This is with 1s sampling, 0 smoothing and 1s delta span:

I had to tear it down because of a faulty TMP1101 for measuring ET. I just received a replacement and will hopefully have time to hook it up this evening.


#55: Post by Bunkmil »

I don't know if this is the right place to ask this question but generally speaking which of these options should be preferred?

Fast sample interval (1s)
Low to no smoothing
High delta span (20-30s)

Fast sample interval (1s)
Some smoothing
Lower delta span (10-20s)

Lower sample interval (5s)
Low to no smoothing
Lower delta span (5-15s)

Or anything else? I know that it's specific to each setup though... I have small and very responsive 3mm probes connected to a phidget 1048 and option A is what I have been using but I am curious to know if I should try another approach.


User avatar
Supporter ★

#56: Post by Almico »

My old setup was

1s sampling
0 curve smoothing
2 delta smoothing
15s delta span.

It worked well. It still works well. But the 15s delta span adds quite a bit of delay to the RoR curve. I'm looking into the past instead of making decisions in the now. If the RTD can give me similar curves with 1s delta span, I would be very happy.

User avatar
Supporter ★

#57: Post by Almico »

I did a quick roast with the new RTD setup. BT is 3mm , ET is 3mm K-type TC.

1s sampling, 0 smoothing, 1s Delta span,

I'm not sure what that BT RoR blip is around 4;00, but the overall difference between the types of sensors is apparent.

Correct me if I'm wrong, but is the dramatic ∆ET curve the result of going from 15s delta span to 1s? This is my curve with 15s DS:

User avatar
Supporter ★

#58: Post by Almico »

FWIW, i just did a roast using the RTD probe for BT with an old profile that used a 3mm TC for the background. Artisan was set at delta span=15 for the old roast. The new roast is set to 5. You can clearly see the difference


#59: Post by tonescout »


I have got some interesting results with my set up - one problem I had was a probe that was not low enough in the roaster as supplied, and this gave too much air turbulence = variation I believe as it sat too high near the top of the bean mass.

The attached profile shows the set up where the ET is the OLD probe position (more like 10 o'clock) and BT is the the new probe position (more like 8 o'clock) on a clockwise drum - I have very low idle noise with machine fan / heater and drum on and experimenting off roast, however during a real roast event I still get a lot of what seems like noise. 1s sampling, 0 smoothing, 1s delta span in one plot shown, and Delta 10s in the other - where the ugly roast profile is more evident now with the new probe location!

Is this just noisy probes or am I missing something with settings? I can use this data for sure to understand the roast now, but it would be amazing to get results like others have posted.

Please ignore the profile crash and flick as this was a roast with scrap/free beans and no profile developed at all - purely run to get the sensors working.

Oh and does anyone else have a few problems with the consistency for the software selection/choice of the two RTD probes, mine are via a VINT hub from TMP 1200 devices, using a Mac; and I am new to artisan (2.0) so it's not clear if I can deliberately select each device by specific part/ID other than the general description in the menu, and checking what has been connected (ET, BT) once up and running.




#60: Post by tonescout »