Have you done something in terms of calibration, or did it go “within a point or so” just out of the box with their calibration? Do you know anyone in possession of a pricier meter to compare the two?
@ mcaillio
Thanks for implementing everything. I went back and edited all my measured roasts.
“Out of the box” it was within a point or two of the Espresso Vision.
I was using Espesso Visions Agtron Gourmet Grind Size Validation Scale:
Their measurements are based upon “the use of a US Standard size 20 mesh sieve such that 70-75% of the particles would pass through (according to SCA cupping protocols).”
I bought a #20 sieve and ran some tests: This worked out to a number 12 setting on my Barratza Encore (YMMV)
I don’t know of anyone with a True Agtron Gourmet machine that could measure and share some whole beans at different roast levels and their Agtron ground values. But, I am looking.
I feel comfortable using this as is. (my OCD wants to go further)
I am not selling roasted coffee, so once I get used to the readings that I am getting I can use them to compare the results of my roasts.
Eventually, I will have a spreadsheet with all of my values and be able to correlate to different roast levels: City, City+, charcoal, etc.
One of the big things that I am workling on is to get very consistent results during the measurement phase. In particular: Grind size and tapping the meter on the table top with the same amount of force so that I always gett the same reading.
Yeah, I was using that, but now that I can just put in my results from Roast Vision, I don’t feel the need to add the Agtron number. The colour is something you should be consistent with in your style of roasting, and as SCA says, even if you’re 6-7 points off, it’s still good. I asked about that because I thought I didn’t do good on the exam coming within a point on Agtron Gourmet scale with my roast. Anyway, with Roast Vision, I’m getting very consistent results between measurements. I think the most important thing there, apart from grind size, is to use excessive amounts of coffee, which are still small compared to other meters. I’ve seen it all over the place even when the tiniest speckle of light passes through. Like two spoons (the one included with Roast Vision) instead of one. By the way, the best thing that comes to mind is that maybe you have a kind-hearted local roaster somewhere around that uses Ligttells, or Agtron, or some of those fancy meters and ask for a favour.
I found that RoastVision has a quirky behavior that you may want to experiment with.
The RoastVision functions by reading the intensity of the light reflected by the grinds from the calibrated light source built into the instrument. When you drop the ground coffee onto the viewing window, RoastVision calibration relies on no light bleeding thru the grinds. With that in mind…
It may be worth it to experiment with placing a cover over the grinds after you note the reading to see if there is any shift caused by incidental light bleeding thru the grinds. If there is a change, that effect from unwanted back-lighting might be managed either by using a finer grind or maybe a little larger sample or maybe just continue with the additional cover.
Bruce
ps- The shift is very small and probably is most noticeable with a very dark roast.
I agree Bruce.
I was thinking about using my black Amercain Express card as a cover…(I painted it, lol)
I think that I am getting a little confused here.
When I talk about the Espresso Vision unit, I am talkiing about the original Espresso Vision meter. However, on their website they are calling their meter: Roast Vision
When I talk about Roast Vision, I am talking about the meter that is availble through Github.
I will start referring to the one on Github as Roast Meter.
Sorry for the confusion.
He had only one meter he was selling and I recall it started out as Espresso Vision… the name of the business. The name of the colorimeter was Roast Vision.
Since then he’s apparently completed his PhD and doesn’t have time for his income generator. Dunno if the design was his or if he was reselling something else. Whichever, Roast Vision is a solid product. And apparently it lives on thru the source you found.
Me
Edit- I drilled down a little further on his site and see that I may not be grinding fine enough… he refers to a ‘very fine espresso grind’. I’ve been using something coarser than that. I use a hand-crank grinder. Getting to an espresso grind would take much more effort than I’m willing to provide. Blocking the back-light is a lot easier for me. Of course YMMV…
On one of his News/Blogs, he points out that he changed the grind size to better and more accurately accomdate the Agtron conversion:
I believe that this was the final recommendation for Grind size. As I said above, I bought a size 20 mesh sieve and dialed in my Baratza Encore (setting 12) to match the 20 mesh sieve. I doubt if every Encore will match that setting for that grind size. (YMMV)
BTW: The electronics for the Roast Meter is EXACTLY the same as the Espresso Vision product, The same microprocessor, the same micro OLED board that measures the reflectivity of the ground coffee. Espresso Vision even lists the components on their website.
Therefore, I used the same grind size on the New Roast Meter.
The finer the better. I almost grind to Turkish. I have all Eureka stepless grinders, so I wouldn’t know a number for it. As soon as I grind coarser the results shift and aren’t consistent between measurements. I don’t think covering the coffee with anything would give you better results. There’s interference all the same, be it from bright or dark colours if they pass through coffee. If you grind finer than 20 mesh sieve, there wouldn’t be significant differences, except for clumping which could create spaces for things to pass through.
I found that Onyx Coffee Lab sells roasted coffee and share the Agtron level that they roast to for each coffee.
I bought five 4oz samples with agtron Values from AG53.4 to AG75.
I then worked on getting consistent results/readings with the Roast Meter.
I used the finest grind setting that I could with my Baratza Sette270. It looks like I am at 250 to 300 microns under a microscope. Turkish grind.
Then I started measuring the Onyx coffees and recording the values that I was getting different readings from the Roast Meter. It was off about 20 points. I went into the program for the Roast Meter and adjusted the led brightness value to 105. (it was 135. It took several attempts to get it where I wanted it)
Now all of the Onyx roasts are close to the Roast Meter except one. I am assuming that roast is off.
Overall I am pleased. For the price, it is a good tool and will suit my needs of having a less subjective analysis for my records.
I found that when I used a coarser grind, there were flakes of chaff that had to be affecting the measurement. by grinding extra fine, the chaff gets ground up.
Super nerdy stuff 
… most of this thread on the Roast Meter should have been a separate thread
Thank you! (very proud of himself, ha ha) “Give in to the Dork side”
I think that you are right. It should probably have it’s own thread.
I ran a couple of more tests and found that I was getting a minor variation of about 2 points sometimes.
I bought another roasted bean that was supposed to be Ag78, but it was only checking Ag75-76.
I ground up some of the beans that I had checked before and were Ag75 and they were checking Ag 73.
I found that the glass window had gotten smudged with an oil film, probably from the oil from the beans. I store my roasted beans in glass jars and often see that the glass get a brown tint to it due to oil from the beans.
I used a cue tip with Windex, and the beans that had been checking good at Ag75 were back to Ag75. I could definitley see that there was some brown on the cue tip.
The new bean that was supposed to be Ag78, checked at Ag77.
Another thing to consider when using the Roast Meter and probably the Rooast Vision.
I have also been using a can of compressed air to lightly dust off the window of the Roast Meter.
Hoping you can help me out Bill – i have all the components assembled and followed instructions on software upload. No errors in the sketch and no issues uploading to Artemis Thing Plus. However, I’m not getting any power to the Micro OLED. Did you encounter any software related issues that would help me get this thing up and running?
Did you install the libraries for the
SparkFun_MAX3010x_Sensor_Library-master
and
SparkFun_Micro_OLED_Arduino_Library-main?
This is where they should be placed
After placing the libraries in the correct location, you need to restart Ardunio so that it can read the libraries.
Also: make sure that you have a good USB cable. I think that you probably do, but I was bit once because the cable was powering the board but there was no data transmission and it wouldn’t load the com port drivers.
@billc thanks for the reply. Yes, I installed the libraries, verified their location was correct and restarted Arduino.
Funny thing about the first cable I tried = even though the serial port was selected it kept saying “not connected.” I swapped out the cable and was able to make sure Arduino said the serial cable was, in fact, connected. Verified/uploaded the sketch and still a blank OLED???
Are the components screwed into the housing?
Can you reverse the order of the components? The wires that go between the components are just a bus. It shouldn’t make any difference in the order.
Do you have a red led lit on the Smartfun and the 3010? Both should be lit when power is applied.
I put together a spare set of components that I had on hand and go them to work as expected. It did take a while though, I need to re-write the instructions to make them easier to follow. More Specific, less generic.
This is the extra setup that I just loaded.
2 red LEDs, and the green box is the sensor that will measure the refectivity of the ground coffee. If you put your finger over the sensor the value will change on the screen.
You do have the Sparfun MICRO LEd library loaded. Correct?
I just have to ask these questions.
Yep, here’s the pics in two dif configurations of second cable running to OLED
I only used the links located in the github assembly guide. Correct libraries are installed.
I am pretty much at a loss as to why the OLED is not lighting up.
Try this:
In Ardunio:
Go To: file, Examples, OLED, Multidemo (or any of the others,
Then verify the sketch.
Then run the sketch
When I do this, the display will run through all of the examples.
If it doesn’t do anything, I am out of ideas.
Thanks for all your help. Still nothing. I’ve submitted a ticket to SparkFun as I believe the only thing left is that the OLED is DOA.
