Barbarblacksheep, I have a hunch, as BDL has suggested, your thermometry is causing you more of a problem than you think. Although, you can adjust if it's consistent. Here's what I noticed about your last post. (I've converted your numbers to F so they'll make more sense to my small brain).
You say your beans are yellow at 318F and you get 1cs at 350F. Have a look at Sweet Maria's chart for bean development on their roaster. I use it as a guide that works pretty well for me, even though on my RK drum, I can't see the beans, but I do have a K thermocouple in the bean mass.
They say they are in early yellow at 327F and get 1cs at 401. That's a 74 degree difference, while yours is only 32 degrees difference. I'm guessing you actually are at yellow much earlier than 318F as you suggest. And, your other temps suggest that yellow would come much earlier. Also, you say you get 1cs at 350F but drop at 361 for only an 11 degree rise. That's uncommonly short for development I think.
So, is it possible your stages are coming much earlier than you think? Perhaps, yellow arrives earlier and 1cs as well. That might explain why you are going too far into the roast. Have you let one of these roasts go into 2cs? I wonder how much temperature rise you would have between 1cs and 2cs. For my gear, I expect a 50 degree rise in Farenheit.

Ahh I see what you're saying.

So, how do you think I should try to correct this?

This is what I'm trying to achieve:


But this is what I'm getting:

barbarblacksheep wrote:So, how do you think I should try to correct this?

Since you're looking for a specific color rather than a flavor profile, type of development or anything else... There's only one way. Use your eyeballs, and Drop at the desired color.

Once you've nailed the color, note the temperature and you can start dropping subsequent roasts at the number -- providing it was drinkable.

It's probably a combination of the picture and my monitor, but the beans in the picture of the roast you posted look awfully damn light to me. Not only that, but a lot of them beans appear very small. Maybe Dropped right at 1stCs, possibly even before, but at most a few seconds later. Not my type of roast, but one man's chlorogenic swill is another's ambrosia.

Rich

It's probably a monitor thing. On my screen the 2nd photo is much darker. But yes the beans in that hopper are very light. The problem is I can't get a roast that light. Straight after 1st c they're already medium.

barbarblacksheep wrote:It's probably a monitor thing. On my screen the 2nd photo is much darker.

Mine too. I apologize for lack of clarity, but when I was talking about the posted beans, I meant those in the hopper.

The problem is I can't get a roast that light. Straight after 1st c they're already medium.

After 1stCe, they're a "City" roast, which is medium by definition. You'll have to drop at some point into the start of 1stC (say 8 snaps into 1st), "rolling 1st," or something else very specific, and easily identifiable by sight, sound and/or smell.

Considering what you're trying to do, there's no substitute for using your eyes and Dropping by sight. If you can't judge color through the sight glass, you'll have to start doing tryer pulls every 15sec or so after 1stCs.

Your thermometer readings seem very unreliable.

Since you're roasting in a 3kg roaster it's safe to assume you have commercial intentions. If that's true you should fix your thermometry and start using real time roaster software. It will give you a lot more insight into what's going on inside the roaster than you seem to have, and greatly aid consistency as well.

Rich

So I tried dropping during 1st crack and I've now got the colour I'm looking for. I'll wait to see what it tastes like. One question, Do you run the cooler in the tray during this rolling 1st crack? If so, won't there be a load of beans that haven't popped yet?
Thanks

It seems to me, your ROR is awfully low. I may be repeating here what's already been said, but, you probably need less air and more heat, especially after yellow. And, again, I suspect you are at yellow earlier than you think. BTW, what I see here and what I use generally, is to have around 20 degree F rate of rise after yellow and than slow it way down after 1cs. Of course, if you are dropping during 1c, you don't need to do that. But, you should probably try to anticipate 1cs and drop your ET a minute before.

What does RoR stand for? EDIT: rate of rise

If the profile shows yellow at 6min, tan at 8min, 1st c at 10min, wouldn't that suggest that things are generally okay? I mean if the ror was too low, I wouldn't be getting to tan untill 9-10mins, 1st c 12min + ?

From this, one can assume the probe maybe giving bad numbers, but if milestones are met at the right times - I could comfortably just taste the results and see if they're good.

barbarblacksheep wrote:So I tried dropping during 1st crack and I've now got the colour I'm looking for. I'll wait to see what it tastes like. One question, Do you run the cooler in the tray during this rolling 1st crack? If so, won't there be a load of beans that haven't popped yet?

There's a good article on cooling in the current issue of Roast Magazine. If you have access, you may want to read it.

Yes, there'll be quite a few beans which haven't cracked. But that's the point. Once they're all cracked, they're "City" and not "Scandinavian. By way of evidence, if you examine the picture of the light roast in the hopper you posted, there appear to be a high percentage of beans which haven't fully expanded.

I'm no expert on the subject, and would love to hear from any of the number of the people who know a more. My experience is that for best results, beans should be cooled within 3min30sec of Drop, but no faster than 1min.

Speaking of lack of expertise, while I love talking to you and helping you out, I'm not the best guy on light roasting.

Rich

Oh.My.God

You are not going to believe this....

I decided that since you guys said the thermocouples were giving bad data, I ought to have a look and give them a clean. I had just finished a roast and pulled the first one out of the drum, yeah it's a bit black, so grabbed a damp cloth and gave it a wipe. I looked across to see how much effect the damp cloth had... none. Eh? I thought. I then looked across to the E temp... it had plummeted. Gulp. Oh god...

Yup, you got it. I've been reading the wrong frikkin dials!

Utterly embarrassed and ashamed, but also quite relieved coz now I can figure out what's the heck's going on!

Ok so got straight into first batch with correct dials and instantly you can see the new problem. The E temp seems way off. By the way, the E probe was caked on with black crud, I gave it a good clean though.

So if you look at this profile, the lines cross over, twice! However, if we were to ignore E value, how does the 'D' temp look? The roast was way too slow (though the beans are a nice colour and well developed) so I'm gonna run it again now with a much higher flame and see what happens.

Ha, what a doughnut