The Summer Conundrum… advice sought!

[DesigningByCoffee]

Morning!
No, they are both what I expect are good roasts (I'm about to crack the 'un-noted' columbian bag - so can't confirm yet! :-).
Not a huge difference in the Columbian, but I could have slowed it a fraction more (see how it sits above the line) - which would have actually fitted it perfectly into my 'theory' - but I chickened out half way and kept the temp up a bit :-)

Just as a comparison - here is the graph of one of my 'summer sour' roasts. As you can see - similar inputs gave the same shape profile - but 1C at 11-12 mins, not 13-14 - and the tasting results show my thoughts…

[insomnispresso]

Loved reading your roast notes

So they are both good ones right? Is the 2C much earlier on in the bad one?

[DesigningByCoffee]

We need some graphs!! Just kidding

Matt, why not also measure time? I think that's what lead to your initial "summer conundrum"

Cheers
Matt

Your wish is my command!
These graphs should answer your time question too - my initial issue was that with warmer weather, the time overall was reducing, giving sour roasts. Those input gun temp tables I posted before show how I deal with changes in ambient to get basically identical profiles (ie same curve, time etc). These profile are from an 18 degree day and a 28 degree (you can see the start temp difference). But after the prewarm they are basically identical.
And the background 'ideal curve' shown is my 'repeated success' curve, Yelta. But I do agree - if you're going to repeat a roast, make it a goodie! And now I can repeat this profile in summer - the goodness is back in the cup!
Cheers Matt

BTW the Peru was a fraction fast (I broke my own rules!) - I should have dropped by the 10deg at all points… live and learn!

[Yelta]

Howdy Yelta
Repeatability is key!

Can't argue with that statement, as long as they're not repeated failures.

[insomnispresso]

We need some graphs!! Just kidding

Matt, why not also measure time? I think that's what lead to your initial "summer conundrum"

Cheers
Matt

[DesigningByCoffee]

Howdy Yelta
Not too complex really - even if it looks it when presented like that :-)
I just add or subtract 10deg to/from my gun input ramp as the ambient increases or decreases - and get an identical profile every time summer or winter. Repeatability is key!
Matt

[Yelta]

Hi all
Just a little update. Like the rest of the country we've been getting some warm ones, so a great chance to try some warm weather roasts!

The findings?
On my covered/insulated corretto with 350g greens, to achieve a consistent roast that matches my preferred profile, I simply need to drop the temp ramp leading to 1C by 10deg on the gun (so 7.5 deg inlet temp) for each 5 degree rise in ambient. Strangely, the settings after 1C remain about the same - maybe because temp by this stage is so much higher than amb anyway? So for example, to achieve the profile attached, the method is as follows:

• So, for a cold weather roast (amb 10deg) the approx ramp looks like this:

190 (gun) pre-warm for 3mins
370 (gun) @ 3mins (DMM)
420 (gun) @ 130 (DMM)
470 (gun) @ 160 (DMM)
Drop to 410 (gun) @ 195 (DMM) just before 1C, then 360 @ 210, maintain till drop.


• For a warmer weather roast (amb 20deg) this ramp looks like this:

170 (gun) pre-warm for 3mins
350 (gun) @ 3mins (DMM)
400 (gun) @ 130 (DMM)
450 (gun) @ 160 (DMM)
Drop to 410 (gun) @ 195 (DMM) just before 1C, then 360 @ 210, maintain till drop.


• For a hot weather roast (amb 30deg) this ramp looks like this:

150 (gun) pre-warm for 3mins
330 (gun) @ 3mins (DMM)
380 (gun) @ 130 (DMM)
430 (gun) @ 160 (DMM)
Drop to 410 (gun) @ 195 (DMM) just before 1C, then 360 @ 210, maintain till drop.


Seems to work really consistently - and the results have returned to the cup :-).
This would obviously need to be varied based on altitude & humidity, also bean mass - but the theory seems pretty solid.

Happy hot-weather roasting all!

Matt

Crikey Matt, all seems quite complex, regardless of season I start the beans @ ambient temp, similar heat gun settings every time (a bit warmer in winter, cooler in summer) let her run to FC (always happens @ 200c) approx 13 mins, if it seems to be getting there a bit fast I knock the HG back by 50c then let it run until finish, depending on the beans somewhere between 225 and 230c.

[DesigningByCoffee]

Hi all
Just a little update. Like the rest of the country we've been getting some warm ones, so a great chance to try some warm weather roasts!

The findings?
On my covered/insulated corretto with 350g greens, to achieve a consistent roast that matches my preferred profile, I simply need to drop the temp ramp leading to 1C by 10deg on the gun (so 7.5 deg inlet temp) for each 5 degree rise in ambient. Strangely, the settings after 1C remain about the same - maybe because temp by this stage is so much higher than amb anyway? So for example, to achieve the profile attached, the method is as follows:

• So, for a cold weather roast (amb 10deg) the approx ramp looks like this:

190 (gun) pre-warm for 3mins
370 (gun) @ 3mins (DMM)
420 (gun) @ 130 (DMM)
470 (gun) @ 160 (DMM)
Drop to 410 (gun) @ 195 (DMM) just before 1C, then 360 @ 210, maintain till drop.


• For a warmer weather roast (amb 20deg) this ramp looks like this:

170 (gun) pre-warm for 3mins
350 (gun) @ 3mins (DMM)
400 (gun) @ 130 (DMM)
450 (gun) @ 160 (DMM)
Drop to 410 (gun) @ 195 (DMM) just before 1C, then 360 @ 210, maintain till drop.


• For a hot weather roast (amb 30deg) this ramp looks like this:

150 (gun) pre-warm for 3mins
330 (gun) @ 3mins (DMM)
380 (gun) @ 130 (DMM)
430 (gun) @ 160 (DMM)
Drop to 410 (gun) @ 195 (DMM) just before 1C, then 360 @ 210, maintain till drop.


Seems to work really consistently - and the results have returned to the cup :-).
This would obviously need to be varied based on altitude & humidity, also bean mass - but the theory seems pretty solid.

Happy hot-weather roasting all!

Matt

[DesigningByCoffee]

Thanks Mal
That's a really helpful overview.
There are just so many variations it's hard to know where to start. Esp when the beans in the supply keep changing! You're almost better off to buy 20kg of one type of bean and work it till it's nailed! I suppose this is where commercial roasters have a faster learning curve - if you do roasts every day (rather than every week) it would certainly speed up the cupping opportunities (and increase the failed batch size too :-)

I'll certainly be interested too to see how that slower start/harder ramp works out - it was certainly a vigorous 1C. I've been pretty sold through personal experience on the s-curve input temp profile - this just really ups the intensity of the curve I've been following.
Let you know how it goes…

Enjoy the storms!
Matt

[Dimal]

Hi Mal
This is an interesting point you make, which I've been pondering the last little while. What basic taste parameters do you use to decide when to adjust a roast profile? Or do you take a new SO bean, just trial 4-5 different 'templates', see which works? So far I have some of basic adjustment ie if it is sour, stretch out the time to 1C (as seen in this thread), if is oily & bitter drop a bit earlier etc. But I wonder if there are any 'markers' pro-roasters use to fine tune a roast for a bean…?

More or less Matt..

I have a number of profiles that I use to start out with a previously untried bean variety. These are based on the bean origins and give me a starting point. I then roast at least three profiles starting with one that takes the batch through to the start of 2nd-Crack, the next batch goes about 20 seconds into 2nd-Crack, with the last stopped about 20 seconds before 2nd. The profile up to the start of Rolling 1st-Crack is the same for each batch as is the profile between Cracks - Only the timing around 2nd-Crack that changes.

After giving the batches a couple of days to degas and flavour development to happen, I then just run a series of simple taste tests using Long Blacks. Which ever of the batches tastes best to me, I then concentrate on subsequent batches based on this profile and play around with it until I'm really happy that I'm getting the best I can from the bean. This is where all the fun is of course... Stretching or compressing the ramp up to 1st-Crack, the transition zone from about 180C until the start of 1st-Crack, post 1st-Crack ramp duration and rate, whether I actually head into 2nd-Crack and if I do, how far along do I go; into Rolling 2nd-Crack or allow the lazy initial cracks to continue for 'x' amount of time... And the list goes on and on...

With regard to "Markers" that the Pro's may use. No doubt there are but I bet that's where the Sample Roasting comes in coupled with Cupping Sessions. Each Pro probably has their own routine and method to suit the way they do things; coffee roasting seems to be such a personal expression when you come right down to it. Would be interesting though to attend some sample roasting and cupping sessions, would be a great opportunity to learn.

Just FYI I also tried a roast today based on one of your suggested profiles from an old post (couldn't find it again if I tried!) which had a 10-12deg ramp for 5mins, then 16-18deg ramp from there to 1C. I was never able to achieve that slow start on my old setup - it always raced away, but the Bosch setup now lets me do it - so I thought - what the hey! We'll see how it goes in a week or so :-)

So thanks in advance!

Wow! That does sound like an old one... Seems spring from when I was reading a lot about how Willem Boot likes to roast. Will be interesting to hear your thoughts once you've had a chance to try it.

All the best mate,
Mal.

[DesigningByCoffee]

Reporting back.
So, after 3 days, how was it in the cup? prepared as a long black, quite drinkable, loads of crema, a little on the thin/watery side for my taste, there is no question, I tend to prefer my roasts on the dark side, perhaps it will improve over the next few days.
I'll drink this lot without complaint, however my next batch will certainly stay in the roaster a little longer.

Hey Yelta
Will be interesting to see how it progresses as it ages.

Just a thought - I found early on with my Bosch setup that a single fixed temp always gave a reducing curve profile - and a bit of a flat taste, just like you mentioned. Not bad - just nothing extraordinary.
I found that as soon as I started ramping the temp input (giving a flatter rise on the profile) - all the lights came on in the cup!
So for my early successful ball park profile with the Bosch, I'd do a basic gentle warm for 2mins on 180 (all fan 3), 350 up until 130(DMM), then 400 until 160, then 450 up till 1C. In the cup, this ramping input seemed to add a lot more interest, plus a lot of additional body and some more 'cocoa' - even when roasted quite light - just before 2C. Before this, non of my roasts were really cutting through soy - but we finally got some taste! Not bad as doppios - but still quite light & aromatic, rather than thick and gloopy.

I also think I had an aha moment about roast depth a few weeks ago, when I roasted darker than usual (about CS9-10) which was too far for the beans I was using (making them oily and bitter) but still had an over-riding sour, under-roasted taste! Doh! So there's more going on that just roasting darker… although some beans, like the mad mexicans, seem to enjoy that zone and handle it quite well. Must be all the tanning oil & tequilas!

So I'm working lots on the zone before 2C - I think this is where a lot of the taste is to be had :-)

Happy roasting over in the hotlands!

Matt

[DesigningByCoffee]

If I am doing a batch for which there is no template available, I just head back to my notes for previous batches of the same bean/blend and attempt to replicate the original roast data, and ultimately create a new Template to use in the future. These, of course, can be manipulated based on taste tests to address any perceived roast shortcomings - Nothing new really...

Hi Mal
This is an interesting point you make, which I've been pondering the last little while. What basic taste parameters do you use to decide when to adjust a roast profile? Or do you take a new SO bean, just trial 4-5 different 'templates', see which works? So far I have some of basic adjustment ie if it is sour, stretch out the time to 1C (as seen in this thread), if is oily & bitter drop a bit earlier etc. But I wonder if there are any 'markers' pro-roasters use to fine tune a roast for a bean…?

Just FYI I also tried a roast today based on one of your suggested profiles from an old post (couldn't find it again if I tried!) which had a 10-12deg ramp for 5mins, then 16-18deg ramp from there to 1C. I was never able to achieve that slow start on my old setup - it always raced away, but the Bosch setup now lets me do it - so I thought - what the hey! We'll see how it goes in a week or so :-)

So thanks in advance!

[Yelta]

Hey Yelta
My FC/2C temps are now back in line with what I'd expect (very similar to yours @ 199-200 1C). I drop lighter - 223 deg which is just before 2C, about CS 8-9. If I go to 225 - I'm in Mexican CS9-10 territory. But I only do 350g batches, which also changes the dynamic somewhat.

Fortunately we don't have to deal with 30-35 degree ambient like you that much! But we have those -5 night roasts to content with :-)

Interested to hear how it goes in the cup.
Matt

Reporting back.
So, after 3 days, how was it in the cup? prepared as a long black, quite drinkable, loads of crema, a little on the thin/watery side for my taste, there is no question, I tend to prefer my roasts on the dark side, perhaps it will improve over the next few days.
I'll drink this lot without complaint, however my next batch will certainly stay in the roaster a little longer.

[Dimal]

Hi Mal
Sorry to hear about the Pneumonia - bummer :-( Glad to hear you're on the mend though.

Thanks mate...

So do you then 'surf' the gun adjustments simply based on you software rate of rise? Does the temp lag when you adjust temps cause any dramas?
Matt

Yes Matt, that's exactly what I do. There is a slight lag of course but I am able to pre-empt that pretty well, just by becoming familiar with the dynamics of the system overall. No dramas what-so-ever. If one was so inclined, you could automate this to an extent where the 'hands-on' element of control is removed, and you just establish Ramp/Soak profiles to do all this for you. I really don't want or need to do this to be honest. Andy's software does a lot more than I actually need, so it's all good.

If I am doing a batch for which there is no template available, I just head back to my notes for previous batches of the same bean/blend and attempt to replicate the original roast data, and ultimately create a new Template to use in the future. These, of course, can be manipulated based on taste tests to address any perceived roast shortcomings - Nothing new really...

To be honest, every roast batch is a new adventure, even if it was one I've completed plenty of times before. Always love the tactility of it all....

Mal.

[DesigningByCoffee]

Hi Mal
Sorry to hear about the Pneumonia - bummer :-( Glad to hear you're on the mend though.

So do you then 'surf' the gun adjustments simply based on you software rate of rise? Does the temp lag when you adjust temps cause any dramas?
Matt