Home Roaster Evaluation

Grinding/brewing equipment used consists of:

1) Mazzer Mini grinder

2) Rancilio Silvia (unmodified)

3) Upgraded La Marzocca portafilter with triple basket

4) Reg Barber tamper

Evaluation Plan

A. All roasts were conducted morning of 6/19, 24 hours prior to the brewing trials. Roasting done on a covered patio, ambient temperature = 90F. For HWP, run two 85 gram batches: roast level 6.5 + roast level 7 (this range has proven to be suitable for delivering roasts in the full city + or slightly beyond). For i-Roast, run two 130 gram batches: one with the “pre-2” pre-programmed roast profile, the second with a manual profile.

B. Video of i-Roast process:

1) Beginning of roast cycle (1.4mb)

2) Toward end of roast cycle (1.3mb)

C. Photos of the roast coffee for the HWP and HIR trials.

D. Taste testing was conducted over the 6/20 to 6/25 time period. In all cases, the pulls were only considered valid if 2-2.5 oz volume was generated in 20-25 seconds. Portafilter and shower screen were thoroughly cleaned using Urnex Cleancaf.

Impressions and Observations for Roaster Comparison

a) Crema production was slightly less for the HWP trials compared to either the HIR or HT.

b) For HT roasts, the Mazzer grinder had to be adjusted to a more coarse setting. Grind setting which works well for either HIR or HWP chokes Silvia when used for HT.

c) Surprisingly, the taste differences were rather subtle which could be partly attributable to the relatively dark roasts used for all trials. Or maybe it just means that all three of these roasters do a respectably good job. All shots were quite good, certainly better than what one gets in most commercial coffee shops here in the USA.

d) My subjective ranking of the shot quality, scale of 1-10, is HWP=8, HIR=8.5, HT=9…where 10 would represent the mythical “God shot.” The HT gets the nod for having slightly better depth of flavor and more complexity compared to the air roasters. But to my palate, the difference is very small indeed.

e) Bottom line: all 3 of these roasters do a fantastic job of delivery top quality fresh roasted beans. From a practical perspective, the HIR is awfully good for less than $200. The HT can deliver a somewhat larger roast batch size and delivers just slightly more depth of flavor, perhaps due to the drum roast configuration. But at $600, is it 3x as good as the HIR? And then there is the ability to set up custom roast profiles, which is a feature that only the HIR possesses. It is entirely possible that after several months of fiddling with the roast profiles, I may meet or exceed the roast quality of the HT.

a) Conclusion: HIR trumps all competition as best value for the serious home roasters.

Addendum: Decaf roasting in the HIR

Following the evaluation effort outlined above, I ran a decaf roast batch (130 grams batch size, “pre 2” roast profile). Much to my dismay, the roast seemed to stall at first crack and never developed properly beyond this point…even though I extended the roast time as much as the machine permitted (15 min max total roast time is the limit). I found this curious since the HWP machine always roasts decaf properly using a roast setting ½ to 1 level lower than what works for regular beans. So why would the HIR under-roast decaf at a roast level which works well for regular? Puzzling indeed.

Three observations helped me understand what the mechanism is behind the slow decaf roasting:

1) Suggested roast profiles have the temp set to maximum (485F) for the first roast phase. This made no sense to me, since in the early part of the roasting the intent is to gradually bring the temperature up to a point just below first crack initiation…to allow moisture to smoothly dissipate.

2) The HIR is rated for 150 grams batch size, but my first trial with regular beans at this condition did not roast uniformly. I noticed that the bean movement slowed down dramatically in the final few minutes of the roast process, and subsequently found the chaff collector screen to be 100% plugged.

3) While monitoring the HIR temperature readout, I noticed that the actual temperature does not exceed 320F during the first roast phase regardless of how high the roast profile was set.

So what’s happening? Simple…for regular coffee (not decaf), once the chaff starts to separate during roast phase #2, the chaff collector starts to clog. This in turn chokes down the airflow in the roast pot, which allows the temperature of the air to increase (max KW rating of the heating element is being applied to a lower air volume). This is why the roaster seems to be limited to 320F during the early roast phase for regular, and why it can’t seem to get higher than this at any time for a decaf roast--since the chaff collector never clogs for decaf, the airflow stays high which prevents the air temp from going higher than ~ 320F.

One tip is to roast decaf after a regular batch—not cleaning the chaff collector between batches. This should work but may not be very repeatable—roast results will depend on to what extent the chaff collector is clogged from the previous roast batch. In fact, I now have some concern over repeatability of regular coffee bean roasting…will roast batch-to-batch consistency be a strong function of chaff volume?

My concept to make the HIR work better for decaf is to simulate the chaff collector clogging phenomenon using an aluminum foil modification. NOTE: this modification is done solely at the discretion and risk of the user, please be aware that this modification is neither sanctioned nor condoned by Hearthware and could void your warranty.

Many thanks to Professor Ted Adlam for creating this evaluation