When roasting coffee I prefer it to be fully developed, without being burned or bitter.  The profile graph in Image 1 is the method I use to do it.  The Caffé Rosto Coffee Roaster I bought, does not have a method to easily vary the applied heat to control the temperature.  So I modified it.  See Image 2, original staight line profile graph.  The idea to implement this profile was obtained from the forums at www.coffeegeek.com.  Specifically as put forth by forum moderator another_jim, aka... Jim Schulman of Chicago IL.  Many thanks Jim.
www.coffeegeek.com/forums/coffee/homeroast/8492.

The reason I purchased a Caffé Rosto was I like the ruggedness of it.  A cousin of mine had one and he brought it to me because it had quit heating.  I opened it up and said WOW, this is built good.  Supposedly he had been roasting back -to- back several roasts at a time, blocking the air flow and adding time to the timer (experimenting) and all I found wrong with it was a blown Thermal Cut-Out, ($1.43 plus tax at local industrial electronics store stocking TCO NTE 8242 at 242ºC or 464ºF).  I was not able to read the code on the thermal cut-out for it was severely discolored.  But later discovered it was supposed to be at 228ºC or 442.4ºF. Suggest if yours goes out, you use part # NTE 8230, to keep it near as engineered.

When first receiving my Caffé Rosto it would not roast to first crack.  Period.  It would not get hot enough to remove the chaf.  Did not put a thermometer in it, because whats the use, not hot enough to remove chaf and this is after adding an extra 13 minutes of full timer.  From just enough beans to prevent them being blown into the chaf collection chamber ( less than 4 oz ) to six ounces ( 6 oz ) by weight and in one-half ounce ( 1/2 oz ) increments between, = NOT Roasting.  Yes, wasted several batches of beans testing, but if it had not been for lack of performance I would not have become interested in putting some variation in the roast time to develop the taste better.  Would have continued on as previously with a Poppery II.  about the same as Image 2.

My "preferred taste" in a profile from my Caffé Rosto is to heat as fast possible to around 350ºF - 380ºF, reduce the voltage applied to the heater which slows the time to "before the end" of the first crack, usually around 420 to 425  410 to 415, and then apply full voltage to force the heat to rise - race as fast as possible toward the second crack.

Update: The width of the "gray area" at the start is variation of ambient temperature, about 45ºF to 85ºF.  I roast outside, covered patio, and relative humidity also causes some variation of total heating time.  Also, the amount of green beans has a large effect on the roasting time.

Exactly where in relation to the second crack the roast is stopped is debatable.  I am convinced when "divots - chards - chips" of the outer skin become apparent (visible) the roast is well into the second crack, how much I don't know nor am I interested, I know it is too dark for me.  When I've tested a new batch of beans to first "chips - divots", which is too strong (dark) for my taste, I record the temperature there, and "tune to five degrees less" on the next roast.  This usually produces very close to my "desired taste."  Note:  Not all beans will chip - divot before becoming too dark for me, in that case I take the end temperature to 455  450ºF and on the next roast tune up or down in 2 -to- 3ºF increments from there to find my desired end of roast setpoint.  If "grassy" taste go hotter, if bitter then not so hot next roast, etc... Still refining this and find African - Indonesia - Micronesia beans are different from Latin - South America type beans.  Remember, this is where it's getting into specific individual taste and that's the reason for home roasting.

Update: Usually find my desired end setpoint within three roasts.  Have yet to exceed four roasts to find end setpoint on any single type bean, it works for me.  To me each type of bean has some difference in characteristics.  For the very first trial of a new type bean, strongly suggest a minimum order size of two pounds (2#'s) if your Rosto is about a four and a half ounce by weight (4.5oz.) best capacity.  Mine will roast more than that amount, but it seriously decreases the brightness if exceeding 5oz. by weight.  A balance between bean capacity and heating capability must be used in my Rosto to achieve the best roast.  A match is necessary.  If you are new to home roasting or changing roasting machines, please be prepared and expect to experiment some with weight - temperature - time.

Had read in various forums the CR120 had a thermostat problem, so e-mailed Steve at www.brightway.com and he sent me a new heater element package within five days.  Then could obtain almost a Medium City Roast, but, was not able to obtain a Full City Roast nor a French Roast or darker.  That's O.K. with me, because I don't like bitter burned coffee.

While waiting for the new thermostat - heater assembly package to arrive, the thought occurred to me why not by-pass (hot wire) the thermostat. It has a thermal cut-out, so should not be a hazard of getting too hot if the air flow is not interrupted.  Shorted out the thermostat and it would then burn the coffee beans if left on too long.  By using a short time of about six minutes like what I had been getting from a Poppery II it produced a very good roast, but not developed long enough to bring out the full taste and smoothness I desire.  See Image 2 above.

Then I decided to see just how much I could modify it in various ways.  Pictures of the various modification steps begin with Image 8 below.  Initial MOD used a "dial thermometer."  Updated (mod the mod) to use a "digital thermometer with a K thermocouple input."

Update: Added a heat shield of "glass - mica wafer" trimmed to fit between heater element coil and thermostat ceramic body to block radiant heat.  It will now heat to at least 460ºF, but stopped there as that is much too dark for me.

CAUTION:  These pages are about the Caffé Rosto Model CR120 and are more for talking about the inherent characteristics of this machine, what and how I modified it to alter certain characteristics of the roast taste.  Unless another temperature measurement method is an exact duplicate of the one used here, expect variation in temperature readings.

Makes one of the best tasting cups of coffee I have ever had the pleasure of experiencing.  When I want a little variation, it's just a small matter of using different types of beans.  Beats "Premium - Gourmet Roasts."  Provided I roast manipulating the temperature over time for a "controlled profile" instead of accepting the "timer produced profile."  Hitting the "brightspot" - "sweetspot" is what it's all about.

WARNING!: Never By-pass a Thermal Cut-Out nor a Fuse nor a Circuit Breaker.

Always monitor closely to adjust or turn-off when any thermostat is by-passed.

Update:  Mike McGinness has placed onLine some great pictures and tips for Variac heater voltage control of the Caffé Rosto.  He shows how to access the main wiring which I am not going to duplicate.  His method provides more overall control for wide variations in line voltage.  Since my home line voltage is over 120 VAC, I only want to slow the roast time.  View Mike's modification in a new window.

Update:  The "gang of resistors and diodes" may be substituted with one single resistor of one-half ohm (0.5Ω) at fifty watts (50 WATTS) dissipation.  Electronic parts store #: NTE-50WMD50 Aluminum Housed Resistor, usually for less than six bucks.  Also available, one-tenth ohm (0.1Ω) NTE-50WMD10 if you want to add some more resistance in series to extend roast time.

When reading the instruction manual which came with the unit, it had a caution about the electrical cord, what I call a lamp cord (2-wire 16AWG = cheap).  Replaced this with a 3-wire 14AWG and grounded the bottom metal frame support and added a connecting ground wire to the bottom of the stainless steel roasting chamber.

I don't like the thought of electrical insulation breaking down and shorting live wires to exposed metal parts that could be touched with the unit plugged in.  Also, on mine, the motor has its own thermal cut-out, but it is mounted in the neutral (low) line side electrical supply to the motor, which means the fuse action of the thermal cut-out could function and open the motor circuit, but not remove electricity from the motor windings.  The unit is mounted on a plastic leg like support with rubber anti-scratch pads, so don't see how it could trip a circuit breaker if certain types of shorts should occur.  With over 40 years electro-mechanical maintenance experience, believe me, shorts occur where they say it is not possible.

Next was how to insert a thermometer.  The glass lid has a screw through it holding on the knob.  Removed the screw-knob, and there is at least a 3/8 inch diameter hole.  Drilled a 1/8 inch diameter hole in plastic top that goes over the chaf collection chamber, and installed the knob there.  If you try this, make sure you use washers with outside diameter at least twice the diameter across the screw head, more is better = fender washer.  Do not want to put concentrated pressure on the plastic to cause cracking - breaking.

Started using a candy - jelly dial thermometer with a 1/4 inch diameter barrel with the pan-side-holder mounted upside down to cock at an angle to put the tip about 1/4 inch into the green beans.  Used some plumbers TEFLON tape held under stretched tension while wrapping around the barrel to fill the hole size difference for no air leak and no chaf escape.  Don't know exactly how accurate everything is, but it produces absolutely great roasted coffee.
Revision:
Used the barrel - stem - tube of the old dial thermometer to hold the thermocouple probe.  Ripped the dial off and punched out the guts.  Squeezed the tip of the barrel - stem - tube down and trimmed corners - edges to allow only the "bead" of the thermocouple to protrude half-way.

To reduce the heating current - voltage I did not want to spend about a hundred bucks or more (>$100.00) for a VARIAC so I decided to use a dropping resister, which I calculated "SWAG" to be about one ohm with a hundred watts dissipation (1Ω @ 100W).  Could not find nor buy one locally.

Experimented it to 4 resistors in parallel at 5.1Ω @ 10 Watts each with two six ampere rectifier-diodes swamping one swing of the voltage, all in parallel with a switch to include the added components if switch is open and components to be by-passed if the switch is closed.
Revision:
Added a fifth resistor and the current drop improved from 91% to 95.5% at 400ºF  of 10.2amps, 10.7amps, 11.2amps.

Using Four Resistors it did require constant attention for 10 to 15 minutes while roasting, monitoring temperature - Flipping the SPST Switch Open - Closed.  Using Five Resistors it's only two flips, reduce the current/voltage by opening the switch at about 350ºF  and return to full current/voltage by closing the switch at toward the end of the first crack.  Until I get a TRIAC circuit with a percentage - pulse width type modulation with zero cross-over firing, it will suffice.  It's very tough flipping a 15 amp switch, I'll probably end up with muscles on my pinky finger in a few years.

I intend to buy a black plastic box to install the resistors and rectifier diodes in to give it a somewhat more appealing appearance.  The aluminum shell box is ugly in my opinion, but it was handy.

As long as the motor lasts, everything else is very minor in cost to replace.  The motor is of much - much better quality than found in hot air pop corn poppers, the fan blade screws on, instead of being pressed on the shaft.  To me the original internal components are all of very high quality.  The whole machine is of very good components.

The mis-placement of the thermostat seems to be the determining factor from machine to machine as to whether it will be a good one or a bad one.  By viewing numerous discussion/forum posts it is apparent the first versions of the CR120 have a wide variation in heating capability over a specified time.  Also, numerous post responses and advice from www.brightway.com seem to put too much spin on voltage variability from home user to home user.  My suspicion still points to how much of the heater coil is exposed to the thermostat ceramic housing, how tight the heater coils are wound in that area = more surface area, with resulting distance between the heater coil and thermostat ceramic body being a great big unknown.  I am experimenting with a glass-mica wafer shield on a new heater element package, have got it completed, but have not tested it in the machine yet, ( as of 11:00 AM EST 07April2003 ).  UpDate:  02JUNE2003;  Now works very good to a measured temperature of 460ºF, which is too dark for me, but had to test it to confirm one of my suspicions.  Remove the shield and it will not heat to 450ºF measured before thermostat begins cycling.  Reinstall the shield and it heats to at least 460ºF and the thermostat has not begun to cycle.  Would like to test on at least ten different machines for widespread data validation, but don't have them.  Since there is almost exact repeatability on my single machine, I'm satisfied I'm on the right track.

All Caffé Rosto's CR120 Model, early versions, should have a very wide degree of heating variance, due to the ceramic body of the thermostat housing support being directly exposed to and in close proximity to the heating wire coil.  A big radiant heat absorber.  Would think the thermostat would be shielded by the glass-mica wafer from the direct heat of the resistive wire coil.

Now that I've got mine working in such a great manner I can not see me ever using anything else.  Less than ten dollars (<$10.00) of purchased parts to do the modification.  To:  My Take on Brewing Superb Coffee.

NOTE 1:  The resistor values may not apply if your supply voltage is different from mine of 121 VAC.  My heater coil measured 9.7Ω COLD, which will change if the NiChrome wire heats a lot, a characteristic, and the four resistors equivalent to 1.275Ω in series with the heater coil when the switch is open.  Add a fifth resistor in parallel and it becomes 1.02Ω which changes the voltage divider ratio somewhat.  At 79 cents per resistor I can do some experimenting.  I do not believe three resistors in parallel of ten watts each will be able to handle the heat-power dissipation, so go to a higher wattage value.  Don't burn down your house to roast coffee if you try some of this, be careful & safe.  I know the pursuit of the best roast is important, but don't let it ruin your life.

       WARNING!: Never By-pass a Thermal Cut-Out nor a Fuse nor a Circuit Breaker.
     Always monitor closely to adjust when any thermostat is by-passed.

NOTE 2:  My household voltage measures a constant 121 VAC, if your supply voltage is less than 115 VAC it may be necessary to use a VARIAC to raise the line voltage.
      View Mike McGinness VARIAC modification in a new window.

NOTE 3:  Expect my profile graphs to change when a new Digital "K" Thermocouple Thermometer arrives.  I almost always use my judgment of sight, sound, and smell for ending the roast.  The use of a thermometer and timer allows me to vary the profile in experiments and duplicate it if I like that type of roast.

NOTE 4:  Since I've bad-mouthed the thermostat placement, how about one about the wire color coding.  The Black Wire of the Heater Assembly is connected to the supply voltage (115VAC) neutral - low side with a wire-nut and one motor lead wire.  The White Wire with the female quick disconnect connects to the Timer High Current Terminal, 115VAC supply high -or- hot side.  This is bassackwards to USA normal wiring practices for 115VAC devices.  It in no way affects the operation of the machine.  It is a great big indicator of design - manufacturing - marketing items pertaining to overall quality.  Whoever had the the original concept design, did a brillant job.  Somewhere along the way cost cutting measures (marketing) seem to have more impact.

      For example, the metal base plate of my unit has a hole in it with a stamped in grounding symbol.  Not wired.  Also, the heater assembly metal chamber, a flared duct, has a tab on it with it a hole drilled into it, but no wire.  The stainless steel roasting chamber is connected to the heater assembly metal duct-chamber by three safety screws, (TORX T-10), but fully isolated from "ground" and could be shorted to 115VAC through insulation breakdown.  Signs - indicators of good to great design, but failure to implement fully.

      Another thing that bugs me is the very short amount of lead wire on the epoxy ( cone ) end of the Thermal Cut-Out.  I've replaced one in a machine owned by my cousin and it sure would have been nice to be able to cut the lead wire at the apex of the epoxy cone instead of cutting through the epoxy to have enough lead wire to a crimp-ferrule over.