Advice Note

April 10, 2014

Advice on an issue affecting a small number of Hottop roasters fitted with HTRI boards**


Description of the issue:

On a small number of B-2K roasters fitted with HTRI boards** pressing eject can result in the Hottop shutting down (control panel reset) without ejecting the beans or cooling. This has only been observed in the following very specific circumstances when all of the following apply:
    1.    Preheat has been completed.
    2.    The roaster is being externally controlled i.e. connected to a computer via a USB cable.
    3.    The external control program (RoastLogger, Artisan or other control software) has the fan set to 0%
    4.    Eject is pressed when all of the above apply

Cause:

Investigation of the issue has highlighted that the inrush current when the rear and cooling fans are started is close to the tripping current of the Hottop 12V DC power supply. With certain roasters and, in particular, certain fans, the inrush current can be sufficient to trip the power supply which temporarily shuts down the roaster and resets the control panel.

Workaround:

It is recommended that the fan is always set on 30% or above at the end of a roast before pressing eject. In testing this issue is resolved if the external controlling program has the fan set on 25% or above when eject is pressed (as the fans are already spinning this limits the inrush current). Most roasters would always have the fan on towards the end of the roast to clear the increasing smoke as the roast nears completion.

Hottop models affected:

This specific issue has been reported by one B-2K user and, in controlled testing, reproduced on two other roasters when fitted with either B-2K or P-2K control boards. In testing we have not been able to reproduce this behaviour on a Hottop B or P roaster. Whilst the problem has not been reported, or experienced in testing, on roasters other than the B-2K and P-2K, it may occur on some other roasters especially if the fans have a higher inrush current than the average.

As most users would always have the fan on towards the end of the roast they may not have experienced this issue even though their roaster is susceptible to it.

Solution:

As of 1-April-2014, firmware used on the HTRI boards** has been updated to provide a soft start to the fans when eject is pressed. This avoids the problem by greatly reducing the inrush current to the fan at eject. If your HTRI board** was shipped to you after 1-April-2014 then it already has the updated firmware.

A plug-in replacement 20-pin CPU chip with the revised firmware is available to any user of an HTRI board** who has experienced this problem on their Hottop roaster, or is concerned that their roasting practices may make their roaster susceptible to this problem. Please contact MLG Properties, LLC or GreenBean to arrange for a free CPU replacement chip (you may be asked to pay postage).

Please note that the only new feature included in the updated CPU chip is the soft starting of the fan upon pressing the eject button.

Further advice based on recent report from a user:

There has been one anecdotal report of similar fan-related roaster resets occurring mid-roast when the user made a step change increase in fan output level from 0% to 40%.  In this case the user was not using RoastLogger but there is no reason to believe the outcome would have been different if he had been using RoastLogger.  The user reports observing this on at least 3 occasions, but resolved the issue by always starting the fan at a low output level at the end of preheat.

This seems to be a rare occurrence affecting a small number of roasters.  Based on this report, users should avoid taking the fan from a stopped condition (i.e. 0% output level) to more than 25% output in a single step at any time during a roast, even though limited controlled testing has failed to show that this practice causes shutdowns on several sampled roasters.

It is strongly recommended that users controlling a Hottop via RoastLogger update RoastLogger to version 2.1.0 8/4/2014 or later.  This update provides a new Arduino Controller menu option to provide a soft start to the fan.  This option is on by default and limits increases in fan settings to a maximum of 25% per second until the fan is on 50% or more when any increase is accepted.   Users should leave this fan soft start menu option selected unless they are certain their Hottop is not affected by the issues described above.

The RoastLoggerTC4 sketch has also been updated to version 3.0 by JimG to provide a similar soft start to the fans.  It is strongly recommended that users who control their Hottop via an LCDAPTER rather than RoastLogger update to version 3.0 of this sketch.

The fan soft start provided by the RoastLogger and separately by the RoastLoggerTC4 sketch can coexist without problems so any users who wish can update both the RoastLogger to 2.1.0 and the RoastLoggerTC4 sketch to version 3.0.

Need to update both the HTRI board** firmware and RoastLogger to version 2.1.0:

The two issues described in this action note both relate to step increases in fan setting causing overload/shutdown of the Hottop 12 volt DC power supply.  Unfortunately, to minimise the chance of these affecting a roaster, it is necessary to implement both solutions discussed.

The HTRI board** firmware revision, distributed on a replacement 20-pin chip, described above addresses only the conditions at eject. It is not possible for the HTRI** firmware to provide soft start of the fan before eject. 

The update to the RoastLogger to version 2.1.0 and/or the update of the RoastLoggerTC4 sketch to version 3.0, to provide a soft start of the fans, can only provide the soft start during a roast before eject is pressed.  The RoastLogger is no longer in control after eject is pressed.


** HTRI boards can be HTC+TC4C, HTShield+TC4 or RL-HT-CTRL

This advice note is issued jointly by MLG Properties, LLC (supplier of TC4C+HTC and TC4+HTShield boards) and GreenBean (supplier of RL-HT-CTRL boards).