USB Hot Top Teaser

Roasters and roasting

Moderators: GreenBean, Gouezeri, bruceb, CakeBoy

Postby mnemonix » Sun Aug 24, 2008 10:19 pm

For the plug & play option, definitely pwm as it can be done in software with no additional hardware beyond the usb interface and the Hot Top board. In the light of Jons information I'll be persuing this first whatever the final outcome as I can have a test rig up and running in next to no time.

My Hot Top is one of the early models, pre digital display of any kind, though I'm toying with the idea of picking up a second one for the project.
mnemonix
 
Posts: 176
Joined: Fri Nov 14, 2003 11:03 pm
Location: Nottingham

Postby zix » Mon Aug 25, 2008 5:43 am

the load we're talking about is < 800W and mainly resistive

a-ha!
Wasn't aware that the HT only used 800W. That is no more than those BBQ heater forks.
‹• Bezzera B3000AL • Strietman ES3 • Chemex • Cona C size • Aeropress • Vev moka • Bialetti Brikka • Espro • Cezve • Bacchi Espresso • Arrarex Caravel •
• HG-1 • Lido 1 & E-T • OE Pharos •
• oven • hot air gun • Behmor •›
User avatar
zix
 
Posts: 1331
Joined: Wed Mar 24, 2004 10:40 pm
Location: Partille/Göteborg, Sweden

Postby GreenBean » Wed Aug 27, 2008 4:05 pm

I am also very interested in this project. I can not understand why they have not provided a usb interface to the Hottop as standard. It would be great both to allow automatic logging of the roast progress and to allow driving the roaster to a given profile.

Whilst I think that human intervention will always be required to achieve excellent results it would be nice if the hardware could deal with the boring stuff. The user could then be left with decisions such as when to override the timing of turning heater power down or up during/after first crack.
Image

Izzo Alex Duetto | Gaggia XD 2 Group | Mazzer Super Jolly | La Cimbali Max | Solis 166 | Dalian 1 kg roaster | Hottop P | Hottop B | French Press (several) | Kettle modded, no really, added digital thermometer |
User avatar
GreenBean
 
Posts: 2215
Joined: Wed Jan 03, 2007 2:15 pm
Location: Chester

update

Postby mnemonix » Sun Sep 07, 2008 1:03 pm

A quick update, work and the need for a new car got in the way of progress somewhat, but I have a working board and code now... just need to test it in situ.

The photo shows my old Hot Top board for comparison and the new USB board next to it. I'm using a couple of AD595 chips for my thermocouple inputs, and an ultra compact implementation of the ATMEGA168 risc chip + FTDI usb controller chip piggybacked on to my prototype board.

(Thanks to Jon for pointing me at the Arduino platform it made getting this off the ground so quick, though ultimately it could be built much cheaper with a diy pcb and barebones ATMEGA168 + usb controller, I just didn't fancy the SMD soldering required for the FTDI chip).

Temperature measurements are oversampled about 5000x a second and seem to be stable to 0.1 degrees centrigrade. I've coded the board to handle pwm for the heater over a 1 second timebase - any shorter would have introduced too large of an error without zero cross detection which my early model Hot Top doesn't have.

Currently I've coded the beginnings of the pc side software to give full manual control over the roaster while displaying the temperature readings and time. Next step is real-time graphing and 2 non manual modes of operation - user defined heater/fan settings over time & user defined profile settings using a PID algorithm to run it.

On the hardware side, it also has the potential to have a profile loaded from the pc and then to run in a stand alone mode for which I'd like incorporate to an LCD display.

More to come soon I hope!
Attachments
P9070764.jpg
P9070764.jpg (128.58 KiB) Viewed 11179 times
mnemonix
 
Posts: 176
Joined: Fri Nov 14, 2003 11:03 pm
Location: Nottingham

RE: update

Postby JonS » Sun Sep 07, 2008 7:27 pm

Way to go, Chris, this is very cool. Glad I was able to help :)
Iberital L'Adri 1grp semi-auto | Macap MXKR Auto | Mazzer Super Jolly | Hottop Programmable
Isomac Zaffiro w/PID | Isomac Gran Macinino | Mazzer Mini
Scace Thermofilter
Custom-built grinder timer
JonS
 
Posts: 488
Joined: Wed May 09, 2007 12:18 pm
Location: Chesterfield

RE: update

Postby zix » Sun Sep 07, 2008 8:19 pm

Good job, Chris! Very very quick.
work and the need for a new car got in the way of progress somewhat

Heh heh... this took you around ten days, yes? If that was me it would have been more like ten years.
‹• Bezzera B3000AL • Strietman ES3 • Chemex • Cona C size • Aeropress • Vev moka • Bialetti Brikka • Espro • Cezve • Bacchi Espresso • Arrarex Caravel •
• HG-1 • Lido 1 & E-T • OE Pharos •
• oven • hot air gun • Behmor •›
User avatar
zix
 
Posts: 1331
Joined: Wed Mar 24, 2004 10:40 pm
Location: Partille/Göteborg, Sweden

RE: update

Postby JonS » Mon Sep 08, 2008 3:48 pm

One thing I neglected to mention when I posted the details of a reverse engineer of the control signals to the hottop board, is that whilst the PWM for the heater is limited to timebase periods that make sense for (ie. are long enough to work with) 50 / 60Hz mains with no zero-cross detection, the control signal for the fan is switching regulated DC, so there are no such restrictions on the timebase. It's probably obvious, but I realised I'd not said it. Chris, shout if you would like me to 'scope the PWM signal for the fan on my "P" model to get the numbers it's using. If you haven't worked out some good ones for yourself yet, that is :)
Iberital L'Adri 1grp semi-auto | Macap MXKR Auto | Mazzer Super Jolly | Hottop Programmable
Isomac Zaffiro w/PID | Isomac Gran Macinino | Mazzer Mini
Scace Thermofilter
Custom-built grinder timer
JonS
 
Posts: 488
Joined: Wed May 09, 2007 12:18 pm
Location: Chesterfield

Postby GreenBean » Tue Sep 09, 2008 10:00 am

Very impressive, keep up the good work. 8)
Image

Izzo Alex Duetto | Gaggia XD 2 Group | Mazzer Super Jolly | La Cimbali Max | Solis 166 | Dalian 1 kg roaster | Hottop P | Hottop B | French Press (several) | Kettle modded, no really, added digital thermometer |
User avatar
GreenBean
 
Posts: 2215
Joined: Wed Jan 03, 2007 2:15 pm
Location: Chester

Postby EricC » Fri Oct 03, 2008 11:36 am

JonS wrote:or even phase angle modulation rather than PWM control of the heater for more precision and response.

Jon


Hi Jon,

Have a look at the bottom of this thread, you were absolutely spot on. 8)

http://www.home-barista.com/forums/hott ... t7820.html

All the Best
Eric
KvdW Speedster, Versalab M3/4, Kafatek Monolith MAX
User avatar
EricC
 
Posts: 692
Joined: Fri Feb 06, 2004 3:31 pm
Location: Liverpool

Postby JonS » Fri Oct 03, 2008 12:32 pm

Ah, thanks for that, Eric :D

"Interestingly" (as in, interesting, but I know there are those who will say it's not strictly coffee) since I originally posted those thoughts, I've been playing with the circuitry in my "P" model hottop. I figured it would be useful to have a proper handle on the differences between these newer boards and the original ones.

Anyway, what I believe I discovered recently, with pretty much 100% certainty, is that my "P" model hottop does not and cannot use Phase Angle control ! It can only use Phase Width Modulation which is synced with the AC mains zero cross for maximum possible accuracy for that method.

The post on HB seems to imply that there's a difference either between "P" and "B" or US / UK hottops that would explain the different findings. I've posted my comments there to see what comes back. I'm pretty sure I was told the internal boards were identical for "P" and "B".

As zix pointed out earlier, PAM / Phase control can generate electrical interference from the switching, so perhaps EU regs are stricter on this ?

Swapping one chip on the internal board would allow my model to do Phase Control, but I'm not convinced any potential greater control that might come with this outweighs the possibility for interference, when given PWM that is accurately locked to the mains cycle, so you can eg. run 100ms PWM periods to get 10% granularity heater control (n out of 10 mains half-cycles)

While I'm posting on obvious differences between the models, I discovered that on the "P" model at least, when compared with the original hottop electronics, the connections run the opposite way on the connector, and require driving to 0v to activate their functions, not 5v like the old one.

Oh, and I also discovered exactly how much interference is generated inside a hottop when I looked at all the signal crap coming out of my thermocouple. Hats off to Fluke, their 51II can really filter a signal to give a clean temperature reading, especially when compared with what the raw signal looks like on a 'scope. The highlight was turning on the hottop's drum motor, and seeing my own temperature readings plummet by something like 18 degC due to induced voltages on the thermocouple...
Iberital L'Adri 1grp semi-auto | Macap MXKR Auto | Mazzer Super Jolly | Hottop Programmable
Isomac Zaffiro w/PID | Isomac Gran Macinino | Mazzer Mini
Scace Thermofilter
Custom-built grinder timer
JonS
 
Posts: 488
Joined: Wed May 09, 2007 12:18 pm
Location: Chesterfield

Postby EricC » Fri Oct 03, 2008 1:40 pm

JonS wrote: Hats off to Fluke, their 51II can really filter a signal to give a clean temperature reading.


No problem.
Good job I have the Fluke 54II :wink:
Very interesting findings there Jon, maybe the EU does have stricter sandards.
KvdW Speedster, Versalab M3/4, Kafatek Monolith MAX
User avatar
EricC
 
Posts: 692
Joined: Fri Feb 06, 2004 3:31 pm
Location: Liverpool

Postby JonS » Fri Oct 03, 2008 1:48 pm

Good job I have the Fluke 54II


Ah, lovely, would quite like one myself. They really are excellent, if perhaps off-puttingly expensive (though not to a man who makes his coffee using a Versalab and a GS3, clearly) 8)
Iberital L'Adri 1grp semi-auto | Macap MXKR Auto | Mazzer Super Jolly | Hottop Programmable
Isomac Zaffiro w/PID | Isomac Gran Macinino | Mazzer Mini
Scace Thermofilter
Custom-built grinder timer
JonS
 
Posts: 488
Joined: Wed May 09, 2007 12:18 pm
Location: Chesterfield

Postby EricC » Fri Oct 03, 2008 1:59 pm

JonS wrote:, if perhaps off-puttingly expensive (though not to a man who makes his coffee using a Versalab and a GS3, clearly) 8)


And after buying all of this kit I am as poor as a church mouse. :lol: :lol: :lol:
KvdW Speedster, Versalab M3/4, Kafatek Monolith MAX
User avatar
EricC
 
Posts: 692
Joined: Fri Feb 06, 2004 3:31 pm
Location: Liverpool

Postby GreenBean » Wed Nov 12, 2008 10:27 am

I have been greatly inspired by the great work of mnemonix and JonS on this project and decided to do some further exploration on my Hottop P model. JonS pointed to some inconsistencies between his findings and those reported on HB. I have added my findings below in the hope that this helps us move towards resolving these inconsistencies.

JonS wrote:Anyway, what I believe I discovered recently, with pretty much 100% certainty, is that my "P" model hottop does not and cannot use Phase Angle control ! It can only use Phase Width Modulation which is synced with the AC mains zero cross for maximum possible accuracy for that method.

I connected an oscilloscope across the heater and can confirm that there is no sign whatsoever of phase angle control. I did not see any sign of phase width modulation either. On my machine at least the heater power remained on (full power, full sine wave) until the set temperature for a segment was reached and then was turned off for the remainder of the segment. I guess that it would turn on again if the temperature dropped below the set temperature but that did not happen whilst I was testing. I repeated this test watching the heater power pin on the connector (thanks for the connector details JonS) with, as you would expect the same result. Anyone can check this for themselves as I found that there are two LED’s on the power board, one at the back of the board and one in the middle. I found that the one in the middle follows the call for heat signal.

JonS wrote:The post on HB seems to imply that there's a difference either between "P" and "B" or US / UK hottops that would explain the different findings. I've posted my comments there to see what comes back. I'm pretty sure I was told the internal boards were identical for "P" and "B".

I checked on the hottopusa.com site and they list an upgrade from the P model to the B model and state that only the control board needs changing (implying that the power board is identical between the B and P Models. I am at a loss to explain the mismatch of findings between the HB site and ourselves.

Armed with the above information I have added phase width modulation to my machine allowing switching from the standard control and PWM control. I have also added an LED to the case to indicate when heat is being called for so that I can easily see what is happening. I have arranged it so that PWM control is only available when the control board requests heat so the original safeguards still apply. So far I have used it in standard mode until the start of first crack and then switching to PWM to lower the heat input and stretch the time to second crack. It seems to work very well and is much less frustrating than turning the set temperature for the current segment up/down to adjust the heat input.
Last edited by GreenBean on Wed Nov 12, 2008 1:32 pm, edited 1 time in total.
Image

Izzo Alex Duetto | Gaggia XD 2 Group | Mazzer Super Jolly | La Cimbali Max | Solis 166 | Dalian 1 kg roaster | Hottop P | Hottop B | French Press (several) | Kettle modded, no really, added digital thermometer |
User avatar
GreenBean
 
Posts: 2215
Joined: Wed Jan 03, 2007 2:15 pm
Location: Chester

Postby JonS » Wed Nov 12, 2008 10:43 am

Good call greenbean, and good to know I'm not going mad. Nice work on the PWM mod, I like it. Still really puzzling about the "differences" between the P & B models and the implications of swapping between them.

I know this topic has otherwise gone a bit quiet. I haven't spoken to Chris (mnemonix) about it for a while. My own efforts on the investigation into the "P" version of the project stalled for a while and I haven't got back to them yet. The main reason they stalled, is, due to the amount of electrical interference generated by the various bits of the hottop, and my use of a grounded thermocouple probe for bean temperature, I ended up with some unpleasant effects on the thermocouple amplification. Eventually, I remedied this by isolating the 3 significant sections of the circuit (USB, Microcontroller + thermocouple amp, Hottop output) so that their power supplies were separate and could float wrt. each other, but I then ended up with a lot less time on my hands to do any more work.

Though basically, I am comfortable I've largely cracked the electronics side of things.

On a slightly related note, I eventually found the correct connectors as used on the circuit boards and cables of the hottops, which I'd previously struggled to source, which has saved me irreversibly modifying anything yet :wink:

Jon
Iberital L'Adri 1grp semi-auto | Macap MXKR Auto | Mazzer Super Jolly | Hottop Programmable
Isomac Zaffiro w/PID | Isomac Gran Macinino | Mazzer Mini
Scace Thermofilter
Custom-built grinder timer
JonS
 
Posts: 488
Joined: Wed May 09, 2007 12:18 pm
Location: Chesterfield

Previous

Return to Roasting - Equipment and Techniques

Who is online

Users browsing this forum: No registered users and 22 guests