coffee-forum.net
Promoting coffee discussion.

Main
Date: 24 Jan 2007 18:06:18
From: Mike Hartigan
Subject: PID controlled roaster?
Just a thought - has anybody considered building a roaster around a
programmable PID? Something like this, for example:

http://tinyurl.com/yuqcjo

It's programmable in 30 steps (vs iRoast's 5) and is accurate to
within 0.2% over its entire range. It seems like it would be a
trivial matter to mod virtually any fluid bed, or even drum roaster
with this PID, an SSR, a fuse, a pair of power switches, a box, and
*maybe* a variac. Simply connect the motor directly to the AC and
use this to control the heater coil. The thermocouple could be
placed anywhere you damn well please and the result would be an
incredibly repeatable profile.

It almost seems too easy - what am I missing?

--
-Mike




 
Date: 30 Jan 2007 02:08:34
From: daveb
Subject: Re: PID controlled roaster?
arguing just for the sake of arguing.





  
Date: 31 Jan 2007 05:32:22
From: Barry Jarrett
Subject: Re: PID controlled roaster?
On 30 Jan 2007 02:08:34 -0800, "daveb" <davebobblane@gmail.com > wrote:

>arguing just for the sake of arguing.
>
>

commenting for the sake of commenting?




 
Date: 29 Jan 2007 05:59:01
From: alt.coffee
Subject: Re: PID controlled roaster?
I have modified my I-roast2 some days before. Firstly It worked good=20
with the pid and SSR . Soon I found the heating-wire cycled to hot and=20
cold quickly (for its low heat inertia) while the fan speed was also=20
waved, though the temperture was stable. I afraid that the heating=20
wire would be dead someday because of so many cycles. One day it came=20
true, the heating wire was open. After I reconnected it , I made some=20
improvment. I changed the output of the pid. Now the pid did not=20
control the SSR directly but the phase of the current.It is easy=20
understood.If you want to control the car speed ,you would change the=20
percent of the throtle but not start or shutdown the engine. I=20
observed the output of the pid would be 50 percent nearly when the=20
temperture was stable .

On 1=D4=C229=C8=D5, =CF=C2=CE=E78=CA=B134=B7=D6, "Dan Bollinger" <danNOboll=
in...@insightSPAMbb.com >=20
wrote:
> >> Dan
>
> > Uh, I don't think you wanted to make that a reply to my message.Uh, Yes=
I did wanted to. ;)



  
Date: 29 Jan 2007 16:21:50
From: I->Ian
Subject: Re: PID controlled roaster?
On 29 Jan 2007 05:59:01 -0800, "alt.coffee" <jiaju.wu@gmail.com >
wrote:

>I have modified my I-roast2 some days before. Firstly It worked good
>with the pid and SSR . Soon I found the heating-wire cycled to hot and
>cold quickly (for its low heat inertia) while the fan speed was also
>waved, though the temperture was stable. I afraid that the heating
>wire would be dead someday because of so many cycles. One day it came
>true, the heating wire was open. After I reconnected it , I made some
>improvment. I changed the output of the pid. Now the pid did not
>control the SSR directly but the phase of the current.It is easy
>understood.If you want to control the car speed ,you would change the
>percent of the throtle but not start or shutdown the engine. I
>observed the output of the pid would be 50 percent nearly when the
>temperture was stable .

Did you run the fan at a constant speed?

--
Posted via a free Usenet account from http://www.teranews.com



  
Date: 29 Jan 2007 14:25:51
From: JimG
Subject: Re: PID controlled roaster?
On 29 Jan 2007 05:59:01 -0800, "alt.coffee" <jiaju.wu@gmail.com >
wrote:

>I have modified my I-roast2 some days before. Firstly It worked good
>with the pid and SSR . Soon I found the heating-wire cycled to hot and
>cold quickly (for its low heat inertia) while the fan speed was also
>waved, though the temperture was stable. I afraid that the heating
>wire would be dead someday because of so many cycles. One day it came
>true, the heating wire was open. After I reconnected it , I made some
>improvment. I changed the output of the pid. Now the pid did not
>control the SSR directly but the phase of the current.It is easy
>understood.If you want to control the car speed ,you would change the
>percent of the throtle but not start or shutdown the engine. I
>observed the output of the pid would be 50 percent nearly when the
>temperture was stable .
>

I did not understand from your previous posts that your concern was
with the life of the heating element. Thanks for clearing that up.

Do you recall the cycle time that resulted in the failure of the
element?

Jim

--
Posted via a free Usenet account from http://www.teranews.com



   
Date: 30 Jan 2007 08:07:43
From: Ian Smith
Subject: Re: PID controlled roaster?
On Mon, 29 Jan 2007 14:25:51 -0500, JimG <jggall01@yahoo.com > wrote:
> On 29 Jan 2007 05:59:01 -0800, "alt.coffee" <jiaju.wu@gmail.com>
> wrote:
>
> >waved, though the temperture was stable. I afraid that the heating
> >wire would be dead someday because of so many cycles. One day it came
> >true, the heating wire was open.
>
> I did not understand from your previous posts that your concern was
> with the life of the heating element. Thanks for clearing that up.

But, if it was a random switching (as seems likely, given that he
doesn't like zero crossing), you'd expect a shorter heating element
life. I'm not sure this anecdote tells us much, other than "don't use
a zero crossing SSR, because a random switching one makes the element
fail".

regards, Ian SMith
--


    
Date: 30 Jan 2007 07:27:20
From: jggall01
Subject: Re: PID controlled roaster?
On 30 Jan 2007 08:07:43 GMT, Ian Smith <ian@astounding.org.uk > wrote:

>But, if it was a random switching (as seems likely, given that he
>doesn't like zero crossing), you'd expect a shorter heating element
>life. I'm not sure this anecdote tells us much, other than "don't use
>a zero crossing SSR, because a random switching one makes the element
>fail".
>

I think his phase angle control scheme is "on" all the time rather
than pulsed. The power output is continuously varied by the power
controller in response to an analog current signal from the temp
controller.

I would have thought that using a short time base with pulsed output
could accomplish a similar outcome, but I am working on my first
popper/roaster right now. So, still gathering information.

Jim

--
Posted via a free Usenet account from http://www.teranews.com



     
Date: 30 Jan 2007 13:28:18
From: Ian Smith
Subject: Re: PID controlled roaster?
On Tue, 30 Jan 2007 07:27:20 -0500, jggall01 <jggall01@yahoo.com > wrote:
> On 30 Jan 2007 08:07:43 GMT, Ian Smith <ian@astounding.org.uk> wrote:
>
> >But, if it was a random switching (as seems likely, given that he
> >doesn't like zero crossing), you'd expect a shorter heating element
> >life. I'm not sure this anecdote tells us much, other than "don't use
> >a zero crossing SSR, because a random switching one makes the element
> >fail".
> >
>
> I think his phase angle control scheme is "on" all the time rather
> than pulsed.

It is now. The reason being as I said - the first one was a PWM
system using an SSR, which was presumably random switching, and which
burnt out the element. Read the thread.

> I would have thought that using a short time base with pulsed output
> could accomplish a similar outcome,

It does, entirely happily, in my roaster. The only failures I've had
are of the SSR, and putting a chunky snubber across the SSR seems to
have fixed the tendency for it to die.

regards, Ian SMith
--


 
Date: 29 Jan 2007 05:52:49
From: jggall01
Subject: Re: PID controlled roaster?
On Jan 29, 3:44 am, Ian Smith <i...@astounding.org.uk > wrote:
>I use something like http://www.astounding.org.uk/ian/roaster/realroast.png
>

Thanks.

Jim



  
Date: 29 Jan 2007 16:28:32
From: I->Ian
Subject: Re: PID controlled roaster?
On 29 Jan 2007 05:52:49 -0800, "jggall01" <jggall01@yahoo.com > wrote:

>On Jan 29, 3:44 am, Ian Smith <i...@astounding.org.uk> wrote:
>>I use something like http://www.astounding.org.uk/ian/roaster/realroast.png
>>
>
>Thanks.
>
>Jim

There's an interesting thread on Home Barista on Espresso Profiles
http://www.home-barista.com/forums/viewtopic.php?t=3115

--
Posted via a free Usenet account from http://www.teranews.com



   
Date: 30 Jan 2007 07:29:14
From: jggall01
Subject: Re: PID controlled roaster?
On Mon, 29 Jan 2007 16:28:32 -0800, "I- >Ian" <someone@nowhere.com>
wrote:

>There's an interesting thread on Home Barista on Espresso Profiles
>http://www.home-barista.com/forums/viewtopic.php?t=3115

Thanks for the link, Ian. That's good info.

I can see now that a ramp/soak controller is going to be needed for
this. Too many steps to manage with stopwatch and manual SV changes.

Jim

--
Posted via a free Usenet account from http://www.teranews.com



 
Date: 28 Jan 2007 19:17:53
From: RoughJaw
Subject: Re: PID controlled roaster?
On Sat, 27 Jan 2007 11:12:45 -0600, Dave S wrote:

> I don't understand the argument for and against zero-crossing PID
> either, but your "year, whatever" certainly helped clear it up.

Since daveb elected to delete that post, most people won't
realize what you refer to. For the record, daveb posted a
one liner in reply consisting of nothing more than
"yeah, whatever"
to alt.coffee" <jiaju.wu@gmail.com > in message
news:1169868296.892494.326610@v33g2000cwv.googlegroups.com.

--
Posted via a free Usenet account from http://www.teranews.com



  
Date: 30 Jan 2007 23:47:31
From: daveb
Subject: Re: PID controlled roaster?
On Jan 31, 12:32 am, Barry Jarrett <b...@rileys-coffee.com > wrote:
> On 30 Jan 2007 02:08:34 -0800, "daveb" <davebobbl...@gmail.com> wrote:
>
> >arguing just for the sake of arguing.
> >
> >
>
> commenting for the sake of commenting?

uh huh. back at ya sport.



  
Date: 28 Jan 2007 16:59:47
From: Dan Bollinger
Subject: Re: PID controlled roaster?
> Since daveb elected to delete that post, most people won't
> realize what you refer to.

I don't think you can delete posts on bbs. I can still read Dave's post.



Dan


   
Date: 28 Jan 2007 17:35:36
From: Dave S
Subject: Re: PID controlled roaster?
Dan Bollinger wrote:
>> Since daveb elected to delete that post, most people won't realize
>> what you refer to.
>
> I don't think you can delete posts on bbs. I can still read Dave's post.
>
>
>
> Dan

It's missing from groups.google.com, but can still be seen in the
"quoted text" of my reply.

Dave S.


    
Date: 28 Jan 2007 19:56:17
From: Dan Bollinger
Subject: Re: PID controlled roaster?
>>> Since daveb elected to delete that post, most people won't realize what you
>>> refer to.
>>
>> I don't think you can delete posts on bbs. I can still read Dave's post.
>>
>>
>>
>> Dan
>
> It's missing from groups.google.com, but can still be seen in the "quoted
> text" of my reply.
>
> Dave S.

alt.coffee isn't hosted by Google. If you used a news reader you'd find that
messages remain.

Dan



     
Date: 28 Jan 2007 21:40:21
From: Dave S
Subject: Re: PID controlled roaster?
Dan Bollinger wrote:
>>>> Since daveb elected to delete that post, most people won't realize
>>>> what you refer to.
>>>
>>> I don't think you can delete posts on bbs. I can still read Dave's
>>> post.
>>>
>>>
>>>
>>> Dan
>>
>> It's missing from groups.google.com, but can still be seen in the
>> "quoted text" of my reply.
>>
>> Dave S.
>
> alt.coffee isn't hosted by Google. If you used a news reader you'd find
> that messages remain.
>
> Dan

Uh, I don't think you wanted to make that a reply to my message. I never
claimed that Google hosts alt.coffee. However, Google is the way some
people (daveb?) access alt.coffee.

Dave S.


      
Date: 29 Jan 2007 07:34:18
From: Dan Bollinger
Subject: Re: PID controlled roaster?

>> Dan
>
> Uh, I don't think you wanted to make that a reply to my message.

Uh, Yes I did wanted to. ;)






 
Date: 26 Jan 2007 19:24:56
From: alt.coffee
Subject: Re: PID controlled roaster?
I don't think SSR with a zero cross function is suitable with pid.
Though it could minimize the influence of an inductive load,it also
would lead to time delay. It is no use for the heating element which is
used in a purely resistance circuit. SSR with a zero cross function act
as a eletric-switch only. I would recommend the SSR with Random
function.
PID with SSR would work properly on the system with high heating
inertia character such as silvia.But if the system has small heating
inertia ,it would be waved.The solution is control the power rate .



On 1=D4=C227=C8=D5, =C9=CF=CE=E79=CA=B100=B7=D6, "I- >Ian" <some...@nowhere.=
com > wrote:
> On 26 Jan 2007 22:26:19 GMT, Ian Smith <i...@astounding.org.uk> wrote:
>
>
>
>
>
> >On Fri, 26 Jan 2007 17:10:35 -0500, JimG <jggal...@yahoo.com> wrote:
> >> On 26 Jan 2007 12:57:43 -0800, "alt.coffee" <jiaju...@gmail.com>
> >> wrote:
>
> >> >With my experience I would say pid with a ssr is not a good design for
> >> >a roaster. On and off would not keep the temperature accurately.What=
I
> >> >modified my IROAST2 is add a phase ajustable module between the pid(p=
id
> >> >output is current:0~20mA) and ssr.So the pid is not simply control the
> >> >SSr on or off but the power rate output to the heating
> >> >element.Temperture can stay at the point very stablely.
>
> >> Why doesn't a PID set with a short time base (say 1 second or less)
> >> accomplish the same thing using a standard zero-crossing SSR?
>
> >It does.
>
> >regards, Ian SMithAn SSR with a zero cross function operates when an A=
C load
> voltage reaches the zero point or its vicinity. This reduces
> clicking noises when the load is input, and minimizes the
> influence of an inductive load, such as a lamp, heater, or
> motor, on the power supply because the inrush current of the
> load is reduced. This can also minimize the scale of the surge
> current protection circuit.
>
> A device with a short cycle time can turn on at any time during the
> cycle.
>
> The two are unrelated.
>
> --
> Posted via a free Usenet account fromhttp://www.teranews.com- =D2=FE=B2=
=D8=B1=BB=D2=FD=D3=C3=CE=C4=D7=D6 -- =CF=D4=CA=BE=D2=FD=D3=C3=B5=C4=CE=C4=
=D7=D6 -



  
Date: 27 Jan 2007 19:41:32
From: Ian Smith
Subject: Re: PID controlled roaster?
On 26 Jan 2007, alt.coffee <jiaju.wu@gmail.com > wrote:

> I don't think SSR with a zero cross function is suitable with pid.

Of course it is. It results in (at most) just under a 0.01 second
delay (half a cycle of the AC). The fact that you have I means that
this will automatically self-correct as soon as it becomes detectable.

> It is no use for the heating element which is
> used in a purely resistance circuit.

It is significant use, because it reduces in rush current.

regards, Ian SMith
--


  
Date: 27 Jan 2007 11:02:12
From: daveb
Subject: Re: PID controlled roaster?

yeah, whatever


"alt.coffee" <jiaju.wu@gmail.com > wrote in message
news:1169868296.892494.326610@v33g2000cwv.googlegroups.com...
I don't think SSR with a zero cross function is suitable with pid.
Though it could minimize the influence of an inductive load,it also
would lead to time delay. It is no use for the heating element which is
used in a purely resistance circuit. SSR with a zero cross function act
as a eletric-switch only. I would recommend the SSR with Random
function.
PID with SSR would work properly on the system with high heating
inertia character such as silvia.But if the system has small heating
inertia ,it would be waved.The solution is control the power rate .



On 1月27日, 上午9时00分, "I- >Ian" <some...@nowhere.com> wrote:
> On 26 Jan 2007 22:26:19 GMT, Ian Smith <i...@astounding.org.uk> wrote:
>
>
>
>
>
> >On Fri, 26 Jan 2007 17:10:35 -0500, JimG <jggal...@yahoo.com> wrote:
> >> On 26 Jan 2007 12:57:43 -0800, "alt.coffee" <jiaju...@gmail.com>
> >> wrote:
>
> >> >With my experience I would say pid with a ssr is not a good design for
> >> >a roaster. On and off would not keep the temperature accurately.What
> >> >I
> >> >modified my IROAST2 is add a phase ajustable module between the
> >> >pid(pid
> >> >output is current:0~20mA) and ssr.So the pid is not simply control the
> >> >SSr on or off but the power rate output to the heating
> >> >element.Temperture can stay at the point very stablely.
>
> >> Why doesn't a PID set with a short time base (say 1 second or less)
> >> accomplish the same thing using a standard zero-crossing SSR?
>
> >It does.
>
> >regards, Ian SMithAn SSR with a zero cross function operates when an AC
> >load
> voltage reaches the zero point or its vicinity. This reduces
> clicking noises when the load is input, and minimizes the
> influence of an inductive load, such as a lamp, heater, or
> motor, on the power supply because the inrush current of the
> load is reduced. This can also minimize the scale of the surge
> current protection circuit.
>
> A device with a short cycle time can turn on at any time during the
> cycle.
>
> The two are unrelated.
>
> --
> Posted via a free Usenet account fromhttp://www.teranews.com- 隐藏被引用文字
> -- 显示引用的文字 -




   
Date: 27 Jan 2007 11:12:45
From: Dave S
Subject: Re: PID controlled roaster?
I don't understand the argument for and against zero-crossing PID
either, but your "year, whatever" certainly helped clear it up.

Dave S.

daveb wrote:
> yeah, whatever
>
>
> "alt.coffee" <jiaju.wu@gmail.com> wrote in message
> news:1169868296.892494.326610@v33g2000cwv.googlegroups.com...
> I don't think SSR with a zero cross function is suitable with pid.
> Though it could minimize the influence of an inductive load,it also
> would lead to time delay. It is no use for the heating element which is
> used in a purely resistance circuit. SSR with a zero cross function act
> as a eletric-switch only. I would recommend the SSR with Random
> function.
> PID with SSR would work properly on the system with high heating
> inertia character such as silvia.But if the system has small heating
> inertia ,it would be waved.The solution is control the power rate .
>
>
>
> On 1月27日, 上午9时00分, "I->Ian" <some...@nowhere.com> wrote:
>> On 26 Jan 2007 22:26:19 GMT, Ian Smith <i...@astounding.org.uk> wrote:
>>
>>
>>
>>
>>
>>> On Fri, 26 Jan 2007 17:10:35 -0500, JimG <jggal...@yahoo.com> wrote:
>>>> On 26 Jan 2007 12:57:43 -0800, "alt.coffee" <jiaju...@gmail.com>
>>>> wrote:
>>>>> With my experience I would say pid with a ssr is not a good design for
>>>>> a roaster. On and off would not keep the temperature accurately.What
>>>>> I
>>>>> modified my IROAST2 is add a phase ajustable module between the
>>>>> pid(pid
>>>>> output is current:0~20mA) and ssr.So the pid is not simply control the
>>>>> SSr on or off but the power rate output to the heating
>>>>> element.Temperture can stay at the point very stablely.
>>>> Why doesn't a PID set with a short time base (say 1 second or less)
>>>> accomplish the same thing using a standard zero-crossing SSR?
>>> It does.
>>> regards, Ian SMithAn SSR with a zero cross function operates when an AC
>>> load
>> voltage reaches the zero point or its vicinity. This reduces
>> clicking noises when the load is input, and minimizes the
>> influence of an inductive load, such as a lamp, heater, or
>> motor, on the power supply because the inrush current of the
>> load is reduced. This can also minimize the scale of the surge
>> current protection circuit.
>>
>> A device with a short cycle time can turn on at any time during the
>> cycle.
>>
>> The two are unrelated.
>>
>> --
>> Posted via a free Usenet account fromhttp://www.teranews.com- 隐藏被引用文字
>> -- 显示引用的文字 -
>
>


  
Date: 26 Jan 2007 23:20:02
From: I->Ian
Subject: Re: PID controlled roaster?
On 26 Jan 2007 19:24:56 -0800, "alt.coffee" <jiaju.wu@gmail.com >
wrote:

>I don't think SSR with a zero cross function is suitable with pid.
>Though it could minimize the influence of an inductive load,it also
>would lead to time delay. It is no use for the heating element which is
>used in a purely resistance circuit. SSR with a zero cross function act
>as a eletric-switch only. I would recommend the SSR with Random
>function.
>PID with SSR would work properly on the system with high heating
>inertia character such as silvia.But if the system has small heating
>inertia ,it would be waved.The solution is control the power rate .
>
>

A zero cross SSR can activate every 10 / 8.3 milliseconds.
The maximum voltage will be reached in half that time.
A typical PID has a minimum period of 250 or more milliseconds.

Almost any heater has more inertia than that.

Controlling the power applied is the correct solution for a roaster,
as you state.

--
Posted via a free Usenet account from http://www.teranews.com



 
Date: 26 Jan 2007 12:57:43
From: alt.coffee
Subject: Re: PID controlled roaster?
With my experience I would say pid with a ssr is not a good design for
a roaster. On and off would not keep the temperature accurately.What I
modified my IROAST2 is add a phase ajustable module between the pid(pid
output is current:0~20mA) and ssr.So the pid is not simply control the
SSr on or off but the power rate output to the heating
element.Temperture can stay at the point very stablely.
On 1=D4=C226=C8=D5, =C9=CF=CE=E71=CA=B122=B7=D6, Ian Smith <i...@astounding=
.org.uk >
wrote:
> On Thu, 25 Jan 2007, daveb <davebobbl...@gmail.com> wrote:
> > 45 amp is fine.
> > "snubbers" are for inductive loads like motors. Heaters need no snubb=
ing.I see, so you're saying a big closely-spaced coil of wire has no
> inductance?
>
> Out of interest, how many poppers have you put SSRs in?
>
> regards, Ian SMith
> --
>


  
Date: 26 Jan 2007 17:10:35
From: JimG
Subject: Re: PID controlled roaster?
On 26 Jan 2007 12:57:43 -0800, "alt.coffee" <jiaju.wu@gmail.com >
wrote:

>With my experience I would say pid with a ssr is not a good design for
>a roaster. On and off would not keep the temperature accurately.What I
>modified my IROAST2 is add a phase ajustable module between the pid(pid
>output is current:0~20mA) and ssr.So the pid is not simply control the
>SSr on or off but the power rate output to the heating
>element.Temperture can stay at the point very stablely.

Why doesn't a PID set with a short time base (say 1 second or less)
accomplish the same thing using a standard zero-crossing SSR?

Jim

--
Posted via a free Usenet account from http://www.teranews.com



   
Date: 26 Jan 2007 17:47:01
From: daveb
Subject: Re: PID controlled roaster?
On - off -- "pulse width modulation" is every bit as good in this app. as
proportional.

dave

"JimG" <jggall01@yahoo.com > wrote in message
news:i1ukr2d5m51srjm9vu0amnvf13uhk6tq3f@4ax.com...
> On 26 Jan 2007 12:57:43 -0800, "alt.coffee" <jiaju.wu@gmail.com>
> wrote:
>
>>With my experience I would say pid with a ssr is not a good design for
>>a roaster. On and off would not keep the temperature accurately.What I
>>modified my IROAST2 is add a phase ajustable module between the pid(pid
>>output is current:0~20mA) and ssr.So the pid is not simply control the
>>SSr on or off but the power rate output to the heating
>>element.Temperture can stay at the point very stablely.
>
> Why doesn't a PID set with a short time base (say 1 second or less)
> accomplish the same thing using a standard zero-crossing SSR?
>
> Jim
>
> --
> Posted via a free Usenet account from http://www.teranews.com
>




   
Date: 26 Jan 2007 22:26:19
From: Ian Smith
Subject: Re: PID controlled roaster?
On Fri, 26 Jan 2007 17:10:35 -0500, JimG <jggall01@yahoo.com > wrote:
> On 26 Jan 2007 12:57:43 -0800, "alt.coffee" <jiaju.wu@gmail.com>
> wrote:
>
> >With my experience I would say pid with a ssr is not a good design for
> >a roaster. On and off would not keep the temperature accurately.What I
> >modified my IROAST2 is add a phase ajustable module between the pid(pid
> >output is current:0~20mA) and ssr.So the pid is not simply control the
> >SSr on or off but the power rate output to the heating
> >element.Temperture can stay at the point very stablely.
>
> Why doesn't a PID set with a short time base (say 1 second or less)
> accomplish the same thing using a standard zero-crossing SSR?

It does.

regards, Ian SMith
--


    
Date: 28 Jan 2007 23:34:46
From: jggall01
Subject: Re: PID controlled roaster?
On 26 Jan 2007 22:26:19 GMT, Ian Smith <ian@astounding.org.uk > wrote:

>It does.
>

I thought it might ;-}

But just to convince myself, I did a little mod to an air popper
today. I did the split wire thing, running the fan on 15VDC and the
heater coils through a zero crossing SSR.

The PID controller doesn't seem to have any problem maintaining the
setpoint. There is very little thermal inertia, and I am cycling
every 0.1 second, but this seems to be adequate resolution even with
an empty "drum."

(The 0.1 second time base is probably meaningless since I don't think
the PID I used can read that quickly - I think I'll slow it down to
0.5 seconds to see how that works).

Now that I can roast little batches at whatever temperature I choose,
I guess I better start looking around to find some roast profiles for
popcorn poppers.

If you have any to share, I'd like to see them. TIA.

Jim

--
Posted via a free Usenet account from http://www.teranews.com



     
Date: 29 Jan 2007 08:44:07
From: Ian Smith
Subject: Re: PID controlled roaster?
On Sun, 28 Jan 2007 23:34:46 -0500, jggall01 <jggall01@yahoo.com > wrote:
>
> Now that I can roast little batches at whatever temperature I choose,
> I guess I better start looking around to find some roast profiles for
> popcorn poppers.
>
> If you have any to share, I'd like to see them. TIA.

I use something like
http://www.astounding.org.uk/ian/roaster/realroast.png

The purple (magenta) line is the target profile. It's in degrees C.
Actually, I tend to roast a little bit lighter than that now, but off
the top of my head, I can't remember what to. To adjust roast I
change the steepness of the last ramp (before the final soak) and keep
the times the same.

regards, Ian SMith
--


    
Date: 26 Jan 2007 17:00:55
From: I->Ian
Subject: Re: PID controlled roaster?
On 26 Jan 2007 22:26:19 GMT, Ian Smith <ian@astounding.org.uk > wrote:

>On Fri, 26 Jan 2007 17:10:35 -0500, JimG <jggall01@yahoo.com> wrote:
>> On 26 Jan 2007 12:57:43 -0800, "alt.coffee" <jiaju.wu@gmail.com>
>> wrote:
>>
>> >With my experience I would say pid with a ssr is not a good design for
>> >a roaster. On and off would not keep the temperature accurately.What I
>> >modified my IROAST2 is add a phase ajustable module between the pid(pid
>> >output is current:0~20mA) and ssr.So the pid is not simply control the
>> >SSr on or off but the power rate output to the heating
>> >element.Temperture can stay at the point very stablely.
>>
>> Why doesn't a PID set with a short time base (say 1 second or less)
>> accomplish the same thing using a standard zero-crossing SSR?
>
>It does.
>
>regards, Ian SMith

An SSR with a zero cross function operates when an AC load
voltage reaches the zero point or its vicinity. This reduces
clicking noises when the load is input, and minimizes the
influence of an inductive load, such as a lamp, heater, or
motor, on the power supply because the inrush current of the
load is reduced. This can also minimize the scale of the surge
current protection circuit.

A device with a short cycle time can turn on at any time during the
cycle.

The two are unrelated.

--
Posted via a free Usenet account from http://www.teranews.com



     
Date: 27 Jan 2007 19:38:21
From: Ian Smith
Subject: Re: PID controlled roaster?
On Fri, 26 Jan 2007 17:00:55 -0800, I- >Ian <someone@nowhere.com> wrote:
> On 26 Jan 2007 22:26:19 GMT, Ian Smith <ian@astounding.org.uk> wrote:
>
> >On Fri, 26 Jan 2007 17:10:35 -0500, JimG <jggall01@yahoo.com> wrote:
> >> On 26 Jan 2007 12:57:43 -0800, "alt.coffee" <jiaju.wu@gmail.com>
> >> wrote:
> >>
> >> >With my experience I would say pid with a ssr is not a good design for
> >> >a roaster. On and off would not keep the temperature accurately.What I
> >> >modified my IROAST2 is add a phase ajustable module between the pid(pid
> >> >output is current:0~20mA) and ssr.So the pid is not simply control the
> >> >SSr on or off but the power rate output to the heating
> >> >element.Temperture can stay at the point very stablely.
> >>
> >> Why doesn't a PID set with a short time base (say 1 second or less)
> >> accomplish the same thing using a standard zero-crossing SSR?
> >
> >It does.
>
> The two are unrelated.

Indeed, but your little essay is irrelevant to teh fact that a PID
with a short time base and a standard zero crossing SSR achieves a
stable temperature in this application.

They don't have to be related for this to be true.

regards, Ian SMith
--


 
Date: 25 Jan 2007 07:16:56
From:
Subject: Re: PID controlled roaster?
Ok, I'll bite. What's a 'snubber'? Your context suggests a relay.
How did the word 'snubber' get attached to that?
Would a suitably rated SSR (45 amp, for example) have survived or is it
more complicated than that?

On Jan 25, 8:06 am, Ian Smith <i...@astounding.org.uk > wrote:
> On Wed, 24 Jan 2007 18:06:18 -0600, Mike Hartigan <m...@hartigan.dot.com> wrote:
> > trivial matter to mod virtually any fluid bed, or even drum roaster
> > with this PID, an SSR, a fuse, a pair of power switches, a box, andOne other thing - I've found SSRs not particularly happy driving a
> popper heater element on the end of a moderate length wire. I think
> the built-in snubber is inadequate, and the SSRs fail after a few
> hundred roasts. Adding a decent-rated external snubber across the SSR
> helps a lot (no failure since doing that, so far).
>
> regards, Ian SMith
> --
>


  
Date: 25 Jan 2007 17:20:33
From: Ian Smith
Subject: Re: PID controlled roaster?
On 25 Jan 2007, ls1mike@gmail.com <ls1mike@gmail.com > wrote:
> On Jan 25, 8:06 am, Ian Smith <i...@astounding.org.uk> wrote:
> >
> > popper heater element on the end of a moderate length wire. I think
> > the built-in snubber is inadequate, and the SSRs fail after a few
> > hundred roasts. Adding a decent-rated external snubber across the SSR
> > helps a lot (no failure since doing that, so far).
>
> Ok, I'll bite. What's a 'snubber'? Your context suggests a relay.
> How did the word 'snubber' get attached to that?
> Would a suitably rated SSR (45 amp, for example) have survived or is it
> more complicated than that?

Heater coil is a significant inductor. Switching it abruptly gives
rise to voltage transients, possibly high enough to damage adjacent
components (the switching components in the SSR for example). A zero
crossing SSR means you won't get transients on the switch on (I
think), but you will on the switch off (when the SSR isn't zero
crossing).

A snubber is a thing that blunts the transients. A simple snubber is
an R and a C connected appropriately. More complex snubbers exist
(you can read wikipedia if you want, or an electronics textbook).
Most SSRs include a snubber internally, but despite that I found my
roaster was frying SSRs, even big ones (I can't remember what I used,
but at least one of them was a very high rating). There's a trade off
- bigger snubber handles higher inductance loads, but allows higher
leakage current when the SSR is off.

The current rating won't directly help - it's a voltage peak with zero
current through the SSR that's the problem. There's presumably an
indirect benefit because a bigger current rating probably means bigger
internal switching components that probably have higher voltage
capability.

Adding an additional external snubber mitigates the transients, and I
can tolerate quite high leakage currents through the heater coil when
the SSR is supposedly off, so that's the solution - a simple but
high-rating snubber connected across the switching output of the SSR

regards, Ian SMith
--


  
Date: 25 Jan 2007 12:11:06
From: daveb
Subject: Re: PID controlled roaster?
45 amp is fine.

"snubbers" are for inductive loads like motors. Heaters need no snubbing.

dave


<ls1mike@gmail.com > wrote in message
news:1169738216.167204.291940@v33g2000cwv.googlegroups.com...
> Ok, I'll bite. What's a 'snubber'? Your context suggests a relay.
> How did the word 'snubber' get attached to that?
> Would a suitably rated SSR (45 amp, for example) have survived or is it
> more complicated than that?
>
> On Jan 25, 8:06 am, Ian Smith <i...@astounding.org.uk> wrote:
>> On Wed, 24 Jan 2007 18:06:18 -0600, Mike Hartigan <m...@hartigan.dot.com>
>> wrote:
>> > trivial matter to mod virtually any fluid bed, or even drum roaster
>> > with this PID, an SSR, a fuse, a pair of power switches, a box, andOne
>> > other thing - I've found SSRs not particularly happy driving a
>> popper heater element on the end of a moderate length wire. I think
>> the built-in snubber is inadequate, and the SSRs fail after a few
>> hundred roasts. Adding a decent-rated external snubber across the SSR
>> helps a lot (no failure since doing that, so far).
>>
>> regards, Ian SMith
>> --
>>


   
Date: 25 Jan 2007 17:22:01
From: Ian Smith
Subject: Re: PID controlled roaster?
On Thu, 25 Jan 2007, daveb <davebobblane@gmail.com > wrote:

> 45 amp is fine.
> "snubbers" are for inductive loads like motors. Heaters need no snubbing.

I see, so you're saying a big closely-spaced coil of wire has no
inductance?

Out of interest, how many poppers have you put SSRs in?

regards, Ian SMith
--


 
Date: 25 Jan 2007 14:06:23
From: Ian Smith
Subject: Re: PID controlled roaster?
On Wed, 24 Jan 2007 18:06:18 -0600, Mike Hartigan <mike@hartigan.dot.com > wrote:

> trivial matter to mod virtually any fluid bed, or even drum roaster
> with this PID, an SSR, a fuse, a pair of power switches, a box, and

One other thing - I've found SSRs not particularly happy driving a
popper heater element on the end of a moderate length wire. I think
the built-in snubber is inadequate, and the SSRs fail after a few
hundred roasts. Adding a decent-rated external snubber across the SSR
helps a lot (no failure since doing that, so far).

regards, Ian SMith
--


 
Date: 25 Jan 2007 13:59:11
From: Ian Smith
Subject: Re: PID controlled roaster?
On Wed, 24 Jan 2007, Mike Hartigan <mike@hartigan.dot.com > wrote:
> Just a thought - has anybody considered building a roaster around a
> programmable PID? Something like this, for example:

My roaster is a home-made microprocessor-based box driving a popper.

http://www.astounding.org.uk/ian/roaster/index.html
http://www.astounding.org.uk/ian/roaster/hardware.html

I intended to make it PID (the algorithm isn't fundamentally
difficult), but actually, after implementing just proportional
control, I found that it works happily with a very narrow proportional
band and simply proportional control - the I and the D did not seem
worth the aggro.

I do correct slightly my roast program - if it's below 0C ambient, I
set the controller to roast to a 1C hotter final temperature. As long
as it's above freezing, I just let it get on with it.

> It's programmable in 30 steps (vs iRoast's 5) and is accurate to
> within 0.2% over its entire range.

Depends what you mean by 'accurate' I think.

However, mine has 256 steps each of which can be a ramp and a hold (so
ramp to 250C over three minutes and hold for two minutes would be one
step). It will store any number of programs within those steps, so
you could have (say) three 80-step programs, or eighty 3-step
programs, or one 200 step, one 10 step, and eleven 4-step, or whatever
else you want. I actually only use about four 4-or-5-step programs
(dark, normal, normal+1C, light) but most of the time I'm roasting the
same beans for the same use.

It's nice having the control and the repeatability. Graphing to a PC
is diverting too. However, I remain convinced that adding I and D
will not improve functionality significantly - the I would simply
avoid the need for me correcting when it's very cold.

regards, Ian SMith
--


 
Date: 24 Jan 2007 17:37:17
From: DavidMLewis
Subject: Re: PID controlled roaster?


On Jan 24, 4:06 pm, Mike Hartigan <m...@hartigan.dot.com > wrote:
> Just a thought - has anybody considered building a roaster around a
> programmable PID? Something like this, for example:
>
> It almost seems too easy - what am I missing?
>
It is too easy, although several people have used them with air
roasters. Drum roasters are harder because of the longer time lags,
etc. in the system, and the fact that you're often trying to control
both a fan and a heater. The placement and design of the sensor(s) are
critical. Jim Schulman, who has done this, has found that wild
variations in air temperature affect the roast negatively, so he closes
the loop on air temperature rather than bean temperature. Some people
have used two cascaded controllers, in a system in which the bean temp
controller produces the setpoint for the air temp controller. Jeffrey
Pawlan has ripped the control out of a Hottop and replaced it with a
whole system driven by LabView on a laptop.

One other issue is that the thermal characteristics of the beans vary a
lot in the course of a roast. Commercial computer-controlled roasters
either use about five different sets of PID coefficients over the roast
profile, as Ambex does, or some sort of more sophisticated modeling of
the system, i.e. a state-space servo or similar.

None of this is to say that you shouldn't try to do it. It just means
that you shouldn't look on it as necessarily an easy or fast way to get
better coffee in the short term.

Best,
David



 
Date: 24 Jan 2007 19:35:35
From: jim schulman
Subject: Re: PID controlled roaster?
On Wed, 24 Jan 2007 18:06:18 -0600, Mike Hartigan
<mike@hartigan.dot.com > wrote:

>Just a thought - has anybody considered building a roaster around a
>programmable PID? Something like this, for example:
>
>http://tinyurl.com/yuqcjo
>
>It's programmable in 30 steps (vs iRoast's 5) and is accurate to
>within 0.2% over its entire range. It seems like it would be a
>trivial matter to mod virtually any fluid bed, or even drum roaster
>with this PID, an SSR, a fuse, a pair of power switches, a box, and
>*maybe* a variac. Simply connect the motor directly to the AC and
>use this to control the heater coil. The thermocouple could be
>placed anywhere you damn well please and the result would be an
>incredibly repeatable profile.
>
>It almost seems too easy - what am I missing?

Check groups.google for mine and others posts on PIDing roasters; it's
easy.

On the other hand, an instance of this Auber controls turkey I tried
out is currently taking up space in the landfill. It has an "advanced"
PID control algortithm designed by the owner. The advance is to
eliminate the P band and go from full on to full off in a 1F range.
This breathtaking advance seemingly makes it a simple thermostat, but
the owner assures me I'm missing it's brilliant subtleties.

Suffice it to say, in a roaster, it did have a programable setpoint,
and cycled the bean temps 60 degrees plus/minus from it in 3 minute
cycles. I've never seen such bad roasts, not even from the nearly as
badly designed Iroars.


  
Date: 24 Jan 2007 21:25:31
From: Mike Hartigan
Subject: Re: PID controlled roaster?
In article <gc1gr2d110i7ubeqe41olitb7ov1c2o0o1@4ax.com >,
jim_schulman@ameritech.net says...
> On Wed, 24 Jan 2007 18:06:18 -0600, Mike Hartigan
> <mike@hartigan.dot.com> wrote:
>
> >Just a thought - has anybody considered building a roaster around a
> >programmable PID? Something like this, for example:
> >
> >http://tinyurl.com/yuqcjo
> >
> >It's programmable in 30 steps (vs iRoast's 5) and is accurate to
> >within 0.2% over its entire range. It seems like it would be a
> >trivial matter to mod virtually any fluid bed, or even drum roaster
> >with this PID, an SSR, a fuse, a pair of power switches, a box, and
> >*maybe* a variac. Simply connect the motor directly to the AC and
> >use this to control the heater coil. The thermocouple could be
> >placed anywhere you damn well please and the result would be an
> >incredibly repeatable profile.
> >
> >It almost seems too easy - what am I missing?
>
> Check groups.google for mine and others posts on PIDing roasters; it's
> easy.
>
> On the other hand, an instance of this Auber controls turkey I tried
> out is currently taking up space in the landfill. It has an "advanced"
> PID control algortithm designed by the owner. The advance is to
> eliminate the P band and go from full on to full off in a 1F range.
> This breathtaking advance seemingly makes it a simple thermostat, but
> the owner assures me I'm missing it's brilliant subtleties.

Doesn't that defeat the whole purpose of a PID?

> Suffice it to say, in a roaster, it did have a programable setpoint,
> and cycled the bean temps 60 degrees plus/minus from it in 3 minute
> cycles. I've never seen such bad roasts, not even from the nearly as
> badly designed Iroars.

Do you attribute those bad roasts to the poorly executed PID? In
other words, would different PID (a 'real' one) be expected to yield
better results?

--
-Mike


   
Date: 25 Jan 2007 01:07:00
From: jim schulman
Subject: Re: PID controlled roaster?
On Wed, 24 Jan 2007 21:25:31 -0600, Mike Hartigan
<mike@hartigan.dot.com > wrote:

>Do you attribute those bad roasts to the poorly executed PID? In
>other words, would different PID (a 'real' one) be expected to yield
>better results?

Sure, I put in a Fuji and it now roasts better than mine (I modded a
friend's P1, and I;ve lesrnt a few tricks since I did mine)