The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: email@example.com (john jardine)
Subject: Re: Heated seat controller, open or closed loop temp.control
Date: 2 Nov 2002 11:07:36 -0800
References: <firstname.lastname@example.org> <email@example.com> <4Diw9.11247$I6.firstname.lastname@example.org> <email@example.com>
NNTP-Posting-Date: 2 Nov 2002 19:07:36 GMT
"Jeroen" wrote in message news:...
> "john jardine" wrote in message
> > "Jeroen" wrote in message
> > [clip]
> I'm familiar with it, and yes, I know it can be damn tough to tune them
> properly. I once made one that had to stabilize the light intensity in a box
> (exactly 38 lux and 512 lux, switchable) for light sensor calibrations.
> Well... that took *much* **much** longer to tune than I ever imagined... :(
> The reason I choosed for the control loop was to compensate for lamp wear,
> mains fluctuations etc.
> Back to the seats:
> From I what I can make up, the problem is that when you turn of the heater,
> heating still continues for a while, so this need to be accounted for (as
> someone once told me that this happens in house heating). And overshoot is
> always a problem. The problem also with the heater is, the characterics
> change depending on who sits on it (or nobody at all). I'll include a timer
> that heats the seat some time before I'm going to use the car, so that I can
> sit myself down on a nice warm seat :)
> But do I need a fullblown PID?? Isn't the I-term just enough? E.g. constant
> switching on/off, period 1 second, and increase/decrease PW by a fraction of
> the error? The P term is not needed, because you can't say 'when continously
> turned on the temp will be so and so much C deg', it also depends on the
> ambient temp.
> Maybe I can make it an autotuning control loop? Digital home thermostats
> seem to do this (not every house is the same). I've E2PROM in the AVR
> anyway. Hmmm, very interesting stuff...
> Are there any other control algorithms?
Apologies for the 'chapter and verse', I just got carried away ... ;-)
No you should'nt really need the full PID. Just using the 'I' seems
like it would work OK. The commercial oven controllers use exactly
that ON/OFF variable duty cycle system. Ideal for your seat, would be
the 45UKP hand sized controller (complete with probe!), I bought last
year to make a small PCB test oven based on an external 500W
Yes, with the AVR, autotuning is a perfectly good option. Lot of
progging though as you need to make a number of tentative control
sensitivety 'tests' before selecting an optimum control strategy.
The old analogue bi-mettalic and the newer digital home T/C's use a
'pre-act' mechanism for switching. In that they are sensitive to rates
of change of the room temperature change. I.e they will act to switch
off earlier than setpoint if the room temp' is rising fast. (The D
method I was on about).
Loads of algorithms or methods !. Most are the type that come from a
'what if I do this' thinking session. None that are especially clever
and most are just mix and match of the basic PID parts.
As you've got EPROM in there I'd be inclined to store a fixed set of
duty-cycles that give (say 4 or more) approx' aiming temps. Then
using the sensor just tweek the Duty cycle from that point. Would
greatly minimise ability of the setup to get out of control. (just
dreamt this one up!, I'll add it to my own list ;-).
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup