Student-resistant, maybe. Then one comes along and shoves the heatsink up against the back of the bench and puts his lab book on top so he has space to eat his lunch. ;)
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Some systems, e.g. the voice coil drive in a hard disk, really are very nearly pure second-order, i.e. displacement is proportional to I*V**2, so that the phase shift is close to 180 degrees out to pretty high frequency. Lightly-loaded, current-driven DC motors are also (nearly) like that.
In that sort of case, you need the controller to have a (band-limited) derivative term in order to have a stable position for the 0-dB cross point.
Otherwise I agree that the D term isn't as useful as it could be.
Probably for the above reason.
In a pure 1-D case, probably.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Putting some delay in my ltspice thermal models is the one bit I'm missing. Is there some simple time delay knob in LTspice? (I've just added some R/C stages w/buffers.)
George H.
Assuming perfect conduction between fluid and plate, your fluids are a current source, heat load. Your job, get rid of the heat... TBH I don't think TEC's help for any sort of real power > 1W How much power do TEC beer coolers use?
George H.
A transmission line.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Right :^) the gizmo I'm thinking of has two sides.
This is a thermal loop that is designed to just hold the magnet at whatever temp it is, when you walk into the lab. It's a pretty small tec compared to the size of the magnets.
George H.
So I can add many miles of transmission line? I'll try that next time... every project has a thermal thing (on average).
George H.
The steamship had a gigantic moment of inertia and a very good DC tach, so the feedback was slow and clean. That's an ideal situation for some derivative.
I simulated that on an HP9100 calculator, and then on a PDP-8. It worked just like the sims.
A delay line might be appropriate for transport delay, then some RCs at the taps. I'd guess that the transport delay isn't very important.
One thing to do is measure the open-loop step response and then booger something in Spice that behaves that way. Then Spice the closed loop.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
It was a cool system. Steam spins up a gigantic moment of inertia (mainly the HP turbine geared way, way up) that spins the prop. That's practically an integrator. Then the difference between hull speed and prop speed loads the shaft with a squared term. Prop torque accelerates the hull. The hull is loaded by the water proportional to speed squared. So horsepower goes as speed cubed.
What was amazing is that it behaved just as predicted. Beginner's luck.
A crusty old Chief looked at my simulation curves (printed sideways on a Teletype and hand-colored) and said "That's just the way an experienced hand would work the throttle valve" so we got the job.
I just did dumb rectangular integration of one chunk of the system at a time. The profs and control students told me that I had to do some fancy runga-kuta interpolations or it wouldn't sim right.
Even if I had 3D thermal sim tools, learning them and setting up situations would be a big deal. It's easier to approximate and build it.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Sadly, there is no thermal analog to an inductor. It's all RCs.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
It's usually crucial in small fast temperature control loops. A pure thermal-mass / Qdot system can be made as fast as you like by applying more gain. TEC capacity limits the slew rate, obviously, but the attainable loop _bandwidth_ is set by thermal diffusion.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Yeah, I stepped in it like that early on too. As the wise man said, "The harder I work, the luckier I get." ;)
Yup, assuming that it doesn't have too many knobs and dials. Distributed systems, e.g. big cold plates with variable dissipation at several points and several temperature sensors, can be way beyond the reach of manual twiddling.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
The heat equation being second order in space but only first order in time.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
rote:
he amount of heat transferred - in watts per amp - can vary quite a bit as you go from heating to cooling. You really wna to measure the temperatures on both sides of the TEC - not only the object whose temperature you are tr ying to control, but also the temperature of the heat-sink on the exhaust s ide of the TEC.
.The problem is that you have to stabilise the feedback loop, and the gain a round a TEC system can vary quite a bit.
If you want your feedback loop to be critically damped - to settle as fast as possible - you do have to deal with the variation in TEC "gain" as a fun ction of the temperature difference across the TEC.
The thermal resistance from the stabilised object to the TEC and from the T EC to ambient are crucial to the performance of the system - the TEC has ma ximum temperature difference it can sustain when cooling with zero thermal resistances and that goes down when you plug in actual thermal resistances.
Two feet by four feet of half inch thick aluminium is a lot of heat sink. A crude heat sink calculation says 0.1K/Watt to air, so your 50W should only have generated 5 degrees of heating.
Obviously, there's going have been a thermal gradient across the plate, but in half inch thick aluminium it won't be much.
-- Bill Sloman, Sydney
This is pretty much the Ziegler Nichols PID tuning procedure, which has bee n around as long as I have.
Ziegler J and Nichols N 1942 Trans. ASME 64 759
Your presentation misses the point that the TEC "gain" in terms of Watts sh ifted per Ampere through the TEC does depend on the temperature difference across the TEC. If you are stabilising a laser diode at a fixed temperature in an air-conditioned lab, this doesn't change enough to matter, but not e verybody has it that easy.
l .
Monitoring the temperature of the exhaust side of the TEC is one way of imp lementing this, with incidental advantages.
-- Bill Sloman, Sydney
ge
hature
ocks
EC's is
maximum
at
,r.
rs on
the
s the
,the time
diodes
by an
for
ne
nd
ion
r things
ing
he
all
ops.
et high
eigh
ader
the
le
increase
e a eof
tst
y,
ce
re
e e nngs.
Newtons Law of cooling is the simplest possible. Real heat sinks are cooled by convection, and if enough heat is being transferred the convection curr ents can become turbulent, and the thermal resistance to air is then domina ted by the thickness of the boundary layer at the heat sink surface, which goes down as the convection currents get faster. Free convection ends up co ming closer to a square law, where the rate of cooling is proportional to t he square of the temperature difference.
Fan cooled systems have a constant - lowish - boundary layer thickness, and you get more cooling per unit of heat-sink surface area. They do conform t o the linear relationship.
-- Bill Sloman, Sydney
I think the first order of business is to just connect up a simple one loop system of pump, heat sink, and standard sized water block and thermal paste and clamp a 50 or 100 watt resistor to the sink, temperature sensor to the water block. Then ramp the power into the sink at a sequence of flow velocities, and plot wrt time.
That should give a baseline for some of the unknowns in the "circuit" and more complex arrangements can hopefully derive from that.
You can have thermal inductors and temperature as a complex number in your model if you're OK with (maybe) violating the second law of thermodynamics.
I meant the water flow delay. That's probably small.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Well, if you're prepared to put in physical nonsense like that, how about a thermal op amp? Could be super useful--a TEC with a gain of 100 dB. ;)
In nonrelativistic quantum many-body theory you can formulate the finite-temperature case as a translation along the imaginary time axis. That turns the Schroedinger time dependence exp(i Et/hbar) into a Boltzmann factor, exp(-E/kT). Of course you can't do the time-dependent, finite-temperature case that way--you have to actually solve the problem. ;)
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.