Posted by Morris Dovey on October 30, 2011, 4:27 pm
On 10/30/11 10:12 AM, amdx wrote:
>> -----
>> * If anyone would like to play along, I could use some help with a
>> combination heater/temperature sensor using an up to 16-inch (406.4mm)
>> nichrome wire threaded through ceramic cores (inside a 1/4x4-inch copper
>> tube that's closed on one end) with one side connected to ground and the
>> other to a SPST relay so that when the relay is powered, the wire will
>> be heated with AC power; and when the relay is unpowered, the resistance
>> of the wire can be sensed to determine its temperature. The sensing
>> circuit needs to produce an output voltage that can be fed to an Arduino
>> A/D converter with a 5VDC Vref and sharing the ground connection. Heater
>> power needs to be something that can be reasonably derived from 120VAC
>> 60Hz mains to provide power in the 250W neighborhood. I just uploaded a
>> photo of some ceramic cores to
>>
>> http://www.iedu.com/Solar/Electricity/Fusion/Core4 (500x400).jpg
>>
> Hi Morris,
> Not sure I clearly understand what you want, but I'll stick my neck out!
> After you have heated the wire then you want to measure the resistance
> of the wire and from that deduce the temperature. I would want to
> totally isolate the AC from the resistance measurement. A DPDT relay
> would do that. (isolate both ends of the wire. I don't know about
> connecting AC neutral to 5v ground, might be ok, but...)
> Unless you cycle the relay for heat/measure modes, I don't see how you
> will know when to stop heating. So, you have the relay on some clock
> cycle, (heat 5 seconds measure temperature 1 second) then when the
> proper temp is reached the relay measures temp only until temp drops
> below some threshold then your heat/measure cycle starts again.
> To measure the resistance of the wire, you will drive some DC current
> through the wire and a series (sense) resistor and measure the voltage
> across the sense resistor. The voltage will probably need to be
> amplified before going to the Arduino.
> There my be a way to sense the AC current to the wire and know the temp
> of the wire but any AC voltage fluctuations to you home may cause
> errors. I'm sure there is a way to compensate for that though.
> Ok, have I got the basic idea?
> Fill me in.
> Thanks, Mikek
You've pretty well nailed it. DPDT relay it'll be, with hot/neutral for
the AC heater power and ground/5VDC for sensing. Some other DC voltage
can be used, but I already have decently-regulated 5V available and
would prefer not to build/buy another power supply if it can be avoided.
The relay power will be software controlled with allowance made for
make/break latencies - and I plan to capture the A/D value while heater
power is off. This approach should allow me to vary the heating duty
cycle and to exercise a (hopefully useful) control of the heating rate -
then continuously monitor the reactor temperature if/after "ignition"
takes place.
If ignition occurs, there will be another latency as heat is transferred
from the containment vessel into the heater tube - but I think that
situation would exist for any affordable strategy.
The change in resistance looks like it'll be small, and I guessed that
an op-amp circuit might be useful to produce a larger temperature swing.
A temperature range of 25-537°C would probably be a sensible design
target. (I'm /aware/ of op-amps, but I've never designed a circuit that
actually /used/ one of the little buggers.)
A linear output would be "nice to have", but isn't essential - I don't
think a table look-up/interpolation is likely to overburden even the
little Arduino.
-----
For my first go at this problem, I plan to start at a very low H2
pressure and increase it incrementally. At each pressure setting I'll
start at a low temperature setting and increase the temperature
incrementally to some safe limit (probably 250°C) and watch for any sign
that ignition may have taken place. If so, then I plan to remove heater
power and check for a self-sustaining reaction. If the reaction isn't
self-sustaining, I'll continue from where I left off in the heating
sequence.
If I reach the safe limit without ignition, I'll let the apparatus cool
down and then repeat at the next pressure increment.
The entire exercise is to discover whether a self-sustaining reaction
can be initiated and, if so, the minimum (pressure,temperature) values.
--
Morris Dovey
DeSoto Solar
http://www.iedu.com/Solar/
Posted by Morris Dovey on October 30, 2011, 10:47 pm
On 10/30/11 4:12 PM, Bob F wrote:
> I would think that thermocouple temperature monitoring/control would make a lot
> more sense that using a relay to disconnect the power so you can measure the
> resistance.
I want to be able to control the amount of power used for heating, and
cycling the power on and off (varying the duty cycle) appears to be the
easiest and most cost-effective method.
If I take this approach, it seems reasonable to use the off time to
sample the resistance directly from the heating element itself.
> The time delay of the relay approach drastically lessens the
> controlability.
I don't see this as a practical impediment. The time required to
initiate an A/D conversion should be negligible and both make and break
times are factored into the duty cycle calculations.
I'm thinking that the latency due to the heat transfer from inside the
heater tube to the reaction chamber (and/or vice versa) is very much
greater - and that latency will be the controlability limit regardless
of the method used to produce or sense heat.
I could improve on this by locating a thermocouple bead inside the
reaction chamber itself, but that introduces other (serious) problems
that I really don't have even a clue how to solve. Don't wanna go there.
> There are plenty of thermocouple temperature controllers
> available.
Agreed. The AD595 mentioned earlier, along with a decoupling cap, would
provide everything needed to allow using a thermocouple - and the µC
provides everything else needed. Its software can provide all of the
control logic provided by the external temperature controller.
--
Morris Dovey
DeSoto Solar
http://www.iedu.com/Solar/
Posted by amdx on October 31, 2011, 1:09 am
Snips
> If ignition occurs, there will be another latency as heat is transferred
> from the containment vessel into the heater tube - but I think that
> situation would exist for any affordable strategy.
> The change in resistance looks like it'll be small, and I guessed that
> an op-amp circuit might be useful to produce a larger temperature swing.
> A temperature range of 25-537°C would probably be a sensible design
> target. (I'm /aware/ of op-amps, but I've never designed a circuit that
> actually /used/ one of the little buggers.)
> > A linear output would be "nice to have", but isn't essential - I don't
> think a table look-up/interpolation is likely to overburden even the
> little Arduino.
> -----
> For my first go at this problem, I plan to start at a very low H2
> pressure and increase it incrementally. At each pressure setting I'll
> start at a low temperature setting and increase the temperature
> incrementally to some safe limit (probably 250°C)
> The entire exercise is to discover whether a self-sustaining reaction
> can be initiated and, if so, the minimum (pressure,temperature) values.
Hey Morris, is there a reason you don't want to stick a thermocouple in
the tube for your temp measurements as others have suggested? It would
sure simplify this section, and, you could get lots of design help on
sci.electronics.design to raise the thermocouple output voltage nearer
5v at 537*C.
Also I suspect you have thought that if the temp sustains itself or
starts to run away, your Ardiuno will sense this and start limiting
hydrogen input and/or pump more water.
Mikek
Posted by Morris Dovey on October 31, 2011, 5:17 am
On 10/30/11 8:09 PM, amdx wrote:
> Hey Morris, is there a reason you don't want to stick a thermocouple in
> the tube for your temp measurements as others have suggested? It would
> sure simplify this section, and, you could get lots of design help on
> sci.electronics.design to raise the thermocouple output voltage nearer
> 5v at 537*C.
I may be able to do that but can't design it that way without first
determining the heater parameters. If the heater wire length takes up
all 16" of the available ceramic core channels, then there won't be any
room for anything else.
So far, everyone has been trying to jump ahead to the temperature
measurement part of the problem, but first we need to design the heater
portion in order to know how much space (if any) will be available for
the thermocouple bead and its wire leads.
Since I don't *know* that there's enough space for the thermocouple, I
pretty much have to plan to make the heater wire do double duty.
There actually is an advantage to this in that it limits the number of
terminals needed at the end of the tube to just two, instead of the four
that would be needed if there were separate heater and thermocouple leads.
"Make everything as simple as possible, but not simpler."
> Also I suspect you have thought that if the temp sustains itself or
> starts to run away, your Ardiuno will sense this and start limiting
> hydrogen input and/or pump more water.
I'm not planning to water cool this thing at all. If the reaction is
able to sustain itself, then the next step (phase 2) will be to coax it
to self-sustain in the 400°C neighborhood needed by the parent project
into which I want to embed the reactor.
The phase 1 testing is to find the (pressure,temperature1) points, if
any, at which ignition occurs and the corresponding minimum
(pressure,temperature2) points where the reaction becomes
self-sustaining. During the initial phase 1 testing the 'safe'
temperature threshold will probably be set down to 250°C to give me time
to get over being terrified. :-)
If a runaway condition is detected (either by exceeding the preset
'safe' temperature threshold or by a significant increase in radiation),
then the shutdown routine will [1] kill heater power (even if it's
already off), [2] shut off the hydrogen supply, and then [3] open a
valve between the reaction chamber and a vacuum tank to remove the
hydrogen from the reactor. A full shutdown should take less than a
second and the vacuum valve will remain open until an explicit "close
valve" command is entered by the operator (me).
--
Morris Dovey
DeSoto Solar
http://www.iedu.com/Solar/
Posted by Morris Dovey on October 31, 2011, 6:55 am
On 10/30/11 10:33 PM, Bob F wrote:
> Actually, Morris - you should look at least at the possibility of using a solid
> state relay to lessen the risk of fused contacts.
Ok. I don't yet know what my heater voltage/current levels need to be,
but I found
"OMRON G3NA210BDC524 Solid State Relay Input DC Maximum Operating
Voltage 24 VDC Minimum Operating Voltage 5 VDC Output AC Output Current
Range with Heat Sink 0.1 - 10 Amps Minimum Output Voltage 19 VAC Maximum
Output Voltage 264 VAC Maximum Input Current 7 Milliamps"
for about $0 with shipping.
That will resolve the fused contacts issue, but leads me to ask a
related question: If/when an SSR fails, does it fail open or does it
fail closed?
If I do that, I'll need the AD595 thermocouple adapter at $7.95+ and a
Type-K glass braid insulated thermocouple at $3.95+
And I need to reduce the maximum allowable length of the nichrome heater
wire from 16" to about 12" to allow locating the thermocouple bead
somewhere near the middle of the heater tube.
That adds up to about $5 in additional cost to avoid fused relay
contacts...
Now can we please go back and figure out the heater itself so there's
something to measure?
:-)
--
Morris Dovey
DeSoto Solar
http://www.iedu.com/Solar/
>> * If anyone would like to play along, I could use some help with a
>> combination heater/temperature sensor using an up to 16-inch (406.4mm)
>> nichrome wire threaded through ceramic cores (inside a 1/4x4-inch copper
>> tube that's closed on one end) with one side connected to ground and the
>> other to a SPST relay so that when the relay is powered, the wire will
>> be heated with AC power; and when the relay is unpowered, the resistance
>> of the wire can be sensed to determine its temperature. The sensing
>> circuit needs to produce an output voltage that can be fed to an Arduino
>> A/D converter with a 5VDC Vref and sharing the ground connection. Heater
>> power needs to be something that can be reasonably derived from 120VAC
>> 60Hz mains to provide power in the 250W neighborhood. I just uploaded a
>> photo of some ceramic cores to
>>
>> http://www.iedu.com/Solar/Electricity/Fusion/Core4 (500x400).jpg
>>
> Hi Morris,
> Not sure I clearly understand what you want, but I'll stick my neck out!
> After you have heated the wire then you want to measure the resistance
> of the wire and from that deduce the temperature. I would want to
> totally isolate the AC from the resistance measurement. A DPDT relay
> would do that. (isolate both ends of the wire. I don't know about
> connecting AC neutral to 5v ground, might be ok, but...)
> Unless you cycle the relay for heat/measure modes, I don't see how you
> will know when to stop heating. So, you have the relay on some clock
> cycle, (heat 5 seconds measure temperature 1 second) then when the
> proper temp is reached the relay measures temp only until temp drops
> below some threshold then your heat/measure cycle starts again.
> To measure the resistance of the wire, you will drive some DC current
> through the wire and a series (sense) resistor and measure the voltage
> across the sense resistor. The voltage will probably need to be
> amplified before going to the Arduino.
> There my be a way to sense the AC current to the wire and know the temp
> of the wire but any AC voltage fluctuations to you home may cause
> errors. I'm sure there is a way to compensate for that though.
> Ok, have I got the basic idea?
> Fill me in.
> Thanks, Mikek