Flying on the Cheap - Instruments

In article et,
Richard Lamb wrote:



When recumbent bike guy was in here, we talked about curing ovens for
composites, and that brought up temperature control units. Those things
use rate of change to avoid temperature overshoot by scaling back power
as the set point is approached.

You do something analogous to this every time you take your foot off the
gas and then begin to brake gently, as you approach a stop sign.

But, the point is, I think you might be able to find an off the shelf
solution. There's nothing terribly sophisticated about a speedometer. In
this case, you want to measure something like degrees per minute instead
of miles per hour.

Good example, but WAY more involved here.

I've fooled around with it some myself, but my experience is in logic design,
not the weird and wonderful world of analog electronics.

Jim's point (as I read it) was that we'd have to notice (and react to) a few
microvolts change - per second - (the RATE question).

But how many mV per second would be considered a threatening situation?
(i.e.: how many degrees per second)

In the automobile example the question might be framed, "What rate of closure
would prompt you to lift the gas pedal and apply the brakes?".

Answer - Depends on how close you are to the thing in front of you.

Jim's response was
It seems like a fairly simple request on the surface, but too many undefined
variables for a simple answer.

There is going to be a bit of research involved before this one can be worked
out.

Richard

My ex has a cute "old saying" that she came up with.

"Two weeks in the lab will save you three days in the library
every time".

Richard,

If it helps you to see this as a very difficult challenge, then go
ahead. I know what the "H" in RAH stands for, but if designing your own
circuit from discrete transistors that you've carved out of wood is
beyond your interest, then why not look around for a plug 'n' play
option? Have you *really* already done that and come up empty handed?
I've got a shelf full of catalogs of temperature control hardware. I
haven't actually looked for one with a rate alarm option, but since rate
is monitored and used by the controller, I don't see why someone
wouldn't offer such an alarm as an option.

If it helps you to see this as a very difficult challenge, then go
ahead. I know what the "H" in RAH stands for, but if designing your own
circuit from discrete transistors that you've carved out of wood is
beyond your interest, then why not look around for a plug 'n' play
option? Have you *really* already done that and come up empty handed?
I've got a shelf full of catalogs of temperature control hardware. I
haven't actually looked for one with a rate alarm option, but since rate
is monitored and used by the controller, I don't see why someone
wouldn't offer such an alarm as an option.

Ok, Smitty.

Nope, not at all interested in carving transistors out of wood.

Since you have all the catalogs, my suggestion is to "go for it".

When you are all done, we can compare "budgets" and judge accordingly.
Budgets?

First, obviously, would be cost.
I'm pretty sure, should he decide to bite, Jim's solution would come in
for under 50 bucks.

Power budget?
We can assume we have a 12 volt DC electrical system. How many amps
will you need? Jim probably wouldn't need over 100 mA or so, if that.

Weight budget? (Oh Sh*t, not that again!)
What say, Jim? 3 or 4 ounces?

And then come the electrical issues!

It's a thermocouple. Output is in the millivolt range.
AND, to monitor RATE of change over a very short time (seconds) we'd
have to get down to the microvolt level!

To be able to use that, it will have to be scaled up - gain of roughly
one million or more? No freakin' way *I* would try that with desecrate
(or even wooden) transistors.

Again, it's a thermocouple!
How do you propose to "interface" to it? Single side ain't gonna get it.
You would wind up amplifying the noise as much or more than the signal!
Differential low noise high gain op amp seem to be the best solution here.

Nope, this simple circuit is way beyond anything this ol' logic mechanic
would be able to do. And, not ashamed to admit it, either.

So, let's see whatcha got there Smitty?

Richard

In article et,
Richard Lamb wrote:

If it helps you to see this as a very difficult challenge, then go
ahead. I know what the "H" in RAH stands for, but if designing your own
circuit from discrete transistors that you've carved out of wood is
beyond your interest, then why not look around for a plug 'n' play
option? Have you *really* already done that and come up empty handed?
I've got a shelf full of catalogs of temperature control hardware. I
haven't actually looked for one with a rate alarm option, but since rate
is monitored and used by the controller, I don't see why someone
wouldn't offer such an alarm as an option.

Ok, Smitty.

Nope, not at all interested in carving transistors out of wood.

Since you have all the catalogs, my suggestion is to "go for it".

When you are all done, we can compare "budgets" and judge accordingly.
Budgets?

First, obviously, would be cost.
I'm pretty sure, should he decide to bite, Jim's solution would come in
for under 50 bucks.

Power budget?
We can assume we have a 12 volt DC electrical system. How many amps
will you need? Jim probably wouldn't need over 100 mA or so, if that.

Weight budget? (Oh Sh*t, not that again!)
What say, Jim? 3 or 4 ounces?

And then come the electrical issues!

It's a thermocouple. Output is in the millivolt range.
AND, to monitor RATE of change over a very short time (seconds) we'd
have to get down to the microvolt level!

To be able to use that, it will have to be scaled up - gain of roughly
one million or more? No freakin' way *I* would try that with desecrate
(or even wooden) transistors.

Again, it's a thermocouple!
How do you propose to "interface" to it? Single side ain't gonna get it.
You would wind up amplifying the noise as much or more than the signal!
Differential low noise high gain op amp seem to be the best solution here.

Nope, this simple circuit is way beyond anything this ol' logic mechanic
would be able to do. And, not ashamed to admit it, either.

So, let's see whatcha got there Smitty?

Richard

I'll pass on your proposed "contest." It's your project; I've got enough
of my own going. It's wonderful that your apparent kinship with Jim is
sufficient to entice him to design a custom solution, from parameters
that, despite his request, you have yet to clearly state (unless I
missed it, or that conversation took place offline.) Now, if you had
*really* looked around to see what was already available, and come up
empty handed, then a custom unit might make sense, provided you can get
all the engineering for nothing, which you apparently believe you can.
Sure, you can buy a big shoebox full of electronic components for your
$50 budget.

But, that's the way some engineers think. Fifty thousand different
screws available off the shelf, and we keep the Brown and Sharpes busy
making custom screws for engineers who design absurdly clumsy and
complex mechanisms with no forethought to how to hold the pieces
together.