Thermocouple gauge voltage vs. pressure calibration

Every fusor and fusion system seems to need a vacuum. This area is for detailed discussion of vacuum systems, materials, gauging, etc. related to fusor or fusion research.
Post Reply
ktu
Posts: 48
Joined: Sun Jul 21, 2002 2:53 pm
Real name:
Contact:

Thermocouple gauge voltage vs. pressure calibration

Post by ktu » Tue Dec 24, 2002 4:42 pm


Hello,

Does anyone have function for the voltage vs.
pressure relationship of thermocouple gauge
KJL-6000 aka DV-6M?

As well, this model of gauge tube uses only three
connection pins - two for AC filament drive and
one one connecting 55 ohms between the pin and
center-tap of AC filament transformer coil.
I don't like to open a working tube, but has anyone
opened and perhaps photographed one? Does it really
have three individual thermocouples? Some sources
state that the two are for temperature compensation
and third for pressure sensing.. Anyone know what
thermocouple materials are used?

Oh yes, I'm building a circuit for measuring with
the darn tubes, and I'm really kicking myself for
choosing this darn three pin tube instead of nice
easy four pin tube with separate thermocouple. :(
Now I have to use the transformer etc. instead of
simple DC circuits. Oh well, at least these are cheaper.

Kristian Ukkonen.

DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 5:13 am
Real name:

Re: Thermocouple gauge voltage vs. pressure calibration

Post by DaveC » Wed Dec 25, 2002 5:02 am

Kristian -
Don't know about the TC tube you have, but... some use the midpoint of the filament to attach the TC directly and then measure the DC current between that point and the filament center tap. Often there is a low resistance adjustment to minimize the AC content to the TC current. I think it is a low cost method that eliminates the need for a DC filament supply.

I am working strictly from memory here answering your question about the TC gage functional response.

The heat transfer from the heated filament at low pressures (below a few Torr) is almost exactly a linear function of pressure, which means the temperature rise of the filament (at more or less constant power) is then inversely proportional to pressure. Lower pressure, higher temperature. The thermocouple response (Seebeck function) is logarithmic in temperature...more or less... so that TC voltage (and thus gage current ) is a approximately proportional to the negative logarithm of pressure.

Perhaps someone has the exact function at hand... it is in the Handbook of Chemistry and Physics, listed as the Thermocouple Equation .

Dave Cooper

User avatar
Richard Hull
Moderator
Posts: 12431
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Thermocouple gauge voltage vs. pressure calibration

Post by Richard Hull » Fri Jan 03, 2003 8:29 pm

All tubes have different characteristics and, thus, different calibration graphs.

A couple of great articles, some detailed hookups and calibration tables are given for a couple of tubes in older issues of the belljar.

http://www.belljar.net

Unfortunately, they are not on line and you may have to contact Steve Hansen about the specific back issues.

The best tube to get is also the cheapest tube (the 1518 - $45.00). These have expanded lower ends (scales open up to higher res at the low micron end) Another plus for the 1518, the average bias current is 15-18ma which means the entire meter project built around the tube can use a 1.5 volt "D" cell. Th DV-6 tube is also "opened up" at the lower end, but might need about double the bias current.

Finally, the output of most tc tubes is designed to function specifically with a special meter designed to read a full scale 10mv (not common). A few are spec'd to a 9.5mv meter input.

I found a couple of meter movements from old TC gauges at a hamfest in a box of meter movements for one dollar each. They were both 10mv rated with micron scales in place. So a fully cal'd TC gauge system was made with a D cell, meter, potentiometer and a $45.00 1518 tc gauge tube.

Ricahrd Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

ktu
Posts: 48
Joined: Sun Jul 21, 2002 2:53 pm
Real name:
Contact:

Re: Thermocouple gauge voltage vs. pressure calibration

Post by ktu » Fri Jan 03, 2003 9:55 pm

Richard Hull wrote:
> All tubes have different characteristics and,
> thus, different calibration graphs.

All too well known. I have the BellJar article
on 531 and 5818 tubes. These have four pins connected
and separate T/C.

> The best tube to get is also the cheapest tube (the
> 1518 - $45.00). These have expanded lower ends
> (scales open up to higher res at the low micron
> end) Another plus for the 1518, the average bias
> current is 15-18ma which means the entire meter
> project built around the tube can use a 1.5 volt
> "D" cell. Th DV-6 tube is also "opened up" at the
> lower end, but might need about double the bias
> current.

KJL-6000 aka DV-6M requires 21mA. I do already have
enough of these KJL-6000s that changing tube is not
an option. Besides, these are the cheapest tube here.

> Finally, the output of most tc tubes is designed
> to function specifically with a special meter
> designed to read a full scale 10mv (not common).
> A few are spec'd to a 9.5mv meter input.

Yes, and 55 ohm resistance of meter.

I've built now two circuits. One uses DC, 21mV to
filament, and measures T/C voltage across 55ohm
from third pin to a voltage at center of filament,
made with opamp buffer (input from voltage divider
across filament). With this I get 0.78mV to 18mV
from 1ATM to vacuum.

The other circuit uses 200kHz excitation for filament
via transformer (secondary with centertap), and 55 ohms
between centertap and third pin of tube, and measures
DC voltage across the 55 ohms. With this setup, I get
0.4mV to 13.1mV. Too much.

With my commercial solid-state meter for KJL-6000 tubes,
operating at 25kHz, I get exactly the 10mV at vacuum
when I measure it from the electronics.

I really can't figure why I get too much voltage, instead
of 10mV with my circuit. Perhaps 200kHz is too
high frequency. I just used one ready IC operating
at that frequency to get the transformer drive,
so it is not easily changeable. Really hate to have
to do the 50% dutycycle oscillator and push-pull driver
from discrete parts.

Kristian Ukkonen.

ktu
Posts: 48
Joined: Sun Jul 21, 2002 2:53 pm
Real name:
Contact:

Re: Thermocouple gauge voltage vs. pressure calibration

Post by ktu » Sun Jan 12, 2003 3:34 pm

Kristian Ukkonen wrote:
> The other circuit uses 200kHz excitation for filament
> via transformer (secondary with centertap), and 55 ohms
> between centertap and third pin of tube, and measures
> DC voltage across the 55 ohms. With this setup, I get
> 0.4mV to 13.1mV. Too much.

So now, after measuring a commercial TC gauge controller,
I get the impression that the tube filament is supposed
to be fed with constant VOLTAGE of 0.38V, and NOT with
constant current of 21mA. There just happens to be 21mA
with 0.38V at 1 ATM, but definitely not at vacuum.

So, can someone confirm that the filament should be
driven with constant voltage?

This would mean that, for example, the Bell Jar article
on TC gauges is perhaps not accurate.. It only uses
a voltage source and a resistor in series with the
filament to set the CURRENT. Current when? - with the
tube at 1 atm or at vacuum.. The resistance of the tube
filament will be different - for KJL-6000 about 14ohms
at 1 atm, and nearly twice that at vacuum, with 21mA
constant current supply. Perhaps 531 has different design,
but not likely.

Too bad there aren't specs available for these tubes
from manufacturers. Not even how to feed the filament.
I've asked KJL about these tubes, and no reply. :(
The Hastings DV-6M specs do not mention this either,
whether CC or CV supply should be used for filament.

Kristian Ukkonen.

Richard Hester
Posts: 1458
Joined: Thu Jul 05, 2001 4:07 am
Real name:

Re: Thermocouple gauge voltage vs. pressure calibration

Post by Richard Hester » Sun Jan 12, 2003 10:55 pm

The Hastings people have been very nice and forthcoming with technical information in the past. They were very helpful to me when I was trying to fix an ailing DV-5M. I suggest that you e-mail them. They post pictures of the meter scales on their web site that might be directly usable.The AC source used in the Hastings gauges is a very simple 2 transistor saturating core oscillator. If I can find my manual for the DV-6M, I might do a scan and temporary post. Once tha approximate operating frequency is known, it wouldn't be too hard to wind up a transformer using a small ferrite core. Other gauges use a conventional 60Hz transformer secondary clipped to a quasi-square wave with back-to -back zener diodes. The 531 gauges I have seen use this principle.
All the schemes I have seen for driving gauge tubes are pretty crude, no doubt as the tubes are not too accurate to begin with, and are generally used for indication purposes only. You won't find anyone on the market that will guarantee really useful accuracy for a thermocouple gauge. Still, it's better than nothing.

User avatar
Richard Hull
Moderator
Posts: 12431
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Thermocouple gauge voltage vs. pressure calibration

Post by Richard Hull » Mon Jan 13, 2003 9:31 pm

I have always used the running current method as most all tubes I have ever investigated specified running heater current at "gauge zero" or max vacuum.

In the case of the 1518, my favorite, the current is usually spec'd at around 18ma. Most other types of TC tubes are in the 20-60ma range, though I believe the 531 and a couple of others is in the 135ma range. What this means is that under no circumstances should this filament current be exceeded.(for it is OVER current that burns out any filament.) Thus, this sets an approximate value for a ZEROing pot and limiting resistor. With a 1.5volt battery the shorting resistance to draw 18ma would be 1.5/.018 = 84 ohms. as the filament will require some drop, we might try a 47 ohm fixed resistor in series with a 50 ohm wirewound pot for a zeroing setup with the 1.5 volt battery. To setup the zero, the pot should be run all the way out to max resistance. Thus throwing 97 ohms plus the tube's filament or heater in series with the battery, thereby drawing far less than the required 18ma. (safety)

A vacuum is pulled on the tube and, hopefully, pulled to at or below one micron. The pot can now be adjusted down in resistance so that the appropriate spec'd meter reads ZERO microns or close to it. (9, 9.3 or 10mv across the TC in the tube) Zero microns is always full scale defection.

At this point, a current meter can be installed in the heater circuit and it should, ideally be at the inked in number on the side of the 1518 tube (~18ma). Pretty straight forward.

Some tube contollers actually have two meters or one meter with a CAL and READ two position switch. In the cal mode a pot, the zeroing pot, is turned to make the needle go to a maked CAL point on the meter face (this sets the specified tube CURRENT). Placed back in the READ position the meter is supposed to be reading correctly +/-10% with air only!!! I personally check all of my TC gauge Cals against a capacitive manometer and adjust the pot to match even though a milliamp or two more or less than the tubes rating. No TC tube can be made tube to tube with a perfectly known cal current.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

Post Reply