Best reference after LTZ1000
Hello.
First of all, my greetings to the partecipants of this mailing list.
I found it very interesting and full of competent people, a rarity nowadays.
I whish to thank you all for that.
Then, let's go to my question.
I am building a precision current source (10uA/100uA) I want to use with an
Agilent 34401A and a RTD PT1000 probe to read temperature with good accuracy
(good is a relative term, actually I would be happy to be limited by the
precision of the probe, a 1/3 B DIN class).
Having already the 34401A I thought it was a good (and cheap) solution
vs. buying a complete thermometer, usually based on less accurate voltmeters
(apart some very very high price). Also, I like to experiment...
Apart the LTZ1000, that's exceeding the 34401A accuracy (and is probably
worth to be used for in-house calibration of it, but that's another matter
I want to pospone for the moment), what would you use, trying to stay in the
<20$ range?
Actually, I bought a MAX6325 of the best grade, some OP177, some good
0.1% 25PPM Vishay resistors, one 10K 0.01% 0.2PPM Z-Foil Vishay resistor for
calibration purpose of the current source...
I would also like to have your opinions about the trend of all manufacturers
who are moving away from buried zeners to go versus bandgap and XFETS, but
looking datasheets I begun to think the reason is more commercial (or
practical: high inital accuracy, low voltage, low power) than metrological...
while they publish exceptional data (the MAX6325 is declared 1PPM/C) seems that
precision instrument makers keep to use (heated) buried zeners... Where is the
truth?
Best regards,
(Mr.) Andrea Baldoni
Ermione s.r.l.
Via Provanone, 9008/D
40014 Crevalcore (BO) - ITALY
PI/VAT IT02745891206
Andrea Baldoni wrote:
Hello.
First of all, my greetings to the partecipants of this mailing list.
I found it very interesting and full of competent people, a rarity nowadays.
I whish to thank you all for that.
Then, let's go to my question.
I am building a precision current source (10uA/100uA) I want to use with an
Agilent 34401A and a RTD PT1000 probe to read temperature with good accuracy
(good is a relative term, actually I would be happy to be limited by the
precision of the probe, a 1/3 B DIN class).
Having already the 34401A I thought it was a good (and cheap) solution
vs. buying a complete thermometer, usually based on less accurate voltmeters
(apart some very very high price). Also, I like to experiment...
Apart the LTZ1000, that's exceeding the 34401A accuracy (and is probably
worth to be used for in-house calibration of it, but that's another matter
I want to pospone for the moment), what would you use, trying to stay in the
<20$ range?
Actually, I bought a MAX6325 of the best grade, some OP177, some good
0.1% 25PPM Vishay resistors, one 10K 0.01% 0.2PPM Z-Foil Vishay resistor for
calibration purpose of the current source...
I would also like to have your opinions about the trend of all manufacturers
who are moving away from buried zeners to go versus bandgap and XFETS, but
looking datasheets I begun to think the reason is more commercial (or
practical: high inital accuracy, low voltage, low power) than metrological...
while they publish exceptional data (the MAX6325 is declared 1PPM/C) seems that
precision instrument makers keep to use (heated) buried zeners... Where is the
truth?
Best regards,
(Mr.) Andrea Baldoni
Ermione s.r.l.
Via Provanone, 9008/D
40014 Crevalcore (BO) - ITALY
PI/VAT IT02745891206
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
A ratiometric conversion using the voltage reference to both set the
current and the DVM reference will negate slow reference drifts.
The stability to a first approximation then depends largely on the
precision resistor used to set the current and the RTD.
The internal current sources used for resistance measurement in the
34401A essentially allow such a measurement.
The drift of the various opamps used in the internal current source will
degrade this somewhat.
Bruce
On Sat, Aug 14, 2010 at 11:43:38PM +1200, Bruce Griffiths wrote:
A ratiometric conversion using the voltage reference to both set the
current and the DVM reference will negate slow reference drifts.
The stability to a first approximation then depends largely on the
precision resistor used to set the current and the RTD.
Very good idea. The ratio of VRTD vs. Vref
would directly be the ratio between RTD and the setting current resistor.
Since I don't really need a precise current source (using <100uA would
allow me to ignore self heating of the RTD) a simple resistor divider between
the RTD and a precision 10K resistor would do the same job using only a, say
1V source at this point...
The internal current sources used for resistance measurement in the
34401A essentially allow such a measurement.
The drift of the various opamps used in the internal current source will
degrade this somewhat.
The 34401A has the VDC:VDC ratio function, it has some limits but it's useable
for this purpose. The internal current source instead (for the resistor value
of interest) is 1mA, too much, it would cause significative self-heat in the
RTD.
I don't understand in which way you mean it is involved in this case.
Best regards,
Andrea Baldoni
Ermione s.r.l.
Via Provanone, 9008/D
40014 Crevalcore (BO) - ITALY
PI/VAT IT02745891206
Andrea Baldoni wrote:
On Sat, Aug 14, 2010 at 11:43:38PM +1200, Bruce Griffiths wrote:
A ratiometric conversion using the voltage reference to both set the
current and the DVM reference will negate slow reference drifts.
The stability to a first approximation then depends largely on the
precision resistor used to set the current and the RTD.
Very good idea. The ratio of VRTD vs. Vref
would directly be the ratio between RTD and the setting current resistor.
Since I don't really need a precise current source (using<100uA would
allow me to ignore self heating of the RTD) a simple resistor divider between
the RTD and a precision 10K resistor would do the same job using only a, say
1V source at this point...
Short term stability of the source is important as the voltage across
the 10k resistor and the RTD aren't measured simultaneously.
The internal current sources used for resistance measurement in the
34401A essentially allow such a measurement.
The drift of the various opamps used in the internal current source will
degrade this somewhat.
The 34401A has the VDC:VDC ratio function, it has some limits but it's useable
for this purpose. The internal current source instead (for the resistor value
of interest) is 1mA, too much, it would cause significative self-heat in the
RTD.
I don't understand in which way you mean it is involved in this case.
The internal current source is derived from the internal reference via a
few opamps and resistors plus some JFETS etc.
To a first approximation its value only depends on the value of a
resistor and some resistor ratios.
Thus the current tends to track drift in the internal reference.
Best regards,
Andrea Baldoni
Ermione s.r.l.
Via Provanone, 9008/D
40014 Crevalcore (BO) - ITALY
PI/VAT IT02745891206
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
Bruce
On Sun, Aug 15, 2010 at 08:47:03AM +1200, Bruce Griffiths wrote:
The internal current source is derived from the internal reference via a
few opamps and resistors plus some JFETS etc.
To a first approximation its value only depends on the value of a
resistor and some resistor ratios.
Thus the current tends to track drift in the internal reference.
Maybe I start to understand.
I may use the 4W ohm measurement, derive a Vref from the force connection
using a stable resistor, for instance 1V from 1K; use a good operational+MOS to
make a current source from this Vref, say 10uA using 100K; measure the drop over
the RTD with the sense connection and voila', the reading directly in ohm x100.
This way, no need for an external source at all and also the meter could read
at full speed because it has not to multiplex between the two inputs like in
ratio VDC:VDC.
Very nice!
As soon as I receive the OP177s I'll try it. I have just to pay attention to
track the heating of the 1K resistor with the 100K, because 1mW is not to
understimate.
The same could be done in reverse, to make a microohmeter out of the 34401A.
Best regards,
Andrea Baldoni