The facility was Rockwell-Collins now a division of Raytheon all the engineering laboratories
have their own cal labs.
have their own cal labs.
I use probes for all AC, AF and RF.
My first design had a AC switch .
Most old VTVMS was only rate to 4MHz after that you use a probe.
After reading the best way to change the AC, AF and RF at probe and most VTVM you either change the probe to without a 1meg resistor or it had switch.
So came up with 3 probes from DC to 250Mhz. I have a design for 1.5 GHz but I do not think I would ever use over 120 Mhz.
I also use probe for 250 volts AC DC keeping the voltage inside the VTVM to 100 volts. I do think I will ever need over 125 volts.
Dave
My first design had a AC switch .
Most old VTVMS was only rate to 4MHz after that you use a probe.
After reading the best way to change the AC, AF and RF at probe and most VTVM you either change the probe to without a 1meg resistor or it had switch.
So came up with 3 probes from DC to 250Mhz. I have a design for 1.5 GHz but I do not think I would ever use over 120 Mhz.
I also use probe for 250 volts AC DC keeping the voltage inside the VTVM to 100 volts. I do think I will ever need over 125 volts.
Dave
Hi RJM1,
Yes, I know they do. I know a few engineers from Raytheon.
In this case, the lab failed you. Not all labs are always competent. I worked in a cal lab, one of 13 labs and the only Canadian lab. They were entirely staffed by ex-military in the US. We were always criticized for lower output, but our work was always in tolerance. Whenever the head lab sent techs in to show us how to do our job, their work was sometimes out of tolerance and had to be redone (we checked their work). Then they sent them up again with their calibrators this time. Again, their work was poor and calibrations and optimizations had to be redone. It turned into a Canada - US thing in the company, but we prevailed.
Calibrations must be done properly and carefully. Readings must be allowed to settle before being taken. Above all, factory cal procedures must be followed exactly and if there is any, and I mean any changes to the equipment, you must confirm the calibration. This means the case being removed or changed. By the way, I passed our internal quality audit. One of the few. The entire staff was qualified on paper, but it comes down to thinking and being aware of anything that may change the readings - then confirming those changes did not affect the readings.
Yes, I know they do. I know a few engineers from Raytheon.
In this case, the lab failed you. Not all labs are always competent. I worked in a cal lab, one of 13 labs and the only Canadian lab. They were entirely staffed by ex-military in the US. We were always criticized for lower output, but our work was always in tolerance. Whenever the head lab sent techs in to show us how to do our job, their work was sometimes out of tolerance and had to be redone (we checked their work). Then they sent them up again with their calibrators this time. Again, their work was poor and calibrations and optimizations had to be redone. It turned into a Canada - US thing in the company, but we prevailed.
Calibrations must be done properly and carefully. Readings must be allowed to settle before being taken. Above all, factory cal procedures must be followed exactly and if there is any, and I mean any changes to the equipment, you must confirm the calibration. This means the case being removed or changed. By the way, I passed our internal quality audit. One of the few. The entire staff was qualified on paper, but it comes down to thinking and being aware of anything that may change the readings - then confirming those changes did not affect the readings.
Hi Dave,
Nicely done. Your drawings are excellent by the way - very impressive!
It's cool seeing germanium diodes used in new product. I haven't looked, but I wonder if there isn't a current (new) low forward drop diode with very low leakage available. Interesting application for a new product if available,
I am very impressed by your work and thought in this. I still suggest you instal some protection diodes on the input of your buffer amp.
-Chris
Nicely done. Your drawings are excellent by the way - very impressive!
It's cool seeing germanium diodes used in new product. I haven't looked, but I wonder if there isn't a current (new) low forward drop diode with very low leakage available. Interesting application for a new product if available,
I am very impressed by your work and thought in this. I still suggest you instal some protection diodes on the input of your buffer amp.
-Chris
I have look in to protection and did put diodes on meter.
But chip LF353N cost me less than $1.00 if fails I have tube full of that chip. The chip goes down to 20pA hard to find something that would save chip.
Dave
But chip LF353N cost me less than $1.00 if fails I have tube full of that chip. The chip goes down to 20pA hard to find something that would save chip.
Dave
It's more a case of a chip failing "funny" where the input is damaged. If you can avoid this easily - you should. The diodes only need to clamp the voltage to either rail so you don't exceed the common mode range destructively. The chip maximum input current (20pA) has nothing to do with it.
Anyway, easy enough to tack a couple diodes from each rail to the input. Most signal diodes, like 1N4148 (silicon signal) would work. A low leakage part would be better. One example, and there are probably better parts for this, 1N4149. 100V, 500 mA - not the interesting specs. 25nA leakage @20VDC and 2 pF junction capacitance 5nA leakage spec is at 75 V reverse voltage applied as a limit.
Anyway, easy enough to tack a couple diodes from each rail to the input. Most signal diodes, like 1N4148 (silicon signal) would work. A low leakage part would be better. One example, and there are probably better parts for this, 1N4149. 100V, 500 mA - not the interesting specs. 25nA leakage @20VDC and 2 pF junction capacitance 5nA leakage spec is at 75 V reverse voltage applied as a limit.
I will look in to that diode
I have look at putting a capacitor on meter to slow down the movement.
The meter is very fast
I would to see a DMM that fast.
Dave
I have look at putting a capacitor on meter to slow down the movement.
The meter is very fast
I would to see a DMM that fast.
Dave
Hi Dave,
Conversion speed on a DVM takes time. You could even use a switch to change the rate of change for the pointer depending on what you're doing.
I think you have something pretty great there already.
A meter movement is great to see trends. It's also easier to see an average on a noisy or variable signal. My 34401A's had to store readings for me to get the basic stats. The newer 34461A and 34465A both have a trend chart on the display as well as the math functions. But either way, a meter movement shows things in real time a digital meter can't quite yet.
Conversion speed on a DVM takes time. You could even use a switch to change the rate of change for the pointer depending on what you're doing.
I think you have something pretty great there already.
A meter movement is great to see trends. It's also easier to see an average on a noisy or variable signal. My 34401A's had to store readings for me to get the basic stats. The newer 34461A and 34465A both have a trend chart on the display as well as the math functions. But either way, a meter movement shows things in real time a digital meter can't quite yet.
Many DMMs measuring DC use different time constants according to the voltage range.
The integration time is very long when the meter is in its lowest DC voltage range (millivolts or microvolts depending on the meter model).
That reduces the "Bounce" from the meters internal noise.
Know Thy Test Equipment
The integration time is very long when the meter is in its lowest DC voltage range (millivolts or microvolts depending on the meter model).
That reduces the "Bounce" from the meters internal noise.
Know Thy Test Equipment
Absolutely!
You can also often choose integration time over x power line cycles, and AC filters for DC measurements. The 34465A has AC filters for AC measurements at 3, 20 and 200 Hz. I ignore the first readings anyway, so variable integration time vs range isn't something I notice typically.
Thank you.
It took a few months to work out the little problems. From January to April of this year. The last was adding adjustments to all ranges.
The most costly part was case $26.00.
Dave
The HP 34401 is very fast because it has a vacuum fluorescent display. A friend of mine bought a new Keithley DMM and the color display is lightning fast. https://www.tek.com/en/products/keithley/digital-multimeter/dmm6500
Reminds me of the Digital RPM meters on cars.
Can't be read as the RPMs are changing, just a blur.
A needle, even with its ballistics is more useful and practical, even if it is less accurate.
Can't be read as the RPMs are changing, just a blur.
A needle, even with its ballistics is more useful and practical, even if it is less accurate.
Hi 6A3sUMMER,
Fully agree. Each display type has it's applications.
Hi astouffer,
Yes. I bought a Keysight 34461A and 34465A. OLED displays (rather have something else to be honest). They are pretty, but do not last and at the cost of these meters they ought to!
One big issue with speed. Someone at Keysight decided that these meters should operate like ATE equipment. Practically, this causes a massive problem. On resistance, these meters can lie - big time! The operator can't tell sometimes and it has cost me. The rate is too high, and when you have reactance in the circuit it can throw off the autorange and also cause incorrect resistance readings. Kelvin or normal is the same. No other meter I have tried has this issue and it makes the Keysight meters unsuitable for bench work. I have talked to them, and their official stance is that they are designed to measure pure resistance only. I'm calling BS on that. In every other way, I love these meters except for the boot time.
My response ... the right answer is worth one hell of a lot more than a fast answer.
Fully agree. Each display type has it's applications.
Hi astouffer,
Yes. I bought a Keysight 34461A and 34465A. OLED displays (rather have something else to be honest). They are pretty, but do not last and at the cost of these meters they ought to!
One big issue with speed. Someone at Keysight decided that these meters should operate like ATE equipment. Practically, this causes a massive problem. On resistance, these meters can lie - big time! The operator can't tell sometimes and it has cost me. The rate is too high, and when you have reactance in the circuit it can throw off the autorange and also cause incorrect resistance readings. Kelvin or normal is the same. No other meter I have tried has this issue and it makes the Keysight meters unsuitable for bench work. I have talked to them, and their official stance is that they are designed to measure pure resistance only. I'm calling BS on that. In every other way, I love these meters except for the boot time.
My response ... the right answer is worth one hell of a lot more than a fast answer.
Hey Dave,
That is the door price for good instruments today. My 34401A's cost me around that (Canadian) and I have three. Check out how much a 34465A is. These are not even close to the most expensive. I'd like a 34470A, and a 3458A - but that is a lab meter and I can't justify it.
That is the door price for good instruments today. My 34401A's cost me around that (Canadian) and I have three. Check out how much a 34465A is. These are not even close to the most expensive. I'd like a 34470A, and a 3458A - but that is a lab meter and I can't justify it.
I saw one the other day that was $8,000.00
I can do a lot of building for that price.
Dave
FYI I look at a HP 410 and was thinking of build to 110meg ohm volt.
I decided that 11meg was right for all my work.
I can do a lot of building for that price.
Dave
FYI I look at a HP 410 and was thinking of build to 110meg ohm volt.
I decided that 11meg was right for all my work.