Thanks for the video. My suggestion is to include a 50 ohm feedthru termination at the meter input. Impedance matching will make the measurements more repeatable. I'd also never trust amplitude calibration of oscilloscopes. They are normally plus or minus a few percent. The 848x RF power sensors (still available with USB varients, but are 50 ohm input and not megohm) also use balanced thermocouple sensors, as do many Fluke products.
I really appreciate all the testing and research you are doing on this meter. I have one, and I believe it is a rev.C (I have to check), and it was not working when I got it. I sort of remember replacing a couple of transistors with generic one (2n2222 or 2n2907/5 I don't recall), so I don't know if mine is calibrated, but I will have to check it out. Thank You again! Bill
Hello. Interesting video. I have bought one HP3400 from the us with a calibration certification and full restortion of it 2 years ago since am servicing Tape Decks and RF IF stages of FM and AM receivers. My question though which still messes me up is, could this meter be used as a VTVM, i.e could someone use it instead of a vtvm as if it is actually substituting a vtvm ? That is a question that could come from your experience in the field.. Thank you in advance.
Quite useful information. I am wondering, if Rigol Oscope have longer collection period (I have noticed it was about 35ms only) would it show more reliable and stable readings? I am asking since HP voltmeter has natural averaging over a few seconds that is much longer than selected Oscope collection time.
The thermocouple is two tiny wires only a few thousandths of an inch in diameter welded to a little heater like a light bulb filament. They are typically in a glass envelope in a vacuum. The vacuum improves life, stability and reduces conductive heat loss. When the filament heats up its resistance goes up. So relationship between heater power and input voltage is non linear because the heater resistance is changing. By itself the heater/TC assembly is a terrible RMS sensor because it is very nonlinear. You could use a nonlinear scale on the meter but it would be awful. The magic of this design is to turn this terrible nonlinear device into a system with a linear scale. The key is to use two matched TC devices. One is used in a feedback loop to drive it's heater with DC until its thermocouple voltage matches the other thermocouple whose heater is driven by the AC input signal. The DC voltage that heats the thermocouple to the same output as the AC voltage thermocouple is the true RMS value. Amazing performance for the sixties.
Thanks for the video. My suggestion is to include a 50 ohm feedthru termination at the meter input. Impedance matching will make the measurements more repeatable. I'd also never trust amplitude calibration of oscilloscopes. They are normally plus or minus a few percent.
The 848x RF power sensors (still available with USB varients, but are 50 ohm input and not megohm) also use balanced thermocouple sensors, as do many Fluke products.
Outstanding series on the HP3400A!
I really appreciate all the testing and research you are doing on this meter.
I have one, and I believe it is a rev.C (I have to check), and it was not working when I got it.
I sort of remember replacing a couple of transistors with generic one (2n2222 or 2n2907/5 I don't recall), so I don't know if mine is calibrated, but I will have to check it out.
Thank You again!
Bill
Aaah, that is interesting that you have one and it's a later model. A very nice meter, I'm glad you got it working!
Wow, you're discovering the old gear. (8^D)
This reminds me that I've gotta get my 400-series HP meter up and running. Thanks!
Hello. Interesting video. I have bought one HP3400 from the us with a calibration certification and full restortion of it 2 years ago since am servicing Tape Decks and RF IF stages of FM and AM receivers. My question though which still messes me up is, could this meter be used as a VTVM, i.e could someone use it instead of a vtvm as if it is actually substituting a vtvm ? That is a question that could come from your experience in the field.. Thank you in advance.
Quite useful information. I am wondering, if Rigol Oscope have longer collection period (I have noticed it was about 35ms only) would it show more reliable and stable readings? I am asking since HP voltmeter has natural averaging over a few seconds that is much longer than selected Oscope collection time.
is it possible that you are missing 50ohm termination in your cable ?
An analog precision voltmeter that can do as well as a good oscilloscope... Impressive equipment indeed. Great video, thank you.
Racal Dana 9303 true RMS can measure up to 2 GHz
@mjlorton :: *In looking on the auction sites for these meters, how do I know if it's a newer solid-state one, and what revision level it is?*
The thermocouple is two tiny wires only a few thousandths of an inch in diameter welded to a little heater like a light bulb filament. They are typically in a glass envelope in a vacuum. The vacuum improves life, stability and reduces conductive heat loss. When the filament heats up its resistance goes up. So relationship between heater power and input voltage is non linear because the heater resistance is changing. By itself the heater/TC assembly is a terrible RMS sensor because it is very nonlinear. You could use a nonlinear scale on the meter but it would be awful. The magic of this design is to turn this terrible nonlinear device into a system with a linear scale. The key is to use two matched TC devices. One is used in a feedback loop to drive it's heater with DC until its thermocouple voltage matches the other thermocouple whose heater is driven by the AC input signal. The DC voltage that heats the thermocouple to the same output as the AC voltage thermocouple is the true RMS value. Amazing performance for the sixties.