How test am503 + a6302 bandwidth

[r0d3z1]

Hello, I have recently purchased a used AM 503b and an A6302 probe. I have already tested the DC and low frequency performance. But I can't find a way to test probe over 500khz. The low frequency test has been done with AFG and 20 turn of wire around the probe.
Anyone know a simple method to test the bandwidth of this setup ?
Also, the BW limit on the AM 503B do nothing when i press it, it became red but there is no difference on the scope signal. In contrast, if i set the 20Mhz limit on the scope the difference is clear. Does it mean the the limit on the amplifier is not working ? Am I doing something wrong ?
Thanks

[alm]

Do you have a signal source that can go up to at least a few MHz? Or something generating a fast pulse?

I tested it with a sine wave oscillator that went up to 50 MHz. I connected a bnc cable to a female BNC connector with a non-inductive (I used metal film) 50 Ohm resistor soldered from center pin to shell, and would clamp the probe around the resistor. This gave me a usable signal at the most sensitivity setting. I would increase the frequency until the amplitude was -3 dB (1/sqrt(2)).

As for noise vs bandwidth limit, not necessarily. It could also be that the noise you see on the scope is dominated by the scope front-end and ADC. How does it compare if you set the input coupling to ground (ref)? Or if you disconnect the BNC cable to the AM503b?

[r0d3z1]

Do you have a signal source that can go up to at least a few MHz? Or something generating a fast pulse?

I tested it with a sine wave oscillator that went up to 50 MHz. I connected a bnc cable to a female BNC connector with a non-inductive (I used metal film) 50 Ohm resistor soldered from center pin to shell, and would clamp the probe around the resistor. This gave me a usable signal at the most sensitivity setting. I would increase the frequency until the amplitude was -3 dB (1/sqrt(2)).

As for noise vs bandwidth limit, not necessarily. It could also be that the noise you see on the scope is dominated by the scope front-end and ADC. How does it compare if you set the input coupling to ground (ref)? Or if you disconnect the BNC cable to the AM503b?

I have tried with a signal generator, a 100ohm ceramic resistor and 20 turn of wire to improve the sensitivity. However with this setup, already at 500khz the amplitude decrease a lot. This is probably due to the parasitic inductance of the setup. So you have simply used a 50ohm through hole resistor at the end of a BNC cable, what voltage do you use to generate the signal ?

[alm]

I don't remember exactly. Probably something like 2.5Vp. In a 50 Ohm load, that is 50 mA peak. As long as the load is reasonably resistive (this should be the case at least up to a few MHz), a voltage probe should give you a good estimate of the current through the resistor.

[r0d3z1]

I have seen that tektronix sell a calibration fixture 067-1686-00. Here is the pictures


it looks like a simple pcb with a track used as multiple turn and 2 100ohm load resistor.

Furthermore, there is this useful answer of joeqsmith in March 06 2015 from https://www.eevblog.com/forum/repair/tektronix-p6240-current-probe-repair/25/. Here is shown some diy test jigs.

[Hydron]

I'd run a sig-gen output (using a single turn only) through the probe using the shortest possible non-coax section, then terminate into a 50 ohm scope input (either built-in or using a 1M input plus a 50R in-line terminator). The terminated scope input gives a 50V/A reference to compare the am503/a6302 output to - even if the sig-gen isn't levelled you'll still get a valid result. Using a few volts of output should give enough signal to test at least the lowest few ranges of the current probe (note that you'll want to test more than one, as there is an amplifier output attenuator switched in/out between the 2mA/div and 5mA/div ranges in addition to the probe input attenuators).

Finally have a look at the service manual on TekWiki: http://w140.com/tekwiki/wiki/AM503 - it has all sorts of testing and calibration procedures listed in addition to full schematics. BTW in my AM503 unit the range switch and some electrolytics needed fixing/replacement.

[r0d3z1]

I'd run a sig-gen output (using a single turn only) through the probe using the shortest possible non-coax section, then terminate into a 50 ohm scope input (either built-in or using a 1M input plus a 50R in-line terminator). The terminated scope input gives a 50V/A reference to compare the am503/a6302 output to - even if the sig-gen isn't levelled you'll still get a valid result. Using a few volts of output should give enough signal to test at least the lowest few ranges of the current probe (note that you'll want to test more than one, as there is an amplifier output attenuator switched in/out between the 2mA/div and 5mA/div ranges in addition to the probe input attenuators).

thanks, you solution has been the simplest, I did so. Here are the results:

Yellow voltage, blue current, it looks right


however on the square wave there is this "ripple" overimposed on the static part of the waveform, is it normal ?


with a slower waveform, a noticeable noise is present, however here the current is very low so probably it is normal.


Could be this noise that cause the "ripple" of the second picture ?





 

[Hydron]

Rise time looks about right, but I don't recall seeing noise that looked like that on my probe - was more just broadband noise. Is the "ripple" on-top of the square wave static, or does it change over time? If static then the frequency response of the amplifier might need tweaking.

I've attached a couple of screen grabs I got when measuring the noise of my A6302 setup at the 1 and 10 mA/div (with 1 div = 10mV) settings on the AM503.