Tektronix 475A calibrator

[maginnovision]

I have a Tek 475A I'm giving away and I wanted to try and get the calibrator signal a little closer to ideal. If anybody has any ideas I'm open to them. The voltage rails are good with very little ripple and both test points show the right signal but also not 50% duty cycle.

[tggzzz]

The duty cycle is completely unimportant. The absolute amplitude isn't expected to be perfect.

All that matters is that the output waveform has clean edges and a flat top.

[tautech]

The duty cycle is completely unimportant. The absolute amplitude isn't expected to be perfect.

All that matters is that the output waveform has clean edges and a flat top.

Yep exactly however on some old scopes you could adjust DC and amplitude of probe Cal output.

However it's nice to have it close as it serves as a quick sanity check that all is in roughly good order.

[maginnovision]

Yea, this is going to someone who has never used an analog scope so I wanted to get it a bit closer to 1kHz 50% duty cycle. They have used a modern digital scope though(siglent 50MHz unit). I never cared until right now. I replaced the original C1505 with a newer 400V DC rated .068uF capacitor and the -PW went down 5us and +PW went down 20us... That might be about all I can do. I just want the future owner to be happy with it.

[MarkL]

You should use DC coupling on your scope and check the compensation adjustment on your probe.  Either could be contributing to the non-flat display.

[tggzzz]

Yea, this is going to someone who has never used an analog scope so I wanted to get it a bit closer to 1kHz 50% duty cycle. They have used a modern digital scope though(siglent 50MHz unit). I never cared until right now. I replaced the original C1505 with a newer 400V DC rated .068uF capacitor and the -PW went down 5us and +PW went down 20us... That might be about all I can do. I just want the future owner to be happy with it.

There's zero difference between analogue and digitising scopes in this respect; it is a probe issue.

The principal use of the calibrator is to allow a user to alter the LF AC response of a *10 "high" impedance probe. You twiddle the trimcap in the probe until the top is flat.

In your picture you are using AC coupling for unknown reasons, but that can sometimes look like an overcompensated probe - it is difficult to tell with such a noisy trace.

[maginnovision]

Ok, here is a less noisy trace with DC coupling although it's only slightly related to the OT(±PW). The issue with him only having used newer digital scopes, for me, is that I have yet to see a digital scope that didn't have 1kHz and almost exactly 50% duty cycle, as the SS I've attached shows. So when he sees the signal walking he may think something is wrong when it isn't. So my idea is to try and adjust it as close as possible and also tell him that it's normal. I haven't even had enough time with it yet to run a self cal(roughly 5-10 minutes at a time, hours apart) on the keysight but I already shorted the test points and adjusted amplitude for the cal signal.

[tggzzz]

To understand how you use a scope calibrator to adjust the probe compensation, see any user manual for either a scope or a probe. Alternatively, just twiddle the probe compensation trimcap and watch the difference. At that point you will understand why the precise frequency and duty cycle is irrelevant.

The trace on the rtb2004 looks like a well calibrated probe.

The trace on the keysight is dreadful; something is very wrong there. I've never seen a trace like that.