开云体育

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of si_emi_01
Sent: Saturday, April 23, 2022 11:43 PM
To: [email protected]
Subject: Re: [TekScopes] 2465 Time-Marker Calibration Procedure

I spent a lot of time calibrating a 2465A for a nephew today. I have all of the Tektronix Calibration equipment to perform the calibrations.

I have also interleaved some responses to your questions as well.

Ross

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Harvey White
Sent: Saturday, April 23, 2022 6:26 PM
To: [email protected]
Subject: Re: [TekScopes] 2465 Time-Marker Calibration Procedure

Interleaved, at least what I can answer

On 4/23/2022 7:50 PM, jimbert4 via groups.io wrote:

1) The procedure calls for a Time-Mark Generator. Is this really just a pulse generator? What is the pulse width?

Tektronix time mark generators can put out narrow spikes with a sharp leading edge (so you can tell where it is). They generally have intervals from 1 ns down, although the 1,2 and 5 ns pulses are more sinewaves than not.

Ross - As Harvey has mentioned, the TG501 (Time Mark Generator), creates a double-exponential spike (Fast Rise, slow Fall - like teeth rising from the baseline or a RADAR PIP waveform). Harvey is correct, the most important characteristic is that the time interval is correct and that they are linear from division one to division Ten. Their amplitude is about 1.5V minimum. The pulse <position> changes based on what Calibration Step you are performing. You will need to provide 2ns, 20ns, 50ns, 100ns, 500ns 1us, 2us, 10us, 50us, 1ms Pulses. The 1/2 width-1/2 height duty cycle is about 30%. Using a Pulse Generator might get confusing. My <TG-501> produces good waveform fidelity out to 5ns, then it becomes a sine wave as Harvey also mentions.

Ross - If you are wondering, when they call out 10 divisions, they are also counting the left most division edge of the first centimeter. So, when you count over to the center graticule, it is actually 6 divisions. When you count the Step where it calls out 6 markers, it ends up being the center vertical graticule. On that same step, it asks you to adjust the waveform separation knob, drop the B-Delayed Sweep waveform down below the Main A-Sweep and you will see two waveforms nearly converging. Center up the first one over the center vertical graticule with the Delay Position first, then merge the other waveform on top of it with the Delta Knob, within 0.2cm. As you can see from the Cal 01 procedure, they want you to do that for all of the similar sections. It is usually easiest to use the Delay Position first then the Delta Delay.

If you are using a Pulse Generator, you may have a hard time with this. You may run into the same issue with the Bench Scope procedure too. You are also going to need an accurately times Sine Wave for the Transient Response test.

2) Is the Time-Mark a positive pulse, i.e. it does not go below gnd in its low state?

Wouldn't matter if it did, all you're looking at is the interval between leading edges.

Ross - Yes, it is a positive going double-exponential pulse except when it is a sine wave as discussed above.

3) What should the amplitude of the Time-Mark be set to?

Enough to give you a good presentation. The idea is time and risetime, not amplitude so much.

Ross - Harvey is correct. The answer is above.

4) The procedure says to adjust the deltaREF control to intensify the 2nd (for example) time marker and adjust the delta control to intensify the 10th time marker. Okay, I understand that. Then it says to superimpose the delayed B Sweep time markers within 0.2 division. Not sure about this. I see the 2 magnified (B?) sweeps, so am I supposed to adjust them to align the edges on top of each other? Is that like a fine control on adjusting the times? Does it matter which delta control(s) I use to make this fine adjustment?

This has been covered in similar questions, you might try 2465 calibration. Since I don't have one, I've not studied the procedure.

Ross - This is answered above. You will do this even when you do the stuff in Table 5-4.

5) The final Time-Mark setting is for 20ns (50MHz). My waveform generator only goes to 30MHz. Could I use a 40ns (25MHz) square wave and count edges instead of time marks?

I suspect so, but it needs to be 50% symmetric as a good guess.

Ross - You will actually need to generate a 2ns waveform for the Transient Response Test. When you move on to CAL 02, make sure that your PG506 is set up to generate a 100kHz Square Wave of the correct Amplitude. You can't just use a DC Supply to generate accurate Voltage Levels during the Automated Calibration Procedure. It has to have an active waveform or it will error (LIMIT), and you can't get past it in a well-behaved way. In addition to that, you will need the Fast-Rise Waveforms. Those are 100kHz are <1V and are about 875ps.

Ross - Lastly, don't get frustrated by it requiring you to go back a step. It saw that the setting was out-of-range. Just trim the Delayed Sweep waveform a little with the Delay Position knob. And, don't operate any know that it doesn't call out for you to move, it may invalidate your Calibration and you get to try it again.

It takes a good two hours to complete CAL 01 through CAL 07. It goes a lot smoother with the right equipment, though.

Harvey

Ross