Tek 465 low bandwidth.

[Zenwizard]

Working on a Tektronix 465 going through the cal document and the bandwidth is very low. It is hitting 3 db down at about 40 Mhz. Instead of the 100 Mhz it is rated for. I have checked the signal generator on a good scope and it is working just fine. The bandwidth is limited on both channels just about evenly so I am starting to suspect it is in the final amp that the issues are showing up. This is a 25XXXX serial number scope non B. This one is the non IC version it is all discrete.  Any ideas on where to start looking to sort this one out?

[David Hess]

The bandwidth limit switch commonly causes this problem.  It should be cleaned and lubricated.

[Zenwizard]

Got the switch cleaned up. When the scope is in 20 Mhz mode the bandwidth is pretty close. When it is in full bandwidth mode it is still stuck at 40Mhz. If pictures will help let me know and I can provide them. Thanks for all the help in getting the lab squared away.

[Chris56000]

Hi!

What signal generator are you using to verify the bandwidth with?

It's possible there could be a rooloff in the capacitance of the connecting cable between your generator and the oscilloscope – I believe there's a special type of cable specified for the bandwidth verification with an extremely low self–capacitance!

It's rare for faults in vertical amplifiers to cause such a drastic loss of bandwidth to the extent of 60%, but if you're absolutely certain you're not getting a major rolloff before your 'scope's input socket, I'd begin with all the final stages decoupling arrangements, as a yukky decoupling capacitor will add the H.T. supply impedance to the loads of the final amplifier transistors!

Chris Williams

[Zenwizard]

Hey Chris, that is why this one is so weird. I am using a SG503 signal generator. I have tested the cable, 10x attenuator and 50 ohm terminator on a Rigol MSO 1104Z. The test in the cal document is set a 50 Hz sign wave to 5 div. Then crank up the frequency till you get 3.5 div of signal. I get 3.5 div of signal at around 40 Mhz and only 2.8 div of signal at 100 Mhz. Scope is not triggered and should not be for this test per the service manual. I have also done the same test with the type 191 with the exact same results. On the Rigol there is 4 Div of display at 101Mhz on the 465 it is about 2.8 Div at the same 101 Mhz. I am still wracking my brain on this one. Also for testing can cal all equipment was given 20 minutes minimum warm up scope was in a cal process so over 2 hours warm up time. If Pictures would help i can defiantly do that.

[particleman]

My 465 manual says set 465 to .2ms both VOLTS/DIV switches to 5mV and the sig gen to 5 divisions of  3 megahertz.
Do you have a fast rise square wave? I think that would tell something also.

[Zenwizard]

I double checked the cal document it has changed it seems. The cal document I am using is for serial number 25xxxx and up. . I am stumped on this one... I do have a square wave generator in the pg506.

[particleman]

What does the rise time from the PG506 fast rise look like? That should at least tell you something.

[macboy]

Since it is only -2 dB difference from 40 to 100 MHz, it is possible that the input compensation networks need adjustment. I would definitely use a fast edge square wave to test the rise/fall time of the scope. If over compensated, you can get a slow rise time and an early roll-off of frequency response. If under compensated, step response rings, and you see peaking in the frequency response.

[David Hess]

Got the switch cleaned up. When the scope is in 20 Mhz mode the bandwidth is pretty close. When it is in full bandwidth mode it is still stuck at 40Mhz. If pictures will help let me know and I can provide them. Thanks for all the help in getting the lab squared away.

What does the rise time from the PG506 fast rise look like? That should at least tell you something.

This was going to be my next suggestion.  The high frequency transient response can reveal a lot about what is going on.

It's rare for faults in vertical amplifiers to cause such a drastic loss of bandwidth to the extent of 60%, but if you're absolutely certain you're not getting a major rolloff before your 'scope's input socket, I'd begin with all the final stages decoupling arrangements, as a yukky decoupling capacitor will add the H.T. supply impedance to the loads of the final amplifier transistors!

The last stages of the vertical CRT amplifier require correct bias for maximum bandwidth.  One version of the 465 has an adjustment for this; during calibration, it is suppose to be adjusted for maximum bandwidth.

[JimS]

In the manual I have (Tek PN 070-1330-00) On page 5-17.
(Short form cal)

Step 7 is - adjust vertical output bias for max signal deflection.
It also says that the adjustment  will effect vertical risetime, aberration, and position effect.
And may be slightly misadjusted for optimum performance of the vertical system.

On page 5-18 there is step 22 Adj low freq response.
And step 24 adjust high freq response.

Page 5-24 is the start of the full calibration/adjustment process.

[David Hess]

There is more than one version of the vertical CRT amplifier in the 465 so you may not have a bias adjustment.  I just mentioned it in the event that you do and if you don't, I would check the bias voltages to make sure they are correct.

[Zenwizard]

I Think I found the missing bandwidth!! I found 2 different variable caps were bottomed out one in each channel... a different one in each channel. the wave form pictured is of a 1 nS rise time out of a pg505 50 Ohm terminated and scope on 5Mv/Div. Channel one 3db down is now 105 Mhz. Channel 2 is 135 Mhz as measured on a SG503. Let me know how I did. I am still learning the art of analog scope calibration.

Sweep rate is .05 us and 10X magnification is on.