Some information on the early troublesome TDS500/500A scopes (I don’t recommend these), and the much nicer in every sense TDS500B/C/D or TDS700/A/C/D's. There are now schematics available for both the early and late oscilloscopes: TDS520 and TDS520B. The TDS520 manual has some fairly decent trouble shooting info. This video especially applies to: TDS744A (Shown), and its' close descendants the TDS754A/754C/754D and the black and white TDS540B/C/D. This info somewhat applies to the TDS600/600A (avoid as well), and the much better TDS600B or later. These scopes can be upgraded from 500MHz scopes to 1GHz scopes but they will need re-calibration which is not a trivial task (but you will get twice the sampling power/rate). I was not clear on the caps that fail, these are electrolytic surface mount capacitors (I use 33uF,35V, 105C, low ESR caps everywhere for the SMD’s; from a reputable supply house). The thru hole capacitors in the power supply or display driver board fail too, but not more than on any other instrument. I constantly misspoke and referred to attenuators and relays interchangeably. The relays sometimes don’t click, an easy fix, re-heat the attenuator pad on the coil +/- posts. All channels should “click” consistently between the channels, the click being the relays: 50/1M/AC/DC/GND/1X/10/100X; and a corresponding change with each change on the display. For the deoxit experiment mentioned, you would use a small gauge insulin syringe to inject deoxit directly onto the relay contact that are just to the right of the hole and the solid line shown. The attenuator is composed of 5 relays, some passives and an active wire bonded chip (no fixing that). Some attenuators failures are cold solder joints but not all the time, some folks have claimed that replacing the relays fixed their attenuators; but my impression is that they (or in fact that one guy?) had done this on only one instrument and hardly a definite answer. So it could be that on attenuators where the relays show appropriate mechanical function, that their contacts are no long up to snuff or that the active component has failed. When you return the ceramic attenuator to is carrier, don’t forget to replenish the thermal transfer paste which will have often dried up; which could be a root cause too (though good attenuators tend to work with or w/o, I’m sure the factory had their reasons to apply the paste). To do general SPC failure troubleshooting of the attenuators. Run the SPC routine 20 or 30 times to fill up the error log. And using the X @ b001 info in the video, start with whatever channel is mentioned the most, if the scope blurts out a human readable “CH1”, count that as well. You can move attenuators around to see if the problem moves as predicted. To get to the attenuators use to take allot of time to figure out because with many attenuator failures the scope rarely simply says “CH1”, rather 0 @ b001, and you have to take most of the scope apart to get to the attenuators, reassemble and so on. Took some time to figure out the error log bit. On some scopes, not all generations, if your failing spc it can help to loop the SPC gpib command "*CAL?". This sometimes fixed the spc issue outright! It’s not just exercising the relays, I wrote scripts to do that and that did not work as well as looping the "*CAL?". Scopes can also have memory errors as well, the later scopes: TDS500B/700 and later simply tell you which chip is failing, U212… The earlier scopes give you a memory address that you have to figure out. To check these scopes out, verify the scope passes all power on self tests, user initiated self test, can reliably pass the Signal Path Compensation routine. All entries in the Cal Menu should show PASS. Now check each attenuator for the following function: 50/1M/AC/DC/GND/1X/10X/100X. All channels should behave the same. It is also usefull to use a DMM to check the actual resistance at the BNC for the 50 and 1M terminations. This family of instavu scopes i.e., Digital Phosphor scopes started the whole fast acquisition thing with oscilloscopes. These scopes are still very capable scopes, great for the hobby bench. Much of this is discussed on the tek forum in greater detail, most folks who have the most experience working on these don’t contribute as that is bad for their business, or their a former Tek employee who has signed a non-disclosure agreement. The information is there, but you have to dig, the search engine there has been intentionally hobbled. Keeping the old scopes alive, in theory, is bad for new sales (therefore marketing too), and the service department guys want to be full-time employees too (I don’t think they service these anymore though, don’t know). Yes, the video rambles and at the end I was interrupted and had to stop, somewhat awkwardly, sorry.

• published: 23 Feb 2015
• views: 4702