Topics

Trace jitter

Reginald Beardsley
 

I managed to get to where I can see an ~1 ps jitter which correlates with the LED, but it is dwarfed by a 5-6 ps jitter with a very slow period. Does anyone know anything about this?

With the scope set a 2 mV/div & 1 ps/div and averaging 4 samples in points mode I see random horizontal dot alignments which appear to move up or down before vanishing. It has the effect of making the rising edge wave randomly. At this resolution it seems to me it's likely to be 1/f noise, but if someone knows of a correctable cause I'd be interested to hear about it.

The lack of schematics makes it difficult to check these. If someone has good relations with the Tek museum, would you please ask if the schematics could be released? My understanding is they got a treasure trove of information from a retiring employee, but that some of it is encumbered by 3rd party IP rights. If the entity that owned those rights no longer exists, it will never get released. No competent lawyer would ever approve releasing it.

The lawyers are likely to respond more quickly if asked narrow questions such as "Can we release the A10 board schematic?" To make that decision, they need only look to see if any other entity has IP rights. If all the design team were Tek employees, it's an easy call. A good lawyers job is to make sure their client doesn't have to go to court.

If the Tek museum made requests which were all separate and board specific, the Tektronix lawyers might fill them as something to do to finish out the day or while waiting for a meeting to start. Asking questions that take less than 30 minutes to answer is much more likely to get a quick response than asking a question which potentially could take weeks to resolve. I know there is at least one board with a 3rd party copyright notice in my 11801. For a lawyer, that's a landmine. Ask for that board schematic last.

Reg

Chuck Harris
 

The copyright notice to which you are referring is certain
to be for software in the EPROMS.

-Chuck Harris

Reginald Beardsley via Groups.Io wrote:

...

If the Tek museum made requests which were all separate and board specific, the Tektronix lawyers might fill them as something to do to finish out the day or while waiting for a meeting to start. Asking questions that take less than 30 minutes to answer is much more likely to get a quick response than asking a question which potentially could take weeks to resolve. I know there is at least one board with a 3rd party copyright notice in my 11801. For a lawyer, that's a landmine. Ask for that board schematic last.

Reg



Reginald Beardsley
 

I think that's probably correct. But a lawyer would have to make *sure* that was all it covered. That takes time whereas "Does Tek own all the IP for this board?" doesn't take much time.

Reg

Clark Foley
 

One unexpected source of jitter on the 11800 series is the timebase delay/position setting. The horizontal position is not phase locked to anything. As such, long delay times (e.g. greater than 1us) result in a wondering trace, horizontally. Verify that the horizontal position is very small, on the order of 10's of nanoseconds before messing with the hardware. The 11800 series gives great results for short delay times but not as you approach 1us.