Why does my analog scope have such a long internal lead going to the vert amp?

[Whales]

A few days back I was repairing my Goldstar OS-7040A:



The vertical amplification sometimes dies on this unit.  It turns out that the culprit was a bad solder joint on a connector.  This conn was on the vertical amp board and is where the signal enters from the front input board.

I'm boggled by the length of the cable that is used to connect these two things.  It's the fat blue cable in this photo:



This cable is about 2 to 3 feet long.  It only needs to be about 10cm long.  It's so long that it's wrapped once around the mains transformer and then once around the whole back of the unit before coming back.

Pulling it apart reveals that it's a shielded twin pair line:



... but it's unshielded for about the same distance as it would take a shorter cable to replace this entire big blue thing. 

I had a go at replacing it with just three cheap F-F 2.54mm jumper cables (blue/green/yellow in photo below).  You can see just how short of distance it is that needs to be spanned:



I decided to compare how the scope performs with the blue monster cable vs my cheap jumpers.  I fed in a square wave at a few MHz and took photos of both cases, then distorted these pictures in the GIMP to make them line up.  Blue is with jumpers, red is with long original cable:



A few things to note:

• My shorter cable gives slightly sharper corners
• My shorter cable gives me a slightly higher amplitude signal display (not visible, visually destroyed when I aligned the two pictures)
• Both attempts had very different phases (not visible, visually destroyed when I aligned the two pictures)

The last point has me thinking: was this wire intentionally long to introduce signal delay?  Otherwise it might have just been dodgy triggering that caused the phase issue.

Any thoughts?

[LazyJack]

Yes. This is a delay line, so you actually see what is happening near the trigger. All good analogue scopes have this, and should not be removed.

[Whales]

Thanks Lazyjack.  That makes sense.

My thoughts are that this is a fixed time delay, so it'll only be useful at some horizontal timebases/conditions?

[PA0PBZ]

Thanks Lazyjack.  That makes sense.

My thoughts are that this is a fixed time delay, so it'll only be useful at some horizontal timebases/conditions?

It's there to even out the fixed delay that your trigger circuit has, if it was not there you would no be able to see the signal that activated the trigger. So yes it is fixed like the trigger circuit delay is fixed too.

[Whales]

Ahah.  I was thinking more along the lines of a 'comfort' horizontal delay, so that the trigger point is eg 1/4 of your screen in.  This would only be useful at very specific timebases, given that the delay is fixed, so it would be a poor design choice unless you had selectable signal paths of different lengths.

[rstofer]

With a delay of, say, 2 ns per foot (depends on cable), all it is going to do is allow you to see the leading edge of the trigger regardless of scale.  It certainly won't displace the trace by 1/4 of the screen at 1 sec/div.

[LazyJack]

Ahah.  I was thinking more along the lines of a 'comfort' horizontal delay, so that the trigger point is eg 1/4 of your screen in.  This would only be useful at very specific timebases, given that the delay is fixed, so it would be a poor design choice unless you had selectable signal paths of different lengths.

No. You'll need a digital scope to do that. This delay line is just to compensate the delay of the trigger circuit. Obviously, the effect is most apparent when the signal is fast rising at the moment of the trigger and at fast sweeps. Without it you will not be able to see the rising (or falling) edge you trigger on. Depending on the scope, it can have up to several hundreds of ns of delay. Scopes above a couple of tens of MHz bandwidth have this, otherwise you won't see the point of trigger on the screen. Which is always where the sweep starts with analogue scopes. Only people already spoiled with digital scopes expect to see pre-trigger display :-)
The line is actually a quite complex distributed LC circuit,  having higher delay than a normal piece of coax (so you don't need a very long one). It is designed as part of the vertical signal path. Usually the line itself has a flat frequency response, but the transmit and receive ends are matched in the vertical circuitry and may contain compensations for the characteristics of the delay line. So you should definitely not remove it. It is there  for a good reason.

[David Hess]

Ahah.  I was thinking more along the lines of a 'comfort' horizontal delay, so that the trigger point is eg 1/4 of your screen in.  This would only be useful at very specific timebases, given that the delay is fixed, so it would be a poor design choice unless you had selectable signal paths of different lengths.

No. You'll need a digital scope to do that. ...

Challenge accepted.

If I put one of my fast timebases for faster triggering in my slow 7000 series mainframe which has a longer delay line, then at the faster sweep speeds supported by the mainframe, the pretrigger view is greater than 100% and the pulse edge is completely off the right side of the CRT.

As a practical matter, the pretrigger view time is usually about half of the delay line length so about 50 nanoseconds on a 100 MHz oscilloscope which would be 1 division or 10%.