[R-390] Ramp Generators

[David Wise]

They are similar but the goal is different.  In an A/D, you're measuring how long it takes the integrator to follow the signal.  In the scope, the ramp itself is the output.  The Tek caps are very good, but they don't need the extreme linearity and stability (leakage, dielectric absorption, tempco) required for a DVM-grade A/D.  I described the Miller Integrator mostly to show there's more than one way to implement a current-source-charging-a-cap.

Dave
________________________________________
From: Barry <n4buq at knology.net>
Sent: Friday, April 26, 2019 11:13 AM
To: David Wise
Cc: Roy Morgan; r-390 at mailman.qth.net
Subject: Re: Ramp Generators

I think that same scheme is used in HP's digital voltmeters (e.g. HP3456A, etc.) as part of the sampling circuit(?).  As I recall, the integrator cap has to be very high quality for it to work correctly.

Thanks,
Barry - N4BUQ

----- Original Message -----
> From: "David Wise" <David_Wise at Phoenix.com>
> To: "Barry" <n4buq at knology.net>, "Roy Morgan" <k1lky68 at gmail.com>
> Cc: r-390 at mailman.qth.net
> Sent: Friday, April 26, 2019 12:16:41 PM
> Subject: Re: Ramp Generators
>
> The canonical Tektronix ramp generator is the "Miller Integrator" used in the
> sweep section of most of their old analog scopes.  Here, the constant
> current comes through a resistor.  It's constant because the voltage across
> the resistor is constant.  The voltage is (nearly) constant because it's the
> input voltage to a high-gain DC amplifier with feedback.  This is your
> generic "operational amplifier" circuit: the output will do whatever is
> necessary (and possible) to maintain the "virtual ground" at its input.  The
> resistor forms the low leg of the feedback network.  The cap, wired from amp
> output to input, forms the high leg.  With a constant current trickling into
> the low end of the cap, the amplifier has to elevate the high end at a
> constant rate to keep the low end from drooping.
>
> Dave Wise
> Tektronix, 1980-1995