At 09:38 PM 1/19/98 -0500, you wrote:
>>* Less than 10nA input bias current
>
>You can get around the bias current limitation by providing another path
>for the bias current to flow rather than through the input terminating
>resistor.  Put another large resistor at the input connected to an
>adjustable voltage and adjust the voltage until the output voltage is
>zero both with the input open and with it shorted.  The input resistance
>is then approximately the two resistors in parallel (one to supply, one
>to ground).  This is a good idea even if the amplifier you use has a
>rather low bias current, as it prevents current from flowing out of the
>scope and into the circuit under test.
>

Mike, thanks for the idea. I had thought of something similar but I didn't
think that it would be possible to compensate like this without causing a
big DC offset. I'll try it.

>>* OK to operate off +5 and gnd (not necessarily input range incl. GND,
>>just
>>doesn't need more than 5v spread from Vcc to Vee)
>
>You may need a negative voltage to supply the bias current, but this is a
>good idea anyway since using a dual-supply op amp opens up a lot more
>choices.  Without a negative supply to the front-end, I'm not sure how
>you intend to be able to measure negative input voltages without lifting
>the input ground from the rest of the circuit ground (a rather bad idea).
>

I did lift the input ground from the circuit ground, but not just with a
resistive divider, I have a slow speed op-amp with a resistive divider as a
reference, the op-amp then feeds a complimentary pair of bipolar
transistors and the negative input of the slow op-amp goes to the output of
the pair, and the output of the pair is also bypassed to ground through .1
uF and 22uF caps. Not the best solution , I realize, but I wanted the whole
thing to operate off a single supply for simplicity.
For scope inputs which do not pass DC, I have a resistor (value not yet
selected) across the input.

>A real simple way to get a few mA of bipolar supplies would be too use
>the voltae converters in a MAX232 style chip.  Linear regulators after
>the converter would be a good idea for consistent performance and to help
>remove switching noise.
>
>>* Less than 10mV input offset voltage
>
>Don't get too hung up on maintaining a zero DC offset through the
>amplifier, it makes it rather complicated.  Many, many useful scopes
>aren't designed to measure absolute DC input voltage (i.e. the VERTICAL
>POSITION dial isn't calibrated).  Digital scopes are lousy as voltmeters
>anyway due to the small number of bits in the converter.
>

Well, I'm not bothering to compensate for offset voltage, so I wanted it to
be low initially. a 10 mV difference times the 10x gain would place a .1 V
error (or 5 percent DC error on the whole 2V scale of my ADC) on the ADC. I
was considering this to be about the upper limit for the contribution from
offset voltage.

>>* Voltage feedback
>
>If you're configuring the amplifier as a noninverting amp with the signal
>applied to the + input and feedback at the - input, a current feedback
>amp will work fine.  They just have a low impedance at the - input
>terminal, requiring the use of smaller resistors in the feedback network.
> With proper feedback maintained, the + input terminal is voltage
>sensitive and high impedance.
>

I must admit, I was opting for voltage feedback mainly out of ignorance of
current feedback op amps, however, I have explored a few current feedback
ones and didn't find them much superior for my application.

>>If not, does anyone know of a buffer that can at least provide a flat
>>gain
>>of 1 and has the rest of these characteristics?
>
>I'd be tempted to try JFET source followers between the input and the
>voltage gain stage.  DC offset may go out the window, but my opinion is
>that that isn't too critical in a scope.
>
Well, I guess I may have to sacrifice the DC accuracy anyway so ultimately,
this may end up being what I do.


>
>>the op amp that I am using.) I have not finished soldering in the
>>components, though, so I could switch op-amps if there was one which
>>was
>>pin-compatible.
>
>I may be suggesting a total re-design, but you should explore your
>present design some more to find out its limitations.  The first unit
>built shouldn't be expected to work really well.
>

Yeah, I realize that, but being a college student, my budget is quite small
and I can't order a new prototype board very often. I was hoping to get
decent, not perfect, performance out of this first prototype. I felt that a
professionally made board (AP circuits in this case) would give me the best
RF performance (double sided w/large ground plane, plated thru hole, using
surface mount for all analog ICs)

>>Also, does anyone know of a way to some how cancel out
>>input bias current without having to know the source impedance?
>
>I touched on this above.  If the amplifier only has to handle AC signals,
>then the bias curent can be automatically cancelled.  Use another
>low-performance op amp to supply slow feedback to supply the proper bias
>current to force the main amplifier's average output to zero.  I think
>this is called a "DC Servo".  If you need to amplify signals near DC
>(which most useful scopes do), then better hope the bias current is
>constant and use a fixed adjustment.
>
>An off-the wall idea would be to use a dual amp, one side connected just
>to measure and amplify bias current.   This would be applied to the other
>one (which is amplifying the signal) to cancel its bias current.  The two
>amplifiers in a dual chip should track each other fairly close.
>

Both of these ideas deserve serious consideration if I do make a second
prototype, but as I said, I don't think I am going to try to implement them
now, being that the board is already etched.

Thanks for all your suggestions and I will try the bias current
cancellation technique which you suggested.

Thanks again (I sure do get lots of great ideas on this list),

Sean

+--------------------------------+
| Sean Breheny                   |
| Amateur Radio Callsign: KA3YXM |
| Electrical Engineering Student |
+--------------------------------+
http://www.people.cornell.edu/pages/shb7
mailto:shb7@cornell.edu
Phone(USA): (607) 253-0315