开云体育

Hi all,

I have a 7904A with a 7A19 plugin (plus lots of slower ones, e.g. 7A24 and 7A26). I'd love to have two channels of > 500 MHz, and I'm contemplating if I should chase another 7A19 or get two 7A29 (and sell the 7A19). The 7A29 will make better use of the mainframe bandwidth and has a variable vertical deflection feature.

What I wonder: how do things look noise wise? I hate blurry traces, and I'm rather surprised how sharp the 7A19 is (much better than the 7A26!). Anyone who could compare them side-by-side?

Any other difference I should know about?

Thanks a lot,
Samuel

I have a few 7A19 plug-ins over and above what I need. Contact me off list if you are interested.

Dave

On Thu, Jan 3, 2013 at 1:57 PM, micpreamp <micpreamp@...> wrote:

I have a 7904A with a 7A19 plugin (plus lots of slower ones, e.g. 7A24 and 7A26). I'd love to have two channels of > 500 MHz, and I'm contemplating if I should chase another 7A19 or get two 7A29 (and sell the 7A19). The 7A29 will make better use of the mainframe bandwidth and has a variable vertical deflection feature.

What I wonder: how do things look noise wise? I hate blurry traces, and I'm rather surprised how sharp the 7A19 is (much better than the 7A26!). Anyone who could compare them side-by-side?

Isn't that just due to the termination on the 7A19? Have you tried
putting a good inline termination on the 7A26?

Anyway, I haven't noticed a difference in trace noise between the 7A19 and 7A29.

On Thu, 3 Jan 2013 14:26:04 -0700, David DiGiacomo
<daviddigiacomo@...> wrote:

On Thu, Jan 3, 2013 at 1:57 PM, micpreamp <micpreamp@...> wrote:

I have a 7904A with a 7A19 plugin (plus lots of slower ones, e.g. 7A24 and 7A26). I'd love to have two channels of > 500 MHz, and I'm contemplating if I should chase another 7A19 or get two 7A29 (and sell the 7A19). The 7A29 will make better use of the mainframe bandwidth and has a variable vertical deflection feature.

What I wonder: how do things look noise wise? I hate blurry traces, and I'm rather surprised how sharp the 7A19 is (much better than the 7A26!). Anyone who could compare them side-by-side?

Isn't that just due to the termination on the 7A19? Have you tried
putting a good inline termination on the 7A26?

Anyway, I haven't noticed a difference in trace noise between the 7A19 and 7A29.

The difference in noise between the 7A19 and 7A29 should be there
(maybe about 30%) but it will not amount to much and I doubt it would
be visible.

The 7A26 is so much noisier because it has a 1 MOhm input impedance.
The higher noise levels involved with high impedance inputs is why you
do not find high bandwidth and high sensitivity in the same amplifier
for a given input impedance. I actually prefer the 7A26 (200 MHz
switchable to 20 MHz) or 7A13 (100 MHz switchable to 5 MHz) to the
7A18 (100 MHz only) because they have switchable bandwidth limiting
which makes a large difference in displayed noise when you do not need
the higher bandwidth.

The real question is why DON'T you see the 7A26 noise?

IF (and that's a big if!) the input impedance of an amplifier were purely
resistive then the thermal noise contributed by the resistor (Vn)is
Vn(RMS)= sqrt(4*KB*T*R*(F2-F1)) in Volts, where
T is the temperature of the resistor (in degrees Kelvin). (70F = 294K)
K is Boltsmann's constant (1.38 x 10^-23)
R is the resistance in ohms
F2-F1 is the bandwidth you are measuring the noise across. In the case of a
Tektronix 7A19 amplifier it would be DC to the full bandwidth of the plugin
(500,000,000Hz - 0Hz) = 500,000,000

So, if a 7A19 had a purely resistive 50 ohm input its thermal noise at room
temperature would be 20uV which is far too small to see on the scope even at
the most sensitive setting of the 7A19.

For a 7A26, with 20,000 times input resistance (if it was purely resistive),
and 2/5 the bandwidth, the thermal noise would be 1.8mV which would be so
great you couldn't help but notice it on the most sensitive scale because it
would occupy more than 1/3 division of the CRT. So why don't you see it?

The answer to my question is simple: the input isn't purely resistive. The
amplifier specs say there is resistance and capacitance at the input. With
1Mohm input impedance the effect of even a small capacitance becomes
significant as the frequency of the signal goes up. Another way of looking
at this is that with a lower 50ohm input impedance, a capacitor will have
little effect until you reach very high frequencies. For high frequency work
1MOhm amplifiers are not a good choice, 50 ohm inputs are better. The
capacitance at the input of the 7A26 kills a lot of the noise and keeps the
trace sharp. The 7A19 will have no visible noise.

Dennis

For a 7A26, with 20,000 times input resistance (if it was purely

resistive), and

2/5 the bandwidth, the thermal noise would be 1.8mV which would be so
great you couldn't help but notice it on the most sensitive scale because

it

would occupy more than 1/3 division of the CRT. So why don't you see it?

On full bandwidth in my 7904 workhorse, one of my 7A26's on DC coupling
measures 1/3 of a *minor* division at 5mV/div (needed a magnifying glass for
that) - so about 0.1mV tangential noise, and it looks just like fuzzy noise.
With the 20MHz bandwidth limit switched in the trace is so narrow it is
impossible to measure the noise.

Craig

Anyway, I haven't noticed a difference in trace noise between
the 7A19 and 7A29.

Thanks!

There is no fundamental connection between amplifier input impedance and noise. I've design preamplifiers with sub-300 pV/rtHz, and they could easily have GOhm input resistance. That's because the noise of the input bias resistor is shunted by the (much lower) source impedance.

The relatively high noise of some of the high-Z plugins seems to be mostly a function of their active parts. I've also noted differences between some time bases (e.g. between two 7B80 I have). Apparently one has higher trigger jitter.

Best,
Samuel

There is no fundamental connection between amplifier input impedance and noise. I've design preamplifiers with sub-300 pV/rtHz, and they could easily have GOhm input resistance. That's because the noise of the input bias resistor is shunted by the (much lower) source impedance.

In my lab (aka basement), the "noise" is mostly RFI pickup.

Did you try connecting a termination to the 7A26?

On Fri, 4 Jan 2013 09:42:53 -0700, David DiGiacomo
<daviddigiacomo@...> wrote:

There is no fundamental connection between amplifier input impedance and noise. I've design preamplifiers with sub-300 pV/rtHz, and they could easily have GOhm input resistance. That's because the noise of the input bias resistor is shunted by the (much lower) source impedance.

In my lab (aka basement), the "noise" is mostly RFI pickup.

Did you try connecting a termination to the 7A26?

On my 7A18 that makes some difference but the noise level is still
noticeably higher at the highest sensitivity where the only difference
is removal of the last high impedance input attenuator. I usually
test this with the input set to GND anyway.

The 7A26 is more difficult to gauge since it uses a switched low
impedance gain stage at the highest sensitivity which adds a
confounding factor.

Both the 7A18 and 7A26 uses the same input arrangement and even the
same dual FET. Besides the bandwidth difference, the 7A26 uses
integrated instead of discrete gain stages and I would expect them to
be noisier anyway.

In my lab (aka basement), the "noise" is mostly RFI pickup.

Did you try connecting a termination to the 7A26?

Yes, I'm well aware that amplifier noise needs to be measured with the input terminated. Doesn't do much if anything to the 7A26.

I'm not talking about severe noise--just the difference between a really crisp and sharp trace, and a slightly fuzzy one.

Samuel

Just for the records: 7A29 arrived, and I can confirm that I don't see any noise difference between it and the 7A19 I already had. Thanks again for your help!

Samuel