Low Noise Front End for P.I.

[JC1]

EE

Wow!! Look at that. And how clever of you Eric to first find
the problem and then measure it.

Extremely useful information there.

[ElectroNovice]

Low noise Front end for P.I

Hi I am a new member to this forum. I hope the following web sites may be of some help to Eric you but as I am very much an electronics novice forgive me if they are irrelevant! It is my first post!
Regards
ElectroNovice

Low Frequency Noise in the Single Electron Transistor:
Instrumentation and Experiments
http://fy.chalmers.se/~starmark/lic.pdf

Low Noise, Low Drift FET Op Amp
http://www.ing.unlp.edu.ar/comunica/doc/ad645.pdf

HIGH SPEED OP AMPS
Driving Capacitive Loads Cable Driving
Single-Supply Considerations Application Circuits
http://www.sss-mag.com/pdf/ad2.pdf

Wideband Gain Block Amplifier Design Techniques
Here is a thorough review of the device design requirements for a
general-purpose amplifier RFIC
http://www.rfmd.com/Published%20Articles/arnott.pdf

Ultralow Input Bias Current Operational Amplifier
AD549*
http://www.ee.washington.edu/stores/...near/ad549.pdf

Comlinear CLC425
Ultra Low Noise Wideband Op Amp
http://eshop.engineering.uiowa.edu/N...7/DS012708.pdf


Wideband, Low Distortion, Unity-Gain Stable,
Voltage-Feedback OPERATIONAL AMPLIFIER
http://focus.ti.com/lit/ds/symlink/opa842.pdf

Designing With Opamps.
http://www.dself.dsl.pipex.com/ampins/webbop/opamp.htm
Release date: August 5, 2005
Operational Amplifiers Offer Industry's Lowest Noise and Distortion for 1.1 mA of Supply Current.Analog Devices has extended its portfolio of general-purpose operational amplifiers (op amps) with the ADA4841 family of single and dual devices. Offering extremely low noise and distortion and drawing just 1.1 mA of supply current, the new op amps are well suited for power-sensitive applications such as portable instrumentation, high-channel-count measurement and medical applications.
http://news.thomasnet.com/fullstory/466224

Choosing the Right Op Amp for Your High-Resolution ADC
http://www.sensorsmag.com/articles/0303/20/main.shtml

Avoiding Op-Amp Instability Problems In Single-Supply Applications
http://www.analog.com/library/analog...5-02/avoiding/

To find the best drive amplifier (buffer) for your ADC, you have to evaluate impedance matching, charge injection, noise reduction, and output accuracy. But what do all these specs mean? Walter Sangalli, Maxim UK Ltd.
http://www.sensorsmag.com/articles/0800/84/main.shtml

Instrumentation Amplifiers (In Amps)
http://www.semiconfareast.com/inamps.htm

Noise Reduction and Isolation
http://www.iotech.com/handbook/chapt_10.html

Analog Devices Introduces Industry's First Instrumentation Amplifier
http://www.radiolocman.com/comp/more-en.html?di=2781

A Designer’s Guide to Instrumentation Amplifiers 2ND Edition
http://www.analog.com/UploadedFiles/..._AMP_GUIDE.pdf

APPLICATION NOTE 3428 Transimpedance Amplifier Buffers
http://www.maxim-ic.com.cn/appnotes....te_number/3428

OPERATIONAL AMPLIFIER BASICS
by Harry Lythall
http://web.telia.com/~u85920178/begin/opamp00.htm

Resistor Types - Does It Matter?
http://www.aikenamps.com/ResistorNoise.htm


MAX6126
Ultra-High-Precision, Ultra-Low-Noise, Series Voltage Reference.
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3623

Op Amp Selection Guide for Optimum Noise Performance http://www.linear.com/pc/downloadDoc...09,C1026,D6539.
HA-5127, HA-5137, HA-5147, Ultra Low Noise Amplifiers
http://www.intersil.com/data/an/an553.pdf


Wideband, Ultra-Low Noise, Voltage-Feedback OPERATIONAL AMPLIFIER with Shutdown
http://focus.ti.com/lit/ds/symlink/opa847.pdf

LT6200: 0.95nV/rtHz Rail-to-Rail High-Speed 2.5-V Op Amp
http://www.analogzone.com/acqp0930.htm

CURRENT FEEDBACK AMPLIFIERS http://www.analog.com/library/analog..._Engineer.html
1/f noise: a pedagogical review.
Edoardo Milotti
http://arxiv.org/ftp/physics/papers/0204/0204033.pdf

Sub-Poissonian shot noise
Marc de Jong
http://www.ilorentz.org/beenakkr/mes...ise/noise.html

Observation of Hot-Electron Shot Noise in a Metallic Resistor
http://www.eng.yale.edu/qlab/papers/..._shotnoise.pdf

Analogue devices Low-Noise, Precision Operational Amplifier
http://www.bgmicro.com/pdf/ics1023.pdf

Hardware and Firmware Noise Reduction Techniques in Embedded Systems http://www.microchip.com/stellent/gr...n/en024545.doc.


Opamp Noise Figure
These equations calculate the noise figure of an opamp using parameters provided by vendors at the operating frequency and gain configurations. Be sure to note units both for the input parameters and for the equations, or you will end up with really bad results.
http://www.rfcafe.com/references/ele...l/opamp_nf.htm



1/f Noise Modeling for Semiconductors
http://eesof.tm.agilent.com/docs/icc...E/NOISEdoc.pdf

9812B NoiseProTM
http://www.celestry.com/pdf/noisepro.pdf

Operational Amplifier Noise Prediction (All Op Amps)
http://www.intersil.com/data/an/an519.pdf

ECE594A notes set 8:
Noise in Electrical Circuits:device noise models.
Mark Rodwell
University of California, Santa Barbara
http://www.ece.ucsb.edu/Faculty/rodw...otes_set_8.pdf

literature electronic noise survey
http://www.ciphersbyritter.com/RES/NOISE.HTM

[JC1]

EE

Hi Electro,

Very good links, I think Comlinear quit making the CLC425, but could be wrong.

I will use the last two links of yours to cut and paste, and will not relink em here. From the last link.



"We thus conclude that a strongly forward-biased pn junction exhibits the full shot noise associated with its average current.

"In zero bias, and therefore with the junction in thermal equilibrium, the two currents I1 and I2 . . . cancel on the average, but their noise components are still independent, and add in a mean square sense." "So . . . we find . . . for low enough frequencies the spectral density

Si(f) = 4 q I0 .

Thus, the noise is identical to shot noise associated with a dc current twice as large as the reverse saturation current of the diode. Since, in fact, there is no dc current flowing under the assumed zero-bias conditions, it is convenient to eliminate I0 . . . by rewriting in terms of the incremental diode conductance. On differentiating . . . we obtain

g0 = dI / dV at V = 0 .

Thus

Si(f) = 4 k T g0

{so the clamp diode/s could be doing something here, even though no current flowing at signal interest time, still I wonder about the noise difference of a diode sitting there, and one that just had 1 amp pulsed though it, maybe needs a moment (us) to quite down.}


next to last link


There are two resistances in the antenna. One is the metal conductor loss... this has the thermal noise of the resistance. {low in voltage, high in current} The second is the radiation resistance. this is the temperature of the place the antenna is pointing...Deep space is at 3.8 Kelvin.
{the radiation of a misstuned atenna can be very high}


So who is the bad guy? The coil, damping resistor, clamp diodes, input resistor, op amp,,,, all the above?

[JC1]

EE

Whoops, again,

The radiation "resistance" of a misstuned antenna can be very high.

sorry

Oh and I guess the bad guy could be someone else.

The Mosfet just had some current dumped though it???

Oh and glass diodes can be light sensitive, including uv, the temperature sensor in my work office is a simple diode used to sense IR, it is glass down inside a hole in a plastic box.

The system is all computerized. when I want the cooling to come on I shine a flashlight on it. On comes the air. Since to save energy we can no longer set room temp and they set it tooo hottt.

the down side is the computer monitors and records room temperatue, they came to the office and showed me the print out of room temperature, there it is wandering up and down a bit, then it pegs at max temp for a bit and then drops back down. They kept trying to figure out what was wrong with my office. to keep from getting in trouble I never told em. they finally gave up, figured it was a loose connection.

So by black diodes or keep em "in the dark" so to speak.

[Ferric Toes]

The AD797, as suggested by treasurediver, looks very good on paper. I'll have to track one down too and do some tests. I remember seeing a detector schematic on this forum that used the AD797 in the front-end, a quick search revealed the post below (by Cossaro Fernando).

http://thunting.com/geotech/forums/s...ad.php4?t=8107

Mark

I got a AD797 today and plugged it into my board. No go, which is puzzling. This one stays in positive saturation too long. Checked my diode clipping, and that is all OK. Put 5534 back in and everything is fine, and a few other IC types, which, except for higher noise levels, all give the right response.

Has the circuit in the above link been built, and worked successfully?

Looks like I will stick with the 5534/8055 combo, which with the lower value resistors seems to be as good as it gets at the moment. As JC1 suggested, there is probably a more exotic way of doing things to get lower noise, but I don't have the time at the moment, to start a brand new approach.

Eric.

[ElectroNovice]

I got a AD797 today and plugged it into my board. No go, which is puzzling. This one stays in positive saturation too long. Checked my diode clipping, and that is all OK. Put 5534 back in and everything is fine, and a few other IC types, which, except for higher noise levels, all give the right response.

Has the circuit in the above link been built, and worked successfully?

Looks like I will stick with the 5534/8055 combo, which with the lower value resistors seems to be as good as it gets at the moment. As JC1 suggested, there is probably a more exotic way of doing things to get lower noise, but I don't have the time at the moment, to start a brand new approach.

Eric.

Other than cost and size is there any reason that you could not use discreet op amps like the John Hardy 990(used in sound recording) or instrumentation op amps like the AD8553 precision in-amp rather than monolithic op amps?

See: http://www.analog.com/en/content/0,2...F66883,00.html and
http://www.johnhardyco.com/pdf/990.pdf

ElectroNovice

[Carl-NC]

Most instrumentation opamps are way too slow... you need a GBW product of >10MHz for a PI frontend.

A discrete opamp would be interesting. Or, better, a discrete first gain stage plus opamp rolled together. Problem is, it would be difficult to get discrete devices with lower intrinsic noise (junction resistances, mostly) than what you can get in integrated processes these days.

- Carl

[Ferric Toes]

So who is the bad guy? The coil, damping resistor, clamp diodes, input resistor, op amp,,,, all the above?

At one stage I disconnected the clamp diodes, but obviously had to disable the transmitter. Only thing left was then the input resistor, with one end to ground. There was no reduction in noise.

When pulsed, there is no evidence that the diodes contribute noise. If you sample at a late delay, you would see less noise than at an early delay, if the diodes are recovering from conduction?

The damping resistor has the coil resistance in parallel, in my case 4 ohms. So, from what has been discussed, it should not materially contribute.

Not sure about the Mosfet. I have seen no evidence of that, but it could introduce noise from the power supply circuits via its capacitance. Running the detector both on a battery or a bench power supply shows no difference.

One other possible source of noise in PI's, is if the charge pump circuitry for the RX power supply has its own non-synchronous oscillator. That isn't the case in my tests, as it is driven from the transmitter drive circuit

Eric.

[Tinkerer]

Scope pics

Ferric Toes,
I apreciate very much your posts. Would you be so kind as to explain the scope pics to me? With the scope at 2uS ?,div. What is that spike of less than halve a division on the second pix?
On the first pix it looks to me that the fly back is overdamped and oscillates. If this is caused by the change of the opamp, it suggests to me that there is additional input capacitance present.
Tinkerer

[Tinkerer]

Could you post scope pics of the AD797 output?
thanks
Tinkerer

[Tinkerer]

GBW of Opamp

Most instrumentation opamps are way too slow... you need a GBW product of >10MHz for a PI frontend.

A discrete opamp would be interesting. Or, better, a discrete first gain stage plus opamp rolled together. Problem is, it would be difficult to get discrete devices with lower intrinsic noise (junction resistances, mostly) than what you can get in integrated processes these days.

- Carl

Hi Carl,

With a gain of 1000 and a frequency of 10k pps, This would be 10MHz bandwith. However my 20 year old textbook says that I should not use more than 5 to 10 % of the GBW of an Opamp.

I guess that modern opamps are a lot better, but what would be a good number % of total GBW to use ?
Tinkerer

[JC1]

EE

Hi Guys,

One more potential "bad guy". The feedback resistor.

Here is a link for the article, Op Amps for Everyone Design Guide (Rev. B).

You have to buy this from Texas Instruments now it appears.
Found link for still free download.

http://www.web-ee.com/primers/files/slod006b.pdf

Chapter 10 talks about noise. In it he puts alot of blame of the noise on a 10 Meg feedback resistor. Don't know if he is right. but might be worth a try.

Of course to keep the gain the same will have to lower input resistor by same factor and if this resistor is used to limit the current through the clamp diodes then something different will have to be done there, like putting resistors in series with the diodes? (depending on coil current)(more resistors? this can't be good? maybe the diodes will block the noise?).

[Carl-NC]

See the "Noise Analysis" thread... the end of the second page has an example of an inverting opamp circuit... output noise is roughly the feedback resistor noise times the gain of the circuit.

This is an example of a non-obvious result of noise, and why you should keep the analysis in variable-land as long as possible, in order to get a feel for what really matters.

- Carl

[JC1]

EE

Hi Carl,

"This is an example of a non-obvious result of noise, and why you should keep the analysis in variable-land as long as possible, in order to get a feel for what really matters."

I'm lost........Who won???? Who got gained up?

[Carl-NC]

Hi Carl,

With a gain of 1000 and a frequency of 10k pps, This would be 10MHz bandwith. However my 20 year old textbook says that I should not use more than 5 to 10 % of the GBW of an Opamp.

I guess that modern opamps are a lot better, but what would be a good number % of total GBW to use ?
Tinkerer

True, you generally want to stay away from the GBW limit of the opamp, or it will quickly degrade signal fidelity. In a PI signal, we are looking at an exponential decay, which looks a whole lot like a simple RC decay. Our sample point is usually beyond the time constant (tau), but for argument let's say we are sampling at 1*tau. If this is 10us, then the required BW = 1/(2*PI*tau) = 16kHz. If gain is 500, then GBW is 8MHz.

This might be pushing things a bit, because the opamp response will be on the same order as the signal decay. Therefore, when we try to look at signal decay variations, the overall decay will have a constant opamp "decay" component that desensitizes the variations. But, then again, I think the 10us sample point is actually a few tau out on the curve (i.e., the decay tau is less), so even a 10MHz opamp should be OK.

Eric, is this the right way to look at this?

- Carl