1. Home
  2. Electronics Design
  3. Sample-and-Hold Differentiator

Sample-and-Hold Differentiator

I need to differentiate a really slow signal (DC to ~2kHz)

I'm toying with the idea that two sample-and-hold's and a steerable diff-stage should do the trick.

Anyone done that before? ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

formatting link
| 1962 | I love to cook with wine. Sometimes I even put it in the food.

Reply to
Jim Thompson
Loading thread data ...

The first finite difference is a good approximation in that situation.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

Why not a conventional differentiator? Series RC into opamp -, parallel RC feedback, rolloff somewhere above 2KHz.

If the s/h circuits drift, you'll get an error component. Given some opamp and some capacitor, it seems to me that using them to make a classic differentiator is about the same as using two to make a pair of s/h circuits, just simpler.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

As you approach zero frequency it becomes crappy.

I have a 70MHz clock available... probably will divide it down, as

70MHz is a bit of overkill, so droop will be a non-existent problem.

I rarely choose "simpler" >:-} ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

Why? The derivative is tiny at low frequencies. All you need to do is keep the DC offset down. The classic problem with an analog diff isn't the low end, it the noise crisis at the high end. Hence the HF rolloff. 2 KHz isn't very ambitious.

You're losing me there. Is the clock the thing that latches the s/h things and flops the diffamp?

The higher the clock frequency, the lower the derivative gain. Try some numbers. For a derivative to work accurately up to 2 KHz, you'll need a pretty high sample rate, 1 MHz or something. A 1 Hz signal, sampled twice at 1 us intervals, is almost no delta signal. No free lunch.

Analog s/h circuits are nasty, if you need precision.

I do whenever I can.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

A CHEAP ADC! like one of those $4 panel meters with 6 digits of accuracy converting at what 5/sec?

then subtract the two digital signals, wait, you said 2kHz, nevermind.

Reply to
RobertMacy

Did you look at TI's ADS1282 24 bit ADC? 24 bit low noise [input 5nV/rtHz] for seismic work. wait, only good up to 500Hz, nevermind.

Reply to
RobertMacy

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

Yeah well, if everyone had the luxury of a million transistors per square mm... :)

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com
Reply to
Tim Williams

I don't think I've gotten quite that high, but I know I've done over

100K devices on a single chip. ...Jim Thompson 
--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

Your lab would obviously be about 2.6 mm square, then. ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

microns, that is.

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

You would have to define "derivative" in a quantized time environment.

It would be dV/dT where dT was a fixed, periodic clock. I suppose there could be a more sophisticated algorithm after the ADC, to improve s/n in various time spans. But that's getting awfully digital.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

Concept works...

Now I have to go back and create a non-overlap timing controller for the switches ;-)

Since my available clock is several orders of magnitude faster than the signal in question I can afford some dead time. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

Yeah, those ideal components are great.

An R-C would do the same thing, but not time quantized.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

Remember I need to get down to zero Hertz. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

It wouldn't be a very powerful lab, but boy would it have some bandwidth! ;-)

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com
Reply to
Tim Williams

An RC differentiates down to zero Hz. It gets better and better as the signal frequency goes down. It's perfect at 0 Hz.

What was the clock rate in your sim? There are no obvious steps in the plot.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

1kHz Signal, 100kHz sampler. Just a concept test. I have 70MHz clock available, which I will divide down and create non-overlap. ...Jim Thompson 
--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

Yeah, but as soon as someone nearby lit off a camera flash...

Reply to
DecadentLinuxUserNumeroUno

A 1 v/s signal, feeding two s/h circuits clocked 10 us apart, gives a

10 microvolt difference. Given that the s/h has a narrow aperature, certainly below 10 us, there will be noise problems. 
--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.