Post by thread:[EE] Looking for capacitance to voltage circuit

Was wondering if anybody can think of a simple off hand way to convert a varying capacitance to a voltage. The capacitance (a humidity sensor) is in the hundreds of pF range, and the voltage output I'd like is 0-5V. Something using an op amp or two would be OK, don't really want to go all out with a PIC. Linearity is a BIG bonus, but it doesn't have to be 100%, a variation of say 5% from linear is perfectly fine. Thanks, TTYL -- ----------------------------- Herbert's PIC Stuff: http://repatch.dyndns.org:8383/pic_stuff/

> Was wondering if anybody can think of a simple off hand way to convert a > varying capacitance to a voltage. The capacitance (a humidity sensor) is > in the hundreds of pF range, and the voltage output I'd like is 0-5V. > > Something using an op amp or two would be OK, don't really want to go > all out with a PIC. I don't think "all out" is necessary; could you use a 555 with the humidity sensor as a controlling element? Variations of pulse length could easily be measured and converted into some usable value with a PIC. I'd think it would be pretty easy to put a software SPI interface onto a 10F (do they have a 10F with that built in?), and have it measuring pulses from a 555 to figure out what's what. If you need more precision, a 555 is a fairly trivial circuit that could be constructed from 2 op-amps (IIRC; I haven't seen it in a couple of years). Actually, this is the point where I ask for more info. Are you tied to a voltage output, or could the sensor unit be an SPI module? How flexible is this? Does it need to be ultra low power? Mike H. > Linearity is a BIG bonus, but it doesn't have to be 100%, a variation of > say 5% from linear is perfectly fine. > > Thanks, TTYL

On Fri, 2005-03-18 at 15:20 -0600, Mike Hord wrote: > > Was wondering if anybody can think of a simple off hand way to convert a > > varying capacitance to a voltage. The capacitance (a humidity sensor) is > > in the hundreds of pF range, and the voltage output I'd like is 0-5V. > > > > Something using an op amp or two would be OK, don't really want to go > > all out with a PIC. > > I don't think "all out" is necessary; could you use a 555 with the humidity > sensor as a controlling element? Variations of pulse length could easily > be measured and converted into some usable value with a PIC. That's the circuit I have at the moment, but for a variety of reasons I need a varying voltage. If you have a simple frequency to voltage converter circuit (one that doesn't use a single chip) that would also work OK for my purpose. (6kHz type frequency, varying a few 100Hz on either side). > Actually, this is the point where I ask for more info. Are you tied to a > voltage output, or could the sensor unit be an SPI module? I need a voltage out, and as simple as possible (i.e. some op amps, or a 555, with some passives). > How flexible is this? Does it need to be ultra low power? Unfortunately not very flexible, but power wise it doesn't matter. Although if it soaks up 10A I might not be to happy... :) TTYL ----------------------------- Herbert's PIC Stuff: http://repatch.dyndns.org:8383/pic_stuff/

A Phase Locked Loop (PLL) will do exactly what you want. Time to do your research homework if you haven't used PLL circuits before. KF4HAZ - Lonnie ----- From: "Herbert Graf" <hgraf@ {Quote hidden}> Was wondering if anybody can think of a simple off hand way to convert a > varying capacitance to a voltage. The capacitance (a humidity sensor) is > in the hundreds of pF range, and the voltage output I'd like is 0-5V. > > Something using an op amp or two would be OK, don't really want to go > all out with a PIC. > > Linearity is a BIG bonus, but it doesn't have to be 100%, a variation of > say 5% from linear is perfectly fine. > > Thanks, TTYL > > --

On Fri, 2005-03-18 at 15:52 -0600, Falcon Wireless Tech Support - KF4HAZ wrote: > A Phase Locked Loop (PLL) will do exactly what you want. > Time to do your research homework if you haven't used PLL circuits before. A little much for my purpose. But thanks for the suggestion. TTYL ----------------------------- Herbert's PIC Stuff: http://repatch.dyndns.org:8383/pic_stuff/

CMOS 555 timer ----- Original Message ----- From: "Herbert Graf" <spam_OUThgrafTakeThisOuTemail.com> To: "Microcontroller discussion list - Public." <.....piclistKILLspam@spam@mit.edu> Sent: Friday, March 18, 2005 2:54 PM Subject: [EE] Looking for capacitance to voltage circuit {Quote hidden}> Was wondering if anybody can think of a simple off hand way to convert a > varying capacitance to a voltage. The capacitance (a humidity sensor) is > in the hundreds of pF range, and the voltage output I'd like is 0-5V. > > Something using an op amp or two would be OK, don't really want to go > all out with a PIC. > > Linearity is a BIG bonus, but it doesn't have to be 100%, a variation of > say 5% from linear is perfectly fine. > > Thanks, TTYL > > -- > > ----------------------------- > Herbert's PIC Stuff: > http://repatch.dyndns.org:8383/pic_stuff/ > > --

At 05:05 PM 3/18/2005, Herbert Graf wrote: >On Fri, 2005-03-18 at 15:52 -0600, Falcon Wireless Tech Support - KF4HAZ >wrote: > > A Phase Locked Loop (PLL) will do exactly what you want. > > Time to do your research homework if you haven't used PLL circuits before. > >A little much for my purpose. But thanks for the suggestion. TTYL Oddly enough, the mailman just arrived with "electronic design" magazine, and on page 79 is an article "Measure relative humidity with a PIC MCU" using a 555 and capacitive sensor HS1101.

----- From: "Herbert Graf" <mailinglist2@ > On Fri, 2005-03-18 at 15:20 -0600, Mike Hord wrote: > > > Was wondering if anybody can think of a simple off hand way to convert a > > > varying capacitance to a voltage. The capacitance (a humidity sensor) is > > > in the hundreds of pF range, and the voltage output I'd like is 0-5V. > > > > > > Something using an op amp or two would be OK, don't really want to go > > > all out with a PIC. > > > > I don't think "all out" is necessary; could you use a 555 with the humidity > > sensor as a controlling element? Variations of pulse length could easily > > be measured and converted into some usable value with a PIC. > > That's the circuit I have at the moment, but for a variety of reasons I > need a varying voltage. If you have a simple frequency to voltage > converter circuit (one that doesn't use a single chip) that would also > work OK for my purpose. (6kHz type frequency, varying a few 100Hz on > either side). <snip> Perhaps feeding the frequency through a filter (might even work with just a cap) and then rectifying it to DC and maybe a small filter cap? scale the voltage with an op-amp.

Mike Hord wrote: > I'd think it would be pretty easy to put a software SPI interface > onto a 10F (do they have a 10F with that built in?), and have it > measuring > pulses from a 555 to figure out what's what. If you're really clever, you don't need the 555. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com

On Fri, 18 Mar 2005 15:54:25 -0500, you wrote: >Was wondering if anybody can think of a simple off hand way to convert a >varying capacitance to a voltage. The capacitance (a humidity sensor) is >in the hundreds of pF range, and the voltage output I'd like is 0-5V. > >Something using an op amp or two would be OK, don't really want to go >all out with a PIC. > >Linearity is a BIG bonus, but it doesn't have to be 100%, a variation of >say 5% from linear is perfectly fine. If you end up using a PIC, a really good way is to use a 555 as a capacitance to frequency converter. You can do a neat trick by switching the sensor and a reference cap alternately to ground using PIC pins, so you can measure teh ratio of the sensor to the ref cap, cancelling out a lot of error sources very easily. PIC10Fxxx and a 555 timer would make a very cheap solution, and allow for sensor calibration, linearisation in software. As humidity changes slowly, you could get an anlogue output of you needed it with PWM output and a simple RC low pass filter. If you want pure analogue, maybe a 556, half in astable mode, half in monostable, using the sensor to vary the frequency of a fixed pulse length (or length at fixed frequency) and a simpl RC lowpass filter on the output.

Olin Lathrop wrote: > Mike Hord wrote: > >>I'd think it would be pretty easy to put a software SPI interface >>onto a 10F (do they have a 10F with that built in?), and have it >>measuring >>pulses from a 555 to figure out what's what. > > > If you're really clever, you don't need the 555. Like: www.dattalo.com/technical/software/software.php http://www.dattalo.com/technical/software/pic/a2d.asm Just tie Vin in to Vcc and let the Capacitance be the variable. Scott

On Fri, 18 Mar 2005 16:21:37 -0800, you wrote: >Olin Lathrop wrote: >> Mike Hord wrote: >> >>>I'd think it would be pretty easy to put a software SPI interface >>>onto a 10F (do they have a 10F with that built in?), and have it >>>measuring >>>pulses from a 555 to figure out what's what. >> >> >> If you're really clever, you don't need the 555. > >Like: > >www.dattalo.com/technical/software/software.php >http://www.dattalo.com/technical/software/pic/a2d.asm > >Just tie Vin in to Vcc and let the Capacitance be the variable. > >Scott With the sort of low capacitances involved with humidity sensors, using the 555 makes a lot of potential problems (pin leakage, capacitance, timing resolution) go away - you can use a relatively high frequency and use the T0 prescaler if necessary. Resolution is limited only by how long you want to wait for the answer.

With only pF to work with, that won't work too well on a Stamp. Simplest is an LC oscillator and MHz, then a divider to get down to a frequency range the PIC can handle. A 4060 osc/counter chain may work for you if you can get it to do R/C osc. Depending on the actual capacitance available, a 74HC14 oscillator (feedback resistor to input, C to ground) makes a stable high frequency oscillator that is easily measured. Robert. Denny Esterline wrote: > How about using a resistor and a PIC and measuring the RC time constant > directly? (it's been done alot, but the basic stamp's POT command explains > it well) If you have to have a voltage output, what about using the PIC to > PWM into a filter cap? > > -Denny > > {Original Message removed}

On Fri, 2005-03-18 at 16:37 -0600, Falcon Wireless Tech Support - KF4HAZ wrote: > Perhaps feeding the frequency through a filter (might even work with just a cap) and then rectifying it to DC and maybe a small filter cap? scale the voltage with an op-amp. > That sounds like it might work, got any guides as to the best way to set something like that up? Thanks, TTYL ----------------------------- Herbert's PIC Stuff: http://repatch.dyndns.org:8383/pic_stuff/

>> Was wondering if anybody can think of a simple off hand way >> to convert a varying capacitance to a voltage. The capacitance >> (a humidity sensor) is in the hundreds of pF range, and the >> voltage output I'd like is 0-5V. >> Can't you just just produce an AC waveform, run it through the capacitor, rectify and scale the result? Two op-amps, I would guess... The frequencies involved for ~100pF might be inconvenient. BillW

On Fri, 2005-03-18 at 17:45 -0800, Denny Esterline wrote: > How about using a resistor and a PIC and measuring the RC time constant > directly? (it's been done alot, but the basic stamp's POT command explains > it well) If you have to have a voltage output, what about using the PIC to > PWM into a filter cap? Unfortunately that's far to complicated a solution. Thanks anyways, TTYL ----------------------------- Herbert's PIC Stuff: http://repatch.dyndns.org:8383/pic_stuff/

Mike Harrison wrote: > As humidity changes slowly, you could get an anlogue output of you needed it with PWM output and a > simple RC low pass filter. > > If you want pure analogue, maybe a 556, half in astable mode, half in monostable, using the sensor > to vary the frequency of a fixed pulse length (or length at fixed frequency) and a simpl RC lowpass > filter on the output. This is the technique used by FM instrumentation tape recorders (fixed pulse width). Gives very high linearity over decades of range. (s/n ~65db 0.2% lin). You'll probably have to scale the LPF output with an op-amp to get the range you want, but it is a purely analog solution. The only big issue will be leakage and how many pF of change you're expecting at what nominal capacitance. Be sure to use a polypropylene cap for the fixed pulse width generation if you want very good linearity. Robert

>> Was wondering if anybody can think of a simple off hand way >> to convert a varying capacitance to a voltage. Circuits that connect to a VOM and measure capacitance are pretty common on the web. Google for "capacitance meter schematic"; some are pretty simple, and could be simplified further for a known and fixed input range... http://earthground.8m.com/indcap.htm BillW

> http://earthground.8m.com/indcap.htm Has anyone been able to make the cap meter work? I have a working inductance meter (yahoo! I can make coils now) but the cap meter does NOT work for me. (I get a sawtooth output?!?) But using two 74HC14 gates, and a resistor and of course a frequency counter or a scope, measure capacitance to an acceptable level (enough for hobby use, ie recycling parts). Christian

Herbert Graf wrote: >Was wondering if anybody can think of a simple off hand way to convert a >varying capacitance to a voltage. The capacitance (a humidity sensor) is >in the hundreds of pF range, and the voltage output I'd like is 0-5V. > >Something using an op amp or two would be OK, don't really want to go >all out with a PIC. > >Linearity is a BIG bonus, but it doesn't have to be 100%, a variation of >say 5% from linear is perfectly fine. > >Thanks, TTYL > > > Complete overkill for you application, but perhaps the coolest part I have seen in a long time Analog Devices AD5934 12 bit impedance converter Network Analyzer. Provides real and imaginary component's of impedance in the audio range, impedances from 100 Ohm to 20 MOhm. Put's out 2 Vdc as an excitation voltage. Dave Bengtson

this is definitely overkill. I ordered a sample and can't wait to check it out. http://www.analog.com/en/content/0,2886,760%255F788%255F62327,00.html The CDC family of converters from Analog Devices are fully integrated converters that address the complex and difficult signal processing challenges of direct capacitance-to-digital conversion. Features • Highly Integrated • Direct Conversion from Capacitive Sensor to Digital Code • Integrated Clock, Reference, Mux, Calibration and Temperature Sensor • No external components required • High Precision • 24-bit Sigma-Delta Architecture • 20 aF (10-18) Resolution • High Accuracy: 2 fF error • Tolerant of ground capacitance and ground leakage currents • Flexible • 2.7 V to 5.25 V • 5 Hz to 90 Hz data rate • Interface to single- or dual-electrode capacitance sensors • Choice of one or two channels • Additional voltage conversion channel • Fully Specified from -40°C to +125°C • Low Power • 1 mA (normal mode) • 1 µA (power-down mode) • Two-wire I2C Serial Interface • 16-pin TSSOP Gus S Calabrese GSC 720.222.1309 VOIP w/vm 303.908.7716 cell no vm 720.221.3795 softphone ichat --> video/audio --> gcalabreseKILLspammac.com http://www.omegadogs.com .....gscKILLspam.....omegadogs.com 4337 Raleigh St near i25 and i70 intersection Denver, CO 80212

Using a 74C14 was mentioned. Use a resistor and the capacitance of the sensor as a first stage RC oscillator. Use the output to drive a pulse generator made using a second gate of the 74C14. Consists of a diode and resistor in parallel for feedback and a capacitor to ground on the input. Couple the stages with a diode. Gus S Calabrese GSC 720.222.1309 VOIP w/vm 303.908.7716 cell no vm 720.221.3795 softphone ichat --> video/audio --> EraseMEgcalabresespam_OUTTakeThisOuTmac.com http://www.omegadogs.com gscspam_OUTomegadogs.com 4337 Raleigh St near i25 and i70 intersection Denver, CO 80212

At 01:54 PM 3/18/2005, Herbert Graf wrote: >Was wondering if anybody can think of a simple off hand way to convert a >varying capacitance to a voltage. The capacitance (a humidity sensor) is >in the hundreds of pF range, and the voltage output I'd like is 0-5V. > >Something using an op amp or two would be OK, don't really want to go >all out with a PIC. > >Linearity is a BIG bonus, but it doesn't have to be 100%, a variation of >say 5% from linear is perfectly fine. Don't know if you already found this: <http://www.talkingelectronics.com/html/CapMeter.html> dwayne -- Dwayne Reid <@spam@dwaynerKILLspamplanet.eon.net> Trinity Electronics Systems Ltd Edmonton, AB, CANADA (780) 489-3199 voice (780) 487-6397 fax Celebrating 21 years of Engineering Innovation (1984 - 2005) .-. .-. .-. .-. .-. .-. .-. .-. .-. .- `-' `-' `-' `-' `-' `-' `-' `-' `-' Do NOT send unsolicited commercial email to this email address. This message neither grants consent to receive unsolicited commercial email nor is intended to solicit commercial email.

On Sat, 19 Mar 2005, Mike Harrison wrote: > With the sort of low capacitances involved with humidity sensors, > using the 555 makes a lot of potential problems (pin leakage, > capacitance, timing resolution) go away - you can use a relatively > high frequency and use the T0 prescaler if necessary. Resolution is > limited only by how long you want to wait for the answer. Or use a capacitance multiplier (Miller effect works fine) and forget about the 555. Peter

Herbert Graf wrote: >> <http://www.talkingelectronics.com/html/CapMeter.html> > > That is PERFECT!! Thank you Dwayne, I owe you a Coke... :) Hmm. That is more complex than your earlier description suggests you wanted. While I agree with the basic concept behind the circuit, there are some nagging issues too. I don't like the lack of bypass cap, especially with 100ohm in series with the supply. Then a 12V supply with 8.2V zener and LED in series doesn't look like a good idea either. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com

On Sun, 2005-03-20 at 09:45 -0500, Olin Lathrop wrote: > Herbert Graf wrote: > >> <www.talkingelectronics.com/html/CapMeter.html> > > > > That is PERFECT!! Thank you Dwayne, I owe you a Coke... :) > > Hmm. That is more complex than your earlier description suggests you > wanted. Not really, as I said I was looking for something that used a couple of opamps or a 555, this one uses just a single common chip (the 74c14), a few diodes and a few passives, about a simple as I was hopping for. > While I agree with the basic concept behind the circuit, there are > some nagging issues too. > > I don't like the lack of bypass cap, especially with 100ohm in series with > the supply. But it's battery powered and very occasionally used, and the odd glitch wouldn't be a problem, so I can understand why someone MIGHT want to forgo the bypass. That said, I agree that it should have a bypass (adding a bypass is something I automatically do anyways) and it's likely the original designer of the circuit would add one if a revision were needed. > Then a 12V supply with 8.2V zener and LED in series doesn't > look like a good idea either. If you read the text, that is there to ensure the battery has enough "omph" to use the circuit (i.e. if the LED doesn't like don't even bother looking at the reading). It's an interesting way of doing things, which assumes the drop across the LED is constant (which of course it's not). Again, I think the author was looking for bare simplicity and allowing for a little Vcc variation was acceptable. In my app I'll be running the thing from regulated 5V, so it's a none issue for me. TTYL ----------------------------- Herbert's PIC Stuff: http://repatch.dyndns.org:8383/pic_stuff/

A few days ago, this thread: "Re: [EE] Looking for capacitance to voltage circuit" asked about humidity sensors, capacitive type, and pics. One solution mentioned was a C -> F circuit with a 555, then to a pic. The recent Electronic Design magazine has this idea with schematic and code. You can see it here: www.elecdesign.com/Articles/ArticleID/9808/9808.html MD

Hello Herbert, Is this circuit work for you? I found that my latest project needs this kind of circuit as well. However it requires the test to be performed under this condition: o---R---C---o OR o---R--+--C--+--o | | +--R--+ In this case, do you think the same circuit will work? I only interested in the C measurement but the R(s) must be part of the whole thing under test, and the connection points allowed is as in above (mark with an 'o'). Thanks & best regards, WH Tan {Original Message removed}