1. Home
  2. Electronics Design
  3. imperfect exponential
  4. Page 6

imperfect exponential

There's a way to make a nonlinear cap in LT Spice.

I think this is right:

Version 4 SHEET 1 1096 680 WIRE 48 80 0 80 WIRE 176 80 48 80 WIRE 320 80 240 80 WIRE 480 80 320 80 WIRE 576 80 528 80 WIRE 624 80 576 80 WIRE 0 144 0 80 WIRE 320 144 320 80 WIRE 528 144 528 80 WIRE 480 160 480 80 WIRE 0 272 0 224 WIRE 320 272 320 224 WIRE 480 272 480 208 WIRE 528 272 528 224 FLAG 0 272 0 FLAG 320 272 0 FLAG 480 272 0 FLAG 528 272 0 FLAG 576 80 CAP FLAG 48 80 RAMP SYMBOL res 304 128 R0 WINDOW 0 49 47 Left 2 WINDOW 3 50 78 Left 2 SYMATTR InstName R1

SYMBOL voltage 0 128 R0 WINDOW 0 68 65 Left 2 WINDOW 3 17 112 Left 2 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value PULSE(0 100 0 100) SYMBOL e 528 128 R0 WINDOW 0 64 53 Left 2 WINDOW 3 58 85 Left 2 SYMATTR InstName E1 SYMATTR Value 1e6 SYMBOL cap 176 96 R270 WINDOW 0 -43 32 VTop 2 WINDOW 3 77 35 VBottom 2 SYMATTR InstName C1 SYMATTR Value q = x / (1+2*x) TEXT 16 112 Left 2 ;1 V/S TEXT 560 112 Left 2 ;1 V / F TEXT 368 8 Left 2 !.tran 0 1 0 TEXT 8 -96 Left 2 ;Nonlinear Capacitor C:V Curve TEXT 48 -56 Left 2 ;J Larkin August 11, 2014 TEXT 480 -96 Left 2 ;Cap is 1F at low voltage TEXT 536 -56 Left 2 ;0.1F at 1 volt

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

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

JPG

Why? If it is a bridge, it is comparing the unknown capacitor with a built- in reference capacitor, and 47uF might be a bit bigger than Booton allowed for.

47uF in plastic film capacitors isn't impossibly bulky, so you might be abl e to bodge up something.

The instantaneous value of a capacitor can depend on the voltage across it at that instant, and the apparent value of a capacitor can depend on the fr equency at which it is measured due to "dielectric absorbtion" which can be modelled by parallel resistor-capacitor segments.

Differentiating a charging curve gives you the apparent capacitance at each voltage, but differentiation is a noise-emphasising process.

The classic AC-excited bridge is another way of doing that "differentiation " where you use lock-in to reject irrelevant noise.

--
Bill Sloman, Sydney
Reply to
bill.sloman

Somebody who knew enough about Ph.D. projects to use the correct contraction could beg to differ. The literature is full of models for imperfect capacitors. The difficulty is that capacitors are imperfect in a variety of different ways.

--
Bill Sloman, Sydney
Reply to
bill.sloman

This is kinda a random thought... and not related to your ltspice circuit.

I was thinking it would be nice if digital 'scopes had a log/linear display function. I often check single pole time constants by finding the 2/3rd's crossing point.... which involves fiddling with gains and such. (I'll want a log freq. axis in the fft display too.)

George H.

ltspice circiut

Reply to
George Herold

Rhode and Schwartz demo 'd a scope that would do this at a "lunch and learn" at work today. It had programmable filters, too. With the appropriate programming it appeared that it could make your signal look however you want it to. I noted that this functionality would be really handy in an EMC chamber. ;-)

The scope, as demonstrated, was something above $50K, but it was 4GHz. It was also a Windows scope and crashed (BSOD) in the middle of the presentation. Lunch was good, though.

Reply to
krw

When we were debugging a bunch of PCIexpress stuff, we bought a LeCroy

4-channel 7 GHz (20 gs/s) scope for $50K.

It does this, but not very often:

formatting link

We were trying to measure the jitter of a couple of circuits last week, and the LeCroy is better than our Tek 11801C samplers. The jitter is under 1 ps RMS. There's an "enhanced" mode that claims 100 fs, but we haven't figured that out yet.

I had a personal beef with Walter LeCroy once, but it's Teledyne now.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics
Reply to
John Larkin

We are programmed to breed. It's pretty powerful.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics
Reply to
John Larkin

The C(v) behaviour is certainly interesting. On the rise the initial capacitance is more like 200uF at 0v falling to around 20uF at 20v.

BTW it appears to be closer to 100uF at nominal 6v and 50uF @ 8v. I wonder what batch to batch variation is like?

The decay curve is similarly corrupted but isn't as obvious since superficially it looks more like an exponential sans kinks but with a rather too fast drop from 20v and a very slow tail from stored energy.

In a large signal curve fitting sense the rise Ceff is 65uF and the fall

40uF but it actually is a highly variable "fixed" capacitor. The answer is biassed by the time spent in each voltage regime.

It would be interesting to check this more systematically by driving it with constant currents of 20uA, 10uA and 5uA to confirm that it is indeed entirely terminal voltage related rather than current based.

The curve will look something like UUUU (exaggerated) instead of VVVV

--
Regards, 
Martin Brown
Reply to
Martin Brown

Here's a 1210 50V 10uF X5R capacitor C-V characteristics:

formatting link

Impressively bad, even at 20% of rated voltage, and they practically disappear at 1/2 or 2/3 rated voltage.

Datasheet:

formatting link 

--
Best regards,  
Spehro Pefhany 
Amazon link for AoE 3rd Edition:            http://tinyurl.com/ntrpwu8
Reply to
Spehro Pefhany

How does the efficiency of such a cooler compare with the miserable performance of a Peltier device? I suppose the Peltier would melt anyway at downhole temperatures.

--
Best regards,  
Spehro Pefhany 
Amazon link for AoE 3rd Edition:            http://tinyurl.com/ntrpwu8
Reply to
Spehro Pefhany

Apply a variable DC at the cap and use an LCR meter to find the (small signal) capacitance? Perhaps with big good decoupling caps -don't know how well the LCR meter deals with superimposed DC.

Pere

Reply to
o pere o

Depends a lot on the conditions. Our modelling was based on the DeltaEC cod es from the DOE group (Livermore or Los Alamos iirc). It showed that with a kilowatt in, we could easily cool anything we liked from 200C to 25C, with a big margin.

The best thing about it was that TA fridges have zero moving parts--they're just stainless steel and copper with a bundle of ceramic tubes inside and a Chromalox element at one end. That would be likely to survive being chuck ed in the back of a truck and bouncing over washboard to the next well. (St irling coolers are notoriously vulnerable to shock.)

Cheers

Phil Hobbs

Reply to
Phil Hobbs

I used to have a TDS 7704A (7 GHz, 4 channels), which ran XP.

My fastest digitizing scope nowadays is a TDS 694C (3 GHz, 10 Gs/s simultan eously on all four channels). I got a super-nice rebuilt unit with a new CR T for $2700.

Boat anchors rule, at least when you're spending your own money!

Cheers

Phil Hobbs

Reply to
Phil Hobbs
020109090005090207010209 
--
This email has not been checked by half-arsed antivirus software
Reply to
Jasen Betts

I'll try to get time to do a real CV curve, maybe with a small sine wave on top of a DC bias. I think we have one c-bridge around here that goes that high and has a bias input; otherwise I can just do it with a function generator and some math.

OK, that's not hard. Connect the 50 ohm fungen across the cap, set a DC bias plus a small sine wave, find the -3 dB frequency.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics
Reply to
John Larkin

One trick is to put two of the same caps in series and apply the DC bias at the midpoint, so the bridge never sees the DC.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics
Reply to
John Larkin

That is about the best it can be in a Bayer mask image since the red channel is uniform and the green channel which carries the detail has the most numerous pixels and highest luminance.

Red on black fine detail would be visibly worse, and a bug in most JPEG decoders makes the degradation on decoding even worse than it needs to be. I ran into this when taking H-alpha images on a Bayer CCD once.

--
Regards, 
Martin Brown
Reply to
Martin Brown
020109090005090207010209 
--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics
Reply to
John Larkin

Artifacts and colour bleeding! Two attributes you won't have to suffer when you use a top quality analogue storage scope (like the Tek 466 for instance).

Reply to
Cursitor Doom

Have you been drinking?

Reply to
Cursitor Doom

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.