Shop OBEX P1 Docs P2 Docs Learn Events
Fastest rise time achievable with a propeller, and Fast ADC? — Parallax Forums

Fastest rise time achievable with a propeller, and Fast ADC?

Im looking to build up a quick LC meter, adafruit has done a little tutorial on how to do one with the arduino, but it just doesn't meet my requirements at all. If I had money I would just buy a decent one, I think a homebrew done right will be a lot better than a cheap Chinese e bay meter.

Basically what I need to know is what the fastest rise time of the prop can be, or if I should use a cmos logic circuit for fast edges. Secondly since im measuring inductors who know what there resonant frequency will be so what is an ADC that can capture in the mhz range easily? Im basicaly using the method you would use with fast pulsing a tank and looking at a scope.

Comments

  • jmgjmg Posts: 15,173
    Basically what I need to know is what the fastest rise time of the prop can be
    ....what is an ADC that can capture in the mhz range easily?
    Sounds complex, given a Prop can only capture every 12,5ns.
    A simpler approach that needs no fast edges nor ADC, is to use a LR circuit as an oscillator.
    https://en.wikipedia.org/wiki/RL_circuit

    The Prop can measure to 40MHz with high precision
    ( see Bean's frequency counter example)
    That gives a ball park lowest L of around 2.5uH in a 100 ohm L-R oscillator, and the LSB will be < 1nH
    You could maybe push lower with a LC circuit, but need to watch for the self-resonant point on the inductor.

    How low do you need to go ?
  • Hmmm well Im trying to measure tiny air core inductors for band pass filtering, the adafruit arduino example doesn't go anywhere near 1nH. I would like to get a broad range maybe a bit smaller than 1nH up to maybe 500/1000uH im not sure is 1000 is even practical.

    My idea was just to fast pulse it and measure the frequency of the ring, I would like to be able to add Q factor, impeadence analyzation, and a lot of other cool stuff in the future so Im kind of looking to do something expandable not just quick and dirty. Ive even thought about using a DDS like to DS1077 to run sweeps, but im not sure what it's rising edge is. I know a prop can pulse at 12.5mhz but what exactly is the fastest rising edge it can do and what how would you accomplish that, timers and PASM, would spin do??

    As far as the ADC side what would I be looking at for a low bit but FAST adc, i dont need 16 bits of resolution here, and 8 or 10 bit would be fine if it could sample fast enough. Im not looking to build something to compete with agilent here, but I have a feeling the stuff on e bay is junk, unless you pay a bit for a brand name.
  • I also know you can use a wheatstone or wein bridge to determine initial impeadence and then do the calculations from there but my understanding is thats how cheap meters work and using Time Domain Reflection is a pretty good method which allows for various other measurements.
  • jmgjmg Posts: 15,173
    Hmmm well Im trying to measure tiny air core inductors for band pass filtering, the adafruit arduino example doesn't go anywhere near 1nH. I would like to get a broad range maybe a bit smaller than 1nH up to maybe 500/1000uH im not sure is 1000 is even practical.

    My idea was just to fast pulse it and measure the frequency of the ring, I would like to be able to add Q factor, impeadence analyzation, and a lot of other cool stuff in the future so Im kind of looking to do something expandable not just quick and dirty.

    Hmmm... Some reality checks ... 1nH and 15pF resonate at 1.3GHz and 10nH and 15pF resonate at 410Mhz
    A 10mm PCB trace is about 7nH

    The problem with ringing is you have few cycles to capture, and you need to resolve to ps

    What frequencies are your bandpass filters ?

    Making an oscillator using the inductor will give the best resolution, as you have many stable cycles to work with. Sounds like you need to test the inductor, at around the frequency it will be actually used at.

    If you need to go >> 100MHz, that's special oscillator design and special dividers.
    There are PLLs around that go to GHz and have the dividers inbuilt.

    If you need < 200MHz maybe a Si5351 can do, as it can synth any Freq and you could sweep a LC with a RF diode probe up to 200MHz
  • jmgjmg Posts: 15,173
    Digging some more around for something that is 'expandable not just quick and dirty', and looking again at PLLs/ Synthesisers...

    Some of those spec good upper limits, but you also need to watch the lower limits when looking for good dynamic range.

    Often dividers / amplifiers that manage GHz+, cannot also manage much below 100MHz

    A part that looks great for high end instrumentation, and that proudly specs DC as the lower f(in) is the Analog Devices (Hittite) HMC70x series.
    HM703/704 look to have lowest Icc and best specs.
    (suggests they are latest generation)

    This one looks quite impressive
    http://www.analog.com/en/products/rf-microwave/pll-synth/hmc703.html#product-overview

    table 28. reg 0fh GPo register
    shows how to tap into the internal signals.

    It has a DC-350MHz in, and a DC-8GHz in, so you would make a some-nH LC oscillator, with maybe a choice of Caps, and can use another LC or RC osc into the 'lower' 350MHz in, and you should easily be able to measure inductors operating from sub 1KHz to GHz, with Bean's Frequency Counter code and some maths.

    With the right VCO, you can also sweep your finished filters :)

    Not a 'cheap' device, but much cheaper than a Ringing capture ADC and with wide general use - would make a good RF Generator and Frequency Counter too...
    Seems to be readily available.

  • Well I have some VFO's using DDS chips and micros that can sweep filters and I also have a zener broadband source ive made. So ive got that nailed I guess I just want to check a 3nH or 500uH inductor is acually the value it should be after warping it
  • You need a grid-dip meter to measure LC resonance at these kinds of frequencies. https://en.wikipedia.org/wiki/Grid_dip_oscillator
  • Have you seen Hanno's method of capturing NTSC signals?

    He uses an 8-bit parallel ADC. I think it's a pretty cool device.
  • thanks ill check in to both of these :). Ive never even heard of a grid dip meter lol, and I was thinking about using the chip in the propscope so obviously his parallel adc may be more cost efficient and easy
Sign In or Register to comment.