Shop OBEX P1 Docs P2 Docs Learn Events
Would anyone be interested in a propeller board geared towards instrument development? — Parallax Forums

Would anyone be interested in a propeller board geared towards instrument development?

I had listed some product/kit ideas I had in another thread and it got me to thinking... I need to learn an EDA package. That is my biggest folly right now. I took about three years of AutoCad so im sure it wont be some huge leap, but I need to move from point to point and dead bugging if Im ever going to get anywhere, I understand good layout theory and mixed signal layout, I have just never made a real PCB, the toner transfer never worked out =p. I have a toaster oven now and a small butane based heat gun for touch up (pretty nice tool for 20 bucks! BenzoMatic 3 in 1 self igniting torch, heat gun, and soldering iron with variable heat this puppys heat gun goes from 200F to 1100F), along with a hot plate, I wish I would have had this when trying to get cramped 0601 caps off board I was modifying!

So I have always heard Duane say propellers are perfect for robots because they can multi task... and he is dead on! I want to make some gear that would have no problem multitasking as long as all the appropriate analog front ends were tied to different cogs. This got me thinking of really small minuscule tasks we do everyday like using two meter to probe current and dc voltage at the same point... this could be done with an ADC and a shunt resistor along with an LCD or serial output.

So here is my idea do any of you guys think a custom prop board made for instrumentation would be a good idea? Something that uses TPS series low noise linear vregs for its power rails, a 6mhz TCXO to keep a reasonably stable time base. Then sprinkle in a 16 bit DAC and 16 or 24bit ADC (whichever is fine with me, one saves money one allows oversampling), a high precision 10v reference connected to a trim pot to set your analog reference voltage, or maybe two resistor networks for multiple voltages? Then add some banna and bnc (or sma, seems a bit to much) sockets to the thing and of course optional Ardweenie style headers. Im not sure what else one could want maybe a display on the board that connects via ribbon cable that way it could be panel mounted, and maybe a select able shunt network? A thermo couple chip, a true RMS converter, a DDS? lol idk it just keeps getting more and more pricey. But you would be able to prototype any test tool you wanted easily and quickly.

Would any one be interested in something like this? and if so how does one get parallax to sell there boards on the site, i know things like the c3 and the arduino looking propeller board arent official parallax boards. This was really just my idea as an easy first PCB project which could help me slap things together later for prototyping

Comments

  • David BetzDavid Betz Posts: 14,516
    edited 2015-10-11 15:26
    Sounds interesting. What about MIDI connectors?

    Edit: Never mind. When I saw the word "instrument" I immediately thought of "musical instrument". That isn't what you're talking about I think.
  • Don't forget that 'pricey' is a relative term. Some people consider $5 to be expensive while others consider $500 to be cheap. You want to target the $500 people.

    I haven't seen a high end board like you're talking about that uses a propeller. Would be an interesting project and a good learning experience if nothing else.

  • For me, being a scientist, it was natural to think about "scientific instruments", such a potentiostat or a spectrophotometer.
    I will be interested, obviously….

    Carlo
  • Well it can obviously be broken out to connect any kind of sensor or interments, Is there any certain connection types or chips that should be on it. I guess I think of as electrical instrumentation but it could really be used for any hi end low noise ADC instrumentation.
  • Yes. I have used a P1 this way a couple times. The big draw would be shipping it ready to go as at least one instrument.
  • A 50khz audio analysis tool?

    Signal data logger in combination with logic analyzer?

    Tricorder?

    With great DACS, one might get good use out of a sensors. Temp, pressure, magnetics, light, position, etc... datalog those to SD and maybe figure out how people can connect new ones in robust ways.

    I do not know as much as I want about the very sensors available what they can do how their best utilize all that sort of stuff.

    Just connecting one to get a deal for what it does with the range of values are interesting to see it operate might really be worth a lot, particularly if there were some common sense display or date of record or chart or something where I can easily take those numbers and associate them the real world things.

  • Well basically I might make this kind of modular the point of it was to help someone learn an EDA package and be reusable in some projects of mine for test equipment. You dont need anything fancy for a logic analyzer just some code.. lol. Adding SD an EEPROM would take care of data logging.

    What do you mean bu a 50khz Audio analysis tool? Like a scope? just plug in to the BNC and use the ADC to log the points on the sign wave. Im not sure how im going to approach this in the end honestly. But the point being if I have to get a 10 of pcb order I wondered if I populated them if it would be of interest to people.
  • One chip I think I really may be putting on it along with some decent opamps is the LT3092 200ma programmable current source.

    You guys have to realize this board is targeted towards mixed signal instrumentation of any kind really (not music). But all the things mentioned above could be done with it if one were to write the code.

    My Issue is im not sure weather making a more genric board with some chips on it is the better route to go, then making expansion boards, or acually throwing connectors and stuff on the PCB. I do think an OLED screen with long ribbon cable would be cool idea, but then again what if the end user wants a 4 inch TFT?
  • Yes - an instrumentation Prop board would be good BUT you quickly get bogged down on what type of instrumentation I/o chips top use. For one company I worked for years ago, I made literally hundred test and fixture setups using the old Basic stamp chip . board. Test and fixture is very different from instrumentation - many checked a bed of nails PCB test set, continuity tests, push button tests, analog tests, electrical noise (S/N ratio) tests and the like. each one was unique and different.

    What I really could have used is a good buffered input /output micro controller, extra memory / micro SD -card, stable 12v0, 5v0 and 3v3 switching power, ultra precision 5v0000 supply. RTC chip, ADC chip with gain amps /voltage divider, AND a really good graphical "LabVIEW" logic editor: the kind with logic and I/o symbols & configurations, and "wire" connections. A glorified PLC type of circuit and software.
  • Sorry, I thought you were looking for more product oriented ideas. A board run to pay for your own is great, but not what I wrote to. Sorry about that. Ignore my post and carry on. :)
  • Hmmm I know nothing about lab view lol :). The idea was a run for myself but if people liked the idea maybe I could start selling modules too. I don't agree with switchers I do however think -/+ low noise linear rails and a stable refernce are a must along with some differential precision type op amps. Im not even sure if you can do a 24 bit ADC without getting in to the noise, using hi end linear rails and battery so switchers are way out of the question, unless there set up for pretracking to keep the linear regs heat down.
  • evanhevanh Posts: 15,918
    When multiple analogue points are in use then electrical isolation on each individual analogue input and output is very valuable in removing common mode noise. This applies to both the signal and the supply.

    One way this is typically achieved on the cheap is for the controller to only have a single analogue point. Then the controller can be commoned to this analogue point and have only the digital side using isolation instead.

    I was once involved in a non-isolated project that ended up requiring unwanted insulation to get anything meaningful at all, and even then it was always borderline too unstable with each finished product yielding different levels of extreme in the instabilities.
  • I you want to make your own PCB's may I suggest expressPCB the software is free and easy to use. The pricing isn't too bad because you can get your boards in a few days. I make my own 8 channel analog input card , 8 channel analog output card. 5 channels of frequency and counters, The system also has 8 channel relays. All are expandable 8 channel, 16 channel etc. This system all runs on the prop dip 40 and I have full Labview VI 's for the afore mentioned items. So by using all of these items and a display for stand alone use you can interface any pressure, volt, amp, flow, vibration senior erc . And then output commands via relays, 0-10 volt or 4-20 ma.
  • Hmmm I started with that software as a way to do schematic entry but I have some issues. I dont think the pricing is good as something like dirtyPCB or seedStudio and OSHPark all though im not sure about turn around time im sure thats a huge plus since I think they are a US company.

    I love there EDA software it is so easy to use!! I have also received PCB's from other forum members using there service and they are good quality. My biggest issue is there closed system, I would definitely use there services if I could use there files to order a one of from OSHPark it would be great, you don't always need 10 boards with fast turn around. If they allowed other board houses I think they would be my go to. I mostly need four layer mixed signal boards that are SMT. I love DIP but and if designed properly will work fine with point to point, but as far as gettint a real PCB done smt all the way of you want low EMC and noise.
  • Hey DigitalBob - That is pretty much what I had in mind. A good solid 8 channel I/o cards, Relay card, Frequency counter - so you can interface all kinds of sensors. Do you have any plans to go into production with these PCB's? How do you integrate LabVIEW with the Prop? Do you / or would you need a Prop graphical IDE to do programming?
  • Hello propguy: right now I produce the boards for in house projects . I have worked out all the VI,s for labview. If you want to run the whole circus analog in, analog out , hz counters, relay I/O , LCD display etc. you will have to use a separate cog for each one. If not it's slows down the visa read in lab view and you will get errors. I can run all with 150-200 millisecond delay to update the labview display without any errors. Note A/C loads on the relay card causes spikes in the USB and will yield lab view errors even with 100% opto isolation in the relays. The solution is using a high voltage snubbing capacitor across the relay coil. This problem only occurs with A/C loads. I run 5 channels of counter / frequency so my update time stays under 200 milliseconds. More channels can be used but your update time will be 1 second or longer.
  • Hello rwgast: You can get really cheap Pcb's if you order out of china just submit your artwork pay and wait a couple of weeks I know a guy that sells robot boards on the internet that uses that method to keep cost low.
Sign In or Register to comment.