Author Topic: Open Source Multimeter  (Read 277133 times)

0 Members and 1 Guest are viewing this topic.

HLA-27b

  • Guest
Re: Open Source Multimeter
« Reply #175 on: July 23, 2011, 01:22:02 pm »
Well the standards are more political than technical for me. Sticking to standards makes it a lot easier  for institutions to contribute. Imagine if we could develop a backplane akin to PXI which is open source (I know it is not literally a standard)- it would be immediate hit with all universities and research institutions. This would encourage them to develop more hardware for it, again open source, and to the benefit of us all. Btw. PXI is not a standard yet mainly because it is not fully open.

It isn't only us amateurs that need this stuff. universities and research institutions need it too. And I am not talking multimeters here. CERN just recently announced their own open source license. There are already several projects under that license, including a very nice clock signal distribution switch.  They sure have one hell of a budget, couldn't they just buy all this stuff? They could but that would have meant firing research staff, and it is not the gear that makes the science, it is the staff. The university I graduated from spent 2M$ in test and measurement equipment while I was studying there. Imagine if that could have been halved? They would have hired twice as much research staff. 

What I am saying is that standards are already out there. If we don't stick to them we have to develop our own, which is arguably impossible because we are talking hardware, not software. They make it easier to contribute to an open source project. They are already well documented - bad documentation is the plague of open source.
 

Offline bobski

  • Contributor
  • Posts: 37
  • Country: us
Re: Open Source Multimeter
« Reply #176 on: July 23, 2011, 03:58:11 pm »
Bobski's modular project looks great, but that amount of plastic would require significant tooling cost and is not really hobby friendly. I can't imagine someone making modules for this design at home.
I totally agree on the financial considerations. The project would need support from a better funded, or at least better equipped, entity. I don't envision the plastic parts being molded by home inventors. Rather, get the project producer to mold the parts and make them available as blank housings on the cheap. To further the project and get a ROI, the producer could go ahead and populate the main housing to spec and sell them along with their own module designs. That would make a standard level of functionality available for purchase with a minimum of fuss, while providing the framework for private inventor types to develop their own designs and further the project as a whole by sharing with the community.

Sticking to standards makes it a lot easier for institutions to contribute.
Sticking to standards (at least at key points in the design) makes it easier for everyone to contribute. Standards are needed if you're going to use a modular, isolated approach. Even if your end design is a physically monolithic device, keeping the circuitry compartmentalized reduces the likelihood of one functional section unintentionally influencing another. A standard set of connections at the borders of that modular isolation make it easier for people (individuals or institutions) to construct their own functional modules and cleanly append them to the core functionality of the meter. The trick is making that set of connections flexible and powerful design-wise, yet inexpensive and easy to work with.
[...] it would be immediate hit with all universities and research institutions. This would encourage them to develop more hardware for it, again open source, and to the benefit of us all.
Yes, this is what I would like to see happen. A large institution take up the cause and deal with the aspects of the design that are impractical for the independent developer to manage.

What I am saying is that standards are already out there. If we don't stick to them we have to develop our own, which is arguably impossible because we are talking hardware, not software. They make it easier to contribute to an open source project. They are already well documented - bad documentation is the plague of open source.
Impossible? I suppose if you want to develop a standard from the electrical and timing parameters up, that's a pretty daunting and expensive task. I don't see why that would be necessary here. There's plenty of existing hardware standards that can be pulled into the project design quickly and easily to create a project-specific standard.
For instance, the input modules on my design. One would obviously need power and ground connections, preferably isolated since we're doing measurements. An analog return line or two might be good... They could be multiplexed with the lines from other modules and fed to a central high quality ADC if the designer doesn't include one in the module itself. Since noise would be a problem on those analog lines, a shielded (coax?) connector of some sort is probably in order. Digital control and data lines of some sort are probably a good idea. SPI and/or I2C maybe?
Throw in some module dimensions and standard connector locations and you have a hardware standard.
 

Offline KuchateK

  • Regular Contributor
  • *
  • Posts: 78
  • Country: us
Re: Open Source Multimeter
« Reply #177 on: July 25, 2011, 04:49:15 am »
Even fancy multimeters are not using anything more than MSP430. This is processing power for SPI, I2C, RS232/485 level.

The standard for such multimeter should describe boards with specific size and connector on the edge that has I2C, SPI, RS232/RS485 (and some other) interfaces and a as much as possible control digital/analog pins. Something that requires more power than some cheap 8-32 bit MCU can handle will scare people off any involvement and development.

To be popular it has to be cheap. When it is popular then it is much easier to create high speed/fancy variations. I'll go back to Arduino as example. It became popular, because initial investment was ATmega8 with crystal and rs232 to a PC. Right now you can also get Arduino running ARM chip and it still works.

We should hit and aim at hobby/student/small business market. Guys with Arduinos that want to measure ESR/power or inductance without spending hundreds of dollars on few multimeters. It should be small modular and customizable (!) companion for cheapest Rigol scope and nothing beyond that.

Even if the tool won't satisfy CERN, it may be great for thousands of students. CERN? They have staff and money to do whatever they want (sometimes) beyond our imagination. We should leave them alone. We should focus on the other end of the market. Beginers trying to learn. When students will start to use something, then institutions will be forced to use it. They have tight budgets and they would love to have cheap open source educational tools.

In my opinion it would be easier to start small and get support from agile companies such as Sparkfun than big slow moving institutions with grants that are nicely distributed between big market players.
« Last Edit: July 25, 2011, 06:19:54 am by KuchateK »
 

HLA-27b

  • Guest
Re: Open Source Multimeter
« Reply #178 on: July 25, 2011, 08:47:47 am »
KuchateK
I follow what you say and I agree with most of the points you make. In that light, there are some concerns that arise. Supposing that we develop an arduino shield voltmeter, I am certain that any kid picking up this voltmeter will stick it into the mains first thing out of the box. We rally can't sell this thing without at least CAT II safety. Therefore we will have to provide it in a box. Who is going to pay for certification? What will happen if somebody really gets hurt? In the country I live in, if you screw up, that's your own damn problem. If you stick an arduino shield into the mains that just means you're being stupid, you've got nobody to blame. But that's my country and that's the way I like it. Europe and US are not like that so we have to figure out how to sell this thing without causing mortal danger and legal issues to ourselves. This is the state of mind I'm in right now.

OTOH developing an arduino shield voltmeter is a quick and dirty way of getting into it, so maybe exactly the reason why we should do it. There are posts on this thread that discuss the theories of measurement in depth and breadth yet no schematics came forward. Maybe it is time to leave the scalpel and get the ax.

 

Offline KuchateK

  • Regular Contributor
  • *
  • Posts: 78
  • Country: us
Re: Open Source Multimeter
« Reply #179 on: July 25, 2011, 03:18:00 pm »
Americans have this stupid idea that if something happens you can blame everyone except idiot who got hurt and his parents that they did not teach him list of "don't do this". Should we care? Hell no!

What can we do to prevent accidents?

If we won't write some warnings then we are to blame (a little). But every mains voltage kit I did had nice paragraph about safety and that is enough for normal people around the world. We are not going forcing anyone to use it.

I think it should be a kit, not a complete product. Then you don't have to bother about some safety issues and certifications. We should not even bother providing enclosures or parts. There are many sources for that. PCBs distributed by some people would be sufficient in my opinion. Good documentation/tutorials plus code is still time consuming. Someone can create kits and distribute them later...

Should we create Arduino shields? NEVER! This form factor begs to be unsafe when it comes to any higher voltages. And I don't really think useful multimeter can be made on (even big) shield.

But... We should use Arduino as a base. As a coding platform and interface to real world it is great. Why? Because you can jump to C or assembly when you need to in any section of your code. Changing default MCU configuration and taking full advantage of its capabilities is easy. If controller would take Arduino code, then we have customization and easy development covered. Arduino would be great connector between some first boards and computer during initial design. We can later jump to ARM or something more powerful, and because they ported Arduino libraries, code can be reused.

There is one more question that needs to be answered. Should we create mains voltage multimeter? Why not make first versions going up to 50V like an oscilloscope? Want to measure something higher? You know what to do :)

I wrote about analog pins on the board connector, but after some consideration, I think that boards should be as independent as possible with something smart built in. Some digital interfaces only providing ready to read data. Without many connections you can then shield them and isolate cheaply.

For example RS485. You have ability to connect multiple boards and external devices (programmable power supply for example). It is popular industry standard and it is quite fast on short distances and can work kilometer away. Microcontrollers are cheap and every board could have some bits without sacrificing cost that much. A chain would connect to a controller that would only handle data display/logging and user interface to modules after querying them what they are. PC would do the same even through terminal. It should be also easy to provide wireless modules for boards.

Seems that there would be need for some clocking signal from controller to each board so they can be synchronized. My knowledge is rather limited and someone having more experience with more complicated measurement gear should write about how they usually work.
« Last Edit: July 25, 2011, 03:58:32 pm by KuchateK »
 

Offline KuchateK

  • Regular Contributor
  • *
  • Posts: 78
  • Country: us
Re: Open Source Multimeter
« Reply #180 on: July 25, 2011, 03:56:08 pm »
here are posts on this thread that discuss the theories of measurement in depth and breadth yet no schematics came forward.
Unfortunately I cant contribute more than talking right now. For me analog circuits are mostly spaghetti ;) Don't be shy people. Show us something.
 

Offline Extech

  • Contributor
  • Posts: 15
  • Country: us
    • Planet Extech
Re: Open Source Multimeter
« Reply #181 on: July 25, 2011, 05:01:50 pm »
I wanted to do a case design for a kick starter video, and i would like people to tell me what they think of the case so I can make it look as good as possible.

Well I think that I am looking at one Gossen range switch,
plus the bottom of the Fluke 28II painted red, with out the holster, and a bit more square !!  :)

Well I have a tip for you,  you are in a good path, but you have to surprise us !!
Yes SURPRISE US !!

Its hard to happen,  but if you are serious in what you do, it will not be that difficult.   
If you like to know what to avoid ?  Look at Extech.   ;)

Nice!  :'(  :'(
 

Offline bobski

  • Contributor
  • Posts: 37
  • Country: us
Re: Open Source Multimeter
« Reply #182 on: July 25, 2011, 05:20:44 pm »
Nice!  :'(  :'(
Haha... Meter snobs? ^_^
I have an Extech as my primary meter. I wish it had a bar graph, but otherwise it seems to be a perfectly good instrument. Fluke and Agilent meters look nice, but as an EE student (when I bought it) I couldn't justify the price: 2-3x that of the Extech for the same feature set.
« Last Edit: July 25, 2011, 05:22:15 pm by bobski »
 

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: Open Source Multimeter
« Reply #183 on: July 25, 2011, 06:31:29 pm »
Here's an example of how a clever start up could take the big companies to the cleaners-
http://cp.literature.agilent.com/litweb/pdf/5990-7778EN.pdf

IMO, they totally blew it with the limited frequencies. With a synthesizer and good interface it would be a killer product, but they probably thought if they did that it would compete with their expensive stuff.
 

Offline bobski

  • Contributor
  • Posts: 37
  • Country: us
Re: Open Source Multimeter
« Reply #184 on: July 25, 2011, 09:32:02 pm »
 

Offline bobski

  • Contributor
  • Posts: 37
  • Country: us
Re: Open Source Multimeter
« Reply #185 on: July 29, 2011, 07:50:53 pm »


In case anyone wants to fool with it themselves:
http://dl.dropbox.com/u/9882625/openMeter%20Sketch%20V2.0.skp
 

HLA-27b

  • Guest
Re: Open Source Multimeter
« Reply #186 on: July 30, 2011, 09:14:58 am »
Mr. Hoffman

What is involved in putting a synthesizer on board? Can you point to example chips/schematics? Also what new capabilities will that add?

I think this is a new idea to most of us so you may need to bring us up to speed.

Cheers
 

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: Open Source Multimeter
« Reply #187 on: July 30, 2011, 08:54:16 pm »
Hi,

What you're looking for are Analog Devices DDS (direct digital synthesis) chips. It used to be simple, but now they have a huge family of them. I'd start by looking at AD9838, which is a simple waveform generator with an I2C interface. They even have the AD5933, which is a complete waveform generator and ADC designed for impedance measurements. Unfortunately it's aimed at higher frequencies and they say it only goes down to 100 Hz. This page would be a good place to start looking at all that's available. BTW, some of them don't even cost that much.

IMO, once you include a signal source in your device, all sorts of possibilities open up because you can then characterize both active and passive stuff (filters, amplifiers, caps, inductors and resonant circuits). You can also use it as a stand-alone signal generator and meter (because of course you'll build in a decent 5+ digit meter and true rms circuit).

People have gone in this direction before to some degree. Look up the Keithley Source-Meters. Too specialized and too expensive for most. The Agilent LCR meters I mentioned earlier IMO could have been much more, but they probably didn't want to drain sales of their more expensive products. The closest thing is a good PC sound card with a program like Visual Analyser, but the signal output from sound cards is less wonderful than they'd have you believe.

Best,
Conrad (my dad is Mr. Hoffman!)
« Last Edit: July 30, 2011, 08:58:52 pm by Conrad Hoffman »
 

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: Open Source Multimeter
« Reply #188 on: July 31, 2011, 05:16:51 pm »
Just to babble a bit, I've no idea how to package this thing- other people here are way better at that. Standards are nice, but can make the physical hardware costs very high. I remember Eurocards, with high pin count on the connectors, but also high backplane costs. There are "NIM bins", Nuclear Instrument Modules, but I don't remember much about that format, only that it was reasonably narrow. No doubt there are many other modular standards. Or, you can put everything in a single box. With a single PCB and all connectors on that single PCB, costs would probably be lowest.

I imagine the basic functions to be 3 major user blocks. First you have a DVM module. Next you have another DVM module that can also be set up to measure current. Third, you have a source (synthesizer signal generator) module. So maybe the box has 3 sets of leads- bananas, BNCs or whatever. So far, pretty simple and just an enhanced DVM.

The key is that the input modules are all fast ADC, feeding a "brain" (either internal CPU or external PC) instantaneous data. Now, with voltage and current, you easily get DC power- that's a no brainer. But it doesn't work for AC. However, if you do instantaneous voltage and current, that does work for AC. So now you have a proper (and high $$ value to the customer) VAW (volt-amp-watt) meter.

Add the signal source, plus the ability to measure instantaneous voltage and current, and you complete the picture, being able to excite the DUT (device under test) and measure its transfer function and impedance at any frequency. IMO, the instrument should work from DC to about 1MHz to be useful.

Hopefully that defines/explains things a bit better.
« Last Edit: July 31, 2011, 05:22:33 pm by Conrad Hoffman »
 

Offline Zad

  • Super Contributor
  • ***
  • Posts: 1013
  • Country: gb
    • Digital Wizardry, Analogue Alchemy, Software Sorcery
Re: Open Source Multimeter
« Reply #189 on: August 01, 2011, 10:17:11 pm »
I'm slowly developing an LCR meter using the AD5933 (see my blog) and it is a pretty good chip in general, but It does need driving from a divider (or another DDS) if you want to lower the minimum frequency, and you need to buffer the drive signal (preferably with variable gain and offset) if you want to measure below 200 ohms or so.


Offline AndyM

  • Newbie
  • Posts: 1
  • Country: gb
Re: Open Source Multimeter
« Reply #190 on: August 23, 2011, 10:15:52 pm »
I've been reading this thread whilst on my holiday (bored without my bench and soldering iron!!).
Whilst an attempt to adhere to standards is understandable I have to say that the simplest construction would be preferable.
As for the front and back panels, I have used FrontPanel Express for many of my projects and I find them to be reasonable (especially for one off panels) and they give a very professional finish.  They even have a range of enclosures that I have also used.
A bench meter with additional functionality as detailed by Conrad gets my vote with a single main PCB and separate pcb's for mounting the external connectors or perhaps wiring the connectors from the front panel down to the PCB.  I don't want it too big (perhaps something similar in size to the HP3478 would be good).  I'd like it to be portable but not permanently so (how about some rechargeable AA's internally and a power brick when used on the bench).
The potential is vast.  It would be great to be able to replace a transistor tester, LED tester, DVM, SigGen etc., etc. with a single instrument.
Love Dave's idea about the thermal probes as I built alot of vacuum tube stuff and often look at the temperature of transformers, chokes and valves.
The synthesis core might also allow characteristics to be plotted when testing semiconductors.

Have I gone too far?????

Andy
 

Offline House91320Topic starter

  • Regular Contributor
  • *
  • Posts: 176
Re: Open Source Multimeter
« Reply #191 on: September 13, 2011, 01:26:44 am »
Just some updates on whats happening with the meter (yes the project didn't just die out). i am currently working on the rev1 of the pcb. As it stands now its going to have a mono glcd, fram or eeprom (the code and board will allow you to chose), usb slave and host (the reason for the host is to be able to plug a usb flash drive in, should mention host and slave are iso to 5kv), and its a it can display and log 2, v channels and 2, A channel + a lcr channel. Any thing more you would like to know just shoot me a question.
 

Offline silversteam314

  • Newbie
  • Posts: 1
Re: Open Source Multimeter
« Reply #192 on: January 07, 2012, 05:09:05 pm »
I just read a blog post about this somewhere and came to check what has happened since Dave talked about this on the vlog.
Is there any progress, or is this project now at a standstill?
 

Offline Bored@Work

  • Super Contributor
  • ***
  • Posts: 3932
  • Country: 00
Re: Open Source Multimeter
« Reply #193 on: January 07, 2012, 06:03:56 pm »
I just read a blog post about this somewhere and came to check what has happened since Dave talked about this on the vlog.
Is there any progress, or is this project now at a standstill?

It was never a project. Dave was very clear about it that he was only doing the sketch for fun. And as you can see from the discussion here, it degraded to what always happened with such ideas. Lots of people coming up with their wishes in the hope that they find someone to do the actual work. Those doing some work quickly find out that it is really hard work and that there is little reward doing this.
I delete PMs unread. If you have something to say, say it in public.
For all else: Profile->[Modify Profile]Buddies/Ignore List->Edit Ignore List
 

HLA-27b

  • Guest
Re: Open Source Multimeter
« Reply #194 on: January 07, 2012, 08:44:19 pm »
Actually this thread inspired me to do something entirely different but still along the lines of open source hardware.
I am hoping to spill my meager beans soon.
 

jucole

  • Guest
Re: Open Source Multimeter
« Reply #195 on: April 22, 2012, 09:29:29 pm »
Hi, this is the sort of thing i'd go for, it's called the Multimogger (Multimeter / Logger); I actually missed out the function generator out, so you have to imagine that's in there too. These could available in different widths allowing for more (expensive) or less (cheaper) features.



The case is simple no-frills, no display either, all that is done in software on the smart device. One thing I thought might be nice was triggers, so i included two outputs, these would be dependent on specific user defined parameters or events. For example when temperature / voltage or current etc on channel X reaches / falls etc Y  then trigger output 1 or 2. You could set it up in the UI like a typical logic analyzer trigger steps, did i tell you it had a I2C bus viewer in there too? (maybe phase 2 ;-)

The UI software logic would handle most of the clever display and control stuff allowing for a minimal hardware specification for each channel plugin (shield?)

All communication is done via bluetooth with the idea it works in conjunction with a smart device or a laptop. 

Also it's got to be really nice and heavy, filled with lots of batteries, I don't want something to slide across the bench every time I hook things up to it.

It's the swiss army knife of multimeters!! and it's putting the MULTI back into multimeter! ;-)


Kind Regards

Jules


 

HLA-27b

  • Guest
Re: Open Source Multimeter
« Reply #196 on: April 22, 2012, 11:13:23 pm »
You mean something like this?



I've been working on mechanical aspects it for a while. It is a very useful concept except all the difficulties in making it actually work ::).
« Last Edit: April 22, 2012, 11:18:55 pm by HAL-42b »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: Open Source Multimeter
« Reply #197 on: April 23, 2012, 02:25:23 am »
I've been working on mechanical aspects it for a while. It is a very useful concept except all the difficulties in making it actually work ::).

Those pesky minor details!  ::)

Dave.
 

jucole

  • Guest
Re: Open Source Multimeter
« Reply #198 on: April 23, 2012, 09:16:39 am »
You mean something like this?

hehe. Now that's just weird! did you sneak into my room while I was asleep claiming to be Darth Vader from the Planet Vulcan and download my thoughts with an 80's Walkman to the sound of Eddie Van Halen?


There seems to be lots of ideas and drive to build such a device, does anyone know of any websites that allow people to contribute to hardware projects in the same way as Sourceforge does for opensource software? if not would sourceforge be a place to start?
 

HLA-27b

  • Guest
Re: Open Source Multimeter
« Reply #199 on: April 23, 2012, 11:29:53 am »
There seems to be lots of ideas and drive to build such a device, does anyone know of any websites that allow people to contribute to hardware projects in the same way as Sourceforge does for opensource software? if not would sourceforge be a place to start?

No such place yet afaik. Actually I'm not sure I've seen a strong collaboration on any OSHW project yet, let alone the need to lower the participation threshold.
The community does not know yet how to handle collaboration when the item in question is an actual physical object to be sold for profit. There is no unit of measurement for contribution.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf