Dangerous Prototypes

Here is a new project we are thinking of attacking. A digital multimeter...
To start of we would make it simple, and easy to build and prototype, to play around with.

Here is what we think would be nice for version 1, but as with all our projects input is both encouraged and in some cases needed :)

*(+/-)Volt,Ohm,(+/-)amp meter (both AC and DC)
*controlled with a microcontroller (with usb)
*Microcontroler will control the ranging, although probably manually at first and auto later.
*possible app to follow it..

those are the basics we want to get right first, once we figure out haw to implement them, sky will be the limit.

below are a few schematics we found on the net of commercial DMMs
DMMSchematics (http://http://www.megaupload.com/?d=Z65I97KJ)

Voltage Input section
To start of the topic I though of sharing a voltage input design that came up..

Most cheap MM use some kind of a voltage divider/ buffer to lower the input voltage to the -200 - +200 mV ADC. Both negative and positive signal go into the ADC and half of the ADC's resolution takes care of either of them.


This is a op amp rectifier circuit that converts and buffers positive and negative inputs to a proportional positive output. The proportion is controlled with resistor R1. (this is also where range changing can be implemented)

This will allow for the full range of the ADC to be used for both positive and negative signals.

this is only suggestion, and any and all input is welcome.
Schematic source (http://http://sound.westhost.com/appnotes/an001.htm)

Also to achieve RMS values of AC signals a dedicated part of the circuit will be required or a single IC like the LTC166 (http://http://cds.linear.com/docs/Datasheet/1966fb.pdf) which at 6USD is not very cheap..(the cheapest true RMS to DC Ive found)

As I already mentioned, custom DMM chip is required if you want to get price close the of the shelve dmm. In any other configuration your dmm will be more expensive and with lower spec then similar of the shelf one.

The best custom dmm chip I found is ES51966. It is very good chip with embedded autoranging, protection etc etc..

VC920 (great dmm by Voltcraft - extremly good performance) schematic show how you can use one (it uses ES51966 for DMM functions, AD636 trueRMS to DC (up to 900kHz) for AC and TIMSP430F135 to control everything calculate stuff and display it on the LCD). I attached the schematic for your convenience.

Problem is, you can't order "one" ES51966 chip. I spoke with them and they are still selling the chips if you want to design DMM that uses them (maybe they in the meantime made something even better). Now if DP want to go into DMM selling business it would be possible to get some but I don't see this on a "open pcb" project as no way ppl can contact manufacturer and get "one chip" for the project.

There are other specialized DMM chips and they are all being sold the same way - in bulk only.

I spent a lot of time deciding to make my own DMM, actually I wanted to make some "turbo ultra fancy universal something" LCRV with super low and super high values, dual measurements, calculations etc etc ... one of the very interesting projects in this area is:
fully featured portable LCR meter (http://http://www.circuitcellar.com/microchip2007/winners/third.html) (third place - microchip competition 2007)
It is LCR meter and not dmm but has some interesting analog parts in it ..

There is another very cool project I can't find link to CORBA or KORBA or something like that

:/ not to hate but I don't like this idea. I have multi meters. everyone has multi meters. why would I want a digital terminal based multi meter...

if it was a super probe, perhaps.

Well there's always a good reason to have a multimeter that you can control fully from the PC... I for e.g. wanted to make "all in one" device that also connects to computer (of course works stand alone too) ... I for e.g. gave up on that idea because after more then 6 months intensive research (more then 100 dmm schematics, talked with 3 different chip manufacturers, got some samples ..) I figured out it would be way more expensive to make it then to just go to a store and purchase it new, not to mention second hand ebay stuff ..

Then I started purchasing equipment instead building it ... (I recently got me this baby :D (http://http://www.hioki.com/product/352250/index.html) that I'm very proud for :D )

Hackaday had a 'add a computer interface to cheap dmm' a couple of months ago, but they didn't mention that it was not isolated.  It's hard to compete for price with a $3 harbor freight multimeter, but they tend not to be very acurate.
besides a computer interface, I would like a clamp-on dc ampmeter, but the hall sensors needed for that don't seem cheap.

The whole story about DMM is that they are very simple devices. There are no unknowns in theory of operation .. Accuracy depends on the parts you use, more accurate parts - more expensive the thing is.. the features are also pretty straight forward, there are no "news" there it's just that - you want more stuff in - you need more chips in, chips are expensive, dmm manufacturers get them in huge quantities and use specialized chips .. if you build 100pcs there's no freaking way you can beat the price ..

In January this year I was making some measurement tool and I needed the ICL7601 .. it was cheaper to get a DMM in a local store that sell PRC stuff and get icl out of it then to purchase the darn chip in a store .. the whole DMM was cheaper then the chip alone!

Our goal is more to find an old multimeter (discrete and analog components) to clone, and maybe bolt a uC onto. We'll never compete with existing multimeters, they're so cheap I've never spent more than $10 on one. I though an open source and usable tool, based on 40/50/60 year old tech (nixie style tools :) ) might be a good learning device and publicity piece.

The initial specs are pretty much - DC voltage and current only. Filip is looking to see what is out there and if the design is even feasible. It sounds like you've been here and the answer is - it's not :) We're just toying around a bit with the idea, there's no intent to actually engage this project yet :D

I wanted to make one for myself for many reasons, did some research and gave up.. it's too much money for no real gain. They either use some Intersil adc with lcd/led/bcd output or they use custom chips.

I also looked at the analog design and those are even worse - the amount of parts required is just huge, and calibration is a complete mess...

That chip I found in Volkraft is the best one I manage to find. I contacted company and told them I'm interested in designing a new line of dmm's and if they are willing to sell me few pieces to test stuff. They actually sent me 2 chips for free but I doubt they'd start selling them in small quantities so not a "solution". I contacted few other companies before I stumbled upon this chip and got no reply at all... so it's a good chip and good company, the only problem is the project would be feasible only if DP / SEEED do some bulk purchase and I'm not sure it would be worth it at the end of the day ..

Making the spec "DC only" ain't a big help as AC is solved by a simple trueRMS 2 DC chip .. there's bunch out there, they cost from few $$ to few tens of $$ but nothing special... again the one in Volkraft is very good AD636 (http://http://www.analog.com/en/special-linear-functions/rms-to-dc-converters/ad636/products/product.html) it is not cheap but works up to 900kHz  and is order of magnitude more precise compared to LTC Filip found (but LTC is order of magnitude cheaper).

Basically all these chips use different ways (switches, relays..) to get input signal to go to -200 to +200mV and then measure that and display .. You can fairly easy put a mcu with adc instead of this adc and process the signal, but except for serial output, what would be the added value ?!

As I said the DMM was only "part" (big part) of the device I wanted to build. I wanted to make real universal meter to have
 - precise 4 wire resistance measurement (micro ohm measurement)
 - precise nano ampermeter and nano volt meter
 - 2x voltage and current measurement (so that I can on the same screen see the input and output power of the psu for e.g)
 - CC and CV source
 - Electronic load
 - LCR meter with proper series and parallel capacitance, inductance and resistance values on different frequencies
 - counter / frequency meter
 - signal generator

I started with dmm functionality first and gave up after research .. what you can today get on ebay.com and ebay.de .. it's just not worth it .. not if you are making yourself measuring devices... if you want to learn - it's fun but it's not going to be cheap learning experience...

There is a big number of dmm and amm schematics floating around. I think Filip managed to get most if not all of them .. maybe you figure out how to make a project out of it, might be interesting, will for sure be fun ... all I'm saying is it ain't gonna be cheap

Thanks for sharing your experience. It is a real reality check. This is exactly why we chose to do this project in the public forums :)

I'd love to see what others think too ...

wrt sanity check, ICL in store is 6eur and cheap prc dmm using it is less then 5eur :)

If I just may say, most of the time if you yourself try to do a project because you think you can do it better or cheaper than current commercial things out there, you will have an incredibly difficult time. My idea when I do these things always starts out like this:

Oh look at that! That is so cool! I wonder how much it costs ...
WHAT !? $600 USD !? It only has like $50 of parts in there!
(Out of frustration) I will make one myself!
Research research, design, think, plan
realize that it will be maybe cheaper yet still high cost, risky to design (cannot do a 2nd prototype due to expense if I made a mistake), and a huge amount of work

BUT, and here is the but, even if you decide to put the project on indefinite hold you still did get something in return.
This is what happened in my case for many many projects, but specifically for my 3d printer project. I saw these 3d printers online and look at what they can do, and for the past few years there are many new 3d printer models, but they are all still expensive. I am thinking of making one myself, so I look on amazon and ebay for steppers and metal rods. Found a good cheap source for each with good tolerances and exactly the dimensions I want. Now I get into designing the thing. Turns out to make this I just HAVE to get the parts made myself because such parts are not sold anywhere which I concluded after many days of googling. So I design a basic part in inventor with the physics simulations and everything, one of the linear bearings, and upload it to ponoko to see how much it costs in various materials. $42 dollars, no way, I need at least three of these guys! So I find a few other online 3d printing companies, and even a few cnc shops online, all of them give high prices, too high for my comfort.

Disgruntled I realize that maybe I just can't make it cheaper, BUT, if I were to make it my design would be much higher quality than many of those out there, exactly what I need, and in my opinion it would look a lot cooler. But, the end all thing for me is how much money it costs to make it vs buying it. Anyways, I realize I don't have enough money for such a thing, what do I do? I either put it on hold or keep working on it. What I did is keep working on it because I love the feeling of properly designing something, and when I am done I can print the design out and put it on my wall somewhere proudly. Then, when I have more money (hopefully, heh) I will be able to make this for myself, and even though it costs more it will still be MY design, with me knowing exactly why that bolt is there and not there, why that is blue and not white, and why it makes that funky sound.

Point being, often your design will not be cheaper. It is you and your quality/knowledge vs the economics of scale, and very rarely if ever will you win. But instead of having a working product you made, you will get a few things in return. You will learn a HUGE amount of new information. My 3d printer thing for example, I learned what ACME screws are, how steppers work, different designs for cnc's and pick n place machines, how important tolerances are, what tolerances today are possible from what companies, a good idea on how much certain parts cost, and most importantly, how to use autodesk inventor. After the smoke clears and dust settles, I am still satisfied with my efforts. I may not have an actual physical thing to show my for work, but in my head there are many many new ideas and information, which for me means it is worth it. So, just because you did not get to build the DMM, no loss, I am sure you gained new things in the process! Besides, you also brought more traffic to Dangerous Prototypes !

After all, in a quote from a member on this forum (I will find you and put you in my sig if you are fine with that), and it went something like this. "If you are having fun with the project, that is all that matters"

Well I have to mostly disagree
 - for e.g. look at my soldering iron driver, they cost 200-800EUR and parts are below 50EUR, what they charge is brand, quality testing, smarts put into the driver. I can't make it for 20-30$ in parts they can but I can make it in 60$ in parts but I can supply "smarts" and "testing" from my own time. So the end product takes some 200$ in parts (failed prototypes etc etc) and lot of time I "don't charge" as I'm actually having fun

 - reprap, I build bunch of those, designed 2 different version, am actually part of reprap core team ... I pushed tens of kilos of plastic trough my extruders ... If you look at commercial machines, it's the same thing as with soldering driver, they charge you for the name and patents they own, the parts for the machine used to be less then 1% of the cost... now we managed to get reprap to become competitive technology so now those companies dropped prices few times ... still reprap development is expensive, Arcol spent more then 15000EUR in development of hot end alone!! You can purchase his hot end for 100EUR (or some similar price), you can even copy it yourself if you have machine shop, but he invested 15kEUR into that development ... look at makerbot, they got 10M$ few months ago - they still havent came up with any serious improvement ... There's one thing to take assembly instructions from wiki and assemble reprap, or order a kit from DP and solder it yourself, and there's another story to design the thing yourself.... I personally invested over 10000eur in reprap project over the course of few years... but I did it for fun

This is all hot water and we all know that very good ..

The question is always what is the target. Mine target with reprap is to have fun, I don't care how much that fun will cost as long as it is fun. I have no other goal then to have fun!! So if I have money - I spend it, if I don't, reprap project waits .. and thats it. Now with DMM, I had a goal to have a good instrument on my desk. That was the primary goal. Secondary goal was to have fun doing that. There was no goal to "learn how to make DMM" as I already said, the principle of how DMM works are known for decades and there's really nothing one could really learn there .. I did research and found that it's cheaper to purchase DMM then to make one (and not "a bit cheaper" but we are talking many times cheaper) so I decided to do other stuff for fun (lately that's the soldering iron driver)....

If we talk about DP, I'd say the primary goal would be to "have fun but not lose money" with secondary goal to "make some money" ... with DMM project IMO that is impossible. You can either lose a lot of money and have fun, or you can start development of a serious DMM with customized chips (ASIC or something like one I shown), you order tens of thousands of pieces, and you start being DMM developer, you compete with PRC companies, with Fluke, HP ... It is a regular business, it could work, it could even bring a whole pile of money to DP, but it's going to be all but fun.

"have fun but not lose money" with secondary goal to "make some money"

That should be our corporate mission statement :) Currently: Make open hardware to help make more and better open hardware.

So I managed to properly understand the DP "mission" :)

FWIW: memorialized in the [s:]forum[/s:] wiki: http://dangerousprototypes.com/docs/Dan ... #Alternate (http://dangerousprototypes.com/docs/Dangerous_Prototypes_mission_statement#Alternate)

:D :D :D

Oh wow, haha :P

I want credit for sparking the conversation that resulted in this! :O

i watched few weeks ago a vidblog by Dave about building a multimeter ...  really interesting.
power measurement is surely fun in a multimeter  etc etc ...

http://http://www.youtube.com/watch?v=NA-MA3PJpKw&feature=plcp&context=C3d2e8dfUDOEgsToPDskICFktv5OjEJjzCGrHfF0ii