Author Topic: Open Source Multimeter  (Read 277128 times)

0 Members and 1 Guest are viewing this topic.

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8515
  • Country: us
    • SiliconValleyGarage
Re: Open Source Multimeter
« Reply #250 on: March 17, 2013, 09:49:22 pm »
Quote
...it just doesn't make sense if one can buy a TQFP-FPGA for 30 bucks...

I wonder why Agilent and Co still use their expensive hybrids in modern devices instead of a low cost fpga or cpld,

fluke uses cyclone fpga's. so do the tek rebranded flukes.

the reason agilent uses those asics is because they have their own waferfab. they can !
that asic was designed many years ago and also does other things besides the logic section of the multislope system. ( display controller , digital isolation link , memory decoder etc. )
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Online Marco

  • Super Contributor
  • ***
  • Posts: 6693
  • Country: nl
Re: Open Source Multimeter
« Reply #251 on: March 18, 2013, 12:01:56 am »
I don't think it's possible to get a comparator fast enough to make a FPGA worth it ... if you want decent offset you're looking at something chopper stabilized and those aren't exactly fast.
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #252 on: March 18, 2013, 12:32:16 am »
Regarding comparators, I think Stoney currently has an LT1016 on the list. That's going by the fact that the datasheet is in the git repo. ;)

I also had a look at that one and compared it to the LT1011 (from the TU Berlin design). Personally I am not yet convinced LT1016 would be better than LT1011 in this application. It does have 1 big thing speaking for it, and that is that the propagation delay has been characterized over temperature. A nice flat 10 ns over temp for the falling output. No such spec to be found for the LT1011. On the other hand the LT1016 has a worse max offset (3.5 mV) than the LT1011 (0.5 mV), and the max offset drift for the LT1016 is not even mentioned. Or at least I couldn't find it. The typical offset drift is the same (4 uV/C) for LT1011 and LT1016.

And as bonus the LT1016 has a differential output, but that's more of a nice to have, not a must IMO.

As for comparator speed, if I understand the dual slope/multislope thingy correctly, you don't care overly much about the speed. As long as it's fast enough. The main characteristic with this sort of low repetition rate zero crossing detector is not so much speed, as it is variations in the propagation delay. For all I care it has a whopping 100 ns prop delay, as long as it is bang on 100 ns over all conditions. It is uncertainty in prop delay that is annoying because that translates into uncertainty of amount of accumulated charge. It's a similar problem to offset voltage, just a different axis. And the slope at which the zero crossing occurs dictates which of the two is the more annoying problem. ;D

This all of course is still assuming I understand the design properly (big if XD).
 

Offline Stoney

  • Contributor
  • Posts: 19
  • Country: de
Re: Open Source Multimeter
« Reply #253 on: March 18, 2013, 07:21:48 pm »
Uhm, well, I just had a look at the datasheet and thought it might be a good point to start from, no special intentions on using THIS specific comparator....we'll see how it turns out once I can make some calculations on the delay/switching delay/speed/clock/timing of the integrator.
The opamp in the integrator is also very critical...it has to be quiet fast, but low drift and low input current/offset current as well! Input "leakage" current into the opamp will introduce errors directly in the runup so should be avoided, and the slew rate comes into play when the voltages are switched for the runup/rundown phases. This is also a limiting factor for the slopes of the rundown, together with the timing resolution and the resulting under/overshoot for the steepest rundown-slope. Low drift is obvious, thou not ultimately critical, since I will null and offset compensate the integrator after each measurement or at a specific rate if the drift allows for it. But still, drift within one measurement cycle should be reasonably low. This will of course depend on the temp. rise, but with a standard, fast opamp, i think the drift is going to kill us. I remember some fast "precision" AD opamp with 50V/µs (the 8-1/2 digit wizards built their own with 100V/µs) wich might be a good candidate...

Anyone had a look at my (very crappy) drawing of the input overrange / VDC buffer stuff yet? I expect plenty of stupid mistakes so please correct me!
« Last Edit: March 18, 2013, 07:24:26 pm by Stoney »
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #254 on: March 18, 2013, 08:42:52 pm »
Anyone had a look at my (very crappy) drawing of the input overrange / VDC buffer stuff yet? I expect plenty of stupid mistakes so please correct me!

I didn't quite follow R5, R6 (near note [9]). This is stil purely input signal conditioning + protection, yes?

If so, why the R5/R6 combo in T912 package there? As in what's the intended function of that LTC1052. I follow things right up to the + input of the LTC1052, and then I don't quite follow your intention.

Don't you just want a boring unity gain buffer amplifier at that point?

Also, don't you want the "To VAC" after you do some buffering with said boring unity gain buffer?

you range switch it, you low pass it, clamp it, and then you x1 buffer it. No x1/x10 precision resistors required so far, unless I am missing a clever trick you have up your sleeve.

 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: Open Source Multimeter
« Reply #255 on: March 18, 2013, 08:53:33 pm »
Quote
The opamp in the integrator is also very critical...

Seems like you haven't watched the posted link (AN-260 A 20-Bit (1 ppm) Linear Slope-Integrating A/D
Converter), as this critical part was already faced there and how it has been solved (page 5-6, A2-A3):

http://www.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=snoa597a&fileType=pdf

« Last Edit: March 19, 2013, 07:31:22 am by branadic »
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #256 on: March 19, 2013, 02:04:28 am »
While we're at it ... possibly something like this for the comparator stage:

edit: reduced size of pic...
« Last Edit: March 19, 2013, 02:10:14 am by mrflibble »
 

Offline Stoney

  • Contributor
  • Posts: 19
  • Country: de
Re: Open Source Multimeter
« Reply #257 on: March 19, 2013, 09:24:44 am »
Hm, the x10 gain would be useful in the lowest range to keep the integration time and get the signal over the leakage introduced errors within the integrator. Otherwise, you would have to integrate ten times longer in the lowest range (200mV FS) to keep the resolution.

I wanted the Vac to be protected by the same overrange protection as the Vdc, otherwise another pair of diodes, more diode leakage, more resistors, etc. I thought about the aliasing error stuff and decided to use a simple first order filter after the initial, "high frequency" filter :) . Maybe, a active second order one with the 1052 could be better, but since the sampling rate will be reasonably low, I don't think we run into problems here. Of course, you don't want a 200Hz -3db bandwidth for the Vac stuff, so it will have it's own buffer, presumably switchable with a small smd reed switch (the meder ones are quiet cool, low EMF, and reasonably priced (4€, don't know the dollar price)).
« Last Edit: March 19, 2013, 10:19:38 am by Stoney »
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #258 on: March 19, 2013, 05:05:29 pm »
Hm, the x10 gain would be useful in the lowest range to keep the integration time and get the signal over the leakage introduced errors within the integrator. Otherwise, you would have to integrate ten times longer in the lowest range (200mV FS) to keep the resolution.

Ah okay. In which case I'll keep it simple and just use a boring unity gain stage there. Especially since for iteration 1 I can live with the longer integration time. And for iteration 2 I am secretly eyeing a multi-slope approach. And in both cases at this point it's a simple buffer without any variable gain.

 

Offline Stoney

  • Contributor
  • Posts: 19
  • Country: de
Re: Open Source Multimeter
« Reply #259 on: March 20, 2013, 09:53:33 pm »
Yep, and after I started thinking about the actual integrator configuration, I realized that one can perfectly adjust this with the "integration resistor"  :palm: So no stupid switchable gain, just a simple unity gain buffer...
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #260 on: March 20, 2013, 10:42:29 pm »
Heheh, that was my reasoning as well. ;) With the change of making it just a unity gain buffer I suddenly agree with the schematic for the most part. :) I'd do the RC filters a bit different, but that is a minor detail that can be worked out later. As long as we can agree on the fact that the "To ADC" has been low pass filtered, clamped and buffered.

Oh yeah, and I would just kick out the auto zero switch near note [10]. Personally I plan to do auto zeroing around the integrator. Definitely not in this input stage, unless I am missing something...
 

Offline Stoney

  • Contributor
  • Posts: 19
  • Country: de
Re: Open Source Multimeter
« Reply #261 on: March 20, 2013, 11:32:19 pm »
Hm, how would you do the filter? Maybe your solution is better ;) (active?)

The autozero should, in my opinion, include the input buffer. Maybe it's placed somewhat wrong...if I would place it just behind the diodes ( between clamps and R4), one could use a low-voltage MEDER one (much, much cheaper than those bloody expensive coto ones and equally low emf). The auto-zero in front of the buffer is certainly not equal to the "each cycle integrator zero" but can be used if requested by the user to really zero the whole unit including the (temperature drifted, its certainly useless if used 2s after power-on) input buffer. The integrator is of course part of the procedure. I'm not really sure if a real auto-zero can be achieved (or makes sense) without grounding the input buffer input...
« Last Edit: March 20, 2013, 11:59:53 pm by Stoney »
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #262 on: March 21, 2013, 12:35:09 am »
Hm, how would you do the filter? Maybe your solution is better ;) (active?)

Well, I don't know about better, but it is even simpler. :) It's passive just like yours. Only difference is that I would only use 1 RC stage, and I would use a slightly different f-3db. Remember that the "we shall use low loooow bandwidth" was from the part of the discussion where delta-sigma ADC was also on the table.

In the meantime I've read a bit more about dual-slope, and I have seen the error of my delta-sigma ways. ;) Dual slope is damn nice in many ways. So I want to try and make a 5-1/2 digit dmm based on dual slope. Not because it makes economic sense, but as a fun learning exercise. If it was to make economic sense I'd better work a bit and buy something on ebay. Definitely more time-efficient than building one. But building one is more fun efficient!  ;D

I would use a single RC filter and limit it to somewhere in the 1 kHz - 10 kHz range. And I'd do that at the point where you have your 100 kHz RC filter. And then I would leave out the 200 Hz Rc filter. Why? Well, lets not forget that our trusty integrator is an integrator. You know, one of them low pass filtery thingies.

In any event, it's not really worth too much discussion. You can put them both on the PCB layout and then make it an optional component fit. You would for example keep your 100 kHz + 200 Hz arrangement, and I would do 5 kHz + passthru. And depending on results we can simply change component values a bit after the initial measurements.

Quote
The autozero should, in my opinion, include the input buffer. Maybe it's placed somewhat wrong...if I would place it just behind the diodes ( between clamps and R4), one could use a low-voltage MEDER one (much, much cheaper than those bloody expensive coto ones and equally low emf). The auto-zero in front of the buffer is certainly not equal to the "each cycle integrator zero" but can be used if requested by the user to really zero the whole unit including the (temperature drifted, its certainly useless if used 2s after power-on) input buffer. The integrator is of course part of the procedure. I'm not really sure if a real auto-zero can be achieved (or makes sense) without grounding the input buffer input...

I'm not entirely convinced I'd want an autozero there. My focus is to keep everything boringly simple for a first go. So if it is a Nice To Have, and not a crucial bit of circuitry ... out it goes! Maybe I will see how it is an advantage when I see it in context with the rest of the circuit.

 

Offline Stoney

  • Contributor
  • Posts: 19
  • Country: de
Re: Open Source Multimeter
« Reply #263 on: March 21, 2013, 10:12:23 pm »
K, I'll leave it in there for now, with the two stages, one can basically run a number of combinations or just put a simple solder bridge in there. Same for the auto-zero, I'll leave it in there as an option  :) Since no other complaints came up, I would call the input conditioning done for now ;) except for:

Let's talk CAT-Protection and over-voltage protection (not overrange, thats already done :P) to finish the voltage input section 8)

I'd love to have CAT-II (for measuring mains stuff), which would be rated 1kV (withstanding 6kV input impulses if I remember it correctly ^^ ).  I don't have access to a 61010-1...but I could certainly use the DO-160 Class 5 as an equivalent :D (just kidding). Maybe someone could post a quick wrap-up what the unit needs to survive in terms of energy and pulse duration....
Littlefuse offers some nice gas-discarge tubes wich might do the job, with low capacitance and certainly no other leakages during normal operation: http://www.littelfuse.com/products/gas-discharge-tubes/high-voltage-gdt/cg3.aspx.
My fist try would be to put a high wattage resistor after the GDT to absorb the energy once it arcs over, together with a fuse in series with the resistor after the tube, to stop the "magic-smoke-and-fire generator" if it fails (obviously sacrificing the unit, but hey, you just managed to blow the protection :) ). I really wouldn't like to see a PTC in a "precision" instrument (all that temperature compensation stuff, and a PTC in the input path?!). But maybe one could use one in series with the high-wattage R instead of the fuse to cool down the current in a constant overvoltage condition, which might be a better idea after all instead of the fuse. I've actually never used a GDT before...so, what do you think? Any chance this arrangement could do the trick? Do we need some additional "constant high energy overvoltage, eg. plug the damn thing in your 5kV generator" protection?

So long,
Lukas

Edit:  :palm: I guess I wrote some BS up there regarding the idea for the protection....basically, It wouldn't do anything to protect the unit...I'll leave it in the thread as a bad example:D
« Last Edit: March 21, 2013, 10:25:47 pm by Stoney »
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: Open Source Multimeter
« Reply #264 on: March 21, 2013, 11:04:19 pm »
What about something like the circuit shown in AN-D16 "High Voltage Protection" by Supertex?

www.supertex.com/pdf/app_notes/AN-D16.pdf

BTW, updated my post concerning ADCs:
https://www.eevblog.com/forum/oshw/open-source-multimeter/msg203717/#msg203717
« Last Edit: March 23, 2013, 02:36:47 pm by branadic »
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #265 on: March 21, 2013, 11:56:24 pm »
I don't know enough about doing a "proper" CAT-II protection, while still maintaining high precision. So if you guys come up with something reasonable I'll be happy to copy that and learn from it. XD In the absence of that my plan is to not measure mains with this thing because I don't care about measuring mains in any high precision. I have my cheapo handheld for that. ;)
 

Offline Stoney

  • Contributor
  • Posts: 19
  • Country: de
Re: Open Source Multimeter
« Reply #266 on: March 25, 2013, 08:10:38 pm »
@banadic: Hm, interesting... the http://www.supertex.com/pdf/app_notes/AN-D11.pdf has some more details on that... and btw, nice ADCs :) Basically a off the shelf multislope thing...might be a good fallback, but I suspect them to be insanely overpriced  :P

@flibble: I certainly don't want my unit with these nice and expensive components and a costly PCB blow up into magic smoke, just because I was too lazy or too stupid to design a proper protection ;D And hey, a protection is simply a must have...one can certainly argue about some, more or less optional, features but not about the protection...

The standard fluke input protection that dave explained in one of his videos (If I remember it correctly) would be a good point to start from...some (most?) of the fluke bench meters are protected CAT I 1kV, CAT II 600V (which means a 4kV impulse protection http://www.ni.com/white-paper/5019/en). And I guess they also use the R-Thermistor-MOV combination on their higher precision bench type multimeters (posted in another thread here in the forum: http://headsplosive.com/2011/11/fluke-8846a-6-5-digit-multimeter-teardown/). Any idea which type/brand they use for the PTC? The MOVs look like standard ones, with the high-wattage R in series with the PTC...Seems like they also used some R's to distribute the voltage equally between the series MOVs there...

Anyone here who did a HP/Agilent/BK/NI/Keithley...etc. teardown and has some insight into their input protection?? The old HP ones seem to use the GDT/Fuse/R's combination in most models...

Tektronics/Fluke 8808: http://s222.photobucket.com/user/Zadpics/media/Tektronix%20DMM4020%20Bench%20Multimeter/Fluke-Input-Protection.jpg.html?sort=4&o=5 Seems like some MOVs + maybe GDTs (??)

Keithley 2000: http://bardagjy.com/?p=1167 Also a MOV/GDT combo with presumeably a PTC (the orange thing I guess) and a series R...
« Last Edit: March 25, 2013, 09:32:46 pm by Stoney »
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #267 on: March 27, 2013, 12:06:17 pm »
Besides input protection, a couple of other things...

- Voltage reference. I am thinking LM399 based, because 1) it is affordable and 2) it has pretty good drift figures so we can get good relative precision.
- 10 Volt calibration source. The suggestion to use Geller Labs 10 Volt standard is actually a pretty good one. It is affordable, has decent performance, and allows us to consider that part of the design as "problem solved". That and it's a nice standalone module that you could use in other things as well.
- zero crossing detector. See the schematic I posted some time ago. Acceptable? In any event, I think I'll make a little breadboard compatible pcb of this ZCD. That way I can easily test it's performance both for this DMM project as well as for a few other things.
- integrator. That's the main (as in highest priority IMO) bit of undefined circuit on the list. I was thinking of roughly the circuit in the TU Berlin schematic. LTC1150 unity gain buffer + integrator around an LT1056 is a decent starting point. As for analog switches the ADG202 is nice. The DG444 will probably do the trick just as well and is a little cheaper.

Does anyone have a favorite good quality yet affordable cap we can use for the integrator?
 

Offline poorchava

  • Super Contributor
  • ***
  • Posts: 1672
  • Country: pl
  • Troll Cave Electronics!
Re: Open Source Multimeter
« Reply #268 on: March 27, 2013, 01:23:09 pm »
Panasonic ECHU series of metalized PPS film smd capacitors seems nice. They are down to 2% tolerance, have a small capacitance variation over temperature range and low leakage. Cost is rather modest.

There are also Cornell Dubilier's silver mica capacitors, available with tolerances down to 1/2 pF. They are THT-only and are a lot more expensive than Panaconics (~2-5€ a piece from Farnell)
I love the smell of FR4 in the morning!
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #269 on: March 27, 2013, 05:44:24 pm »
Thanks! I did a bit more reading, and looks like metallized polypropylene film is the ticket. So for example Vishay MKP1837 or MKP1840 series.
The reason for MKP being that it has pretty low dielectric absorption (0.05% typical for the two series mentioned). And that's the one you want to keep low for the dual slow (or multi-slope for that matter) integrator.

So right now I've got those two MKP series on the list, cap value in the range of 0.01 uF to 1 uF, depending on choice of integration times and whatnot.

Probably for the dual slope approach you need something closer to 1 uF, while for a multislope approach you can get a way with a smaller 0.01 uF cap.
« Last Edit: March 27, 2013, 05:49:28 pm by mrflibble »
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #270 on: March 27, 2013, 06:07:45 pm »
Also, a bit of Bob Pease fun on the matter: http://portal.national.com/rap/Application/0,1570,28,00.html
 

Offline Stoney

  • Contributor
  • Posts: 19
  • Country: de
Re: Open Source Multimeter
« Reply #271 on: March 27, 2013, 08:39:15 pm »
Thanks for all the input on the integrator so far...hope I can start with some detail drawings soon...I've actually started with the concept for the multi-slope, but didn't get far because of some looong days at work -.-

But I managed to think about the protection stuff. I've updated the repo, so please go for it and have a look at the schematic (pretty basic) and the thoughts gone into it. After that, I think it's finally time for some integrator schematics :)

Btw: Couldn't get the PTC out of the signal path  ::)
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: Open Source Multimeter
« Reply #272 on: March 27, 2013, 11:58:18 pm »
Quote
Thanks! I did a bit more reading, and looks like metallized polypropylene film is the ticket. So for example Vishay MKP1837 or MKP1840 series.

As far as I now the better choice is polystyrene and even better is a teflon type capacitor. Last one is for sure expensive:

http://www.soundlabsgroup.com.au/p/AU-0U68-600V-Aura-Teflon/0.68uF+600V+Aura+Teflon

What the hell...? And if you were J.W. you wouldn't care and select from a hundred the one with the best characteristics, right? The cap itself for sure exceeds the price for a 31bit ADC ADS1281 or ADS1282, so again I wonder if the slope adc principle with all the cirtical parts is reliable enough and if it wouldn't be worth to combine something like STM32 with such a high resolution ADC?

By the way, I was playing around with a TDC eval kit (time to digital converter) a few weeks ago and as long as you are able to give a start and at least on (or two) stop signal this guy is a better choice for whatever into time into digital converter because of its ps-resolution ring oscillator and this for a low cost, space saving one chip solution with the ability for temperature compensation. The maximum samplerate is somewhat like 200kHz under best conditions:

http://www.acam.de/tdc-gp22/

Another idea could be to use their CDC and measure the "capacitance" of a varicap in series with another low temperature drift cap. This guy is at least nothing but a TDC with some integrated switches and resistors and discharge time is measured.
The varicap is pre-loaded by the input voltage I want to measure, so the discharge time is somewhat proportional to the input voltage. If the whole input is designed to measure a differential capacitance instead of a single ended (how ever this could look like e.g. a double varicap) I don't need a high stability voltage reference at all and most of common-mode effects are already compensated.
Max. sample rate is 500kHz, so enough values to average and this guy was spend a DSP which means full polynomal compensation for temperature and ... and ...

This are just my thoughts while brainstorming the complete problem of a high resolution DVM for a good buck. And we have learned that it is easier to make (portable) high precision time measurement (thanks to DCF77, GPSDOs or cheap second source RB-references) instead of voltage measurement "without" the reference long-term stability task. Contrary to the mentioned the slope adc still needs the high precision voltage reference.
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #273 on: March 28, 2013, 04:54:29 am »
As far as I now the better choice is polystyrene and even better is a teflon type capacitor. Last one is for sure expensive:

http://www.soundlabsgroup.com.au/p/AU-0U68-600V-Aura-Teflon/0.68uF+600V+Aura+Teflon

What the hell...? And if you were J.W. you wouldn't care and select from a hundred the one with the best characteristics, right? The cap itself for sure exceeds the price for a 31bit ADC ADS1281 or ADS1282, so again I wonder if the slope adc principle with all the cirtical parts is reliable enough and if it wouldn't be worth to combine something like STM32 with such a high resolution ADC?
What expensive cap? All I see is an audioph00l website with an overpriced cap. Doesn't mean that the cap we will actually use is going to be super expensive.

And AFAIK for this application teflon is indeed best and most expensive. Then 2nd best is polypropylene and after that polystyrene. With those last two being not too expensive, which is why I chose from that set and not teflon. And having read that Bob Pease article I thought it might be fun to grab a bunch of different types and just measure what kind of soakage you get for various integration speeds. Buuuut, since I also know how easy it is to accumulate side-track projects I decided to write that idea down in the column of "Fun but not required" things to do. It definitely is fun, and I might just do that in the future (and use the information to improve cap choice) but not now...

Quote
By the way, I was playing around with a TDC eval kit (time to digital converter) a few weeks ago and as long as you are able to give a start and at least on (or two) stop signal this guy is a better choice for whatever into time into digital converter because of its ps-resolution ring oscillator and this for a low cost, space saving one chip solution with the ability for temperature compensation. The maximum samplerate is somewhat like 200kHz under best conditions:
*edit: timing related thread derail removed*

At any rate, I had promised myself I wasn't going to put any more time into the timing aspect until we get some hard numbers on timing resolution requirements. :P So chop chop, hard numbers!

Quote
And we have learned that it is easier to make (portable) high precision time measurement (thanks to DCF77, GPSDOs or cheap second source RB-references) instead of voltage measurement "without" the reference long-term stability task. Contrary to the mentioned the slope adc still needs the high precision voltage reference.

Yeah, and if those bleeping Chinese guys would just lower the Rb price a bit I'd have bought 2 already. :P

And unfortunately no-one has developed an affordable virtual photon counter yet with single photon resolution, so I think we're stuck with the voltage reference a while longer.
« Last Edit: March 28, 2013, 05:46:18 am by mrflibble »
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: Open Source Multimeter
« Reply #274 on: March 28, 2013, 05:00:46 am »
But I managed to think about the protection stuff. I've updated the repo, so please go for it and have a look at the schematic (pretty basic) and the thoughts gone into it.
Cool, I'll have a look at it. Incidentally, you've got a PM.

Quote
After that, I think it's finally time for some integrator schematics :)
Couldn't agree more.  ;D
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf