--FREE-- 24 Bit ADC's(just make me a driver!)
Philldapill
Posts: 1,283
I have some 24-bit·ADC's that I need to make a driver for. Frankly, I don't have the time or knowhow to make an effective one. I do, however, have a couple of these chips to give away for free(I'll ship them free, too). I've looked on the object exchange for any similiar chips, but can't find one.
The IC is ADS1211P. It's a 24-pin DIP and has an internal microcontroller that can set various settings onboard.
If anyone is up for it, I'll send you two if you'll make a good driver. Any takers?
The IC is ADS1211P. It's a 24-pin DIP and has an internal microcontroller that can set various settings onboard.
If anyone is up for it, I'll send you two if you'll make a good driver. Any takers?
Comments
Have nice holidays.
Ale
Your requirement is vague. Why did you choose that particular ADC? High resolution certainly. There are so many options out there now in sigma delta at high resolutions, from Linear Tech for example, and Microchip and others too. Attaining the high resolution depends critically on circuit layout and noise sources. Imagine a stack of paper over 1 kilometer in height, three times the height of the empire state building, and take the relative thickness of one sheet of paper. That's one bit.
The first step in working with any ADC is to get it to do anything at all. Sometimes this is easy, but sometimes there are ambiguities and it takes quite a bit of reading between the lines of the data sheet and associated head scratching. This is a relatively complicated ADC with lots of possible modes. It is the sort of thing one would take on only if through a careful evaluation it has been found to be the best or only option for the project at hand. It is a lot of work. The inducement of free chips is not that strong, because, as you pointed out, TI is liberal with samples. But maybe someone will like to solve the problem for its own sake, better than crossword puzzles or Sudoku!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
50A/.01A = 5000 ~= 12 bits (4096)
Might be easier to just find a 12 bit ADC that already has a driver here obex.parallax.com/objects/search/?q=adc
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I'm new to the propeller!
My better half, who bestowed its name, reminds me to mention that L.saurus is the product of Famous Artists Schooling.
(b) I have the feeling that you will have to do more than just getting the voltage drop at a resistor.
This resistor is your main issue. How do you calibrate it? Temperature will influence its resistance in most terrible ways. You have to get its PRECISE temperature, this most likely will be the limiting factor of your accuracy. You will also have magnetic fields at 10A, with additional impact.
Post Edited (deSilva) : 12/25/2007 9:44:48 PM GMT
What you are doing would solve my problems with the memsic 2D. Would you like me to "sweeten the pot for you?"
Rich
deSilva, as for the shunt. It is still up in the air as to what I will be using. Any suggestions? Remember, this will measure anywhere between 0-50A.
I have in front of me 200amp 100, 50, shunts, these are rated 200, 100 and 50mv at max current, respectively.
They are others available with resistive manganin metal elements in the range from 50 to 500amps all at onlt 50mv at max’I
The accuracy of this type of shunt is well within 0.25% across the range. The Temp co_ef is very close to zero.
All these units are readily available and specifically designed and calibrated for instrumentation purposes.
Best to use a twisted pair from the shunt to a precision inline buffer amp with noise filter ahead of the ADC.
Ron
When you say they are calibrated there seems not much to be left to improve the 0.25% (= 9 bits) ?
As Phil needs (but will never get
To find a low noise Pre-Amplifier will now be the next challenge
They are manufactured to within 0.25% accuracy, the user would be left to handle the offset. The elements would be stable enough, temperature wise I would think, but one would have to do some homework on that. Personally I wouldn't bother.
There are many amplifiers out there specifically designed for shunts. Zetex, Maxim and others make them but I'm partial to Linear Tech, and they recently came out with a CAZ (zero drift) version with <10 microvolts offset and common mode up to 100 volts. With an amplifier of that sort, a 12 bit converter could be adequate. It is true you can do this with a high resolution converter directly, but not on the high side.
Whether on the high side or on the low side, it is important to add good filtering to protect the inputs and to keep out noise and especially ferrite beads for RFI.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
(a) V0 = R(T,B)*I How well do you now the resistance as a function of temperature and magnetic field?
(b) V1 = p(V0) As V0 the order of 5µV/mA, what noise and uncompensated drift is expected from pre-amplification
(c) V2 = w(V1) Assuming that the pramplifier is close to the shunt, what error is introduced by the wiring, PCBb layout?
(d) DIG = ADC(V2) What additional error is introduced by the ADC.
The ADC might not be the accuracy determining factor
Resolution to 12 bits can be achieved with some degree of difficulty.
If your aim is lump all the losses with the Solar Array on one side of the shunt (i.e net power source) And the all the losses on the other side of the with the battery bank (net load) that’s fine.
Set up the system, using a controlled light source……….controlled light source
Collect the data.
Now connect your back box between the Solar panel and the shunt and collect the data again, with the same light source
Forget calibrating the shunt, we have just done it. It’s the controlled light source you now have to find !
I remember walking down a country road and asking a Old Irish gentlemen the best way to get Dublin.
In a flash he said, “If I were gowin to Dublin, tis to be sure, I’d not start from ere”
In the real world.
Replacement of Storage Batteries in our case 2000ah every two years is expensive. A charging/management regime, which shortens the life of batteries, would not be a good thing.
I came into this, discussion because we derive 90% of our power requirements from Solar, and have done for 5 years. Phil is working on a Solar Power Project.
Ron, I'm not quite sure how your calibrated light source comes into play. Also I don't think you're saying that charging/management system would necessarily _shorten_ the life of the batteries--quite the opposite? My experience is only with panels meant to power instrumentation, not 2000 ah systems!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
Controlled light source. Yes it would have to be measured accurately across the array.
Simple test
Load = battery bank (Check Specific gravity of electrolyte in each cell, and measure temperature)
Power Sourse = Solar Panel excited by controlled artificial light
Shunt in series with load and power source.
Log power in over fixed time period.
We have a reference
Now discharge the Battery until we get back to the same Sg @ temp.
Add in series the black box under test and repeat the test.
We are not concerned (I don’t think) with how efficient the array is, of for that matter what the acceptance rate of the battery is, as long they are the same for both tests….
Re- battery charging regime.
The following parameters need be considered.
Bmax Max voltage in boost mode 13.5 –16.5 v
Emax Equalization voltage 14.0 –17.00 v
Etim Equalization time 0 – 2.0 hours
Efrq Number of days between equalization 20 – 150 days
Absv Absorption voltage 13.5 - 15.5 v
Atim Absorbtion time 0 – 4.0 hours
Fltv Float voltage 13.5 – 15.00 v
Hyst Hysteresis when using PWM mode 0.1 – 1.0 v
Brtn Voltage below which return to Boost 11.00 – 13.00v
Chrg Current limit 1- 20(40)A
Bfrq Maximum number of days between boost 1 - 20
Tcmp Selection of temperature profile. 0 – 8 ( I have 8 set up)
Some but not all of these parameters have a significant effect on battery life..They must (should) be available on “Intelligent Solar Controllers and set for the various battery types.. Lead Acid, Glass Mat, Gel, etc e…. Addition tweaking may be required depending upon the environment.
Bank switching is essential in order to maximize available power. Capture techniques are different story
<http://www.allegromicro.com/en/Products/Categories/Sensors/currentsensor.asp>
Nick
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Never use force, just go for a bigger hammer!
The DIY Digital-Readout for mills, lathes etc.:
YADRO
@phil, don't be put off by anyone go for it.........Online Solar Australia is worth a look.....Down load the PL40 manual from Plasmatronics, youl find them on OL Aust
There's lot do in Alternative energy, we need a new thread for it. Make your dream work !
I'm not sure if we're talking about the same thing... From what you say about the trace, it is on the pcb. I highly doubt that a PCB trace can handle 50 amps. Do they make sensors that measure this much?
>I highly doubt that a PCB trace can handle 50 amps. Do they make sensors that measure this much?
No, I don't sratch that!
My answer was more general than necessary.
> Now, the question is, how would I physically put that together and caliberate it?
With some known current?
An other simpler approach is to buy a amperage clamp meter. Some to have an analog output (0..5V or such). You'd have to pay attention to buy a DC-A-meter. You should get them for around 100$. They aren't very fast (some kHz) but it should be good enough for what you do. This also solves the calibration issue and isolation of your measurment system and the high-current stuff that might make problems.
HTH,
Nick
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Never use force, just go for a bigger hammer!
The DIY Digital-Readout for mills, lathes etc.:
YADRO