There are a couple of designs out there that I find odd. This one, features no on-board caps and filters, a tiny land, etc. The adjustable Sparkfun board features a tiny load path and a giant nearby GND path nearby. I don't get it for safety reasons. The pololu board at least features some serious high-voltage lands though the distance to the low-voltage components seems questionable.

For all I know, this chip puts up a lot less resistance than the manufacturer warrants in their documentation. And 10A through the 30A version may be perfectly safe, generating no heat in your application, etc. I am not criticizing your use of this chip. I was merely constrained by a design that requires high accuracy, low insertion losses, and the ability to withstand long duration exposure to high currents.

This sensor design with its SO8 frame and a mere two leads each carrying that current gave me the willies. That does not mean it doesn't work. I simply wanted something with more meat and some other attributes.

Please also note how the Allegro eval board offers a wide land for the leads going to the sensor and very heavy-duty connecting posts. It also features double-sided high-voltage PCB traces and 'stitched' construction. In other words, Allegro is using the copper on both sides of the board for the power signal and the conductors are thoroughly bonded to each other using lots and lots of vias to reduce insertion losses and improve thermal heat dissipation from the chip.

I am not convinced that many of the commercially-available boards out there go to the same lengths to ensure a wide, thick path or use the same extra-heavy copper trace thickness (2oz copper is uncommonly used, costs extra) as the Allegro board. Just something to keep in mind as you source your gear. Relying on conductors attached to the terminals to dissipate heat (as some design apparently expect to) seems like bad engineering practice.

On the other hand, if the chip stays super cool no matter what currents are being passed, then so much the better for you.

All I will point out is the size of the eval pcb heat sink vs what the commercial resellers typically design and sell. The datasheet mentions a typical internal resistance of 1.2 mOhm for the acs712 and a 5x over current survival.

The datasheet also mentions a atypical 2oz copper thickness for the pcb used on the eval board. The FAQ also shows die temperatures assuming optimal conditions (ie eval board) reaching 165 C at 20amps in hot ambient conditions. I expected my device to potentially get hot inside as it did not feature active convection and the enclosure is small.

So, by all means go and use this sensor but between the relatively high error and the heat issues, if was not suitable for me. For those measuring short impulse loads in particular, this chip family features very attractive attributes like small size and relatively low cost. I went for the LTSR series from LEM instead. It's leads are significantly beefier, you get a reference voltage for differential measurements, etc.

30A continuous wouldn't be a sensible application for these sensors.

My interpretation of the datasheet is even more conservative than that. I am happy to be wrong, but my recollection of the datasheet and the engineering examples was that the sensor would require a substantial heat sink not to turn into blue smoke at sustained amperages well below the maximum. The breakout boards do a good job minimizing the PCB around the chip (good from a $$$ perspective) but at the very real risk of frying the chip if it encounters sustained loads.

Since my application has the possibility of such loads (think dryers, water heaters, and like products that can pull 5kW), I simply gave up on these chips. They may work well for some applications (like detecting abnormal current draws on motors necessitating the shutdown of an IGBT before it blows up) but for power measurement, they offer neither the accuracy nor the power capacity of other solutions.

I briefly considered using these chips, as others have. In the end, I dumped them on account of the inability on my part to see how these chips would not turn to crispy fritters during sustained high loads. The heat sinking requirements are 'interesting'.

I much prefer the LTSR series from LEM - excellent stable signal output, and a reference output for those of us who use differential ADCs. If you don't need the reference output, use the LTS series instead. Tamura also makes some nice ones.

Hall effect-compensated current sensors of this sort are more expensive, for sure but offer almost no insertion resistance, virtually no current limit, etc. The only 'downside' is the working voltage - i.e. they operate on 5V whereas I am making the transition to 3.3V systems. So I use a voltage divider to bring things down to a level that makes my Teensy 3.0 ADC happy.

Presumably the design of the lightning connector was driven by the above observations, ie as connectors get smaller it becomes more and more convenient for users to eliminate orientation issues. Obvious visual cues (such as the oblong shape of a IEC power connector) are not as obvious when things get tiny.

Just shipped a 2lb 11oz 4"x4"x5" package domestically - $11.49. Perhaps my FedEx shipping account (with employer discount) would have been cheaper but I doubt by much. My guess is that the big shippers (i.e. the Amazons of the world) are getting subsidized by us small fry.

I have left a customer service complaint with Paypal, let's see if that has the desired effect or not. However, I am not too hopeful that they'll actually do anything.

As best as I can tell, their bottom line is to get more business and someone like me that happens to only have a couple of transactions per month is not as valuable as a commercial customer that likely sells more merchandise in a day than I buy in a year.

I also find it ironic that the seller continues to misrepresent the parts he's selling, i.e. this appears to be a deliberate and ongoing attempt at defrauding customers, not a one-time mistake.  Maybe the BBB can nudge him into being more truthful about the parts he's selling?

[not]

So, I just finished my first dispute ever with Paypal. I suppose I have been lucky over the years and the plethora of transactions that this is the first one that I've had to dispute. I have to say, I am underwhelmed by the response, however.

As a quick history, our refrigerator condenser HX fan failed a while back. I rigged a temporary cooling fan, then looked up the replacement part number and found a web site claiming to have the part in stock. I ordered the part, opened the box and found something different from what was described.

The part was not genuine replacement part since the motor was not the same type as the one I pulled from the fridge. Instead of a PSC, it was a lower-efficiency and louder SPM motor. It also did not feature the molex plug necessary to hook up to the refrigerator wiring harness without modifications. It clearly was not a official replacement part as implied by the web site nor an equivalent substitute.

So, I contacted the seller who replied that the part was in fact a inferior substitute because the OEM part was no longer available. Thankfully, the part I pulled from the fridge had the OEM name and part number on it, so I looked up and contacted them and found out that they very much still make the part, 5 are in stock, VISA and MC are accepted, etc.

But even with all this information, the best that Paypal can do is refund me the original cost of the transaction while requiring me to return the part at my cost. Hence, PayPal is aiding fraudsters by giving victims of fraud a further disincentive to dispute dodgy transactions. Needless to say, I am not pleased and I will hence reduce my transactions with Paypal accordingly. If given a choice, the credit card looks like a much better alternative thanks to better dispute resolution systems being in place over there.

Thank you, your code confirmed my previous results but its a lot simpler!

...The nice thing about the ARM Cortex M4 used in the Teensy3 is that it has a barrel shifter...

Is that like having stainless braided brake lines on motorcycles?

I bet this is going to come up as more and more folk are transitioning to 32-microprocessors like the Due... libraries will need to be updated to reflect the different lengths of bytes, etc. since an int on a Due can hold a greater range of numbers than on a Uno. Anyhow, I wonder if someone could look over this code for a MCP3421 ADC, which offers multiple resolutions, ranging from 12-18 bits that I am trying to use on a 32-bit micro with the Arduino IDE (Teensy 3.0)

The ADC uses a I2C connection, if 18-bits are used, three data bytes are sent, otherwise, only two. The databytes then have to be re-assembled into the signed numbers they represent. My code is an attempt to be flexible, hence I declared a int32_t as the ADC value being computed. As best as I can tell, the code below is OK for the positive portion of the range. However, negative numbers appear to be coming out incorrectly. Could someone more knowledgable see if I got the code right?

Code:

int32_t Read_Thermistor_Value (byte Bits)
{
  uint8_t Hi=0, Med=0, Lo=0;
  uint8_t reads=2; // default number of bytes that have to be read for every conversion
  int32_t ADC_Value=0; // return value

  if (Bits==18) reads+=1; // add another round of reads if 18 bit mode is used.

  Wire.beginTransmission(MCP3421_address);
  Wire.requestFrom((int)MCP3421_address, (int) reads);
  if (reads ==3) Hi = Wire.read(); //only read top byte if 18 bit conversion is used
  Med = Wire.read(); //Otherwise, just read two bits
  Lo = Wire.read();
  Wire.endTransmission();

  //here use signed int32_t and right + left shifts to carry the sign and append the value as needed
  //for 18-bit operations, only the last bit of the Hi byte is relevant. The remaining bits are sign bits
  if (Bits == 18) ADC_Value = (((Hi << 30) | (Med << 22)) | (Lo << 14)) >> 14;
     
  //for all other resolutions, the returned value is left shifted based on the resolution to preserve the sign,
  //then right shifted back, i.e. if 16 bit resolution is needed, left shift Med 24 bits, while 12-byte values
  //would require a 28-bit left shift.
  else ADC_Value = ((Med << (40-Bits)) | (Lo << (32-Bits))) >> (32-Bits);

return ADC_Value;
}

Shield

Just remember all those graphics program's that were written on 8 bit micros running at a quarter the speed of an arduino that were produced in the 80s.

Brings back fond memories of the C64/C128 and the amazing work people did in assembly to make that processor sing, speak, and display some pretty amazing graphics.

As for whether 8-bit CPUs are obsolete, there is a place for them, so why not use them? As others have pointed out, the ability to drive strong signals without having to resort to external transistors is a good one. But is it life-changing? No... just that especially for beginners the ability to interface directly with an MCU is a big help because it potentially eliminates a link or two in a chain of things that can go wrong.

8-bit microprocessors will remain for as long as manufacturers are willing to make them. My guess is that the low end of the market is slowly eroding as manufacturers upgrade their manufacturing lines and obsolete old MCUs in the process. As an example, look at the DS2423, a chip with a dedicated following - discontinued with no substitute. By eliminating that chip Dallas/Maxim also guaranteed the death of the 1-Wire platform for many other applications - counters are very useful!

Margin pressure will also play into this, it's pretty funny to see 8-bit MCUs selling for more $$$ than computationally much more powerful 32-bit MCUs. At some point, the market will hit the tipping point and the bulk of development will go to 32-bits even if a 8-bit MCU was perfectly adequate.

Cyclegadet,

I suggest you take a look at the Analog Devices and TI web sites. Both offer a wealth of information re: op-amps and from the looks of it, your application is tailor-made for an op-amp. You could go for a G=1.5 circuit that simply shifts the signal from +/- 0.7Vrms to 1.5VDC +/- 1.5Vpp. That would result in a signal from 0-3VDC, well within the specs of the Due ADC. Plus, you gain the benefit of a low-impedance source, i.e. one that will make the ADC inside the Due very very happy.

Nice chips/breakout boards. I'm still looking for easy to use bipolar ADC to measure +-10V range. Too bad these ones are unipolar.

May I suggest the use of a custom-made  ADC driver chip from Analog Devices for this purpose? The 8275 will not only take that bipolar input and reduce the output to 0-4VDC, it'll also create a nice signal centered around 2VDC, perfect for differential ADCs with unipolar inputs.