0
\$\begingroup\$

I want to measure the voltage drop across a shunt resistor, amplify it to 50 times the original value, and feed that (0 - 5 V) to an ADC. I'd prefer to do the "high side" current sense method (rather than having the shunt between load and ground), but might have to change that to get it all to work.

This is driving me mildly insane. I've spent many hours on this by now, but I still can't find a single IC that meets all my requirements - just most of them.

Anyway, my requirements are:

  • Less than $7 or so (I only need 2, so no "per 1k" prices!)
  • Able to tolerate the common-mode voltage of ~15 V (either side of the resistor with respect to ground)
  • Able to measure up to ~1.5 A (always less than 2) with good precision, preferably with precision at the low end as well (5 mA = 0.25 mV, 1.5 A = 75 mV, 2 A = 100 mV)
  • Rail-to-Rail
  • 50 gain, or adjustable via resistors
  • DIP package/through-hole mounting. This one appears to remove about every single solution that otherwise works.

The requirements are absolute, perhaps except for the price... Still, I don't want to pay too much for this one part (current sensing) of the project; other ICs are involved in this as well, so the total for just the current sensing will turn out quite high.

All I can find based on this is the Maxim MAX472, which is obsolete and not available. Its replacements are surface-mount only, and I'd rather avoid making my first (after a small test) PCB double-sided, as long as routing will be theoretically possible with a single-sided board.

\$\endgroup\$
6
  • 1
    \$\begingroup\$ The last point on the list makes it a shopping question, too localised to be useful; you should change your constraints, we can't help you in this. \$\endgroup\$
    – clabacchio
    Feb 20, 2012 at 21:32
  • \$\begingroup\$ @clabacchio True enough, I removed that point. Still, that means that potential answers may be useless to me, though perhaps useful for others later on. \$\endgroup\$
    – exscape
    Feb 20, 2012 at 21:39
  • 1
    \$\begingroup\$ DIP package? Really!? That and insistance on a particular price when you are only getting 2 makes it hard to take you seriously. Think about the "many hours" already spent looking around and asking here. If the perfect chip costs $10, that means you're not going to get? That would make no sense to not spend the extra $6 after all this work and if it solves the problem. Get real. \$\endgroup\$ Feb 20, 2012 at 21:48
  • \$\begingroup\$ I've looked for AD620, it seems to be in DIP package at Farnell, but it cost $1 more than the SOIC8 version...why do you need DIP? \$\endgroup\$
    – clabacchio
    Feb 20, 2012 at 23:32
  • \$\begingroup\$ This site requires that questions should be useful for the global audience of the internet. Choosing current sensors is a valid question, and there are topics that we'd be happy to educate you about such as, as you noted, high- vs low-side monitoring, gain, common-mode voltage, rail-to-rail nature, etc. Deciding what you need is a perfectly valid question. However, once you arrive at that decision (which you have) and begin mandating specs like "In stock at Mouser", "<$7", and "DIP Packaging" remove that opportunity and make it applicable to only you. \$\endgroup\$ Feb 21, 2012 at 2:09

1 Answer 1

3
\$\begingroup\$

How about the AD626.

  • $9 in ones :(
  • Tolerates +-24v common mode range
  • Rail to rail
  • Gain of 10 or 100. (you could always use the x10 and add a x5 op amp afterwards if you really need x50 and not x100.
  • Available in DIP package from Digikey

Or the AD629 - $7.31 in ones :) - Tolerates +-270v common mode range ! - Gain of 1 so you'll have to spend a few cents adding extra amp stages. - Available in DIP package from Digikey

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.