Electronics > Projects, Designs, and Technical Stuff
ACS711 not outputing full voltage range.
djacobow:
What is the fault/ output doing? A short current spike from the drill could be sending the unit into fault mode, which it recovers after current goes down and stays down for a bit.
Would be interesting to test with a less noisy load, perhaps a resistive heater driven through a variac.
guitchess:
Well I started the morning changing loads. Didn't try a resistive load, just different. a single speed blower motor. Probably not much less noisy. It did reveal a few things.
Thank you and everyone for your input. I think I may have it sorted out. Unfortunately, it shows my lack of experience/technical knowledge.
I think my issue is that I had DC on the brain. When I measure the VIOUT on AC mode, it appears to work properly. However, it measures 0v with no passing current. When current is applied, the voltage goes up. Measurements show that it is an accurate 55mv/amp as well.
To my way of thinking this is right because VIOUT goes the same direction as the measured current. I think the only reason any VIOUT was showing with the drill is because of noise or possibly the variable speed circuit. With the blower motor that I tried this morning, there was no reaction from VIOUT when measuring DC but when measuring AC, the voltage was right on.
How does this theory sound to more experienced designers?
Unfortunately, this thesis requires a version 3. Which, will have to include a rectifier and an op amp.
Again, thanks for all the help. Any tips, input, or advice is always appreciated.
guitchess:
--- Quote from: djacobow on October 24, 2019, 03:01:29 pm ---What is the fault/ output doing? A short current spike from the drill could be sending the unit into fault mode, which it recovers after current goes down and stays down for a bit.
Would be interesting to test with a less noisy load, perhaps a resistive heater driven through a variac.
--- End quote ---
I haven't measured the fault output yet. I really had ruled that out because the datasheet said that the IC had to power off to reset it. I will try that direction though. Thanks.
guitchess:
Well here's V.3. It feels a bit overkill to use an op amp for a precision diode, but unless I'm calculating wrong I would have to be pulling 3 amps just to get over the voltage drop of a simple peak detector.
I'm going to try to breadboard this up and find out if I have my head screwed on straight. Feel free to let me know if you think is not. :D
IDEngineer:
--- Quote from: wraper on October 24, 2019, 02:14:52 pm ---As you are connecting drill, it likely produces wide band noise. And coupled through internal capacitance of ACS711, quite likely it can make IC go nuts. Hence I suggested placing ferrite beads. Using noise suppression capacitors might help as well.
--- End quote ---
On this particular topic: I used several models in the ACS series a few years ago during a study of brushed DC motors. Peak startup currents were on the order of 70A with sustained currents in the low 10's of amps - in other words, far more than the ~8 amps mentioned by the OP. If you want to talk about noise, I can post some amazing scope photos. There aren't many noisier "signals"!
The ACS devices worked flawlessly. They revealed an amazing amount of detail on the commutation as the brushes made and broke contact. Absolutely no "resetting", etc. If the current exceeded the range of the particular device its output would saturate, but when the current came back down the output signal would follow faithfully. The one caveat here is that the ACS's we were using did not have the "fault" output. They simply saturated as described. (The internal current path will sustain ~500% of the device's rated current so you can saturate them to some degree without damage.)
Based on my experience, I'd trust the ACS is working properly - unless the "fault" thing is the cause. There are standard versions without the fault output, and since it seems you don't care about the fault output anyway, why not use a similarly rated version without the fault feature? That would eliminate one possible source of the problem.
EDIT: Studying the functional diagram in the posted spec sheet, it's hard to see how latching the fault output would affect the analog signal path.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version