Recently I purchased the AD8232 evaluation board and it has been working perfectly.
After seeing that this worked, I purchased some AD8232 IC's from digikey and populated my own PCB that had the exact same schematic as the evaluation board.
I unfortunately see nothing coming from the output of the AD8232. Just a zero volt signal.
I've tried both 2 lead and 3 lead designs.
I've ordered more AD8232 components in case I received a bad batch of components.
I've tried both reflow and hand soldering.
Even with all these variations, I could not get anything to work. A couple of times a board would work for a few minutes and then return to outputting a zero volt signal.
At this point I assumed that maybe the component was just extremely sensitive to stray impedance, so I decided to copy the exact layout.
I ordered a different eval board from Spark Fun (SparkFun Single Lead Heart Rate Monitor - AD8232 - SEN-12650 - SparkFun Electronics), tested it to see that it worked, and then copied the layout and made my own PCB using Advanced Circuits (www.4pcb.com).
Unfortunately, this still didn't work.
After reading around the forum it doesn't seem like anyone else is having this problem...
Does anyone have an idea why I can't get this IC to work on my own boards?
Soldering lead-less packages can be tricky. Have you tried washing the board after soldering? There are several components that are high resistance value. Make sure there are no evident assembly…
To verify if you have damage the chip, you can try several things. For example, you could use a DMM to measure the current consumption; if you read 170uA or so, most likely you have a working…
Thanks a lot for the support Gustavo,
I was able to properly and successfully test the AD8232 on my board now! I'm assuming I must have failed at soldering it the first 3 times. It's just a bit harder…
Soldering lead-less packages can be tricky. Have you tried washing the board after soldering? There are several components that are high resistance value. Make sure there are no evident assembly defects and then start debugging with a voltmeter. But when you do, you're going to want to use a meter in which you can disable the 10M input impedance (this is very often fixed in handheld meters but not in benchtop boxes). The reason for this is that the input impedance of the meter can load down the nodes you want to check.
For example, have you checked the REFIN input? SDN? It is very important to make sure the part is not in shutdown. Then, take a look at the REFOUT terminal to see if it has the same voltage as REFIN. Also, you should check the output of the instrumentation amplifier at the IAOUT pin, where you should observe the input signal amplified by a factor of 100. If you do see a signal here, then it is possible that the problem is with the opamp circuit. I would remove components to make it a simple follower and add them back.
Try a few of these things and let me know if you get stuck.
Thanks for the advice Gustavo,
I was very careful with the soldering this time and washed the board and I am now getting an output signal.
Unfortunately, the output I'm getting is a square wave of 20Hz that goes from ground to 3.3v (this is vdd for me).
I put a probe on the IAOUT pin and I see heart rate there!
When I put a probe on the SW pin (this is after the passive hpf that is after IAOUT), I see this 20Hz signal being interjected on the heart rate signal.
I have a LPF with some gain after this, so my guess is that the 20Hz signal is causing saturation after the amplification so the heart rate is no longer noticeable.
any thoughts as to why 20Hz noise is coming from the SW pin? My AC/DC pin is grounded and my FR pin is pulled high. I've triple checked these connections.
I'm happy to hear you have a signal at the output of the InAmp. I don't think you have a problem with the leads off detection or FR, although it was a good idea to check that. I would say that you have some noise pickup, but if it does not show at the output of the InAmp, then the chances for pickup at that node are really small (unless there is a mistake such as the AC coupling capacitor not being well soldered, for example). One other thing you could check is that the values for all the passives are what they should be. Sallen-Key filters can become unstable, especially when you have gain, if for example, you have the wrong capacitors. If the amplifier is oscillating, then you would see a fraction of that signal in the SW node as you describe. An easy way to check would be to eliminate the 10nF capacitor in the filter and see if the 20Hz oscillation goes away.
I am currently doing something similar, I have made a few boards with this chip and for one reason or the other I kept thinking it was my design that was faulty. I too purchased the test board from SparkFun and was able to get a signal... I tried your previous suggestions (Washing the board, I checked RFIN, IAOUT etc) I cannot get any signal out... All I get is a 60 Hz noise signal coming out of IAOUT and obviously nothing readable at Signal Out...
I am wondering if you could answer a few concerns I have regarding the chip... The datasheet mentions its MAX temperature is 140 degrees Celsius, what are your recommendations to solder this chip then? since most solder pastes only start to melt at 250 or so Celsius. I am currently soldering it with a hot air rework station and using solder paste that melts at 250 Celsius.
I'm assuming this is what's causing the problems with the chip since I have tried everything else to debug the chip and I still get no proper signal from it...
A note to make, is that although I do not get an EKG signal from one of my boards, (while wearing 3 electrode configuration) I do get a signal response when I move one foot up and down as an output... more like an EMG signal response... however the placement of the electrodes is in both hands and the elbow (for reference).
Also the circuit configuration that I tested is exactly as the data sheet's configuration for 3-lead electrodes, same components and same everything.
I would greatly appreciate your help and comments regarding this issue, I believe I'm in the last stage of development or so I believe, I'm just being held back by not being able to receive a signal output.
Thank you kindly
To verify if you have damage the chip, you can try several things. For example, you could use a DMM to measure the current consumption; if you read 170uA or so, most likely you have a working part. Moreover, if you flip the shutdown pin, you should see this current go well below 1uA. You can also look at the REFOUT pin, which should have the same voltage as the REFIN pin (+/-1mV).
Assuming you haven't damaged the chip, then I would suggest to try to drive the inputs with a function generator instead of going directly into your hands. There are also ECG signal simulators in the market that you can use, but to see if the signal is there, a simple function generator would work. Using an isolated power supply (such as a battery or a lab power supply) connect one of the inputs to REFOUT, as well as the common of your generator; i.e. REFOUT will become the signal ground for the instruments. Then, connect the signal terminal of your generator to the other input and probe IAOUT with a scope. Apply a 10mVpp sine wave at the input (around 10Hz or depending where is the center of your band) and you should see 1Vpp at the output. If you have gain on the Op Amp, you may want to use less than 10mVpp, maybe 1mVpp and probe the output of the Op Amp too.
If you can get all this to work, you need to consider whether the contacts or electrodes are good enough. The fact that you can detect movement tells me you you won't have trouble with any of the tests above. So, what is most important is to twist the wires from the electrodes to the board, place the board inside a shielded box, and keep noise sources away from your setup. Another problem could be your power supply, so you could try using a battery (if you haven't already). The In-Amp has a gain of 100, which makes it very sensitive to external noise pickup, but this is what you need if you want to get a good signal from the hands. But in my opinion, testing on a live subject is the last thing you want to do during the development stages. You want to make sure the device is properly operating before you plug yourself (or someone else) to it.
By the way, the 140C limit is the max junction temperature (as generated by the chip) and is imposed by the package material. Exceeding this limit could induce package delamination. However, all of our packages conform to JEDECT J-STD-020. This means that you can apply up to 260C for 30 seconds, with a ramp-up time not to exceed 3C per second. So, for soldering you can exceed 140C for a short period of time (assuming moisture has been controlled) but during normal operation, you should not exceed a 140C junction temperature. I know, confusing, but that is how it works for most IC plastic packages.