ADS1248: device not entering "sleep" low power state with START pin low after conversion completes

Richard,

When the START pin goes low, the ADS1248 should complete the current conversion and then go to the low power state.

My guess is that in your system, the internal reference is still on. In your system, I would check the VREFCON[1:0] bits in multiplexer control register 1. You can find the writeup on page 59 of the datasheet. The internal reference control for these bits may be set to 01, for which the internal reference is always on. If you set these bits to 10 or 11, the internal reference is on when there is a conversion in progress and turned off when the SLEEP command is issued or after the START pin is taken low.

Just as a check, I measured the current of the device while converting at 5SPS, with a PGA=1. The AVDD current is 222uA (compared to the 225uA listed in the datasheet). When the internal reference is turned on, this current jumps to about 389uA.

If I set the internal reference to be always on, and take the START pin low, the current drops to 168uA. The datasheet lists 180uA, while your measurement is 174uA. This is close to the current required for the internal reference, so my guess is that the internal reference is always on. If I set the reference to shutdown with the SLEEP command or the START pin transitioning low, then I see the current drop to 0uA.

Again, check your reference settings. If this doesn't solve your problem. Let me know, and we'll think about other possibilities.

Joseph Wu

A2D_no_sleep_5_24_17B.7z

Richard,


I don't see anything that could be wrong at this point. You're right that it looks like the VREFOUT droops and shuts off after START goes low. Are you monitoring the current with a multimeter? Are you able to send any command to the device that shuts everything down? How much is your current when everything is on?

I was able to open the Saleae file to look through the command set to see what you're doing. However I don't see anything that would cause an extra draw of current. I did have a few questions about the sequence.

Looking at the logic analyzer, it looks like you set up the devices, issue a sync command and then wait for the data ready to read the data. It looks like you repeat the sequence 8 seconds later. Here's a close up of the WREG sequence:



The sequence is 40 06 13 00 60 02 CC FF, then 4A 01 07 8C. In real terms, this is a WREG starting at 00, of 7 bytes, setting inputs to AIN2/AIN3, VBIAS off, reference on, but shutdown with SLEEP. CC FF FF written to the offset calibration register. Then there another WREG to the IDAC register for two bytes. This sets the IDAC magnitude to 1.5mA, and the outputs are IEXC1 and the other is disconnected.

After that, you send 04 04 to SYNC the device, and then there is a read in about 50ms after the conversion completes. This repeats itself after 50ms.

First, why do you have an offset calibration setting for CC FF FF? That seems like a very large negative offset. It won't change the current, but it is a bit unusualy large for the offset calibration.

After the WREG commands, can you put in a RREG of the entire register map? I'd like to make sure that the device understood the WREG. There might be something else to see in the register map.

Once you get the RREG, post another Saleae file, and I'll look it over.


Joseph Wu

Hi Joseph,

I'm not sure if I can share the schematics, but I'll check on it.  I have the ability to externally switch off the power to the A/D via S/W.  I measure the current with a "uCurrent Gold" https://www.eevblog.com/product/ucurrentgold/ with its output going into the Saleae A/D.  This is the bottom trace on the screenshot png.  The scale is 1mV/1uA.  I see a baseline additional current of ~ 147uA when the A/D is powered and has taken a reading but should be sleeping versus ~0uA when I switch off the power to the A/D via S/W after taking a reading.

Hi Joseph,

I believe the offset calibration is not too big, -52 decimal?  It is the value the was obtained by doing a SELFOCAL.  I will attach a Saleae file that includes a register dump in addition to the standard 8s A/D sample.

Thanks,

Richard

A2D_no_sleep_5_25_17A.7z

Richard,

I was able to duplicate the setup, but didn't see any problems with the current. I used the ADS124S08EVM with ADCPro as the controlling software. It's not a perfect duplication of your setup, but it is very close. Here is the setup:

You can see the ADS1248EVM on a breakout board in front of the MMB3 controller board. I use an Agilent 34401A for the current meter.

With the software, I set the device to AIN2/AIN3 inputs, PGA1, 20SPS. The reference is on, but is shutdown when the device is shutdown. IDAC current is 1.5mA, where IDAC1 is set to IEXC1, and IDAC2 is disconnected.

At startup, the current is about 198uA. When I turn the reference voltage on, current goes to 389uA. After I turn the IDACs on, the current goes to 517uA (note that the IDACs are not connected to anything at this point). Then I connect IEXT1 to AVSS to shunt the current and the current goes to 2mA.

Using the START pin, I can pull it down to ground and the current returns to 0.1uA. When I pull up the START pin, it begins converting again and the current returns to original 2mA. I still don't see any extra current.

I don't have total control over what I write to the ADC through the software. I can only do so much with ADCPro. However, I do grab some snapshots of the Saleae file. To show what is written to the registers of the device.

Note that I use 40h in the MUX1 (02h). In your program, you use 60h in register. This should give the same result. If you want, you could try changing it, but the two bits that define the reference behavior are same.

In your system, try measuring the current as you start up the device, and start to program the settings. If you can, try to match the currents in the datasheet. There might be some small variation in current because I didn't try to tie the inputs to mid-supply. This might cause the PGA (even in gain=1) to have some extra current load.

joe wu

Hi Joe,

Thanks for duplicating my setup and running this test.  This is quite helpful!  The fact that you have proven the same code and register settings works on you H/W allows me to focus on the H/W differences on my end what may be contributing to the excess current.  I'll let you know what I find...

Thanks,

Richard