PGA309EVM-USB: Gain/offset settings do not seem to be properly written, in "Block Diagram"

HI Benjamin,

I will get a hold of the PGA309EVM and attempt to reproduce the issue.

Please allow a couple of days.

Thank you and Regards,

Luis

Hello,

Thank you for your reply.

I add that I finally succeeded to change the Output Amp gain value by modifying an EEPROM file in "Addr8". Then, the value in the Block Diagram changed, but the output voltage didn't.

Hello again,

I bring some more information. I used a data acquistion to observe the signals.

On I2C, the PGA309 is looping to get the register values from the EEPROM (alternatively from register Addr0 to Addr15 and from Addr16 to Addr27). The signals seem to be compliant with the protocol detailed in the PGA309 user guide.

By the way and as you may know, the TPs SDA1 and SCL1 are inverted (cf. circuit diagram sbor010).

On One-wire, there is no data on the test point Vout_F1as long as I don't I click on "Write All Reg". When I click on it to change the settings, the signals also seems to be compliant with the protocol detailed in the PGA309 user guide, figure 4-6 (I can send you the trace if you want).

What I think is that the data are sent from the microcontroller to the PGA309 via the One-wire communication, but as soon as the value are received, the PGA309 is automatically reset by the EEPROM with the I2C communication.

However, I still don't understand, why even when I change the gain/offset value in the EEPROM thanks to the file I load in the "EEPROM" section of the software, the value at the output of the PGA309 does not change...

Hi Benjamin,

One possible issue, is that you are loading the "4wire_4p096_diode.txt" model file and attempting to change settings in 3-wire mode (PRG connected to VOUT).   On my side, if I use the "4wire_4p096_diode.txt" model file, I leave JMP7 set to NC, and JMP8 set to ONE to PRG, for 4-wire configuration.I am able to go in the Block Diagram and modify gain settings, and see the output and registers update (read and write) properly; and I can see the device output change accordingly as I change gains, use different input voltages on the emulator. 

Can you start by attempting the 4-wire mode first and see if this works for you properly?

Please, keep in mind the PGA309 checks both the proper flag to start reading the EEPROM, and will also verify valid checksum while reading the EEPROM settings. Therefore, if you have manually modified the EEPROM settings without updating the checksum with a valid value, the PGA309 will not load any of the settings.

Please refer to the following ppt document that details how to complete the calibration with the PGA309 in 4-wire votage output, 3-wire voltage output, and IOUT mode:

Calibration = USB_DAQ + PGA309EVM_updated.pptx

If you are interested in using the device on 3-wire voltage mode, please load the following pre-cal file: V_3wire_1-5V_pre-cal.txt

and the following model file: V_3wire_1-5V_model.txt

On the ppt file, chapter b) for 3-wire voltage output mode uses the default settings with the sensor emulator, the ppt documents also the proper jumper settings.

Thank you and Regards,

Luis

Hello Luis,
Thanks for this reply!
I actually tested the other configurations (3-wire mode / "4wire_..." file, 4/4, 3/4, 3/3), with the jumper placed accordingly with the Table 5 of the document sbou084b, but none of them are working :-(

I already tried to follow the ppt document before and proceeded to a auto-calibration, but that's not really relevant with what I'd want. As a matter of fact, I don't really care of the kind of calibration model I use, since I just want to test the gain/offset settings. By the way, is it mandatory to use the auto-calibration every time I have to run an emulated sensor? I noticed the "Absolute Bridge" values displayed in the window "Sensor Emulator" were the same, whether I performed the auto-calibration or not.

I created a new Model File using the pre-calibration file and the model file you sent me and following the ppt document. It's a shame there is not the configuration of the "Block Diagram" (in order to not be in over or undervoltage at the output) or at least the "Output Mode" one (but I guess it's obvious) presented in the document.
I set the Output Mode to "Vout (3-wire)".
The jumpers are placed as follow : JMP7 set to "Vout to PRG" ; JMP8 set to "ONE to Vout".
When I click on "Run Sensor Emulator" and I set the "Load" to 25% ("Temperature" set to 0°C), the "Absolute Bridge" value is set to 9.54mV (confirmed by multimeter between Vin1 and Vin2) and the Out Meas value is 127mV (also checked at the connector P1 output).
Then, I change the value of the "Output Amp Gain" from '3.6', to '6' and click on "Write All Reg" aaaaaand... The value stays at 127mV :-(

When I click on "Read all Reg", the value of the "Output Amp Gain" is set back to 3.6.
It's a shame that everything seems to work fine, appart from the controls in the "Block Diagram" :-s

When I look at the register 4 of the PGA309, this latter is set to 0xB619 (shame since it means that the OWD is enabled and I believe I need it set to '0' for a 3-wire communication).
As I cannot change it with the "Block Diagram" I try to change it in the EEPROM window.

N.B. : About that, indeed I did not change the checksum in the register 14 during my previous tests, then I had systematically the value 3.7V at the output (high impedance level, I guess).  Thanks for the tip!

The file  "MODEL_TEST_MN_7_SN_3_DC_3220.txt" seems to be the selected model. Then, I copy the file and I just change the value of the Addr8=46617 (0xB619) to 13849 (0x3619) to disable the OWD (and the Addr14 accordingly) and I save and load the new file in the software.
This time the modification really change the state of the OWD checkbox to "unchecked" in the "Block Diagram" section, but when I try to change the values of gain/offset (I don't forget to click on "Write All Reg"), the output voltage stays at 127mV : the value of the registers of the PGA309 still cannot be changed.

Then, I change the value of the output gain to '6' in the EEPROM configuration file (Addr8=22041 (0x5619)). The output voltage also stays at 127mV.
I get the same issue when I try to reduce the Output Gain or change another gain/offset value...

I can't think of any other test to run and I begin to think there is a problem with my evaluation board :-(

Hi Benjamin,

When using "Test mode", the PGA309 Vout is immediately active (vout is on), and therefore, PRG pin needs to be independently connected to establish communication.  On the PGA309EVM standard multi-cal routine, when using the device connected on 3-wire (PRG shorted to VOUT), the software cycles the supply to disable vout momentarily to establish digital communication through the PRG pin.  Once the device has been programmed, the calibration routine continues.  The time interval after power-up is the only opportunity that an external communications controller can initiate digital communication with the PGA309 through the PRG pin and trigger a one second delay in the internal state machine. After VOUT is enabled no further digital communication is possible, unless power is cycled...

The PGA309EVM software uses the PRG 1-wire interface to communicate with the device on 4-wire mode (also on 3-wire mode by toggling the supply, establishing digital communication, but not on test mode). I believe the SDA/SCL 2-wire interface on the EVM software is only used for direct EEPROM programming.

The block diagram,  will not work on "test mode" with PRG shorted to VOUT since VOUT is immediately turned on. The EVM software uses the one-wire PRG pin to communicate with the device, and I believe you will not be able to use "test mode" with PRG shorted to VOUT.  If your application requires no EEPROM due to current consumption concerns, with PRG connected to VOUT (three-wire) an alternative solution may be to use the PGA308 instead. 

If you have already programmed the EEPROM with a three-wire (PRG to Vout) configuration, and wish to re-establish communication, you could exercise the "test mode" on the block diagram on 4-wire mode, you will need to reload the EEPROM contents for 4-wire mode, cycle the PGA309 power, and then exercise the device.  Attached is an example of an EEPROM file for 4-wire interface. It may be also possible to communicate with the PGA309 device using the 2-wire SDA/SCL communication while on test mode, but I don't believe this function is not supported by the PGA309EVM software.

[General Info]
Date=9/4/2019
Time=6:55 PM
[Paths]
Model path=/C/Program Files (x86)/proj_PGA309EVM-USB/Model_Files/4wire_4p096_diode.txt
Pre-Cal path=/C/Program Files (x86)/proj_PGA309EVM-USB/Pre_Cal_Files/Vout_4wire_4p096_diode.txt
Sensor Emulator path=/C/Program Files (x86)/proj_PGA309EVM-USB/Sensor_Emulator_Files/nonlin2.csv
[EEPROM Data]
Addr0=21577
Addr2=0
Addr4=0
Addr6=1792
Addr8=1555
Addr10=0
Addr12=5133
Addr14=35478
Addr16=3599
Addr18=15176
Addr20=65535
Addr22=3702
Addr24=159
Addr26=64813
Addr28=3775
Addr30=172
Addr32=60156
Addr34=3849
Addr36=187
Addr38=60264
Addr40=3922
Addr42=200
Addr44=60377
Addr46=3996
Addr48=211
Addr50=60489
Addr52=4069
Addr54=224
Addr56=60602
Addr58=4143
Addr60=239
Addr62=60710
Addr64=4217
Addr66=249
Addr68=60824
Addr70=4290
Addr72=263
Addr74=60935
Addr76=4364
Addr78=277
Addr80=61049
Addr82=4437
Addr84=288
Addr86=61156
Addr88=4511
Addr90=301
Addr92=61270
Addr94=4584
Addr96=316
Addr98=61380
Addr100=4658
Addr102=325
Addr104=61492
Addr106=4732
Addr108=339
Addr110=61603
Addr112=4835
Addr114=355
Addr116=61738
Addr118=32767
Addr120=0
Addr122=11523

Thank you and Regards,

Luis

Hello Luis,

Thanks for this new reply.
For your information, all the tests described in my last message before this one were performed in Normal Mode.

What I understand from your statement is that I can program the PGA309 in 3-wires, but only during a little time interval before it reaches Stand-Alone Mode (25ms, according to what is described in chapter 4.10 of the User's manual sbou24b).
Furthermore, I cannot also used 3-wire communication in "Test Mode" for the obvious reason that Vout will be constantly enabled.
Then, I guess the only way to perform a 3-wire configuration with the PGA309 is to follow the state-machine described in the Figure 4-16 in the document sbou024b.
The chapter 4 is really confusing about the definition of One/Two/Three/Four-wire communications modes. Maybe a table, which summarizes the whole communication modes (or a better order in the subchapters) could make it easier to understand.

I had already figured the MCU could not program the PGA309 with the I2C communication out, but thanks to have confirmed it. And I confirm to you I could not communicate with the PGA309 in Test Mode either.
It's a shame none of these details are mentionned in the User Manual of the PGA309EVM-USB or the ppt document... I lost a lot a time to find all by myself :-s

In any case, thanks for this new file. I'm going to try the last test you suggest me to perform (and also to follow the state-machine for 3-wire communication), but not today. I'll try to perform it by the end of the week and I'll let you know the results.

HI Benjamin,

Glad to hear the solution works for you.   

Thank you and please let us know if you have any questions. Have a nice day as well.

Kind Regards,

Luis