Analog-Digital ICs

[BelaRincewind]

Due to the Arduino Nano not having a DAC on its own and the Arduino Due, which was used in the prototype, the idea is to use an IC carrying a DAC with multiple channels.
As there are a lot of possible choices, this is the place to save ideas, ICs which could be used and requirements the IC must fulfil.
The IC should be available at Mouser so the website's link to the IC should be attached.

[BelaRincewind]

https://www.mouser.de/new/texas-instruments/ti-dac70504-dac80504-dacs/
DAC-IC that could be used. Features 50MHz SPI Communication, 5V Output, Gain factor setting, 4 channels, samplingrate of around 1MHz.

[leloup314]

In principle the IC looks good but take a look at the package: it is exactly what we talked about. Very difficult to solder manually so I would recommend taking a different footprint

[leloup314]

This is a nice DAC https://www.mouser.de/ProductDetail/Analog-Devices/AD5686RBRUZ?qs=NmRFExCfTkENVENei53ArQ%3D%3D

[leloup314]

We can also think about having an external 4 Ch DAC and an external 8 Ch ADC, both with 16 bit resolution and SPI interface.
Then the Nano simply just does the communication via SPI (which is quite fast) and the calculation. Would probably be a nicer solution.
One could use such an ADC https://www.mouser.de/ProductDetail/Maxim-Integrated/MAX1168BCEG%2b?qs=d3kjJb5aE2YvByMCteG6qw%3D%3D

[BelaRincewind]

After browsing again myself, I find the same/similar chips as suggested.
I would propose using following ICs.
DAC: AD5686BRUZ

• 4 channel, 16-bit, SPI, 25MS/s

ADC: MAX1168BCEG+

• 8 channel, 16-bit, SPI, 200kS/s

[leloup314]

ADCs which also include multiple gains:

• https://www.mouser.de/ProductDetail/Microchip-Technology/MCP3464R-E-ST?qs=W%2FMpXkg%252BdQ5bjobVCBkOgQ%3D%3D
• https://www.mouser.de/ProductDetail/Texas-Instruments/ADS130B04QPWRQ1?qs=Rp5uXu7WBW%2F83WRIW0ul1A%3D%3D

[BelaRincewind]

I think the ADS130B04-Q1 from Texas Instruments is a good choice.

[BelaRincewind]

Currently, one can choose between 10-10000nA as full-scale for the IU-conversion. As this won't change, I propose a 120kΩ resistor as 'gain'. 10µA would this way use the full range of the ADC (1.2V) and 10nA would use ~2/3 of the gain 2 setting (±18.75mV). Using this setup would also allow negative beam currents if necessary. Otherwise we could half the gain and shift the incoming voltage to use the full ADC voltage-range. I don't think this is needed as the resolution of 16-bit is probably still far greater than the effective resolution.

[leloup314]

Sounds good to me :) ICs should be ordered next week

[BelaRincewind]

In the data sheet of the ADS130B04 I noticed, that AGND to DGND should be between -0.3 and 0.3V. Is that an issue? I remember we discussed that the analog ground should be -1.65V and AVDD 1.65V so that a bipolar signal can be digitised?

[leloup314]

Well-spotted! This will be a problem. Luckiy, the ADS130B04 states that either analog input can be the following
Analog input voltage | AINxP, AINxN | min: AGND – 1.6 , max: AVDD + 0.3 V|
Therefore, we should be able to operate the amps at +-1.65V, configure the output as we planned to be at +-1.2V maximum, and just operate the ADC with 0 to 3V3, analog and digital. Then, the maximum we have is AGND -1.2V which is within specifications