Programmable Electronic Load - ADC and DAC BoosterPack test

WIP

This blog documents the characteristics of the DAC/ADC BoosterPack of the electronic load that Robert Peter Oakes, jc2048 and Jan Cumps are designing.

We have the same DAC/ADC BoosterPacks so we can compare the measurements of each ADC and DAC channel. Check the main document for anything related to hardware and firmware.

Test Setup

Peter has created a GUI that we can use to shoot SCPI commands. We can control the DACs and read the ADCs.

Measurements are performed with and without the power driver BoosterPack installed. We'll use 5V to power the DACs and ADCs in both situations.

P1 and P2 on the bottom of the DAC/ADC BoosterPack have to be open. Else you'll inject 5V onto the LaunchPad 3V3 rail and you'll damage components.

It's not critical if you switch on the power supply before or after you connect the LaunchPad to your computer. The circuits are isolated.

All measurements use P3.10 as GND reference and P3.1 as reference voltage.

Setup without the power driver BoosterPack:

Mount the DAC/ADC BoosterPack on the LaunchPad.

Use a +5V power supply. It should be a decent one, current is approximately 5 mA.

Connect with DAC/ADC BoosterPack P4.2 and with P4.1.

Setup with the power driver BoosterPack:

Mount the DAC/ADC BoosterPack on the LaunchPad and the Power Driver BoosterPack on top of that.

On the Power Driver BoosterPack, bridge pins P2.4, P2.5 and P2.6 to tie both opamp inputs of opamp 3D to ground.

Use a +12V power supply. It should be a decent one, current is approximately 50 mA.

Connect with Power Driver BoosterPack P1.1 and with P1.2.

Measurements

In the table, the first column shows what we're measuring, the second column the SCPI command to execute with Peter's GUI.

The next columns are the measurements for Peter, Jon or Jan with and without the Power Driver BoosterPack.

The last column is for comments. We can indicate odds, errors, learnings.

Data Point	SCPI	Without	Without	Without	With	Measurement method
		Jon	Peter	Jan
Supply	N/A			4.989 V	11.88 V	DMM
DAC/ADC V+	N/A			4.989 V	5.035 V	DMM
VREF	N/A			2.046 V	2.046 V	DMM
ADC A GND	DEVE:ADC1?			2		SCPI output
ADC A VREF	DEVE:ADC1?			10918		SCPI output
ADC A	DEVE:ADC1?				1138	SCPI output
ADC B GND	DEVE:ADC2?			12		SCPI output
ADC B VREF	DEVE:ADC2?			10920		SCPI output
ADC B	DEVE:ADC2?				1	SCPI output
ADC C GND	DEVE:ADC3?			2	2	SCPI output
ADC C VREF	DEVE:ADC3?			10920	10919	SCPI output
ADC C	DEVE:ADC3?				26868	SCPI output
ADC D GND	DEVE:ADC4?			8	3	SCPI output
ADC D VREF	DEVE:ADC4?			10919	10919	SCPI output
DAC A 0	DEVE:DAC1 0			0.000 V	0.001 V	DMM
DAC A max	DEVE:DAC1 65535			2.040 V	2.040 V	DMM
DAC B 0	DEVE:DAC2 0			0.005 V	0.005 V	DMM
DAC B max	DEVE:DAC2 65535			2.046 V	2.046 V	DMM
DAC C 0	DEVE:DAC3 0			0.000 V	0.000 V	DMM
DAC C max	DEVE:DAC3 65535			2.039 V V	2.039 V	DMM
DAC D 0	DEVE:DAC4 0			0.000 V	0.000 V	DMM
DAC D max	DEVE:DAC4 65535			2.039 V	2.039 V	DMM

To resolve
With the power driver installed, ADC C and D act strange (probably also A and B, but I didn't inject a voltage in them because they are connected to opamp pins). In particular strange for ADC D, because that one isn't routed. Strange. I will try to decrease the frequency of the sample speed and the i²c speed to see if it's related to that ... Follow up: the cause is that OPAMP 3C injects -0.8V into ADC A. All measurements are off when that happens. As soon as I force ADC A positive (by tying it to ground or a positive voltage) all measurements are correct.

To resolve

With the power driver installed, ADC C and D act strange (probably also A and B, but I didn't inject a voltage in them because they are connected to opamp pins).

In particular strange for ADC D, because that one isn't routed. Strange.

I will try to decrease the frequency of the sample speed and the i²c speed to see if it's related to that ...

Follow up: the cause is that OPAMP 3C injects -0.8V into ADC A. All measurements are off when that happens. As soon as I force ADC A positive (by tying it to ground or a positive voltage) all measurements are correct.

We have a separate document to analyse the summing node at zero.

Top Comments

Parents

Jan Cumps over 8 years ago

For some reason it's working now. I'm not comfortable because I don't know the reason.
I've removed and plugged in everything - but I've done that umphteen times before.

With the board:
(I didn't drive ADC C because that's a diret output from an opamp without any resistor between opamp and ADC so I don't want to fidle with that)

DEVE:ADC1 = -0.006
DEVE:ADC2 = 0.000187500
DEVE:ADC3 = -0.00468750
DEVE:ADC4 = -0.00431250

DEVE:ADC1 = 2.03662
DEVE:ADC2 = 0
DEVE:ADC3 = 2.03888
DEVE:ADC4 = 2.03944
- Cancel
- Vote Up +1 Vote Down
- Sign in to reply
- More
- Cancel
Jan Cumps over 8 years ago in reply to Jan Cumps

For some yet to understand reason: if I tie ADC A to ground then all others work fine. odd ...
- Cancel
- Vote Up +1 Vote Down
- Sign in to reply
- More
- Cancel
Jan Cumps over 8 years ago in reply to Jan Cumps

owkey. There's -0.799V on that pin and the ADC doesn't like that at all. Once I push that ADC pin to 0 or higher, all channels work fine.
So we can now look at the opamp that's feeding ADC A.
- Cancel
- Vote Up +1 Vote Down
- Sign in to reply
- More
- Cancel
jc2048 over 8 years ago in reply to Jan Cumps

Do you have the current sense resistor attached?
- Cancel
- Vote Up +1 Vote Down
- Sign in to reply
- More
- Cancel
Jan Cumps over 8 years ago in reply to jc2048

no. Nothing from the off-board power components. I haven't purchased them yet. So the inputs for the sense opamp are floating -> output can be anything -> may be the cause for the woes ...
- Cancel
- Vote Up +1 Vote Down
- Sign in to reply
- More
- Cancel
Jan Cumps over 8 years ago in reply to jc2048

There's good news. If I tie both inputs of the current sampling opamp 3C (the one that feeds back into the summing node) to ground, the negative voltage disappears.
I now get a fairly stable reading on all four ADCs.
It's very likely that this turns out to be a non-issue.
- Cancel
- Vote Up +1 Vote Down
- Sign in to reply
- More
- Cancel

Comment

Jan Cumps over 8 years ago in reply to jc2048

There's good news. If I tie both inputs of the current sampling opamp 3C (the one that feeds back into the summing node) to ground, the negative voltage disappears.
I now get a fairly stable reading on all four ADCs.
It's very likely that this turns out to be a non-issue.
- Cancel
- Vote Up +1 Vote Down
- Sign in to reply
- More
- Cancel

Children

Robert Peter Oakes over 8 years ago in reply to Jan Cumps

Good, I was hoping it was not my design , once I saw you mention the - voltage it seemed to indicate a floating current sense input, glad it was something simple
- Cancel
- Vote Up +2 Vote Down
- Sign in to reply
- More
- Cancel
Jan Cumps over 8 years ago in reply to Jan Cumps

Some test points:
The 5.036 V on the ADC3 (measured with a DMM) is sampled as 5.0364

With all inputs of opamp C tied to ground and the DAC A set to 0 V, the voltage at the summing node (measured with a DMM between R3 and R4) is 0.018 V
This results in a voltage going out of opamp B into ADC A of -0.212V (also measured with a DMM on the pin of ADC A). The ADC reads -0.2130 V.

I've been running the device for many hours now. Things aren't drifting
- Cancel
- Vote Up +2 Vote Down
- Sign in to reply
- More
- Cancel
Jan Cumps over 8 years ago in reply to Robert Peter Oakes

Robert Peter Oakes, I've made a document to investigate the summing node of the design: Programmable Electronic Load - Analyse the Summing Node Zero Point
Let's check what drives this away from 0V and if we can compensate that ...
- Cancel
- Vote Up 0 Vote Down
- Sign in to reply
- More
- Cancel
Robert Peter Oakes over 8 years ago in reply to Jan Cumps

As soon as I get the parts and build mine up, I will be in testing things out

from a summing junction perspective, think of it as simply a pair of resistors where one end is the input and the other is the feedback, the middle should always return to zero, pretty simple when thought of that way

of course we have added op amps etc in the feedback loops etc just to make it more interesting

Peter
- Cancel
- Vote Up +1 Vote Down
- Sign in to reply
- More
- Cancel