1
\$\begingroup\$

This is going to be a long question because I plan to include a lot of context that I feel is important to grasp the full scope of my issue.

I am currently a senior year university student hard at work designing PCBs for my senior design project. My group and I have settled on designing a multimeter from scratch.

After doing some component analysis and research, we have landed on the Maxim Integrated MAX134 Digital Multimeter IC for our data acquisition chip. Part of the requirement for our project included incorporating a Cypress programmable system on a chip (PSoC), specifically the Cypress PSoC 63 BLE module (CYBLE-416045-02) for our microprocessor.

Since the DMM circuit runs off of +/-5V supplies, and the logic level of our PSoC is strictly 3.3V, we needed to include level-shifters between our DMM circuit and our PSoC. For this we used the Maxim Integrated MAX3001E 8-channel level shifter, as well as another 4-channel one by a similar name.

I should note that much of the design decisions I made for the DMM circuit application was taken from the MAX134 Evaluation Board Schematic and Datasheet, which I will link below, along with the MAX134 chip datasheet.

We got the boards fabbed and assembled them ourselves to begin verifying the integrity of the circuits. We put the PSoC and DMM circuit each on their own individual breakout boards, using jumper wires to connect the two for testing purposes.

Now I will begin explaining the core of the issue. To limit as many variables as possible, we started by simply connecting power to only the DMM board on it's own using a programmable power supply to measure the power draw of the board as well as make sure that all data that would be leaving the board would be properly voltage-translated.

One of the things that caught my eye is the current draw of the system was upwards of 80mA, which corresponds to almost 0.8W. Specifically, the 5V source was drawing about 70mA, and the -5V source was drawing 10mA. The 3.3V source was only drawing about 1mA, which is what I would expect, considering it was only powering the level shifters. The DMM circuit power draw seemed particularly high, and after measuring the temperature of each of the IC's with my finger, it was clear that the MAX134 chip was consuming most of the power.

We measured the voltage coming to the breakout pins from the output of the level shifters, and they were a very clean 3.3V. However, the logic coming from the MAX134 chip seemed to be in some sort of voltage-reduced state, all of the data pins outputting a voltage between 3.7 and 4.1 volts.

Next I wanted to measure the voltage reaching the MAX134 from the pins. On the address logic headers, I supplied 0b101 to the board. This voltage reached the level shifters just fine, but after being shifted to the new level, it was once again reduced to around 3.7 to 3.3 volts, with the logic low state measuring around 0.5 volts.

This is the crux of my issue really. I'm unsure whether to hook up my microprocessor board to try to communicate with the MAX134 chip, when it seems that either the level shifters or the MAX134 chip itself seem to be having issues. I was looking at the data sheet for the MAX3001E chip, and supposedly receiving these voltages after translation is "within threshold" but it realistically it could be below the threshold for the MAX134 chip to recognize. This could also be a power-supply issue as well, or perhaps a board assembly issue since it was hand-assembled using solder paste, possibly leading to a short somewhere (but continuity checks seem to debunk this possibility).

All of this, not even to mention the fact that it seems that no matter what control register address I attempt to index in the MAX134 returns anything other than 0b1111 in the data lines.

Has anyone had logic level shifting issues in the past? Could it be an error in assembly or design?

Any and all help is appreciated. If you would like to see my schematic or PCB document, I am happy to share.


Level shifter data sheet: https://datasheets.maximintegrated.com/en/ds/MAX3000E-MAX3012.pdf

MAX134 chip data sheet: https://datasheets.maximintegrated.com/en/ds/MAX133-MAX134.pdf

MAX134 Evaluation Board Datasheet: https://datasheets.maximintegrated.com/en/ds/MAX134EVKIT.pdf


EDIT

All boards were designed using Altium Designer.

Screenshot of Board Layout Document (MAX134 in orange, level shifters in blue, bandgap reference in red, external AC to RMS converter in green. 5 IC's in total): Board Layout Document

Screenshot of Schematic Document:

Board Schematic Document

\$\endgroup\$
19
  • \$\begingroup\$ Step 1: beep connections and make sure pinout is OK. Step 2: scope the power line of the chip. Step 3: try to replace the chip (if possible). Also, I take it the negative rail of MAX134 is -5V and its output may be referenced not to GND but to -5V. Did you take that into account? \$\endgroup\$
    – Ilya
    Feb 12, 2022 at 22:09
  • 1
    \$\begingroup\$ This sounds as if the 5 V supply is shorted to the 3.3 V supply. Anyway, show the schematic. \$\endgroup\$
    – CL.
    Feb 12, 2022 at 22:10
  • 1
    \$\begingroup\$ Can you please show your board’s schematic? The symptoms you describe sound as if the max134 interface bus isn’t being enabled. I’m familiar with both the max134 and the max300x level translators, but there may be some subtle detail in how they are connected. The logic levels should be much closer to the power supply rails; that’s related to higher than normal supply current. \$\endgroup\$
    – MarkU
    Feb 12, 2022 at 22:26
  • \$\begingroup\$ Did you build a prototype before sending the board design out for fab? That would usually be part of the design process for a capstone design course, and it would help tell you whether the problem is in the overall design or in the PCB. \$\endgroup\$ Feb 12, 2022 at 23:16
  • \$\begingroup\$ @Ilya -- I'll try scoping the power line of the chip on monday when I'm back in the lab, but the power supply seemed to have fairly stagnant current and voltage, not much flakiness in the connection either. Also tried re-soldering the powe supply headers early on, but no avail. As per the datasheet, the output logic is taken in reference to DGND, which in our schematics' case is ground. Thanks for the reply. \$\endgroup\$ Feb 12, 2022 at 23:59

1 Answer 1

1
\$\begingroup\$

For those wondering, I found a solution to my issue. Indeed, the package information on the datasheet is wrong. The correct pin layout, quizzically, is only explicitly described on a Maxim Integrated technical forum post posted in my last comment. All in all, ADI/Maxim needs to get their datasheets reviewed, as my group and I are very unlikely to complete our project before the quarter is over at this rate due to this obvious issue.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.