Announcement

Collapse
No announcement yet.

MAX134CMH+D A/D Converter: Faulty or OK?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    MAX134CMH+D A/D Converter: Faulty or OK?

    Hello Badcaps.netters,

    I have a problem with a DMM, where there is a malfunction with the 4V measuring range, and the diode test function. The meter has a MAX134CMH ADC IC in a MQFP-44 pin format. See pinout diagram and schematic diagram of DMM circuit below.

    When there is nothing being measured on the 4V range, pins 24-28 should be at 0V, but instead pin 24 has a voltage that alternates between 1.84V and 0V, and pin 25 has a steady 2.5V. Pin 24 is the input from the positive test probe lead, and pins 25-28 are connected to attenuator switches inside the chip, and to a resistor network externally. The 40V range and up all work OK.

    I can only think that the presence of the voltages on pins 24 and 25 indicate that there is a fault within the ADC IC, and that replacing the IC should bring normal functioning to my meter. Am I correct in my thinking?

    Cheers,

    Krisby.
    Attached Files

    #2
    Re: MAX134CMH+D A/D Converter: Faulty or OK?

    I can't see anything that would give stray voltage at those range divider pins. The IC's internal schematic just shows cmos switches there that are controlled by the MCU I think. It's strange the voltage is going up and down, like it's autoranging or something?
    Attached Files

    Comment


      #3
      Re: MAX134CMH+D A/D Converter: Faulty or OK?

      Hello redwire: Thanks for responding.

      I hope it's not the MCU that is at fault. I emailed MAXIM regarding the MAX134, but they weren't all that helpful, though they did concede that the MAX134 ADC could be at fault. I emailed Protek, the manufacturer of the Protek 506 DMM, and they sort of agreed that the fault may lie within the ADC IC.

      I have two of these DMMs, and that's why I know that the MAX134 pins 24-28 should be at 0V when there is no voltage being measured. I can only assume that the voltages on pins 24 and 25 are coming from within the ADC IC. Why they are doing so is a mystery to me, as I know nothing about the innards of the ADC IC.

      On the meter's display, when I have selected the DC voltage measurement function, but not measuring anything, it displays 3.999V and 00.00V, alternately. When I manually select the 40V range, the meter is working normally.

      The other problem with the meter is with the diode test function: This function is meant to initially display the word "OPEN", along with the test voltage of about 3V. If a diode is tested, the display is supposed to return either "GOOD" and the breakdown voltage, or "SHORT" and 0V. When I select the diode test function on my meter, it displays "OPEN", but also shows a voltage of 3.999V, which seems to be related to the 3.999V showing on the 4V range of the voltage measurement function.

      I've ordered a MAX134 on Aliexpress, and plan to replace my meter's ADC IC with the new one, and hopefully the problem might be fixed! Assuming and hoping that the Aliexpress version IC performs as expected.

      Still, I'd like to know more about the internal structure of the MAX134, and thus maybe gain some insight as to what could be wrong.

      Comment


        #4
        Re: MAX134CMH+D A/D Converter: Faulty or OK?

        I think pin 25 stuck at 2.5V when measuring DC volts means the IC is damaged.
        I don't think it's stuck in Ohms function because ZD6 is only 1.2V for pin 40, not 2.5V so where is the 2.5V coming from?

        Comment


          #5
          Re: MAX134CMH+D A/D Converter: Faulty or OK?

          Actually, I am now seeing 2.99V at pin 25. All the associated components with the subject IC test OK, so I presume that the voltage at pin 25 must have its origin within the IC. The same goes for pin 24, which should be at 0V when no voltage is being measured at the probes, instead of alternating between 1.8V and 0V. Curiously, if a voltage less than 4V is measured at the probes, The LCD display will alternately show (at 4Hz?) 3.999V and the actual voltage being measured. 3.999V would be at the top of the 4V range, so it seems to be an issue associated with the range selection? But, if that is the case, is that range selection done wholly within the ADC IC, or is the MCU involved?

          I note on page 8 of the data sheet a section named "Digital Interface", and explains the communications between the MAX134 and the microprocessor, so I wonder if there might be an issue there of corrupted data?

          In that schematic of the MAX134 that you attached (Figure 1 Input Section page 6 of datasheet), there are circles with "X"s in them; do they represent unnamed components?

          Comment


            #6
            Re: MAX134CMH+D A/D Converter: Faulty or OK?

            I looked at the MAX134 datasheet, the circle with an X represents a CMOS switch like a CD4066, that the MCU can control. So the MCU is doing the auto-ranging by reading the A/D and deciding which switches to have on/off.
            If it keeps going back and forth, the MCU must think the range is over-range, then move up a range then sees under range and it just keeps flipping back and forth.

            I did notice the IC has the usual oddball power configuration, with two or three grounds.
            Digital GND is 5V less than V+ (which is battery positive terminal +9V).
            Analog GND (COM jack) is 3V (internal zener) less than V+ (which is battery positive terminal +9V). So the AGND is floating and if that was way off the A/D converter would have trouble as well.

            I think the chip sees 9V between V+ and V-, genrates AGND at 3V and DGND at V+ minus 5V. Check the power +V to AGND is 3.0V, and +V to DGND is 5V.

            Comment


              #7
              Re: MAX134CMH+D A/D Converter: Faulty or OK?

              Pin 30 is named COM and it shows a voltage of 0.00V, and Pin 29, DGND, is showing -1.9V. On the other side of the IC, Pin 9 is also DGND, and it also shows -1.9V What does it all mean?

              I don't understand what you mean, redwire, when you say, "Check the power +V to AGND is 3.0V". If AGND is the COM at Pin 30, then the voltage would be 0V, how could it be 3.0V? I've got the negative test lead of my other meter attached to the GND of the subject meter, and I am probing with the positive test lead, so if I probe anything connected to the subject meter's GND, it will read 0V.

              How should I be testing the AGND (pin 30) and the DGND (pin 29)?
              Last edited by Krisby; 10-18-2020, 12:19 AM.

              Comment


                #8
                Re: MAX134CMH+D A/D Converter: Faulty or OK?

                When you have a ic chip that has “DGND” and “ AGND” these are reference points and are different connections you need to use the right reference point depending on what voltage you are measuring

                This can make your life missable if forget this point
                9 PC LCD Monitor
                6 LCD Flat Screen TV
                30 Desk Top Switching Power Supply
                10 Battery Charger Switching Power Supply for Power Tool
                6 18v Lithium Battery Power Boards for Tool Battery Packs
                1 XBox 360 Switching Power Supply and M Board
                25 Servo Drives 220/460 3 Phase
                6 De-soldering Station Switching Power Supply 1 Power Supply
                1 Dell Mother Board
                15 Computer Power Supply
                1 HP Printer Supply & Control Board * lighting finished it *


                These two repairs where found with a ESR meter...> Temp at 50*F then at 90*F the ESR reading more than 10%

                1 Over Head Crane Current Sensing Board ( VFD Failure Five Years Later )
                2 Hem Saw Computer Stack Board

                All of these had CAPs POOF
                All of the mosfet that are taken out by bad caps

                Comment


                  #9
                  Re: MAX134CMH+D A/D Converter: Faulty or OK?

                  Originally posted by Krisby View Post
                  [...] I don't understand what you mean, redwire, when you say, "Check the power +V to AGND is 3.0V". If AGND is the COM at Pin 30, then the voltage would be 0V, how could it be 3.0V? I've got the negative test lead of my other meter attached to the GND of the subject meter, and I am probing with the positive test lead, so if I probe anything connected to the subject meter's GND, it will read 0V.

                  How should I be testing the AGND (pin 30) and the DGND (pin 29)?
                  Portable digital multimeters, their input COM jack is not at 0V potential as far as the multimeter IC is concerned or their battery. This is true whether a Fluke 87 or ANENG as well. It's good to understand this when troubleshooting them but a little difficult to figure out.

                  The reason for this is the A/D has be able to read negative voltages (for -DCV and ACV) which a chip cannot do below it's ground pin. The lowest point an IC can measure is its V- or VSS pin or the most -ve potential.
                  The fix is to float up the A/D converter ground (AGND) to a midpoint say 3V, this way it can read down to 0V and up to say 6V, to be usable with a sine-wave at +/-3Vpk centered at 3V. So AGND is always biased up and is connected to the input jack COM. Not to BATT(-). It's very hard to notice this on schematics. Protek schematic has three ground symbols- all separate at its "Note 4".

                  So if you use BATT(-) as your 0V reference point, you would measure 9V at BATT(+) and 6V at AGND (and input COM), and 4V at DGND for the MAX134 as I read the datasheet.
                  The MAX134 keeps DGND at 5V below +V (so 9-5=4) 4V to BATT(-), with an internal voltage regulator.
                  AGND is 3V below +V (so 9-3=6) 6V to BATT(-), with the 3V internal zener diode. I was meaning the reading "between" those points.

                  Note that Protek 506 MAX134 V- is really it's ground, same as BATT(-).
                  But the AGND (input COM) is biased to float at ~6V.

                  When a multimeter IC gets zapped, the regulators/zener and A/D bias usually get blown and the IC can then float at some oddball voltage and sorta work but readings are wonky.
                  There is a slight chance a protection zener is shorted like ZD2, 3, 4, 8, 10, 11 and upsetting the AGND bias point.
                  Last edited by redwire; 10-18-2020, 12:49 PM.

                  Comment


                    #10
                    Re: MAX134CMH+D A/D Converter: Faulty or OK?

                    Thank you redwire for expanding upon the workings and the power/voltage requirements of the MAX134 ADC IC. I've printed out what you've said so I can study it in detail and see if I can make some relevant voltage measurements.

                    Comment


                      #11
                      Re: MAX134CMH+D A/D Converter: Faulty or OK?

                      Adding to what I have said above, my understanding is that, measuring from BATT(-), I should see around 6V at the AGND (input COM, pin 30) and about 4V at DGND (pin 29), but I'm getting 0V and -1.9V respectively.

                      I think that before I noticed this fault with the meter, I had tested the output of a DC to AC inverter, which was about 240V AC. This may have caused the problem, because I had not manually set the appropriate range, relying on the meter to auto-range.

                      I shall test all Zeners mentioned by redwire.

                      Comment


                        #12
                        Re: MAX134CMH+D A/D Converter: Faulty or OK?

                        I have tested, in situ, all 15 Zener Diodes named as such on the boards of the Protek 506 DMM, with the diode test function of my other meter. With the exception of two, they all test "GOOD" with the breakdown voltage within limits and when the test probes are reversed, they all test "OPEN".

                        The two anomalies are named on the board as ZD12 and ZD15, which test "Good" in both directions. Maybe that test result is because of other componentry that they are connected to.

                        ZD12 is technically not a Zener Diode, but a LM385 1.2V micropower voltage reference in a TO-92 package. On the schematic, it is depicted as ZD6, and connects to the COM pin 30 of the MAX134. I have previously removed it from circuit to test it, because of the two-way reading I was getting on the diode test, but it tested OK when off the board.

                        The other Zener that gave a two-way "GOOD" reading is ZD15, which I cannot find on the schematic. On the board, its Anode is connected to PTC1 and C1 (I cannot find C1 on the schematic either). and the Cathode is connected to two contacts of the rotary switch. With the movable part of the switch absent (this is the top PCB, removed from the meter), I cannot detect that the Cathode of ZD15 is connected to anything else, so that must mean that ZD15 is faulty? When I test with the NEG test probe on the Cathode, and the POS probe on the Anode, I get a "Good" with a reading of .680V. When I reverse the probes, I get "GOOD" and .838V I shall remove one of the Zener's leads from the circuit board, then test it.

                        Comment


                          #13
                          Re: MAX134CMH+D A/D Converter: Faulty or OK?

                          Whoops! I've made a mistake! The Cathode of ZD15 is also connected to PTC2 and SG1.... when I pulled one of ZD15's leads off the board, that Zener tested normal. So, seems like there is no problems with the Zener Diodes.

                          As it seems that the MAX134 A/D Converter has a fault within it, and I have ordered a replacement, is it likely that the MCU also has a fault? I hope not, because the Samsung MCU has umpteen pins and is probably expensive.

                          Comment


                            #14
                            Re: MAX134CMH+D A/D Converter: Faulty or OK?

                            I don't know if it is relevant, but: When the meter is on the 4VDC range, and the display is alternating between showing 3.999V and 0V, if I short out X3 (connected to pins 15 and 16), by pressing my finger on the two leads sticking out the other side of the board, then the display stabilizes at 3.999V! When I remove my finger, the alternating voltage display returns. Pin 15 is named OSC1 and 16 is named OSC2, so, presumably X3 is an oscillator.

                            What, if anything, does it mean?

                            Comment


                              #15
                              Re: MAX134CMH+D A/D Converter: Faulty or OK?

                              The multimeter seems to have completely lost the AGND bias voltage if it's reading the same as BATT(-). There's supposed to be a 3V difference between BATT(+) and AGND.
                              Measuring ohms between AGND and BATT(-) would likely show any short, although where it might be is another issue.
                              I could not see any components between AGND (COM) and BATT(-), sometimes there is a capacitor or diode. I thought a shorted zener could be to blame but they all look fine. The rest of the circuit RMS, freq, beeper etc. cannot cause AGND to short to BATT(-). That kinda leaves the multimeter IC as the fault.

                              The crystal oscillator is sensitive, touching any traces or leads will make it stop. I think it's a side-effect the oscillator might not be getting proper operating voltage right now. The MCU too- it's supposed to get 5V but be careful it must be high because the 3V AGND voltage drop (from BATT(+) is not there. I would be using a weak battery to power the multimeter then and not overvoltage the MCU. You can measure between DGND and BATT(+) is supposed to be 5V for the MCU.

                              Comment


                                #16
                                Re: MAX134CMH+D A/D Converter: Faulty or OK?

                                I get a reading of 0.2 Ohms between AGND and BATT(-)

                                Where does AGND and DGND come from, or go to?

                                On the schematic, there is an arrow coming from the DGND at pin 29 of the MAX134, pointing to a "cartouche" with some letters in it, but as it is on the crease of the original schematic, I can't read it.

                                I can't see on the schematic of the MCU, which pin/s are supposed to be supplying 5V.

                                I've tried to find a pinout drawing of the MCU, which is a Samsung S3F84DBXZZ-QXRB TQFP100, but every link to the datasheet turns out to be for an 80 pin MCU, not the 100pin MCU that is used in the subject meter.

                                For what its worth, I've measured some voltages at CN9, which is one of the two 12pin connectors joining the two boards. According to the schematic, the first four are named as:

                                1 (a circle with a + in it): 2.9V
                                2 V+: 2.9V
                                3 Vss: -5.69V
                                4 DGND: -1.9V

                                Does it mean anything?

                                Comment


                                  #17
                                  Re: MAX134CMH+D A/D Converter: Faulty or OK?

                                  Uh that seems like a short between AGND and BATT(-)

                                  From the MAX134 datasheet:
                                  "Analog Common is derived from a zener and is nominally 3.0V below V+"
                                  "...MAX134 also generate a Digital Ground voltage, which is nominally 5V below V+ and which will remain in the range of 5V +/-10%..."


                                  I did find this (different) Protek 506 CPU schematic. There is tantalum capacitor C65 2.2uF between BATT(-) and AGND, might be shorted but at this point I think it's the multimeter IC that is shorting those two grounds.

                                  Your voltage measurements seem weird, which point is your reference?
                                  Circle with + is switched 9V BATT(+) that Q2 does so it (CPU) can turn off the multimeter.
                                  V+ is raw BATT(+) after the mechanical on-off switch.
                                  VSS I can't find where it's from.
                                  Attached Files

                                  Comment


                                    #18
                                    Re: MAX134CMH+D A/D Converter: Faulty or OK?

                                    Thanks, redwire, for supplying that clearer schematic of the CPU (U1). Is there a clearer schematic image of the MAX134 (U2) and rotary switch connections to be found?

                                    So, are you thinking now that there may be a fault within the CPU (U1)?

                                    Curiously, I checked out the temperature function on the meter, and it was reading well over 200 deg Celsius for the ambient temp! By turning the temp adjuster VR4, all the way, I got the temp reading down to 159 deg C! Is this strange anomaly in any way relevant to the problem at hand?

                                    Comment


                                      #19
                                      Re: MAX134CMH+D A/D Converter: Faulty or OK?

                                      I found that schematic at https://www.elektroda.pl/rtvforum/topic360033.html (run through google translate) but nothing more other than Excel file about calibration.

                                      I think the CPU is OK although likely running at over 5V so I would wait until the new MAX134 comes in. I'd just use sharp side cutters and carefully cut the pins to get the old IC out, and then a quick wipe with the soldering iron to clean off the pads.

                                      The A/D is pretty much crippled with the bias voltage being gone. It can read some voltage values OK but not ones that are -ve or close to the rail- there it would read incorrectly.

                                      Comment


                                        #20
                                        Re: MAX134CMH+D A/D Converter: Faulty or OK?

                                        Interesting website, that "elektroda".

                                        Yes, I just have to wait until the new MAX134 shows up (assuming it is not junk). Cutting the pins off with small sharp side-cutters is effective, but I'm thinking of making a purpose-built 4 sided copper "iron" which I can melt the solder on all 44 pins at once, then pull the IC up & off with a bit of glue on a stick. Well, that's the idea.

                                        redwire, I hope you are looking forward to an Alberta winter! I spent a couple days once up at Grand Prairie, but that was in the Summer.

                                        Comment

                                        Working...
                                        X