|
Panel Instruments
Q1. What is the maximum input voltage to a module?
Q2. The meter is reading '1' or '-1'. What is going wrong?
Q3. The readings are unstable. What is going wrong?
Q4. So what's all this common mode input stuff anyway?
Q5. What are the scaling requirements of a module?
Q6. How do I use multiple meters with the same PSU?
Q7. How do I use 2 meters to read Voltage and Current with a common PSU?
Q8. I am using a dc/dc converter to obtain an isolated supply for the meter, but the display is unstable, help?
Q9. Help! I've tried to follow all your suggestions, but I still can't get my meter to work. What do I do now?
Single Meter Application Inquiries
Multiple Meter Application Inquiries
Counters
Q1. I have a sensor with quadrature outputs and I need to count pulses up in one direction and down in another?
Panel Instruments
Q1. What is the maximum input voltage to a module?
A1. Our modules are generally configured for 200mV input. All meters can be scaled using resistors to read higher inputs. The product data sheet has a table showing Ra and Rb values. Some meters have provision for fitting these components onboard. In this case we recommend that the maximum level applied should not exceed 60V. This restriction is due to the track spacing on the PCB.
Q2. The meter is reading '1' or '-1'. What is going wrong?
A2. There are three possible causes:
1. The most likely cause is that the meter is being used with its inputs beyond their operating range. Check the product data sheet and make sure that the inputs are well within the power supply rails. A good practical test for this on a non 'S-type' LCD meter is to replace the power supply with an isolated 9V battery. If this works, it suggests that your existing power supply is causing the meter's input to exceed the common mode input range. Remember that only an 'S-type' meter can have its inputs referred to the same supply as the meter.
2. The meter's voltage reference has not been connected. This will depend on the meter in question, but check the application notes on the back of the product data sheet, especially the required link connections.
3. The input voltage is too large!
Back to top
Q3. The readings are unstable. What is going wrong?
A3: The answer to this would fill a book. Consult the Application Notes on our website. Avoid placing the meter near to cables with large noisy currents in them. Check that ground loop errors are not causing signal offsets. Try an isolated battery supply and fit a 10_F capacitor between signal common and V+.
Q4. So what's all this common mode input stuff anyway?
A4. It depends on how much you want to know. In the simple case of measuring a differential input using a fully isolated power supply or battery, you can pretty much ignore such errors. If you need a full spec run down on function, supply & input limits, then the best sources are IC manufacturers' data on 71XX series from Maxim, Intersil, TelCom, etc.
Q5. What are the scaling requirements of a module?
A5. The scaling resistors Ra and Rb should be metal film types, 1% tolerance and have a temperature coefficient of 50ppm/°C or less to minimize drift. Note: you normally have to cut a link (on the module's PCB) that is shorting out Ra. When scaling resistors have been fitted, it will be necessary to recalibrate the module. Please note that if potentiometers are used for scaling, their temperature coefficient will be greater than that of fixed value resistors. If measuring high voltages please remember that a standard leaded resistor typically has a maximum operating voltage of 250V (sometimes less).
Scaling for above 200V should always be done externally to the meter. Metal oxide resistors often have a higher breakdown voltage than metal film, at a penalty of a higher temperature coefficient, (actual specifications should be carefully checked prior to use). Also bear in mind that scaling resistors should be of an adequate voltage rating for any transient, or fault conditions that may be experienced.
To change the scaling, use the following method.
Example, a 0 to 10V signal needs to display 0 to 500 (500 equates to 50mV). Assume Ra is 1M. In our example, Ra must drop 9.95V and therefore the current will be 9.95_A (Rb has very little effect once above 2V range). The value of Rb is therefore 50mV/9.95_A = 5025R. The nearest preferred value would be 5K1. Recalibrate by adjusting the CAL potentiometer on module.
For current measurement simply measure the voltdrop across a resister (shunt). Ra is short circuit and the shunt resistor is fitted in the Rb position.
Care must be taken that differential inputs (INHI, INLO) are not taken beyond their common mode operating range. For a standard meter this means a floating supply (e.g. battery) must be used. An S-type meter can use the same supply and have a common ground connection.
Back to top
Q6. How do I use multiple meters with the same PSU?
A6. First you need to be fully familiar with both your own circuit and the common mode input range of the panel meter(s) in question. Refer to the IC manufacturer's original data, (Maxim, Intersil, TelCom, etc), for full details of common mode input limitations.
Once you have this information to hand you should be in a position to work out the relative levels of each part of the circuit you wish to measure. Keep in mind you no longer have a simple case of an isolated differential measurement, you need to be aware of the absolute levels of the connections you intend to make.
The most frequent mistake made when the negative inputs are commoned together, is to wire all the meters as fully floating types. In such a situation, only one meter's Com connection should be used to establish the reference level for the negative measuring input. Dependent on the module type, you may need to cut links between Com & In Lo on the other meters to achieve this.
(The negative voltage reference connection will still need to be made on each meter. Typically this is Com to Ref Lo to Ref -. Depending on model type and default links made, some or all of these connections may already be in place). Refer also to the Application Notes on our website for further guidance and application examples.
Q7. How do I use 2 meters to read Voltage and Current with a common PSU?
A7-1. Non-S-type LCD meters, see example in Figure 1. Note that this is either powered from an isolated power supply or a battery in the range of 7.5 to 14 volts.
A7-2. LED and S-type LCD meters, see example in Figure 2. This may be powered from the supply being measured and is typically in the range 4.5 to 5.5 volts.
Back to top
Q8. I am using a dc/dc converter to obtain an isolated supply for the meter, but the display is unstable, help?
A8. DC/DC converters typically need a minimum load of 10% to become stable. In addition, because of the converter's high frequency switching, spikes may be present on the supply to the module. The manufacturers will probably detail additional filtering which should improve the situation. Remember that standard requirements such as short leads, decoupling capacitors and ferrite beads may improve the situation.
Q9. Help! I've tried to follow all your suggestions, but I still can't get my meter to work. What do I do now?
A9. Below follows a brief trouble-shooting guide. It would be beneficial to work through these basic checks, before requesting assistance from one of our application engineers.
CAUTION: For brevity we have omitted detailing all procedures in full. However, ensure that any wiring or connection changes are made with the system fully powered down. The same applies to continuity & short circuit checks of wiring etc.
If you have a spare meter, try substituting this one, but if it shows the same symptoms remove it immediately as, if you have a serious wiring or application problem, you can easily destroy your spare unit. All meters are calibrated and subject to at least one full functional check before dispatch, (in fact many are also subject to an extended burn-in), so it is extremely unlikely you could have received two faulty units.
Voltage measurements need to be made with unit(s) in-situ and powered up. We recommend using a quality battery powered digital multimeter, (to ensure reliable, isolated measurements), for most basic checks. However be aware that these tend to average a.c., and you may need access to an oscilloscope, to see the true “picture”. At all times ensure that safe working practices are rigorously adhered to.
In particular beware of hazardous voltages, especially in installations involving multiple meters!
Back to top
Single Meter Application
1. Is the meter the correct model number?
Is there anything visible on the display?
If not, check the wiring and polarity of your power supply.
Is it switched on?
Have you verified the wiring against the latest copy of the Data Sheet?
Is the meter plugged-in the correct way?
Try using a digital multimeter to verify the polarity and supply voltage level to the meter when in-situ.
Is it the correct level for the meter you have installed?
Does this change substantially when the meter is removed?
If so, most likely you either have:
- A serious wiring error. Switch off immediately & check it.
- Damaged the meter. Try swapping it, if you have a spare.
- A poor connection to the power supply. Try removing the meter, and substituting a resistance to represent the load, (e.g. 10K for LCD “s-type”, 50K for non “s-type”, or 27R for LED meters).
If the supply level measured at the connections to the meter still drops substantially when the (dummy) load is applied, this would confirm poor connection(s), dry joint(s) etc. If not, this would point to one of the other two issues above.
If all appears O.K., but there is still no display then the meter has probably been damaged.
Check also that the meter you are using does not have Digital Hold or Display Inhibit, in which case check that the voltage present on the Hold/Inhibit pin does not correspond with the level required to enable this function.
Try also using the (Display) Test pin, (if available).
Finally try removing the meter from the application, and plugging it into a simple test circuit, as described in 7., below.
2. Is there some kind of reading on the display, or is it just random
segments?
If it appears that segments are random, check that:
the meter is plugged in the correct way.
the power supply in particular and ancillary wiring to meter are correct. (see section 1., above).
If readings appear to be valid identifiable numbers, even if unstable, “overrange,” too high, too low, or not appearing to respond to inputs, it at least suggests that the polarity of the power supply is correct.
Make sure that the Power Supply Voltage at the module's connection pins is within the limits specified on the product data sheet, and proceed to the next section of the fault-finding procedure.
If however, nothing obvious comes to light from these checks, try removing the meter from the application, and plugging it into a simple test circuit, as described in 7., below.
3. Check the voltage reference connections. In general for the majority of (standard) Lascar DPMs the reference voltages (VREF HI VREF LO) should be:
- 3.5 Digit types: 100 mV
- 4.5 Digit types: 1 V
Back to top
(Other more specialist types have different reference voltages).
However the key thing here is stability. The reference voltage should be stable and not alter significantly with power supply, input, or temperature variations etc. (Assuming of course all of these remain within the specifications quoted on the relevant module's data sheet).
If you cannot measure any voltage across the voltage reference inputs (generally termed REF HI & REF LO), or this is not stable, then this is almost certainly the cause of your problem.
If the reference voltage (100mV typ) is present and stable, then go to 5., below.
If the reference voltage appears to be around twice its correct level, (characterized often by the DPM reading around half of the correct value), then check that COM has not inadvertently been connected to 0V (DGND etc). Most often this happens when customers try to use the fully floating application circuit for an application where one of the inputs is already referenced to the DPM. (e.g. A single-ended input to a “S-type” meter).
4. Check the level of COM. The Analogue Common connection, on most Lascar DPMs, has a dual purpose. (Although its original function was to provide access to the integrator reference level of the internal A/D converter to avoid Common Mode Input Voltage errors, this is no longer necessarily the case).
However:
- It does serve to set the input and reference levels in fully floating applications.
- Also for many panel meters it forms the default internal reference voltage.
Check the voltage between COM and V+. For most meters and applications, you would expect to see between 2.8 and 3.6 V, between these two pins. (Note: for some applications and some 4.5 digit meters, COM is deliberately pulled down, often to 5V below V+, but not necessarily).
Should you not be able to measure any voltage between V+ & COM, this is indicative of damage to the DPM, or possibly a connection short circuit between these two pins. Switch-off & check immediately.
Assuming COM is at the right sort of level (2.8 to 3.6 V below V+), then check to see if REF LO is also at this level. Also for meters with a REF- output, this again should be connected to the same level, for most applications.
Be aware that it is often difficult to tell if two points are at the same voltage level with a modern digital multimeter, if one is floating, so either switch-off and check for continuity, or refer the measurements to a different known level (e.g. V+).
Assuming COM, REF-, & REF LO, appear correct, verify that the reference voltage between here and REF+ is at the correct level, (as detailed in section 3., above). Now check to see if this same voltage is reaching REF HI. If not, then, dependent upon the meter type, you may need to make a link, or hardwire to the connector.
[Aside: Due to the large variety of DPMs that are produced, both current and legacy types, it is almost impossible to give generalized connection advice, applicable to all types of meters. However starting with the most configurable types, in essence what we are trying to do with the reference circuitry is this:
a. Bias the reference circuit by connecting REF- to COM.
b. Connect REF LO to this reference point (COM).
c. Tap-off the reference voltage from REF+ of the divider chain.
d. Connect this reference voltage (REF+) to REF HI.
Hopefully this will give you the necessary information to verify whether the reference circuitry is connected & functioning correctly.
5. Check the input levels. Try shorting the inputs IN HI and IN LO together, (make sure that at least IN HI is isolated from your application circuit before you try this). Does the meter read zero, (or close to zero)?
Wild fluctuations would suggest that the inputs, although zero'd with respect to each other, may be floating with respect to the DPM. For fully floating applications, typically IN LO needs to be connected to AN COM to establish the correct level.
Other possible causes are noise on the power supply and/or the inputs.
(Try adding a capacitor of, say, 10uF between V+ and Com).
Assuming zero to be O.K., remove the short. Reconnect the module and measure the signal with an external digital multimeter. Is it within the allowable input range of the Lascar DPM?
If not, it will display “overrange”. This is typically a “1” or “-1”, (although on some DPMs this is characterized by flashing zeros or display segments).
Note: it may be that the DPM you have features on-board scaling, or you have scaled this yourself. If the meter zeros correctly but shows overrange with a normal signal, measure this signal. Is it greater than the default full scale reading for the meter?
What happens if you reduce the input to below the default full-scale reading of the meter?
Does it start to respond correctly again?
Most likely in this case, the meter has either not been scaled, or a connection error has been made. Check that the Link (typically denoted link La) across the series-scaling resistor (typically denoted RA) has in fact been broken. Check also that the parallel scaling resistor (typically denoted RB) is not open circuit.
Usually you can trace this signal across the scaling network with your digital multimeter (typical input impedance 10M), in contrast to the recommended scaling impedance for the DPM of 1M.
Back to top
6. Power Supply / Noise Problem?
Assuming none of the above checks have already pin-pointed the cause, check the power supply to the DPM. Is the power supply adequate? (You may need to check this with an oscilloscope for a LED meter) Is there a noise problem?
This is often characterized by the meter reading the correct value most of the time, but occasionally the reading jumping significantly, before resuming a constant level, or just general instability, or sometimes failing to fully zero, when the inputs are shorted.
Try connecting a capacitor (10_F) between V+ and COM.
Next, try disconnecting the power supply to the meter and replacing it with a battery fed supply. If instability disappears, this would indicate that noise is being transmitted through your power supply. Ideally we would recommend that a good quality linear power supply is used.
If the meter starts to respond correctly, having previously been reading incorrect values, this suggests a wiring issue, in that the inputs to the DPM are referenced to your DPM power supply, other than by the DPM itself. Check the wiring, or try replacing the PSU with an alternative fully isolated type.
7. Test the meter in isolation from the intended installation.
This is the recommended procedure when all else fails. In fact, it is often a quicker approach than progressively working through the full fault-finding procedure with the unit in-situ.
Remove the meter from the application. (If you are using socket connectors as opposed to direct soldering, you can simply unplug it.)
Now (again using connection sockets), connect the power supply pins (typically V+ & V- or V+ & 0V), to a battery, (or failing that, a completely isolated linear power source). A PP3 / MN 1604 (9V battery) is ideal for non-“S-type” LCD DPMs.
Referring to the Lascar Data Sheet for the product in question, make the necessary interconnections for the fully floating application mode.
Typically this would be COM to REF- to REF LO, with REF+ to REF HI.
Inputs would be to IN HI & IN LO, whilst IN LO would also be connected to COM (or 0V for some “S-type” meters).
Now try shorting IN HI to IN LO. Does the meter now zero?
Next, apply an input corresponding to half of the meter's full-scale reading.
Typically this will be the same as the voltage across the reference inputs (VREF HI and VREF LO), unless the meter has been scaled. Does the meter read 1000, or close to this value?
Try reversing the inputs. Does the meter display the same magnitude but with a negative sign?
Try also halving and doubling the input voltage levels, and reversing the inputs at each level. Does the meter appear to respond in a linear fashion?
If yes to the above, then this is a good sign that the meter is functioning correctly and the next step is to start substituting it back into the intended application.
If not, and assuming the meter is otherwise unmodified, then this is indicative of a fault. However, it usually worth performing a few basic checks, as outlined in 1. through 5., above, to verify that wiring and other voltage levels are indeed correct.
Having proven the meter's function to be O.K., start substituting connections back to those of your original circuit, in phases, starting with IN HI & IN LO.
By progressively re-introducing first the connections and then the meter back into the intended environment, the nature of the wiring / application fault should become obvious.
8. Transients, etc.
While not strictly part of the basic fault finding process, the subject of transients is worth a quick, or passing, comment. Although this is probably one of the most difficult areas to design-out, here are some suggestions.
- Be conservative with voltage breakdown ratings.
- Try to keep all connection wires short.
- Keep signal inputs away from other circuitry, especially digital, mains, or high frequency signals.
- Decouple power supplies and signals to power supplies.
- Where longer signal connections are necessary, twist the signal wires together, as this will tend to cancel out any noise that may be picked-up.
- Ensure that input voltages cannot exceed power supply rail voltages, even under fault conditions. If necessary, use high-speed diode rail-clamps and series limiting resistors to provide additional protection.
- Do not assume that because a meter appears to function O.K. for a period of time under steady-state conditions, that this verifies the circuit / installation under all conditions.
Back to top
Multiple Meter Application
1. Try physically unplugging all meters, except one. Does this single meter now function correctly?
If so, most likely your circuit has exceeded the common mode input voltage range of one or more of the meters, or you have inadvertently dragged the reference voltage down, (characterized by the meter under-reading the expected value).
Now try plugging the DPMs back in, one at a time, until one or more fail to respond correctly. This should give you a pointer to the problem area.
2. If not, remove this meter also and replace it with one of the other meters. Does this fix the problem?
If so, it is likely that the first unit had connections or on-board links incompatible with the current single-meter test set-up. Try replacing the meters one at a time until one or more cease to read correctly. This should give pointers to the wiring or circuit design that has lead to the problem.
3. If not, then having tried all meters individually, proceed with fault finding as detailed for a single meter in the section above. - Hint: try fault-finding on the meter that has the simplest interface and connections.
4. None of the above is a substitute for proper design & analysis of your circuit/interface to the DPMs, rather it may give you a pointer to a problem, or overlooked aspect of your design.
Back to top
COUNTERS
Q1. I have a sensor with quadrature outputs and I need to count pulses up in one direction and down in another?
A1. Simply connect one output to the count input and the other to the counter's up/down pin. Be aware that your sensor's phase relationship only holds good for given rotation speed ranges. Small and slow movements backwards and forwards can give rise to false readings.
You can contact us at:
Martel Electronics Corporation
PO Box 770
Londonderry NH, 03053
Sales: sales@martelmeters.com
Support: support@martelmeters.com
Phone: 800-821-0023
Fax: 603-434-1653
Thank you,
support@marteltesttools.com
|
