Click here for Blue meter component placement showing actual component values

Click here for Blue meter copper side layout for identifying solder bridges and damaged tracks




So your just-built Dick Smith Electronics ESR meter doesn't work properly!!?

This fault-finding information is generally applicable to the original and Mark 2 Dick Smith ESR meters
and the Anatek Blue meter, except that its PCB is totally different.


Click here if you'd prefer an experienced technician to get your meter working for you.

Click here if you don't have a variable DC power supply to set up VR1.

If your meter's working OK, Click here for hints on using it!


Have you had an accident with your Dick Smith ESR meter?

If you've connected your meter to something like a 100uF electrolytic cap charged to 300V, unfortunately the massive spike has probably destroyed the microcontroller (IC2). In addition, C6 (47uF 50V bipolar) is likely to have been blown largely open circuit, and I'd suggest checking C5 (0.1uF ceramic disc), and D3/D4 (1N4002) which might be leaky or shorted. If the meter still works but sits there displaying some number (maybe '20' or similar) instead of '- ' with the test leads separated, it's possible Q6 (BC338) has been damaged.

If only the right-hand display illuminates with bright random segments, probably the microcontroller (IC2) is dead. If you're in North America, you can get hold of a replacement from Anatek Corp in New Hampshire ( http://www.anatekcorp.com ) or John's Jukes in Vancouver Canada ( http://www.flippers.com ). If you're in Australia, I can help you out. Please click here to go to my contact page.


Meter board diagrams to help faultfinding! (Not applicable to the Blue meter)

Click here to download a diagram of the top of the ESR meter board containing circuit references of all components. A big "thank you!" to Martin Pickering for making this very helpful diagram available.

Click here for a copper side diagram to help identify any solder bridges or track damage etc. Many thanks to Bill De Rose for kindly providing this artwork, and to Jacques Carrier for suggesting it. :-)



First the basics


Incorrect mounting of test lead banana sockets...
This is quite a common problem. If your meter immediately reads well under one ohm after flashing 'EA' at switch-on, check that the plastic insulating rings are correctly installed between the solder lugs and the metal front panel, as shown in this drawing.





Soldering and PCB faults...
I've repaired a lot of ESR meter boards (in person and by remote diagnosis) for people who've had trouble getting them to work correctly, and in about 95% of them the fault has been due to soldering problems.

This is such a common problem that in future I'll be charging $$$ for getting meter boards working when the fault is caused by defective soldering.

The tinned copper tracks on the PCB are very closely spaced in many places, and a common problem has been tiny 'whiskers' of solder bridging the tracks, especially where they pass component solder pads, in particular around Q3, Q4 and Q5 (all BC328) on the left-hand side of the board.

Because the component leads and the PCB pattern are all the same shiny silver colour, sometimes people overlook an unsoldered joint. And of course the traditional cold solder joints can happen! The small solder pads are easily lifted by excessive force on components, and I've seen a fine break in a PCB track due to a manufacturing defect (once).

The way to identify these kinds of problems is to download and print out the PCB copper side diagram (see 'Meter board diagrams to help faultfinding!' section above), then hold the component side of the PCB up to a bright light and carefully compare your board against the diagram, using a magnifier if necessary. Some solder bridges look a lot like part of the PCB pattern!


This is a problem area of the circuit board (not applicable to the Blue meter)
If your meter display is freezing on '.00' or doing other strange things, please check this area of the circuit board where the copper pattern is very closely spaced and quite a few people have accidentally made solder bridges which tend to look like part of the pattern and are easily overlooked.







Misplaced components...
It's easy for your mind to wander when you're performing a boring task like inserting components onto a PCB (I know all about that!), and this can lead to components winding up in the wrong place, or the wrong way around.

Resistors
One of the most common problems is misreading the colour code on resistors!! Under artificial light, especially the fluorescent kind, many resistor colour bands can appear to be a different colour to what they really are, especially in the case of those awful blue-bodied 1% metal film resistors. Also, I'm told that about 10% of males have some degree of colour blindness which can also cause problems. Please check the resistor placement under natural light (and ask your wife/girlfriend/daughter etc to double check!) if you have any doubts.

Diodes
Please make sure that the correct type diodes are in their correct positions, and the right way around. The 1N914 / 1N4148 diodes are tiny little glass ones, while the 1N4004 type are the larger black-bodied kind, with much thicker leads.

Transistors
There are four different transistor types in the meter, and it's easy to mix them up. Also it's possible to insert their leads into the wrong holes- check to make sure the flat part of the transistors is facing in the right direction.


Specific faults, and what to check...


Abnormal Display

Right-hand display shows weird character (and right-hand decimal point LED might be on), but only while button is pushed: The microcontroller is probably not running... check around XTAL1 and C11 / C12 (both 27pF).

Both displays are working, but show weird characters: Almost certainly a short between PCB tracks, probably around the pins on the 28 pin IC socket holding the displays, or around IC3 (4094).

Some or all segments of one display don't light: Try swapping the two displays... if the fault moves too, you have a faulty display. In this case, please e-mail me and I'll arrange to get you a replacement. While you're waiting, you can probably get by with another common-cathode 0.5 inch display such as an NTE3079, though it will probably be quite dim.

Some incorrect segments light dimly on the right-hand display: Q12 (BC338) might be low in gain, or R28 (4.7k) is the wrong value.

Right-hand display shows '8' then '-' at switch-on, and left-hand display never lights: Q12 (BC338) or one of its connections is probably open circuit.

Meter reads '.56' when connected to the 5.6 ohm test resistor: LED2 (the right-hand one) is probably installed backwards, or has a bad connection in the socket or on the circuit board.

Decimal point LEDs don't light up: The LEDs could be installed backwards. Check their orientation against what's shown in the kit notes.



Power Control Problems

Meter won't stay switched on after releasing the button: The fault is in the area of Q2 (BC338). Check for solder bridges, bad joints, wrong orientation of Q2, and that R3 = 15k and R29 = 2.7k.

Meter switches off instead of zeroing: It will do this if the test lead circuit resistance is higher than 0.99 ohms. Check that the nuts on the banana sockets are tight, and that the banana sockets and plugs aren't oxidized. A solid squirt of CRC Contact Cleaner on the plugs then some serious jiggling in their sockets can help a lot.

Meter is permanently on, and displays '??' after the button is pushed: Check for solder bridges in the area of Q1 (BC328), Q2 (BC338), R1 (10k) and R2 (4.7k). Also check that Q1 and Q2 are the correct types.

At switch-on, meter permanently displays 'EA' and cannot be switched off, and shorting the test leads has no effect: There's probably a problem around Q9, Q10 (both BC558), Q11 (BC338), R19, R20, R21 (all 10k), R22 (470k), or C10 (0.47uF). Possibly a solder bridge in the area, or a wrong-type transistor in Q9, Q10 or Q11 position.



Incorrect Readings and Display 'Freezing'

Meter will read some resistors correctly, but gives absurd readings on others: This is a common fault! Carefully check the values of R6 (10k 1%), R8 (1k 1%) and R10 (100 ohm 1%). In the original ESR meter kits, Dick Smith Electronics was supplying an 82 ohm 1% and a 5.6 ohm 1% resistor in the kits to help with calibration.... the 82 ohm 1% looks very much like a 1k 1% under artificial light, and a lot of constructors have accidentally installed the 82 ohm 1% resistor as R8. [The Mark 2 meter kit has a 68 ohm resistor instead of 82 ohm to stop this problem]. If that's not the problem, check for solder bridges in the area of Q3, Q4 and Q5 (all BC328), where the PCB tracks are very closely spaced (and where a lot of constructors have made solder bridges).

Meter shows '20' or similar with test leads separated: This is often a symptom of Q6 (BC338) being defective, or a problem in that part of the circuit such as a solder whisker between its collector and base leads.

Meter shows '90' or similar with test leads separated and self-test function gives "F6" fault code: These are the typical symptoms of trimpots VR1 (10k) and VR2 (200 ohm) having been swapped.

Meter is inaccurate measuring all resistors: Check that R16 (180 ohm), R17 (6.8k) and R18 (680 ohm) are all in their correct positions. If you're also having difficulty in setting up the low battery voltage warning, make sure you haven't accidentally swapped trimpots VR1 (10k) and VR2 (200 ohm).

Varying readings and switching off instead of zeroing: This might not be a fault in the meter itself at all. As Doug MacDonald found out the hard way, the wire could have become loose or detached inside one of the test probes and/or banana plugs.

Meter permanently reads '.00' and won't switch off: First make sure you installed the banana sockets with their insulating rings between the mounting nuts and the front panel... otherwise the test leads will be shorted to the metal front panel! If that's not the problem, there's probably something wrong in the pulse amplifier circuit consisting of Q7 (BC548) and Q8 (BC558) and their surrounding components. Check that Q7 and Q8 are the correct types and the right way around, and that D5 and D6 (both 1N914 / 1N4148) are the right way around. Also check carefully for solder bridges and bad joints. You should measure about +0.6V on pin 8 of IC2 (Z86E0408 / Z86E0412); if you don't, then there is definitely a fault in this part of the circuit.

Meter flashes 'EA' then '- ' for a moment, then 'freezes' displaying '.00' and won't switch off: Possibly a problem in the area of Q11 (BC338), e.g. misoriented leads, solder whisker.

Meter shows 'EA' until momentarily connecting the test leads, then freezes on a very low reading: Look for an open circuit or solder whisker around R5 (2.2k), R6 (10k 1%) and Q3 (BC328). The meter board where I saw this fault had a detached solder pad on R5.

Meter flashes 'EA' then '- ' but 'freezes' displaying a low value after test leads are momentarily connected: Check for solder whiskers, fractured PCB tracks or other problems around R7 (2.2k), R8 (1k 1%) and Q4 (BC328). If it's displaying '.00', see next paragraph.

Meter flashes 'EA' then '- ' but 'freezes' displaying '.00' after test leads are momentarily connected: Check for any solder whiskers, fractured PCB tracks or other problems around R9 (2.2k), R10 (100 ohm 1%) and Q5 (BC328). In one case where a meter was doing this, it turned out that R10 measured about 500k ohms!

ESR meter comes on with 2 zeroes, no decimal points.: Check that R11 and R25 (220 ohm and 220 kohm) have not been interchanged. (Thanks Kevin, of Kevin's TV service).

After being connected to a large charged capacitor, meter reads low especially on the 0 to 0.99 ohm range: It's possible the large current pulse has damaged C6 (47uF 50V bipolar) and blown it partially open circuit. If you don't have another ESR meter to check it with, you could buy one ;-) ..... or temporarily short out C6 and see if the readings return to normal.

When you short the test leads to zero the meter, you get a reading of maybe 2 ohms which varies wildly as you move the leads: If you bought your kit from John's Jukes and installed the protective MOV device across the test leads, this is probably causing the problem. Disconnect it and see if the meter works normally. If so, you'll need to install a pair of 1N5404 or similar power diodes in inverse parallel where the MOV was, to give added charged capacitor protection.


Getting technical...

Here are some important waveforms taken from a correctly-working meter
with a x10 (10 megohm impedance) probe and the test leads open-circuited.



IC2 pin 10

This is the reference voltage waveform, critical to measuring ESR/resistance and battery voltage. It's a slow signal which must be measured with a 10 megohm probe, and will tend to look like a jumping dot on an analog 'scope. If it's different to this, look for a fault around Q9, Q10 (both BC558), Q11 (BC338), and their surrounding components.









IC2 pin 8

This is the output voltage from the pulse amplifier, and should be seen as several bursts of pulses per second. It's important that its 'base' voltage is close to +0.6V, and the peak voltage is close to +5V as shown here. If it's different to this, look for a fault around Q7 (BC548) and Q8 (BC558) and their surrounding components.









Waveform on test leads

This is the voltage you should see on the open-circuited test leads, occuring as several bursts per second. If the peak voltage is much different from the +500mV as shown here, look for a problem in the area of D3, D4 (both 1N4002 or similar), Q3 (BC328) etc.











IC2 pin 15

If the meter's operating more or less correctly, the microcontroller will go to the top (10 to 99 ohm) range and be driving Q3 (BC328) with this waveform. If it's absent but you see this signal on either pin 16 or 17 instead, then it probably indicates a fault (such as a wrong resistor or solder bridge) around Q3 or Q4 (both BC328).









Found a faulty transistor??

It's possible your testing has revealed a defective transistor in your kit; it's rare but it has happened. If you live in the UK or Europe, you shouldn't have any trouble finding a replacement.... but if you're in North America you'll find them hard to get. Mouser Electronics sells them cheaply (thanks to TVSteve for this information), otherwise they can be replaced by types readily available in the US:

The common 2N3904 will replace the BC338 and BC548, and the 2N3906 will replace the BC328 and BC558.... but note that the 'flat' on the 2N3904/6 will face in the opposite direction to the 'flat' on the 'BCxxx' it's replacing!!




Phew!! I hope this page has enabled you to find and correct the problem(s) in your ESR meter, and that it's working OK now. If you're still stuck, I strongly suggest that you ask another person with electronic experience to check for anything abnormal like misplaced components or soldering errors. It's amazing how someone else will often spot a problem immediately.

If this doesn't help, send me an e-mail telling me what the trouble is, and we'll get it working one way or another. It will save time if you give me as much detail as possible and the results of any measurements you've done. :-)


- OR -

Ben Cook in Western Australia is set up to give people a helping hand to get their kits working. He only asks that owners pay for postage, any components needed and a small charge for his running around. You can e-mail him at benok@iprimus.com.au.






Back to the home page...
Back to the ESR meter page...




Last update: 6th March, 2011