Identifying a board component out of Clare A433

[Guineafowl]

This board is out of an appliance test station (Clare A433). The unit tests the Earth circuit, insulation resistance, load current and also has a HV flash probe to check for flashover from live parts to metal casings.

The insulation resistance function is faulty - it’s meant to apply 1 kV ac between L+N and E (ground) and measure the leakage.

I’ve narrowed the problem down to this board:UIKEYINPUTRIGHTARROW

The component PD1 is open circuit. Here are some details of the numbers, but they are hard to read:

Is PD1 a power diode? Do the (obscure) numbers ring a bell? The best I can make out is CP2060A.

Many thanks.

 

[NascentOxygen]

You say PD1 is o/c. Did you establish this using a multimeter set to kilohms? If so, it might be that the voltage the meter applies is too small to cause a perfectly good diode to conduct, and leading to a wrong diagnosis.

 

[Asymptotic]

Does PD1 have the dimensions of a DO-27 (or other standard) device form factor? There isn't a polarity band, but perhaps (if it is a power diode) the manufacturer placed the part number text at the extreme end to serve this purpose. CP2060A isn't ringing any bells, but insofar as the circuit board is a 'Clare' it may be an in-house manufacturer's mark ("CP" for CP Clare).

 

[Guineafowl]

You say PD1 is o/c. Did you establish this using a multimeter set to kilohms? If so, it might be that the voltage the meter applies is too small to cause a perfectly good diode to conduct, and leading to a wrong diagnosis.

I tried the diode check function and resistance mode both ways. Auto-ranging meter.

Does PD1 have the dimensions of a DO-27 (or other standard) device form factor? There isn't a polarity band, but perhaps (if it is a power diode) the manufacturer placed the part number text at the extreme end to serve this purpose. CP2060A isn't ringing any bells, but insofar as the circuit board is a 'Clare' it may be an in-house manufacturer's mark ("CP" for CP Clare).

It seems to be like a DO-14. Looking at the pictures, I wonder if it’s upside-down - you can almost read VD907AD

Thanks to both for your input, by the way.

 

[Borek]

I wonder if the symbols printed on PCB don't tell us something. Note they are different under resistors and diodes (and shape of the ones under diodes can mark polarity).

 

[Guineafowl]

I wonder if the symbols printed on PCB don't tell us something. Note they are different under resistors and diodes (and shape of the ones under diodes can mark polarity).

You’re right - the diodes are marked with a form of arrow, as on a schematic. If only there was as much care taken over printing the component numbers!

 

[wirenut]

You’re right - the diodes are marked with a form of arrow, as on a schematic. If only there was as much care taken over printing the component numbers!

I don't know if this helps, but there is a board for sale on erazushop.top (http://www.erazushop.top/other-test...pment-flash-test-unit-a203-0352912-p-7747.htm) for 26.00 GBP

 

[Tom.G]

Nice photos!
How about some of the other side of the board? One of the whole board and one of the region around PD1 and nearby components.
That may give us some clues about the function of PD1 and needed characteristics.
It doesn't look like a multilayer board but if it is, some shots with the board backlit to show buried layers would be helpful.

Cheers,
Tom

 

[jim hardy]

It doesn't look like a multilayer board but if it is, some shots with the board backlit to show buried layers would be helpful.

Yes, indeed. With just a little imagination i can see PD1 connected across 7 and 8

suggesting it's perhaps some sort of avalanche overvoltage suppressor .??

Is my guess about 7&8 right ?

old jim

 

[Guineafowl]

Wirenut - yes, I saw that very one while searching the board number. Nice to know there’s a spare about...

Tom.G and Jim Hardy - Been rather busy of late, but I’ll update soon. A brief test recently showed the unit works (I think) as normal now. It could have been a simple loose connection, then. I’d still like to get to the bottom of that component, though, so more photos and details to follow...

An interesting device and an ebay bargain - just £22. The Clare A433 is a ‘comprehensive test station’ for periodic, post-repair and type testing of electrical appliances, both 110V and 240V. Made c. 1988.

Since there are frustratingly few details online, I’ve looked over it with my meter and cobbled together a sketch of its features:

(L = live, active, hot, phase; N = neutral; E, Earth = ground.)
1. 500mV ac continuity test - for verifying Earth pin connection to casing?

2. Earth bond test - 8V ac and up to 36A through the Earth path. Measures down to 0.05 ohm. Heavy current ensures any feeble connections or wispy wires are uncovered.

3. L+N to probe 0-5000V ac - checks for flashover between live conductors and any metal parts, which are probed with the ‘flash prod’. Voltage variable with what looks like a variac-type rotary brush contact. Trips (with buzzer alarm and neon lamp) at a selectable 5, 10 or 15mA level. I will use this with care, since 5000V ac at 15 mA sounds rather lethal.

4. L+N to E 0-5000V ac - insulation resistance. Measures leakage to Earth at a suggested setting of 1 kV. Scale in megohms and leakage in mA.

5. Load test - runs the appliance and measures current, scale 15 or 1.5 A.

 

[Guineafowl]

Hi all,

Having said the machine works, I’m now not so sure. Here it is, on the bench:

The third meter is labelled “Megohms at 1 kV ac”. The top scale runs from inf to 0.4 (Mohm) non-linearly, and the bottom runs from 0 to 2.5mA linearly.

This, I think, is measuring the insulation resistance between L+N and Earth for an appliance plugged into the test socket (off photo, to right). You press the button marked “FLASH...IR”, dial the second meter scale to 1 kV with the black knob, and read off the insulation resistance value from the third meter.

However, I can’t get the readings correct. If I measure the voltage between L or N and Earth with my multimeter, it reads 1000V ac, and the scale correctly reads the 10 M ohm internal impedance of the multimeter.

But putting, say, two 1 megohm resistors between L or N and Earth gives a reading of 1 Mohm. Three resistors gives 1.5 Mohm. It’s always half. If I twiddle the meter cal pot shown here to the right of the lower blue cap:

... I can alter the scale reading, but that throws off the 10 Mohm reading from the multimeter.

Also, with only 1 Mohm resistor between L or N and earth, the machine trips, suggesting a leakage current >5mA, when at 1 kV the leakage should be 1mA. From this, I assume there’s a problem with the measurement circuitry, rather than the meter itself. The very same resistor is measured correctly by my other 1 kV (DC) insulation tester. Is this an AC/DC thing? Peak voltage at 1 kV is 1414 kV. This doesn’t give enough extra current for the trip, but could the resistor be breaking down? They are standard 1/4W.

Another problem is the “set inf” pot under the meter is very jumpy - when you turn it either way, the needle jumps up before settling. Could this be a clue to the fault?

In summary: The voltage applied is measured correctly (1 kV), but the current measured and resistance calculated is off for values below 10 Mohm. The meter control board shown above tests OK on a component level, but the tripping is handled by the board which was the original subject of this thread. Here is the backside:

Legs of the suspect component (mystery diode) marked with some chalk. The frontside:

... has a meter cal pot for the kV scale (working), and a trip cal pot. The tripping at 1mA fault is out of range of the pot.

A long-winded post, for which apologies. Sadly there is no schematic, service manual or even user manual online.

 

[NTL2009]

However, I can’t get the readings correct. If I measure the voltage between L or N and Earth with my multimeter, it reads 1000V ac, and the scale correctly reads the 10 M ohm internal impedance of the multimeter.

But putting, say, two 1 megohm resistors between L or N and Earth gives a reading of 1 Mohm. Three resistors gives 1.5 Mohm. It’s always half. If I twiddle the meter cal pot shown here to the right of the lower blue cap:

...

Also, with only 1 Mohm resistor between L or N and earth, the machine trips, suggesting a leakage current >5mA, when at 1 kV the leakage should be 1mA. From this, I assume there’s a problem with the measurement circuitry, rather than the meter itself. The very same resistor is measured correctly by my other 1 kV (DC) insulation tester. Is this an AC/DC thing? Peak voltage at 1 kV is 1414 kV. This doesn’t give enough extra current for the trip, but could the resistor be breaking down? They are standard 1/4W.

Yes, you are probably exceeding the resistor in two ways. First, power. P=E^2/R. So that is 1 W for 1000V and 1 M Ohm. You are also probably exceeding the voltage rating for the resistor, which can happen before the power rating on high R values.

The fact that it reads the 10 M of your meter, which is (I assume) designed for that voltage, tells me it is working (mostly) correctly.

 

[Guineafowl]

I think I’ve sorted it. For a start, I dismantled and cleaned up the three contactors that run the various functions.

A very important rule of troubleshooting is to check, check and check again. Two of the 1M resistors, which I had measured with multimeter and with a 1000V dc insulation checker, had broken down to about 90 k or so.

Replacing these gave proper readings again. I ‘calibrated’ the trip by measuring the leakage current, with resistance values just either side of a tripping value:

At 1000V, for 5 mA a resistance of R=V/I = 1000/0.005 = 200,000 ohm. Using a 180k gave a trip, but 220k didn’t. I fine-tuned the pot which I had fiddled with by simultaneously measuring the leakage current.

This still leaves the mystery of PD1, but I guess Jim is right that it’s some sort of transient supressor. If I had a spare I could test the breakdown voltage, I suppose. Many thanks for all the replies.

 

[Baluncore]

PD = Protection Diode.
Voltage protection across the supply should be high resistance below the protection voltage.

PDs come with both symmetric and asymmetric voltage characteristics. Protection diodes tend to fail short circuit, while ZnO varistors fail open circuit and so do not inherently stop the circuit functioning.

 

[Guineafowl]

PD = Protection Diode.
Voltage protection across the supply should be high resistance below the protection voltage.

PDs come with both symmetric and asymmetric voltage characteristics. Protection diodes tend to fail short circuit, while ZnO varistors fail open circuit and so do not inherently stop the circuit functioning.

I’ve never heard of them - are they made with a very high forward voltage? What about the reverse voltage? The tendency to fail short circuit suggests some sort of crowbar.

 

[Baluncore]

I’ve never heard of them - are they made with a very high forward voltage? What about the reverse voltage? The tendency to fail short circuit suggests some sort of crowbar.

https://en.wikipedia.org/wiki/Transient-voltage-suppression_diode

When a point contact signal diode fails it blows the contact whisker and goes open circuit.
When a power or protection diode fails it becomes a metal conductive lump.
A protection diode is not a deliberate crowbar. The operating mode is different to the failure mode.

Protection diodes can be designed for symmetrical AC or asymmetric DC signals.

A ZnO varistor is gradually consumed when it conducts, so it has a lifetime limited by the total energy suppressed.
A protection diode is not consumed by conduction.