{"id":134,"date":"2025-12-26T11:22:33","date_gmt":"2025-12-26T11:22:33","guid":{"rendered":"https:\/\/spec-electronic.pl\/?p=134"},"modified":"2026-01-07T17:10:49","modified_gmt":"2026-01-07T17:10:49","slug":"measurement-of-pcba-cleanliness","status":"publish","type":"post","link":"https:\/\/spec-electronic.pl\/index.php\/2025\/12\/26\/measurement-of-pcba-cleanliness\/","title":{"rendered":"Measurement of PCBA cleanliness"},"content":{"rendered":"\n

The device was created in my spare time during the COVID\u201119 period. You know those moments when you sit at your computer, lose track of time, and your mind starts wandering? That happened to me too. Eventually my thoughts landed on a simple question: <\/strong>How clean are PCBs coming from an EMS after assembly?<\/strong><\/em> (And if you think that\u2019s a strange question \u2014 well, I\u2019m the kind of person who gets curious about electronics, not cheese production.)<\/strong><\/p>\n\n\n\n

In my work, long\u2011term and flawless PCBA operation is a major concern \u2014 especially when failures appear without a clear explanation, such as issues caused by contamination or humidity. One part of this challenge is understanding how clean a PCBA really is when it comes straight from the factory.<\/p>\n\n\n\n

A \u201creal expert\u201d might simply define an upper limit for PCB surface conductivity in mS\/in\u00b2 (or \u00b5S\/cm\u00b2) and call it a day. But in practice, measuring this value is not that simple. There are specialized machines that can do it \u2014 you submerge part of the PCBA, or the entire board, in ultra\u2011pure water and measure the conductivity of the fluid.<\/p>\n\n\n\n

This leads to the real question I wanted to solve: how can you measure extremely high resistance values without relying on expensive equipment?<\/p>\n\n\n\n

Means to measure current<\/strong><\/p>\n\n\n\n

\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 T<\/strong>he first method is straightforward: measure the voltage drop across a shunt resistor. Since the voltage drop across a known resistor is proportional to the current, Ohm\u2019s law allows you to calculate the resistance of the circuit.<\/p>\n\n\n\n

The second method is to measure the magnetic field generated by current flowing through a conductor, using a Hall\u2011effect sensor or a current transducer. For Hall\u2011effect devices, Allegro offers excellent compact sensors such as the ACS750. For current transducers, companies like LEM provide reliable solutions.<\/p>\n\n\n\n

Today\u2019s microcontrollers are inexpensive and highly accurate, which means we can measure time very precisely. If the measured time correlates with current \u2014 for example, by observing how long it takes to charge a capacitor \u2014 we can also determine the resistance of the circuit. There are many demonstrations of this technique using discrete components, and they illustrate the concept well.<\/strong><\/p>\n\n\n\n

The last method I\u2019ll mention is the transimpedance approach (TIA). It is similar to the shunt method, but instead of measuring the voltage drop across a physical resistor, the \u201cresistor\u201d is an operational amplifier configured as a current\u2011to\u2011voltage converter. In practice, this behaves like measuring across a resistor that is effectively approaching 0.00…0 \u03a9.<\/p>\n\n\n\n

So naturally, I decided to start with the most exotic method.<\/strong><\/p>\n\n\n\n

The TIA approach appeals to me because it is complex, challenging, and time\u2011consuming \u2014 in the best possible way. Another reason I chose it is that it eliminates the need for a high\u2011voltage source. Measuring something like 1 G\u03a9 often requires applying several kilovolts, which is dangerous and can easily cause arcing between PCB traces. In my application, I don\u2019t need the exact resistance value; I only need to verify that the circuit can withstand a specific test voltage. High\u2011voltage methods are great for underground power cables, but not for delicate PCB assemblies.<\/p>\n\n\n\n

In my design, the test voltage is only \u00b12.5 V, which is perfectly safe \u2014 but it also makes the measurement extremely sensitive to external noise and EOS\/ESD effects. Remember, we are trying to detect signals in the picovolt range. Yes, picovolts.<\/p>\n\n\n\n

What to build<\/strong><\/h2>\n\n\n\n

When researching possible solutions, I found many interesting high\u2011end circuits online \u2014 some of them brilliant, but many relying on components that are practically unobtainable. My first choice was the ADA4530\u20111<\/a> from Analog Devices, but COVID\u2011era shortages made it impossible to buy. Even today, availability is limited and the price is far too high for a DIY project (around 25 USD per chip).<\/p>\n\n\n\n

A more realistic option came from Texas Instruments: the LMP7721.<\/a> It costs roughly one\u2011quarter of the ADA4530\u20111 and is actually available.<\/p>\n\n\n\n

Datasheets for these components are a goldmine of knowledge. The op\u2011amp I selected includes excellent explanations and application notes, which significantly reduced the design effort. Of course, the lower\u2011cost part has its drawbacks \u2014 for example, it lacks the built\u2011in guard\u2011drive circuitry found in the ADA4530\u20111. This means the guard signal must be generated using an external op\u2011amp. (The guard signal is a special potential used to block noise from external sources.)<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

10M resistor that is above first from left op amp is input for my application ( in real schematic there is jumper so you can switch it from 10M\u03a9 to 1G\u03a9 \u2013 it suppose to give you 2 scales one form +\/- 30pA and second +\/- 3nA ).<\/p>\n\n\n\n

Values that are on simulation diagram and voltage source for input 1 AA battery (1,55V) current that flow at input is 40,7pA \u2013 this value is translated to 40,7mV at the end \u2013 my DMM can easily read this value up to 3 decimal places so its good for me.<\/p>\n\n\n\n

Key component in this simulation is precision for resistor and capacitor in this stage. Also that is not\u00a0 obvious \u2013 mechanical board for this device. Input stage have to be isolated from rest of PCB by slots \/ milled cuts around op amp input. Those yellow ovals around trace.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Last but not least \u2014 the wires used to connect the test samples. The LMP datasheet notes that for very low\u2011current applications, a special triaxial cable should be used. This type of cable consists of an inner signal conductor, a surrounding guard conductor, and an outer grounded shield. Unfortunately, triaxial cable is expensive, and the connectors are also costly and difficult to source. Roughly speaking, one meter of this cable can cost around 150 USD \u2014 definitely not ideal for an average DIY project.<\/p>\n\n\n\n

Verification circuit<\/strong><\/p>\n\n\n\n

            Real life measurements started from checking circuit, cleaning PCB fev time in IPA and ultrasonic bath. Next –  noise of system by it self(PCB was closed in reused metal cookie box with one hole for BNC output cable). For \u201cstandard reference\u201d I chose few high resistant components from ROHM (1G\u03a9).<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Small table was created in Excell and input of device was closed by end cap for SMA connector. Power up and 5 min of gathering data (my DMM have math functions like min max and avg). As you can see we have small offset (5,9mV) from beginning but we can resolve this issue later.\u00a0 DMM was set in DC range.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Last position in table can be explained by standing up from desk during measurement. Theoretically device is shielded from environment but DMM is grounded so some electrostatic field was created and detected.<\/p>\n\n\n\n

So next step \u2013 checking by \u201cstandards\u201d 3x brand new 1G\u03a9 1% and one 10G\u03a9 2% from old soviet union (1980?). Values are display below.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Resistors from 1 to 3 supplied by one AA that have 1,5539V \u2013 so there is deviation form theoretical 1,5539nA to little bit hig \u2013 1,57nA it mean ~1,22% off. For first run it great result. For vintage 10G\u03a9 result was not so great from expected 153,9pA my DMM read 183,86pA. Desition was to not take to account this value because component is old and I cant confirm its true value.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n
\"\"<\/figure>\n\n\n\n

After adjustment of output offset second measurement for the same sample was perform. Values are in table below. As you can see there are very close to expected simulated values. Deviation from calculated values is ~0,63% <\/p>\n\n\n\n

In the end some useful links and key words that is nice to google like .:<\/p>\n\n\n\n

SIR \u2013 surface resistance RV \u2013 volume resistance<\/p>\n\n\n\n

https:\/\/www.circuitinsight.com\/pdf\/understanding_sir_ipc.pdf<\/a><\/p>\n\n\n\n

https:\/\/eprintspublications.npl.co.uk\/571\/1\/CMMT48.pdf<\/a><\/p>\n\n\n\n

https:\/\/www.researchgate.net\/publication\/317416157_The_Effect_of_FR-4_Laminate_Materials_on_the_Surface_Insulation_Resistance_of_Wave_Soldering_Fluxes\/link\/5939cf0f458515320630a778\/download<\/a><\/p>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

The device was created in my spare time during the COVID\u201119 period. You know those moments when you sit at your computer, lose track of time, and your mind starts wandering? That happened to me too. Eventually my thoughts landed on a simple question: How clean are PCBs coming from an EMS after assembly? (And …<\/p>\n

Measurement of PCBA cleanliness<\/span>Read More »<\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-134","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/posts\/134","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/comments?post=134"}],"version-history":[{"count":1,"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/posts\/134\/revisions"}],"predecessor-version":[{"id":143,"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/posts\/134\/revisions\/143"}],"wp:attachment":[{"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/media?parent=134"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/categories?post=134"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/spec-electronic.pl\/index.php\/wp-json\/wp\/v2\/tags?post=134"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}