Insulation testing is designed to measure how much ‘leakage’ of voltage is present between two points. Typically this is done on systems with high voltage present, such as a BEV, where you need to ensure that insulation between the high voltage source and chassis ground has adequate resistance. A test voltage which is the same or higher than the vehicles high voltage battery is used to test the insulation. For example if a vehicle has a 400V battery but the meter has the ranges 125V, 250V, 500V and 1000V, you would select 500V using the RANGE button on the Pico meter.
There will always be some leakage but you are looking for the value to be as high as possible. From the ECER100 regulations there must be a minimum of 500Ω for every 1V when a system has both AC and DC voltage present in the system. Therefore for a vehicle fitted with a 400V there must be at least 200,000Ω, 200kΩ, of resistance between the high voltage source and the chassis. As long as this is over the minimum resistance then it is deemed as acceptable. You will find though that vehicle manufacturers specify different results for example, Toyota look for values of 1MΩ or higher.
A number of factors can cause the result to alter. Resistance can alter based on temperature and humidity which will explain why taking a measurement on the same component at different times of the day may give different results. Some tooling may have built in temperature compensation and may give more stable readings but this doesn’t mean that those without temperature compensation are any less accurate.
How long you keep the applied voltage connected to the component under test can also affect the results. Generally speaking if the insulation is good, then the longer the test is carried out the resistance should continue to rise. VM’s may have a set time they wish the test to be carried out and it’s always best to check the technical information first. If you look on the internet for insulation testing times you will usually see results from the industrial sector stating measurement times anywhere from 30s to 10mins. There are a number of tests and analyses that can be calculated from these long measurements such as dielectric absorption ratio, amongst others, seen here - https://electrical-engineering-portal.c ... ance-tests.
For automotive though where the VM hasn’t specified a test time, we then observe the rate at which the resistance values change. If you see that the measurement shoots up to above the minimum resistance and passes 1MΩ, then my opinion that would be that the insulation is good. For a reading I would continue to hold the test till the reading starts to settle and use that result as my measured value. There isn’t really a set time for this but the benefit of the Pico insulation meter is that you are in control of how long the test is carried out.
You can also lock the meter so you don’t have to continue holding the test button. To do this, switch the meter on by rotating the dial to insulation. Select the test voltage by pressing RANGE and then press and hold the MODE button which has the word LOCK above it for around 3 seconds. On release you will now see the word LOCK and a padlock has appeared on the LCD screen. When you now press the insulation test button, it will continue to apply the test voltage to the component without having to hold the test button down. Once the measurement has settled take a note of the measurement or if possible press HOLD which will keep the result on the display. Steve Smith does a great job of demonstrating the insulation tester in the following video - https://youtu.be/9iq5y9y39iY/. This video is also available through the EV guided test in PS7 Automotive.
We have to be mindful of the amount of power required to generate the high voltage required for the test. Bear in mind that the Pico insulation meter is battery powered and prolonged testing at high voltages can cause the batteries to drain rapidly. You will notice this by keeping an eye on the test voltage in the upper right corner of the LCD screen. If you have set the voltage to say 500V but the applied voltage is less, then it could be the meter isn’t able to maintain this voltage. If you’re seeing the test voltage drop, it could be a sign that the batteries need replacing and will of course change the measurement result. This can also cause discrepancies between tools. One meter manufacturer state that the battery voltage should be at 75% or above before performing the test. Whilst the Pico meter doesn’t give you actual battery level, there is a low battery warning.
Some insulation tips that we picked up from a high voltage course at ZF involved testing a motor at the inverter. Typically we will take the red test lead on insulation positive output of the meter and apply that to the high voltage part and then use the black lead on insulation negative to apply to the component housing or chassis ground as per the image below.
The problem here is that the test voltage from the meter being pushed through the HV+ will not get to the windings in the motor as the IGBT’s will be open, vehicle off, and the diodes will block the path. Therefore, it’s essential that you repeat the test on HV- or switch the test leads around and apply the red insulation positive to the motor housing and then the black insulation negative to the HV+. If the motor is suspected then I would encourage this test.
I hope this helps in some way.
Having a battery of 400V and doing insulation testing at 500V. Without doubt, with any system is subjected to an insulation test, makers have taken into account some extra margin, however, never in my career of more than 20 years, we ever tested systems with a voltage higher than the maximum operating voltage. I'm only talking about d.c. "megger" test, as if you go a.c. testing, we go in a different territory.
Cold or hot testing. Well, this one is obvious, one would want to perform an insulation test while Equipment Under Test is in hot conditions, as this will obvious catch the insulation in its worst condition and will give more down to earth values.
R100 allows the insulation resistance to be demonstrated by calculations, I'm pretty convinced that VMs have this already build in, for example calculating inverter output current against battery current, or such similar comparison. Any leakage would be pretty obvious. I would be surprised if in the long list of possible DTCs, this fault code would not be an possibility.
I would be extremely cautious to just hook up the test leads and go for it, without having an actual understanding of how stuff is made and connected. Imagine subjecting a control unit with reverse polarity for 500V. Hopefully they're reverse polarity protected, but who knows!
Another important safety aspect, also indicated in the manual (but a bit in a wrong way) and that cannot be stressed enough, after the insulation test, do not immediately remove the test leads from EUT, let the energy accumulated in the EUT to discharge through the meter's circuits. I know a few chaps getting some nasty shocks due to this.
Interesting subject. But will be more interesting when time will come for fuel cells. As their insulation resistance is just about 0 ohms by nature.
Not a problem and of course open up a discussion. This is what the forum is all about. My understanding may be different from someone else’s and we all know it’s impossible to know and retain everything. We should be questioning and challenging as this is what keeps us moving forwards.
When carrying out an insulation measurement we are testing for a breakdown in insulation, or isolation, between a high voltage circuit and the vehicle chassis and/or the housing of the component under test. To do this we should match the voltage the component would be subjected too. Many insulation testers come with predetermined test voltages and so if we can’t match the voltage we must go higher as going lower would not subject it to the same voltage during normal operation.
Within the ECER100 regulations we find the following passage -
“1.1.1. Measurement instrument
An isolation resistance test instrument capable of applying a DC voltage
higher than the nominal voltage of the tested-device shall be used.
1.1.2. Measurement method
An insulation resistance test instrument shall be connected between the
live parts and the ground connection. Then, the isolation resistance shall
This can be found in the document here -
https://wiki.unece.org/download/attachm ... pdf?api=v2
We can find further information on the Fluke website with the following quote,
“The idea behind using an insulation tool is to verify the coating on the outside of the wire has not lost integrity and shorted to ground or another wire.
A common question when using an Insulation tester is what Voltage level to use. The general rule of thumb is whatever the circuit or motor is rated for you double that voltage when using the Insulation tester. Example if the system is 480V you would set the Insulation tester to 1000V.”
https://help.fluke.com/hc/en-us/article ... on-Tester-
In all my years of working around high voltage vehicles, I have never been asked or instructed to carry out an insulation test while the vehicle is still live. Insulation measurements would always be done after the vehicle was safely powered down and the battery isolated from the rest of the high voltage circuit. This is typically done by verifying the absence of voltage at a suitable test point after the main contactors. For household electrical work it does appear that you can do insulation testing on live conductors which involves joining the live and neutral together. If you did this on a high voltage battery you would have a dead short. If you have an example where this is a measurement you have been asked to do then please let us know. As I said, we don’t know everything and if there is an application where an insulation measurement is required to be taken on a live high voltage circuit then I would like to know more.
You are quite right that the ECR100 regs state that insulation can be done with calculations and most VM’s will have a reading through the serial data for this measurement. That being said I would always want to qualify and validate any measurement that is being calculated by an ECU. As you’ve mentioned, DTC’s will typically come with additional data. For example Toyota used information codes to help identify a zone in the vehicle such as battery, inverter, motor etc. This of course helps speed up the diagnosis. When it comes to rebuilding a battery though, you would have to put everything back together, refit the battery and then power the vehicle back up in order to use the vehicles built in insulation measurement. If for some reason the insulation measurement failed and logged a fault within the battery, you would then need to shut the system down and remove the battery to investigate. Carrying out insulation measurements prior to refitting high voltage components mitigates potential issues when powering back up.
I completely agree that those in possession of an insulation tester without the understanding of how and where it should be used are similar to those with a PowerProbe and not knowing how and where it should be used. The potential for something to go wrong is of course possible. Personally, I would like to think that anyone thinking about diagnosing a high voltage vehicle will have had the correct training and not just having a go.
You’re absolutely right and I should have mentioned this in the original article. Any voltage that is used by the device should be dissipated by the meter’s internal circuitry. The Pico meter will display what voltage has been applied in the top right corner. Once you have stopped the test, the meter will display the measurement and automatically reduce the voltage back to 0V. If the meter is used where the probes are connected to the high voltage positive and high voltage negative, then you will start to charge the capacitors. This then has a real safety concern as you say and carries the risk of an electrical shock with the potential to cause serious injury. In the event that this happens, you will be able to exit the measurement screen by pressing Exit and you will have a live voltage reading in the top right corner. If the voltage was still above 30V and an attempt to perform another measurement, the insulation tester will warn the user and not permit the test to be carried out.
Very much looking forward to seeing more fuel cells on the market. Fundamentally they are still an electric vehicle with a similar battery & motor setup. Insulation testing I would imagine is similar in the sense that we are looking for a breakdown in the insulation barrier between the high voltage circuit and the vehicle’s chassis but not having worked on one, yet, I'm not entirely sure what measurements would be required.
What a time though to be in the motor industry with so much changing so rapidly. Trying to contain all that information is impossible as we've said, which is hopefully where this forum helps.
I hope some of this helps with some of the points you have made about insulation testing. We’re here to learn as much as everyone else and if there are different applications, techniques, regulations in other markets then we would like to hear about them.