Renault Zoe | Renault Zoe HVAC

When it comes to EV’s we often get a little caught up in all the systems that are different. The large traction battery, 3 phase motor, charging etc. We must not forget that the systems we worked on with ICE vehicles are still there on BEV’s. 

One which is now more important is the HVAC system. No longer is heating or cooling the cabin to keep us at a comfortable temperature or to demist the windscreen its sole purpose. Thermal management of the high voltage battery is critical to ensure its maximum efficiency in all temperatures. Having the battery at the right temperature also means we charge at faster speeds so it is important to make sure it is working at its best.

This particular Zoe had an issue where following a repair due to the vehicle not charging, the air conditioning compressor when operated would run up to maximum speed and then pop the safety relief valve on the pump, venting the gas to the atmosphere. Not good for the environment and your wallet, especially with the price of R1234yf refrigerant! Of course, how can the two be related and the answer is they can’t but we’ve all heard the customer that states - “ever since you…” The original issue was in fact an insulation fault on the air conditioning pump which was rectified. Once the pump had been replaced, the system regassed and the vehicle was checked to confirm the repair, the fault emerged!

As always, a good understanding of how the system works is critical to diagnosing correctly and the Zoe system isn’t the simplest to understand as it also works as a heat pump.

The Renault Zoe has been around for a while and due to its popularity there is a lot of information around. When it comes to EV’s, you can’t go wrong with the information available from HEVRA. With the subscription, they provide fact sheets on a growing number of BEV’s which include detailed information on the various different systems - including the HVAC system on a Zoe. The images used of the system in both heating and cooling modes have been kindly provided by HEVRA for use with this case study.

Here, you can see an overview of when the system is in cooling mode.

As you can see there are some additional items that may not be present in a conventional ICE vehicle. First off is the interior condenser which is fitted inside the vehicle near the bottom of the heater box. We then have a fixed orifice tube before getting to the traditionally found condenser at the front of the vehicle. You will also notice 2311 next to the FOT. This is a bypass solenoid which in cooling mode is set to open so allowing the hot, high pressure gas to flow easily through to the front condenser where it is cooled to become a high pressure, high temperature liquid. We then see a 3 way diverter valve that in cooling mode directs the refrigerant towards the interior evaporator and also to the evaporator in the battery where over another fixed orifice tube, the pressure drop changes the refrigerant back to the a low pressure, low temperature liquid and then back into a gas. You will see by-pass valves for both of the evaporators which will allow the vehicle to select what needs more cooling. For example during charging, if the battery temperature needs to be cooled down the HVAC can cool just the battery without also cooling the cabin. This all then passes back through the receiver drier to ensure it is definitely a gas before starting the process again at the compressor.

In heating mode the main difference is that the bypass valve 2311 is now shut forcing the high temperature, high pressure gas over the fixed orifice tube which now uses the front condenser as an evaporator. This means the change from gas to liquid takes place in the interior condenser as cooler air is directed over it and so warms the cabin. As the refrigerant exits the front condenser as a low pressure, low temperature liquid, there is no reason for it to pass over the battery or interior condenser. This means the 3 way valve diverts the refrigerant back towards the receiver drier to start the process again. 

To prevent any unintentional loss of refrigerant to the atmosphere, the system had the gas removed. Initially investigation of course involves a scan tool, one of the least intrusive methods of diagnosis. Checking for fault codes there were a number of historic fault founds but nothing permanent, one of which was for high pressure. Checking in the live data we found something out of the ordinary. The refrigerant inlet and outlet temperature sensors were reading 482C and 472C respectively. Also to note is the compressor speed reference which was reporting 65,535 rpm. Given the fact the AC was switched off this was clearly wrong. 

Our first check was to determine if there was an issue with the temperature sensors. If they are reporting this type of temperature then would a fail safe for the HVAC system be to put the compressor into full speed for max cooling. It seemed logically so off came the bumper.

Referring back to the scan tool though we noticed that the readings had now changed to the more expected values. We were chasing a wiring issue? Looking at the wiring diagram, both temperature sensors share power supply. Connecting Pico to the shared power supply we could determine a possible open circuit making use of the mask and action feature in PS7. Including the LIN communication on the AC pump we could monitor any possible changes in the signal whilst we manipulate the harness to check for any wiring issues.

After spending too long checking the wiring harness, we were unable to get a fault to occur. As with anything, when finding something you think is wrong, first check your measurement. With the scan tool connected we noticed something odd when accessing the live data for the HVAC system. Turns out you do not need to have the ignition on in order to communicate with certain ECU’s. However, without the ignition on none of the sensors are powered. What we believe happened was, as with other vehicles, just by unlocking the doors wakes up a lot of the ECU’s. When connecting the scan tool we could access the HVAC ECU so we potentially didn’t switch the ignition on and didn’t suspect anything else till checking the live data. As we started removing the bumper in preparation for testing, the vehicle eventually shut itself down properly, losing communication with the scan tool. As all the doors were unlocked and open already, instinctively we switched the ignition back on to get communication back up with the scan tool only to find that the values were now correct. Always something to learn when diagnosing or trip you up! 

Unable to get the vehicle to fault and yet to confirm the issue, the only thing next was to put some gas back into the system. A vacuum was performed first to ensure no leaks and once happy, half the refrigerant was placed into the system. This did mean the pressure was slightly lower than normal but it should be enough to get the pump to operate for us to determine what was happening.

With the machine still connected we run the HVAC system, starting with it in heating mode where the interior condenser is used to heat the cabin. Given the way the heating circuit operates using the gauges wouldn’t have helped as the high pressure port is after the FOT meaning it’s on the wrong side to measure the pressure. This meant using the scan tool to get the live data from the pressure sensor. As the pump continued to work all seemed to be OK then all of sudden the relief valve opened with the pressure sensor reporting 11 BAR. This should be well within the operating limits of the compressor and shouldn’t be causing the relief valve to open.

Switching to cooling mode, the high pressure gauge was reading the same as the pressure sensor which is as expected as any blockage to the bypass valve 2311 would have meant these would have been different. The overall pressure was lower than normal but again given the lower quantity of gas this was to be expected, what we weren't expecting was the relief valve opened at 5 bar! The only likely cause for this is a blockage between the compressor and the pressure sensor. How can we prove this is the case?

With the new 4425A we have been able to create a temperature probe, TA395, as we can provide power to the probes. If there was a blockage, the most likely place would be in the interior condenser. Given we have high pressure, high temperature liquid from the compressor, should it meet a restriction there will be a pressure drop meaning one side of the blockage will be hot and the other side will be cooler. Should the liquid/gas be free to move unrestricted the temperature should have a minimal drop across the condenser. There will be some temperature change as the heat will be dissipated from the surface of the condenser but there shouldn’t be a large difference.

Unfortunately, I only had one temperature sensor with me which meant having to perform the measurement twice. But saving all captures in Pico meant we could analyse post capture. First measurement was taken at the pipe closest to the bulkhead by using a battery clip to hold the temperature sensor body against the pipework.

You may have noticed that there is a pressure measurement on this trace. More on that later but it was important to note the pressure to ensure we didn’t get to the point where the relief valve opened. Whilst this reading isn’t the actual pressure at the pump based on the location in the circuit, we had previously seen the relief open at 5 bar. Making sure the pressure didn’t pass 5 bar meant no gas would be vented and lost. 

Running the HVAC system in cooling mode we had a max temperature of 50C at this pipe. Having the peak pressure here allows us to use that as a reference point bringing out the time rulers to measure 1 minute before this point and measuring between the rulers.To help with the comparison between captures we can add an additional measurement for the rising rate of the temperature between these rulers. Here we see that the software is showing a rising rate of 321.8C/ks!

Moving the temperature sensor to the pipe furthest from the bulkhead.

Here we see that there is a considerable difference between the temperature of the two pipes as the pressure increases to 4.5 bar and looking at the rising rate here we have 69C. This is compared to 321.8C/ks from the other pipe. 

To confirm the suspicions the same tests were carried out on a known good vehicle.

From these measurements we can see that there difference is much much smaller with the max temperatures being around 4C difference and the rising rates nowhere near the 328C/ks mark. What is quite interesting is that the pipe furthest from the bulkhead is actually hotter which is the opposite for the faulty vehicle. 

This is enough evidence to inform the customer that the interior condenser needs to be removed for inspection. It’s the most likely cause but considering how easy it is to get to, it is worth pulling it out first in case it is a blocked pipe. 

As mentioned we do have a reading for the refrigerant pressure. This was captured using a custom AC filling hose and adapted with a Foster coupling. The new BNC+ pressure transducer has been confirmed as applicable with R134A due to its stainless steel construction. Given that R1234yf and R134A are similar, I thought I’d see! I will add that this is purely experimental though and Pico does not condone the use of accessories outside of their intended purpose. This was more for convenience than anything else though as the pressure sensor would have been a much safer option but being as it was buried under the battery tray and behind the inner wheel arch it seemed like a good excuse to try an alternative method. As you can see from the results above it isn’t too bad with the one issue being the max pressure which for this transducer is 200psi (13.7Bar). Given the high side of the air conditioning system could get over this it isn’t recommended. That being said, there may well be an application for being able to graph the actual refrigerant pressure through the charge filling ports. This was something Steve also experimented with in this forum post.

I hope this helps.