TLDR: The APS alternator performs pretty close to advertised performance but requires driving or higher than nominal idle to supply power. If you are looking for serious charge capability we'd recommend this product.
Firmware and Settings
The Wakespeed has a LOT of flexibility and thus many settings. We highly recommend spending timing reading through the content on their technical web page paying especial attention to the "Communication and Programming Guide".
For optimal performance, you will need to determine some specific details for your alternator. Thankfully Wakespeed has released an ios app that will produce a base configuration file specific for the APS 185 amp alternator. The configuration it produced for us is here. We annotated the meaning of the configuration parameters, and noted required changes in this version here.
Because we wanted our Victron system to be able to read data from the Wakespeed, we needed to upgrade its firmware to the latest beta version (2.5.0). This version is available on their technical page. Details about this ability can be found in this github comment thread.
We updated the Wakespeed configuration file (manually) to add a few new parameters for firmware 2.5.0 and made a few changes for our specific configuration and conservatism. The modified configuration file is here with annotations here. Aside from changes related to our configuration, we reduced the alternator temperature limit from 120C to 100C for conservatism. One motivation is that this alternator is directly lugged to a 2/0 cable with an insulation rating of 105C. Another is that for any electrical equipment, cooler operation equates to longer life.
We used the Wakespeed Universal Configuration Utility (a batch file) to update the firmware and upload the initial configuration. We're a Mac household so we used a free Windows 10 image running on VirtualBox to run the utility. It was relatively painless (we did need to play with some USB settings to ensure the virtual machine was passed the Wakespeed USB device by default from the host machine).
Initial Light-off
We started the engine and observed via our Victron system a decent charge current with RPM above base idle. Success!
Logging Capability
We fired up our loggingPi and recorded CANBus data with the Wakespeed operating and got to decoding. Our code to log Wakespeeed data is here. The script we run for testing, which logs BMS data as well is here.
Initial Findings
With data inn hand we could now get a good appreciation of how the system was running. We noticed that reported Engine RPM looked off, and we felt that the system wasn't charging as hard as expected.
Based on some help from the forum we determined that our pulley ratio parameter was incorrect. We had assumed (and measured incorrectly) a 6.5" crank pulley diameter which threw off the ratio. Using the 6" diameter from the forum feed back put our reported engine RPM right in line with expectations.
We increased the "normal derate" parameter by 20% but did not see any change in performance. We theorized that the system must be operating in "small" mode as we had not changed the "small derate" parameter. We confirmed this by monitoring serial data from the Wakespeed and looking at field drive. Upon "opening the patient", we found a DIP switch was placing the system in "small" mode. We reduced the "normal derate" parameter back to the default ~62%, set the DIP switch for "normal" and retested.
For the following tests, the vehicle was being driven normally, usually a bit of start and stop to get to the highway, followed by cruising speeds between 55-70 mph.
The tests started with a SOC around 80% and terminated after charging terminated with SOC at %100.
The measured current is that measured by the BMS into the battery. For the tests, solar power and all switchable DC loads were secured leaving ~0.7 amps (~20 watts) of base draw for the Wakespeed, Victron Cerbo, loggingPi, router, main contactor, BMS, etc. We also found that the field drive appears to be about 3 amps or so once the Wakespeed becomes active. So the data shown below is likely low by about 2-3 amps.
Ambient temperatures were about 85-90 F.
46% Field Drive Test
Testing with in "small" with "small derate" parameter (field drive) set to 0.46 (and active) demonstrated the following performance:
The modified configuration file is here with annotations here. Raw test data is here.
62.5% Field Drive Test
Testing with the "normal derate" parameter (field drive) set to 0.625 demonstrated the following performance:
The modified configuration file is here with annotations here. Raw test data is here.
84.0% Field Drive Test
Testing with the "normal derate" parameter (field drive) set to 0.84 demonstrated the following performance:
The modified configuration file is here with annotations here. Raw test data is here.
Later Findings Regarding High Voltage Cutoff Events
On 2022-7-10 we observed a high voltage cutoff event upon starting the van. The system then restarted without resetting SOC to 100% and charging resumed occured via solar limited by max voltage of 28.1. Some cell balance occured prior to the SOC reset to 100%. The next charging event occured on the 12th, once SOC dropped to > 40% where we have SOCH set. No alternator charge occured. Upon inspecting the unit it was flashing an error code corresponding to BMS notifying of cutoff. Had to power cycle BMS to resume operation as the high voltage cutoff disable alternator charging until reset (with present settings).
Over time, cells have drifted based on limited ability to balance with the new BMS firmware. This resulted in another high voltage cutoff event on 2022-9-28 (with no SOC reset) while driving with alternator charging. Essentially, a single cell is hitting the high voltage threshold prior to charge being limited by the charge voltage limit (28.1V presently).
To solve this problem, we've made use of the CERBO generator start/stop function of relay one. The CERBO gates the engine run signal to the Wakespeed 500 based on battery SOC. We've tested it successfully. It is currently set to enable alternator charging when SOC drops below 80% and disble when SOC reaches 95%. It also disables alternator charging if it loses communication with the BMS. This should:
- Eliminate high voltage cutoff events due to high alternator charge rates
- Allow more balancing to occur as solar tops off the battery at lower charge rates
Future Testing
Idle thermal performance is still undetermined and will be tested in the future. Inital testing suggest we can sustain about 4KW charge rates steady state while idling.
Timeline
Note: This doesn't reflect full time effort. Its sporadic and limited by kiting, other projects, family visits, and van availability.
- Finalized Wakespeed settings, Updated Wakespeed firmware, Uploaded Wakespeed settings, Initial operation (6-16-2022)
- Completed Logging Code (6-17-2022)
- Identified Engine RPM discrepancy (6-20-2022)
- Tested v1.1 46% field drive (6-21-2022)
- Identified "small" mode, configured for "normal" mode", Tested v1.2 62.5% field drive (6-22-2022)
- Tested v1.3 84% field drive (6-23-2022)
- Implemented CERBO gating of engine run signal to limit charge (9-29-2022)