Building a Rotary Inverter Controller with a Barth lococube mini-PLC Part 2: The AssemblyFollow article
Assembling the control system and enclosure for a vintage aircraft rotary inverter.
In part one of this series, we took a look at the vintage rotary inverter, designed a suitable control circuit to control the current through the inverter field coils, and explored the Barth lococube mini-PLC to see what it was capable of doing. In this post, we will be designing the control hardware and a suitable enclosure to contain the system.
Current Control PCB
In our first post, we had designed and simulated a high-side current source based upon two operational amplifiers and a MOSFET for the output pass stage. In the simulation, this circuit appeared to work exactly as we hoped, but in practice when we constructed the circuit on a piece of stripboard it went into self-oscillation and no adjusting and tweaking the circuit would stop this. This is often the case with simulations and it is always worth prototyping the circuit before committing to a full PCB!
We instead opted to swap the design to a simple four-transistor regulator which trades efficiency for ease-of-assembly but does not exhibit the self-oscillation issue that we saw in the op-amp based circuit. The simulation results showed that this circuit would perform as expected, and give us enough output current to drive the field coils of both motor and generator.
Stripboard was again used to prototype this circuit, and a questionable clamping method used to affix the transistors to a heatsink for testing. The results of the testing showed that the circuit would indeed work as we expected — at this point, we committed to schematic capture and a suitable PCB layout.
During the schematic capture process, we opted to swap the fixed value resistors for 25-turn variable resistorsso that we could adjust currents flowing within the circuit to set the output current range.
With the PCB fully assembled we moved onto testing the circuit using our favourite RS Pro electronic loadset in constant resistance mode to stand in for the field coils. The transistors were again clamped to a heatsink, and the circuit was left on test for some time with no adverse effects noted. The potentiometers were then adjusted to give the desired output currents, and then the circuit was put aside whilst we designed the rest of the inverter.
Now we have a circuit capable of controlling the current flowing through the inverter field windings, we need to design a control system around this. The heart of this system are two Barth lococube mini-PLCsthat will monitor the output voltage and frequency and in turn adjust the current flowing through the field coils, as well as performing other system functions, such as fault monitoring and startup sequencing.
The first step is to pick out suitable power supplies to feed the control systems — the field coil control circuit needs 36V (to be able to provide enough voltage to push the current through the field coils) at approximately 3A, and 24V at a budgeted 2A for the rest of the controls which includes the PLCs, panel meters, contactors and indicator lights. We settled on the RS Pro 36V 155W PSUand the RS Pro 24V 52W PSU as these both provide the necessary output voltages and currents at an affordable price point.
To monitor the output voltage, frequency and current, five-panel meters from the Trumeter APM series were picked: the APM-VOLT-ANO, APM-FREQ-ANO and APM-AMP-ANO as these meters offer USB configuration, and two outputs — one of which can be configured to provide a 4-20mA current loop. We used these as not only do they provide a rather satisfying LCD display that mimics traditional moving coil meters, they also provide an output that is easily usable by the PLC — only requiring simple signal conditioning to convert into a 0-10V signal.
One point worth noting is that the APM-AMP-ANO meters cannot directly measure currents where the circuit voltage exceeds 60Vdc or 30Vac, so a current transformer is necessary. In this case, the output current of the rotary inverter is stated to be no more than 3.1A so a 5:5 current transformerwas picked.
Designing an Enclosure
Our original enclosure plan was to select a 19” rack-mount enclosure such as this and then mount the rack enclosure inside an aluminium extrusion frame that has the inverter located on a top plate.
After a rather terrible mockup sketch was produced we decided that even a 5U enclosure would not give us enough room to achieve an appealing controls layout, so instead, we opted to go for a custom-made front panel that will bolt directly to the extrusion frame.
The second iteration of the control panel was more in line with what we envisaged, featuring a row of five meters and all the user controls below. The design was then further refined, spacings tweaked, connectors relocated and additional controls added.
A render of the front panel was produced before committing to making a physical prototype in MDF; we can produce laser-cut panels in-house and it is less expensive than getting a machined aluminium panel made only to discover something is missing or wrong!
All the parts were fitted to the MDF prototype, and spacings around switch contact blocks and the panel meters were confirmed to be okay. The next logical step was to get a nicer front panel made up out of aluminium, then assemble the control system.
The control system can be viewed as two mostly independent control systems, each controlled by one PLC — this helps to simplify the requirements for each control system, with only some signalling taking place between the two. One PLC will handle regulating the output frequency which requires controlling the motor field windings, and the second PLC will control regulation of the output voltage by adjusting the generator field windings.
Power sequencing also needs to be managed with power to the motor field coils applied before the armature is enabled, otherwise, the motor will sit there and act as a large heater — not good for the life of the motor and could even potentially result in overheating.
Once the motor has started, the generator field coils can be turned on, which will start producing an output voltage and frequency. At this point, the voltage and frequency can be stabilised, and then the output can be enabled.
If the output goes outside of the desired voltage and frequency range, the controls will also be able to disconnect the output to protect both the inverter and the device. The PLC is also able to monitor the temperature of the inverter and can disconnect the load if it gets too warm — there is an integrated centrifugal fan to cool the inverter so it would be wise to leave it running at no load to cool off.
Prototype Control Panel Video
In the video below, we are demonstrating a basic operational PLC program that will control the motor field coil, in response to the frequency of the output.
This post has taken a look at the design of the control system, and the construction of the enclosure. In the next post, we will finish the assembly and test the operation of the rotary inverter.