Design project: An AC auto-boost unitOctober 2013
The need for this unit came about because our Hoover top-loader washing machine would stop working whenever the AC mains voltage went below 234V. I could not find an actual fault in the washing machine and it operates perfectly normally if the mains is high enough. I have assumed there is a design problem with the Hoover. It is designed in Australia, where the AC grid used to be nominally 240V, although it is officially 230V now. Even then, a 240V appliance refusing to work below 234V is a bit poor. In New Zealand we have a nominal 230V system which has a tolerance of plus and minus 6%. This means the mains voltage could be as low as 216V and as high as 244V, which normally creates no problem, apart from with the Hoover. At our residence, the voltage swings routinely between 223V and 248V, so the upper end is strictly outside tolerance, however the upper end is not the problem.
Auto-boost unit design
The Hoover draws a maximum of 4 amps when washing and spinning. I wanted to provide only just enough boost because creating excess AC voltage is likely to be imprudent. So, the design provides for just 12V boost, which kicks in if the input side goes below 234V. Once the AC input reaches 241V, the unit switches back to bypass. I used a toroidal 12V+12V transformer, rated at 160VA; wired the secondaries in parallel and connected one side of the secondary winding(s) to the incoming mains in the sense that provided boost. The parallel connection means this transformer can deliver 13A, so is larger than is needed, but the next size down (80VA) I regarded as marginal for this application. Usually I prefer not to wire secondaries in parallel because a slight difference in voltage between secondaries causes additional transformer heating, however a well made toroidal type can be relied on to minimise this problem.
The schematic is below. I should add that this information is provided as a design idea rather than for a constructional project. I did not make a pcb for this one-off unit.
The design is fairly conventional; relying on a sample of the input voltage to drive a comparator, which then operates a relay. The N/O contact of which selects the boosted winding, while the N/C contact selects bypass (direct). The main transformer is drawn at upper right, with the relay contact shown below the transformer. The following briefly describes the control section operation:
Control PSU section
A small 8VA 12+12V pcb transformer provides power for the control circuitry while it's other secondary is used as the sample voltage. The AC from winding S1 is rectified by a small bridge rectifier, filtered and applied to a LM317T adjustable regulator. The dc rail was 5.5V using standard resistor values. The comparator chip will work from 5V to 7V so there was no requirement for exactly 5 volts. Because the input dc to the regulator is about 16V, there is plenty of headroom should the AC input reduce to below even 200V. To provide an stable comparator reference, an LT1009 2.5V reference IC is used. Remaining circuitry around transistors Q1 and Q2 is to create a short power-up delay for the opto-coupler. This is to give the comparator time to stabilise after power-on. The delay time is around 3 seconds and avoids the relay being pulsed briefly at power-up.
Winding S2 of the control transformer is fed to another bridge-rectifier then filtered in a two-stage filter C4 and C5. While full-wave rectification seems overkill, these small lossy transformers appear to work better with symmetrical loading, especially in the presence of flat-topped mains waveforms. There is about 19Vdc at C5 for nominal mains input. A potential divider and wire-wound potentiometer drop the 19V to near 2.5V with a fairly long discharge time-constant of some 13 seconds with C5. The charge time constant however, is only 22ms, so that the sample voltage reacts quickly to voltage increases, but slowly to decreases.
I chose to use an AD790 precision comparator, because it is an 8 pin DIL, does not create glitches on the rail, has some built-in hysteresis and has a latch feature. Comparator circuits are notoriously tricky when a slowly changing voltage is being compared to a fixed reference, as in this case. It was initially difficult to avoid all chatter of the comparator output until the sample dc was provided with two stage filtering, additional hysteresis was used and the latch pin of the comparator was strobed. Although the AD790 has some hysteresis, the 3.3M resistor from output (pin 7) to the non-inverting input (pin 2) adds more, so that the mains has to change by 7V at the input for the comparator to change state.
Strobe and latch
The AD790 has a latch function. When pin 5 is high, the comparator functions normally, reacting to the state of its inputs, but if held low, the comparator output will not change state from that state which it was last in. This is a useful feature for sample and hold designs. In this case, I strobed the latch pin with a short (5ms) pulse every 2 seconds. This means that the output of the comparator cannot chatter or change states at less than 2 second intervals. I chose 5ms since that is shorter than the nominal relay pull-in time. The 2 seconds could be increased if desired. The strobe is created with a TLC555 chip; the low-going pulse from the TLC555 is inverted by Q3 before driving the latch pin of the AD790.
I had a 230V relay to hand with big enough contacts, so used that in conjunction with a triac optocoupler IL420. This optocoupler has a sensitive input, requiring a minimum 2mA through the LED. In this case I made the LED current around 7.5mA. That also meant I could wire an external LED in series to act as a 'boost' indicator. Because the AD790 comparator has a symmetrical output; that is it can both source and sink, I added an orange (in bypass) LED as well. To finish, I added a switch to provide either auto operation, manual-bypass and manual-boost modes.
This information is not intended to be a step-by-step constructional project, although experienced constructors could easily assemble it. The dangers of building and testing such mains projects cannot be over-stated. I use an isolating transformer on the test bench and triple-check that things are un-plugged before making changes. I used up a couple of my lives when a technical trainee nearly 40 years ago, but have not been caught out since.
Axino-Tech Consulting & services; October 2013
Make a comment using the form on the contact page