User Tools

Site Tools


SSR safety

The SSR has been sourced from domestic suppliers due to the influx of counterfeit relays coming from China. It is hoped this will reduce the odds of using a relay that is dangerously under-specc'd.

Without a heatsink the example SSR should not be loaded with more than 10A (2.4Kw @240v). The SSR uses the aluminium case as a heat-sink in this application, but the enclosures thermal capacity is unknown.

Use of the ACS to internally switch more than 10A should be thermally tested under controlled conditions. Mounting the ACS to an external heatsink or thermally conductive machine chassis should be expected to mitigate this, but test before commissioning.

3Kw (13A) thermal load test

This is a test performed under uncontrolled conditions and should only be considered an estimate. Authenticate the results before applying the ACS to 3Kw (approximate UK domestic ring main maximum appliance load) equipment.

The ACS was wired to a multi-socket extension lead and several electric fan-heaters were plugged into it. The input mains plug was fitted with a standard 13A fuse and plugged into power via a generic device power-monitor. The ACS was placed on a non-conductive wooden surface in a separate room from the fan heaters.

All subsequent mentions of the ACS temperature are taken to mean the temperature of the outside of the ACS aluminium enclosure at the middle of the bottom edge beside the flange.

  • The starting temperature of the room was X°C.
  • The starting temperature of the ACS was X°C.

The ACS was pre-loaded to approximately 10A, the maximum specified load of the SSR to operate with no heat-sink.

  • After XX minutes the ACS temperature was X°C.

The load was then incremented upward by turning up the fan heaters.

  • At XXXXWatts after XX minutes the ACS temperature was X°C.
  • At XXXXWatts after XX minutes the ACS temperature was X°C.
  • At XXXXWatts after XX minutes the ACS temperature was X°C.
  • At XXXXWatts after XX minutes the ACS temperature was X°C.

From this is is suggestive that the ACS in it's basic configuration is likely safe/unsafe for 3Kw loads.

Current sensor choice

Q: Why not use one of the ubiquitous ACS714-based Arduino modules instead of the less available WCS1800?

A: The ACS714 sensor is rated for 240v applications, but the boards they are supplied on are not. This is apparent in under-rated terminals used, undersized traces and copper thickness, poor conformal coating and other electrical isolation design issues. In particular the pins of the ACS714 are too close together to allow the minimum 2.5mm isolation required. Those advertising 240v applications generally also only specify for intermittent use.

The modules also use varying mounting systems, often with no facility for grounding.

The WCS1800 requires ordering in from China, but is simple, compact and around the same price-point as the ACS714 boards. It is also requires no mains-side connections to install. While it suffers from some minor non-linearity across its range, absolute measurements are not required for this application, nor a high level of accuracy.

Arduino stability

The Arduino variant R3 Uno is used as it has both through-hole solder points in addition to the shield headers, as well as extra 3.3v and ground pins available, meaning no additional breakout shield is required and allowing it to fit inside the existing housing.

There has been some anecdotal suggestion that the capacitors used on this variant may be of low quality and make them unsuitable for prolonged use in 24hr applications. If you experience stability issues under extended use you may wish to try replacing the capacitors with ones of known quality. The PoE board however should minimise the potential effects by providing pre-filtered power.

designs/acs_notes.txt · Last modified: 2019/03/16 22:56 by sci