Ok, here also my 2 (Euro)Cents ;-)
Forewarning: Here in germany regulations are a bit elaborated regarding PAT tests and the duties of a company of the need of regular testing. I myself have (besides being a sysadmin) the qualification of testing appliances, as my employer is a company that does consulting in this area as one of the main business topics...
(So I am being a "befähigte Person" according to TRBS 1203 and I do the inhouse PAT tests)- the requirements that those tests are to be done are mandated by insurance (Berufsgenossenschaft), so the standard there is actually calleed DGUV Vorschrift 3, the technical way HOW to conduct those tests are written in electrical standards (VDE), currently VDE 0701-0702, which has been recently split into VDE 0701 and 0702, respectively.
(DIN EN 50678 VDE 0701:2021-02 and DIN EN 50699 VDE 0702:2021–06)
I do NOT know how the legal stuff works in UK and what is mandated by local regulations, and how well a repair cafe fits in this, so I can only give some general tips and considerations for this.
First: It is a good thing that the people running the cafe care about safety, as sometimes some very questionable things can happen during a repair or simply do not get noticed. And if they (what nobody hopes for) come in a situation to explain to a judge why some repaired appliance killed or injured somebody, some solid paperwork is the only thing that helps explaining things.
Second: Knowledge WHAT to do is always required, and more important than a fancy tester. I also have seen lots of protocols where people conducting tests simply chose the very wrong steps to do.
This said, there can be multiple ways to conduct this. If they aim for having some serious proof that they did the correct thing, the devices should have a recent calibration, and some course how to do this is never wrong.
If they would have to give proof that they did some PAT test, a protocol is a MUST-Have. This can be manually written (here in germany there is something the ZVEH developed) as template, but they can make their own- some simple Excel sheet would be sufficient. Print it out, maybe manually fill in the measured values, and sign it. Important: The PAT tester(s) have to be included in the protocol with serial number, so that in case of trouble one can recreate the way everything has been measured. Also the device under repair has to be able to be clearly identified, like by model and serial number.
The tests itself are basically the same for the past 30+ years, only some slight adjustments in the tolerated measurements have been made, so basically it all boils down to:
- Check earthing of the touchable parts, if present. This has (in germany) to be conducted with a current of min. 0.2A and should yield a resistance <0.3 Ohms.
- Check isolation of all touchable parts, if not grounded. This should be done with 500VDC, and the tester has to limit the current (I think to about 1mA to protect the person doing this).
- Check isolation of L and N against PE, if present
- Active tests then involve firing up the device and measure the leakage current via PE, if present
- then the current at metallic parts that could be touched and are not grounded has to be measured
Thats the standard program, if there are other safety mechanisms present, like an embedded RCD in a hoover or similar, or some protection thermal switch in a water cooker, this also has to be tested if they function.
For ease of use, a PAT tester that can do all of it is nice, but it is totally legitimate to use different testing devices. This specifically applies if there is a low volume of appliances that are to be tested.
At work, where I do mostly the initial testing of new gear like freshly bought Notebooks (resp. their Power supplies) and desktops, or the freshly bought new coffee machine because the old broke down, I use an appliance tester that is controlled via USB from my Notebook and has an application on it that stores the measured values in a database and creates fancy protocols with it.
Especially in a corporate environment, where the same piece of hardware will get several tests over the course of its lifecycle, this is a big timesaver due to automation.
At home, where I do some tinkering and private repairs (I repair stuff that I will then use to repair other stuff...) I managed to get my hands on some old isolation tester about 40 years old, but perfectly working- Last calibration in 2015, and aside from a slightly damaged tip (which I will replace with a standard 4mm jack to put various standard probes on it) it is in perfectly working order.
With this isolation tester, I can test isolation with 500VDC (in reality it puts out 660VDC), and I can measure low resistances up to 4 Ohms with 0.2A current.
->For electric equipment like extension cords thats all.
Leakage currents are a bit more difficult to handle, especially when according to EN61010 some frequency-dependant characteristics of the current has to be taken into account. Here some leakage current clamps might be acquired for decent money that have the so called "50 Hertz filter", that dampens frequencies below and above 50 Hz to simulate the body sensitivity. With an adaptor, the Device under Test can be put on a wooden board for isolating it from the ground and then the PE leakage current can be measured via clamp- Touch current basically could be made the same way, as you connect the touchable conducting parts of a device (that are not grounded) to external PE and then measure the current.
Voltages from power supplies can be checked with a reliable multimeter.
Important: All of them should have calibration protocols and be noted in the PAT protocol.
If someone here could share some knowledge about the situation in UK, how PAT testing is done there, and which electrical standards apply?