Advice on setting up a home electronics lab - Living in the UK

[IanMacdonald]

If working on valve/tube gear you should ideally have a fully insulated bench surface (wood is OK if dry) and an insulating floor or mat where you stand or sit. Definitely don't stand on concrete.

The thing you absolutely don't want to have, when doing any powered tests, is any part of your body earthed. No matter how careful you are there is always the risk of touching something with 500v on it, and yourself being earthed is the difference between, 'Ouch' .. and being put in a 6ft hole in the place the current through your body returned to. 

It follows that earthed test equipment is a personnel hazard when working on exposed high voltage. Don't touch metal knobs or panels, or earthed scope probes if you can avoid it. I really wish scope manufacturers would change their ways over this. Having an earthed probe in your hand is about the most dangerous thing you can do when testing live high voltages. Particularly if it's in the other hand, not the one which gets the shock.

There have been several deaths on stage through a  performer holding an earthed metal microphone in one hand and playing a guitar in the other connected to an amplifier whose earth had come off and touched the live. It's the hand-to-hand shock which is most likely to kill since it goes through the heart muscle. Wireless mics have been a huge safety benefit here. Eliminating the one earth prevents the circuit being closed if the guitar becomes live. As have sensitive fuses in guitar leads. (but they need to blow with ~20mA or so to be any help)

Also, an RCD DOES NOT protect you against this risk when working on amplifier guts. It cannot sense current 'lost to earth' if that current came from a transformer secondary. (Though it would help in the stage performer situation)

Fire: In my 40+ years in electronics I've never seen a soldering iron start a fire. They generally don't get quite hot enough to ignite things. Hot air guns are a somewhat worrying issue though; there have been quite a few reported cases of them going out of control and becoming red or even white hot. I'd advise making sure any such tools are unplugged when not in use. (On some models the on/off switch is not a proper mains switch and should not be relied on)

Safety is achieved by understanding the risks.

[OwO]

I just prefer to not deal with mains ever. If a PSU is bad just replace it (or fix obvious faults like bad caps).

[bd139]

Sensible. Replacement is usually cheaper in time and energy. I must have spent 6 hours fucking around with a Tek 2235 PSU recently. That’s a DS1054Z in cash for me.

[Berni]

That depends on what sort of PSU it is.

If its a usual 12V 5A thing that runs as soon as you give it mains yeah replace it unless its a easy fix.

If for example its more like the PSU from my old Tekteonix AWG2041 then not so much. Just the PSU in there is about the size of a small toaster and has 3 or 4 extra PCBs sticking out of the main PCB and everything surrounded in complex bent aluminum to get the heatsinking on to all the stuff and hold the extra boards. There are like 6 or more supply rails, some linearly post regulated, there is a standby rail and the PSU turns on via a power on signal and provides a power OK signal back. Googling the original manufacturer of the PSU module retrievals they went out of business by now, there are no schematics and ebay doesn't find any spares.

In such a case its worth going to pretty long lengths to fix such a complex unobtanium PSU. In the case of my Tek signal generator the PFC transistor blew up and has taken out some of the PFC components with it and blew up the soft start circuit that's before it. Got it fixed and was well worth the effort.

But yeah only attempt repairing a SMPS if you know what you are doing, they are dangerous things.

[mikeselectricstuff]

Forget about the isolation transformer. It will render the RCD useless. Better get a differential probe; these aren't very expensive nowadays. MicSig has a model which offers excellent value for money.

While an isolation transformer defeats the effects of an RCD placed before the transformer, you can always add a second RCD after the transformer.

An RCD after an isolation transformer will be useless - as the transformer secondary is floating, all the current from it will inherently be balanced, so nothing that happens after the RCD can cause an imbalance to trip it.
 

[andyowenwest]

Thanks to everyone who has commented

There's been some really good advice here that will definitely be useful to myself and others in the future

I'm due to move in around 3 weeks time. My plan is to document the process of building a basic lab and extract some top tips from here and start building a home-lab blog

Ill post back when I start making progress!

Cheers

Andy 

[HalFET]

In my case I have an RF lab so I also have a cheap VNA and SDR receiver. If I felt like spending more money *maybe* I'd get a Rigol spectrum analyzer. As it stands there is no more than $1000 worth of equipment in the lab:
- Soldering iron and misc materials (flux, solder, wick, etc) - $20
- Hot air station - $40
- Power supply - $50
- Scope - $300
- SDR - $100
- VNA - $300

That adds to $800 total. If I wanted to measure phase noise of a circuit I don't buy a $5000 spectrum analyzer, I build two such circuits and mix the signals and look at the phase noise of the result. I have built quite many RF products with this minimal lab layout.

Get another power supply is my main advice, always get one power supply more than you think you need. And when working on valves I also like to have a period scope, mostly because if I do inadvertently blow it up somehow (haven't done so yet though), it'll be a pretty easy fix. Also, wouldn't you want a signal generator if you are going to be working on audio? I have three signal generators and I'm still always short a signal generator when building circuits.

But a few non obvious ones that are nice to have:

• Solder fume extraction (you can buy old units for less than 50 bucks and it's totally worth it).
• High brightness light you can aim at things, if possible one of those magnifiers with a ring light around it.
• A stereozoom microscope with a long work distance.
• A wire rack.
• Storage boxes for components.
• Bench vice with turnable head.
• A voodoo doll with a skeleton made out of discarded resistors that were soldered together.
• A system to put your rolls of wire and cable on. I currently have three pieces of MDF bolted together as a U bracket with a piece of PVC tubing for this.
• An old computer you don't mind blowing up for interface testing.
• Boxed 1:1 isolation amplifier so you can feed things in and out of your computer's soundcard safely.
• Isopropyl alcohol and acetone.
• A pyrex oven dish or something else protect your desk from things that might go hot/catch fire.
• CO2 fire extinguisher
• Bucket of sand
• Safety goggles
• Dremel or column drill
• Scotchbrite pads to quickly remove soldermask.
• Hacksaw
• Scalpel
• An old AM radio (you damn well know why )
• Kapton tape
• Widlarizer/hammer
• A benchmeter which can measure currents above 1A for longer than ten seconds.
• Glue and hot snot gun
• Self-healing cutting mat

Ok, maybe the voodoo doll isn't strictly necessary, but I definitely recommend the solder fume extraction, safety goggles, CO2 fire extinguisher and a bucket of sand (for lithium battery containment).