Current CE Requirements for EU sales

[amplifiednation]

Hello my amp building friends

We have been expanding distribution to EU...does anyone have the requirements to declare your product is complaint with CE?

I know you have to add a screen to the back panel of the amps...what else we got?

[R.G.]

For guitar amps and musical equipment, at one time it was conformance to IEC60650, certified by a formal testing lab, or self certified. The standards change over time, and may be different now.

Self certified sounds handy, but can be revoked by any EU testing lab finding any non-conformance, and they do look. I suggest the testing lab approach.

It is not as simple as a few construction hints, and not always obvious what to do from buying a copy of the standard.

The standards are copyrighted and sold, and cost a few hundred dollars each, so it's critical to pick the correct one that applies currently; there are many such standards. Testing labs have experts who have current knowledge on what does and does not pass the requirements for the CE mark.

Finding out how to get safety standard conformance is not as simple as getting a few tips on network forums. If you're serious, buy some expert help. It's cheaper in the long run.

[Stevem]

It's not easy, so get help!
For instance Canada requires the power trans primarily and secondary wires to be a certain color!

[pdf64]

Start getting a technical file for each product put together. Everything needs documenting, eg the original design goal, designs / choices made and rationale behind them, specifications for key components and rationale behind them. It’s a big job in itself.

[amplifiednation]

Thanks guys!

I just need to send an amp over to demo right now.

I’m already planning on going through the actual testing

But could we ‘ball park’ this?

Do the jacks need to be enclosed?

[R.G.]

Sure.

The best ball park advice I know is to get a copy of "Practical Guide to the Low Voltage Directive" by Gregg Kervill and follow its advice. You can decide which points you're willing to play quick and dirty with. Also excellent is "Electrical Product Safety" by Tzimenakis and Holland. My copy is copyright 2000, so standards have probably changed since it was published, but it remains an excellent general guide.

If you really don't have time for the specifics, (1) get the safety grounding right (2) ensure all touchable metal is continuous to the safety ground wire (3) high pot test the unit (4) ensure that user can't touch any surface that can cause a burn by touching (5) make the enclosure from wood > 1/2" thick so internal flaming liquids or electrical arcs can't set the enclosure afire (6) seal or divert openings so poking a wire through the opening can't touch a live conductor. The resulting unit will almost certainly not pass real testing, but might keep a demo unit from killing someone until you can get it back. If you do go the quick and dirty route, do ensure that you get the demo unit back; don't leave it there to cause future liability. Remember that in some countries, causing injury by failing to make a unit safe enough can be a criminal offense, not just grounds for civil liability. It's this last consideration that drives a lot of my view that you have to take this seriously.

Better yet, find a testing lab nearby and pay a few hundred to a thousand dollars for an engineer to do a quick advisory review that's less than a full inspection test.

Here are some specific things that I know have been required to pass certifications.
> Every single user-touchable metal part on the outside of the unit, including painted parts, must show continuity to the AC safety ground wire. Paint and coatings are specifically disallowed as insulation, as they wear through. It is possible that ancillary items like protective screens that pass 4200Vac high-pot to safety ground might pass under the reinforced-insulation rules.
> Use only UL/CSA/TUV/ETL/etc listed parts for any circuits involving the AC mains
> High-potential (called "hi-pot" in the trade) test the whole thing for 4200VAC between the AC mains wires in the line cord to any metal the user could theoretically touch, including poking their fingers through penetrable openings.
> Arrange the enclosure and all openings so that poking a finger or stiff wire through any opening cannot contact a conductor with a voltage of more than 42Vdc or peak AC. This is the enclosed-jack thing you refer to, and using enclosed jacks is one way to keep a wire poked through a jack from contacting live wires. Another is to make sure that the jacks have no hot wires behind them, or using some kind of barrier.
> Wire the safety ground wire on the AC entry to a single bolt/stud on the chassis used for no other purpose; the transformer mounting bolts are not to be used for this. There is a specific way of stacking the conductors that has passed testing: star washer, then ring terminal crimped to safety ground wire, then another star washer, then the nut holding this down.
> Do the AC wiring in such a way that the safety ground wire will break last if the AC cord is pulled out
> There are specific force pull requirements for pulling the AC wires without causing a hazard, but in general if the line cord does not enter the chassis in something like an IEC power entry socket, you should be able to lift the unit by the line cord without causing any possible hazard, including abrading the AC wire insulation.
> In general, ensure that no single component failure can cause an electrical hazard or fire. This includes any single wire breaking.
> One subtle implication of this is that AC mains wires should be cable-tied or dressed in a way that any one wire breaking at the end cannot contact secondary conductors. So cable tie AC wires in position or to one another so that one wire breaking does not result in a live wire flopping about. A more subtle instance is that the AC mains switch should be secured so that it can not rotate and twist the AC wires until they break or short. Ditto fuse holders, etc.
> Hmmm. Fuse holders. The incoming live wire to a fuse holder should go to the fuse holder terminal that breaks first and makes last as you re-insert a fuse so that replacing a fuse never leaves the exposed fuse end live.
> Creepage and clearance distances greater than 8mm between all AC mains conductors and secondary conductors, including especially on PCBs, turret/eyelet boards and point to point wiring.

This list just goes on and on. But it's a quick list, and many other points can prevent certification and at least in some cases allow your unit to kill someone now or in the future.

[pdf64]

I seem to recall that a specific named person (the ‘agent’) who’s importing the stuff into the EU has to bear the criminal liability for its CE compliance.
Also need to think about the EMC and ROHS directives, the same applies.

[batali]

You would have to have some pretty crazy lab equipment and test gear to do this without getting in trouble. Much better to have an legit place do this for you. There are things like no primary wires touching the chassis and must be sleeved, staked thread studs for grounds, specific lock washer/washer setups. No 3rd pin ground tied at the same place as audio grounds, they tests for emissions and the interference on other gear. There are books on this and every year there is a new thing they add to CE. Makes sure you have 0 lead on the board and all parts are ROHS compliant, no vactrols etc. Average cost is around $3500 an amp for testing.

[amplifiednation]

This is exactly what I needed! Thank you so much!

Sure.

The best ball park advice I know is to get a copy of "Practical Guide to the Low Voltage Directive" by Gregg Kervill and follow its advice. You can decide which points you're willing to play quick and dirty with. Also excellent is "Electrical Product Safety" by Tzimenakis and Holland. My copy is copyright 2000, so standards have probably changed since it was published, but it remains an excellent general guide.

If you really don't have time for the specifics, (1) get the safety grounding right (2) ensure all touchable metal is continuous to the safety ground wire (3) high pot test the unit (4) ensure that user can't touch any surface that can cause a burn by touching (5) make the enclosure from wood > 1/2" thick so internal flaming liquids or electrical arcs can't set the enclosure afire (6) seal or divert openings so poking a wire through the opening can't touch a live conductor. The resulting unit will almost certainly not pass real testing, but might keep a demo unit from killing someone until you can get it back. If you do go the quick and dirty route, do ensure that you get the demo unit back; don't leave it there to cause future liability. Remember that in some countries, causing injury by failing to make a unit safe enough can be a criminal offense, not just grounds for civil liability. It's this last consideration that drives a lot of my view that you have to take this seriously.

Better yet, find a testing lab nearby and pay a few hundred to a thousand dollars for an engineer to do a quick advisory review that's less than a full inspection test.

Here are some specific things that I know have been required to pass certifications.
> Every single user-touchable metal part on the outside of the unit, including painted parts, must show continuity to the AC safety ground wire. Paint and coatings are specifically disallowed as insulation, as they wear through. It is possible that ancillary items like protective screens that pass 4200Vac high-pot to safety ground might pass under the reinforced-insulation rules.
> Use only UL/CSA/TUV/ETL/etc listed parts for any circuits involving the AC mains
> High-potential (called "hi-pot" in the trade) test the whole thing for 4200VAC between the AC mains wires in the line cord to any metal the user could theoretically touch, including poking their fingers through penetrable openings.
> Arrange the enclosure and all openings so that poking a finger or stiff wire through any opening cannot contact a conductor with a voltage of more than 42Vdc or peak AC. This is the enclosed-jack thing you refer to, and using enclosed jacks is one way to keep a wire poked through a jack from contacting live wires. Another is to make sure that the jacks have no hot wires behind them, or using some kind of barrier.
> Wire the safety ground wire on the AC entry to a single bolt/stud on the chassis used for no other purpose; the transformer mounting bolts are not to be used for this. There is a specific way of stacking the conductors that has passed testing: star washer, then ring terminal crimped to safety ground wire, then another star washer, then the nut holding this down.
> Do the AC wiring in such a way that the safety ground wire will break last if the AC cord is pulled out
> There are specific force pull requirements for pulling the AC wires without causing a hazard, but in general if the line cord does not enter the chassis in something like an IEC power entry socket, you should be able to lift the unit by the line cord without causing any possible hazard, including abrading the AC wire insulation.
> In general, ensure that no single component failure can cause an electrical hazard or fire. This includes any single wire breaking.
> One subtle implication of this is that AC mains wires should be cable-tied or dressed in a way that any one wire breaking at the end cannot contact secondary conductors. So cable tie AC wires in position or to one another so that one wire breaking does not result in a live wire flopping about. A more subtle instance is that the AC mains switch should be secured so that it can not rotate and twist the AC wires until they break or short. Ditto fuse holders, etc.
> Hmmm. Fuse holders. The incoming live wire to a fuse holder should go to the fuse holder terminal that breaks first and makes last as you re-insert a fuse so that replacing a fuse never leaves the exposed fuse end live.
> Creepage and clearance distances greater than 8mm between all AC mains conductors and secondary conductors, including especially on PCBs, turret/eyelet boards and point to point wiring.

This list just goes on and on. But it's a quick list, and many other points can prevent certification and at least in some cases allow your unit to kill someone now or in the future.

[R.G.]

I'm glad it helps you get things into perspective.

I have to be very clear about this though; my advice to you is to fully comply with all safety regulations wherever you sell amps. Anything less is you deciding to take a chance on legal liability for you/your company and also taking a chance on exposing other people to hazards.

If you decide to do less than full compliance, it is your personal decision, not my advice.

[Guy77]

Two Rock amps are CE and RoHS certified, at least they say so on their website.
https://www.two-rock.com/warranty-information/

Cracking one of these open and taking a look at how they wire their AC Main lines and their AC heater lines is a great place to start too.

I attached a pic of a Two Rock I had on my bench a while ago (Bloomfield) and they wire their mains fuse (the only fuse in their amps) differently than the standard way that involves going from the Hot AC line from the IEC plug to fuse. You can see in the pic they go from the HOT IEC wire directly to the MAIN power switch and then that goes to the fuse and then the fuse goes to to terminal board where other transformer wires are attached too. I am sure this was constructed in a fashion to meet CE approval. The heater lines also use large 2-3 watt 100 ohm balancing resistors. This amp used solid 18 AWG for everything on the power supply section and solid 20 AWG everywhere in the preamp. The input guitar jack was also enclosed. The second large brown switch you see to the left of the MAIN power switch is their 1/2 power switch.

heaters.jpg

I am curious as to how they wire their MAIN fuse. I don't have this amp any longer.

Our local CSA office charges $5000 cdn for CSA testing and approval certification.
My brother in-law used them to get CSA approval for ICE bath tubs (electrical powered) he has started building and selling.

Cheers
Guy

[R.G.]

I attached a pic of a Two Rock I had on my bench a while ago (Bloomfield) and they wire their mains fuse (the only fuse in their amps) differently than the standard way that involves going from the Hot AC line from the IEC plug to fuse. You can see in the pic they go from the IEC directly to the MAIN power switch and then that goes to the fuse and then the fuse goes to to terminal board where other transformer wires are attached too. I am sure this was constructed in a fashion to meet CE approval.

Very interesting photo. Thanks.

If there was a testing lab review, and the testing lab approved it, they approved it. A valid testing lab report approving the unit to standard is likely to be a defense to "reckless endangerment" suits in the USA and possibly so in the EU.

However, I have personal experience with units being failed specifically for the fuse being wired after the mains power switch. This makes me less sure that it's wired in a way that a different testing lab would approve it for the CE mark. If I were running Two Rock amps, I'd take special care of the approval documents I got from the testing lab in case I had to produce them some day.

But I am limited by my experiences, as are we all. Perhaps this is a point of contention and interpretation among the testing labs. The safety standards never say "do it this way, that is then safe". Instead, they say "if it does not withstand at least this much failure/stress/spacing/misuse, it is not safe". So the labs have to make judgement calls.

I also wonder if there is a fuse built into the IEC entry module. That's pretty common, and has been certified by labs many times.

As I said, interesting point/photo.

[Guy77]

I attached a pic of a Two Rock I had on my bench a while ago (Bloomfield) and they wire their mains fuse (the only fuse in their amps) differently than the standard way that involves going from the Hot AC line from the IEC plug to fuse. You can see in the pic they go from the IEC directly to the MAIN power switch and then that goes to the fuse and then the fuse goes to to terminal board where other transformer wires are attached too. I am sure this was constructed in a fashion to meet CE approval.

Very interesting photo. Thanks.

If there was a testing lab review, and the testing lab approved it, they approved it. A valid testing lab report approving the unit to standard is likely to be a defense to "reckless endangerment" suits in the USA and possibly so in the EU.

However, I have personal experience with units being failed specifically for the fuse being wired after the mains power switch. This makes me less sure that it's wired in a way that a different testing lab would approve it for the CE mark. If I were running Two Rock amps, I'd take special care of the approval documents I got from the testing lab in case I had to produce them some day.

But I am limited by my experiences, as are we all. Perhaps this is a point of contention and interpretation among the testing labs. The safety standards never say "do it this way, that is then safe". Instead, they say "if it does not withstand at least this much failure/stress/spacing/misuse, it is not safe". So the labs have to make judgement calls.

I also wonder if there is a fuse built into the IEC entry module. That's pretty common, and has been certified by labs many times.

As I said, interesting point/photo.

Hi R.G. thanks for chiming in , I hoped you would.

Yes I think what you said makes perfect sense in this amp:
The safety standards never say "do it this way, that is then safe". Instead, they say "if it does not withstand at least this much failure/stress/spacing/misuse, it is not safe". So the labs have to make judgement calls.

In this amp the IEC had no fuse in it.
In regards to the 18AWG wire I see now that it also prevents the wires from moving to touch other components if wire is broken, like you mentioned. Stranded wires would probably more easily move around if broken at an end.

Guy

[R.G.]

Good point. The 18 gauge solid doesn't flop around much, even with one end loose.

There's a lot of "this worked, let's do it again" in safety work. Given the reluctance of the safety standards people to say "this works and is safe", the testing labs have a real incentive to recognize what has passed before, and what is a glaring failure in their experience. This is what amplifiednation was looking for, I think - the list of "do this, it is safe".

I have participated in a number of formal product safety certification campaigns (the use of "campaigns" is not an accident here) and at times decided that the safety lab's goal in life was to find something to fail us on. The design engineers also developed a mental list of "do this, it's safe", and a list of "if you do this, they will fail you" and applied all of their previous tricks to pass certification in their designs with only one trip through the lab before ever submitting. The safety lab guys would invariably find some detail to make us fix. Maybe we were both playing a game.

The game cost about $10,000 per trip through the lab back in those days. It's amazingly more affordable today, with certification labs doing tests for what, $3500 to $5000 for a test, in lower-value and inflation-cheapened dollars. There were fewer certified labs back then, so maybe the competition has been good.

Before the CE mark in the EU, European nations each had different safety laws. Getting all that into one set of IEC standards was a great step forward. The CE era also marked a change in legal status. Pre-CE at some point, a certification by a national testing lab was a legal defense (in that country, at least) to liability for safety issues. With the Low Voltage Directive and CE marking, that's no longer the case. Much like in the USA, certification merely shows that you looked at safety and have some valid reasons to show in court that you're not reckless in ignoring safety issues. On the other hand, not certifying can be presented as evidence of willful disregard of safety issues by plaintiffs; certainly in the USA that could happen. My understanding of EU law on safety is that you can still be on the hook for damages, but (like so many legal issues, probably) not punitive damages for recklessness. It depends on what the judge and jury say. I am not a safety engineer, nor a lawyer, but I have been close enough to both of those to want to keep a wider distance from them.

I'm largely ignorant of any cases where a testing lab certified, then someone got hurt, and the testing lab then being sued for "unreliable certification" or false certification for a fee, but that might exist. I believe that fraudulent certification exists, just by looking inside some products and seeing things that got me failed in certification tests before.

For my personal work, I decided a long time ago that I would play it straight with safety. I don't want my mistakes to kill anyone or burn down their house. That's even worse than legal liability.