On 1/2/2024 11:13 PM, Jeff Liebermann wrote:
> On Tue, 2 Jan 2024 04:54:16 -0800 (PST), "
funkma...@hotmail.com"
> <
funkma...@hotmail.com> wrote:
>
>> Considering it's what I do for a living, I can give excruciating details of regulatory design requirements, for example:
>>
>>
https://www.iecex.com/dmsdocument/1589
>> "4 Capacitive Evaluations
>> 4.1 Overview
>> The ignition capabilities of all capacitive circuits under fault conditions must meet the requirements of IEC
>> 60079.11. Group II resistor protected capacitors cannot be assessed from the reference curves."
>>
>> "Piezo electric devices
>> The voltage generated by an impact on a piezo electric crystal needs to be considered in the ExTR
>> and compared again the acceptable energy level for the group."
>>
>> "5 Inductive Assessment
>> 5.1 Overview
>> The ignition capabilities of all inductive circuits under fault conditions must meet the requirements of IEC
>> 60079.11."
>
> Intrinsically safe in explosive atmospheres. I'm glad I don't have
> your job although I think it might be fun to run some tests. Almost
> as much fun as me learning how to use explosive forming to make
> parabolic dish antennas.
I have fun on a regular basis:
- we recently had to do battery testing of AA cells. This required
thermocouples at three points on the battery to measure temperature
increase of the cell discharging through a .5 ohm load. Most of them got
so hot that it detached the adhesive holding the thermocouples in place
and several spectacularly spewed dielectric inside the vent hood. Of 7
samples we tested only one met the HazLoc thermal limit requirements.
- Internal testing requirements for zener clamps in IS barriers:
increase current thought the device in situ until failure, sometimes we
get flames.
- Testing of plastic housing materials: Standards require a 30 day soak
at 20C over the operating temperature of the product at 90% RH, after
which the sample is transferred for a cold soak a 5C under the minimum
operating temperature for 24 hours. The sample is removed and
immediately impacted with a 1Kg steel ball from 70 cm height. Eye
protection is required.
- Long-term soaking of plastic materials in various hydrocarbon fuel
blends at elevated temperature followed by stress-to-failure physical
testing
- our Fuel-Depot gantry-mounted equipment is explosion-proof. Testing
requires measuring pressure from igniting an air-fuel mix inside the
enclosure. That pressure reading is then multiplied 4X and the enclosure
is tested hydro statically at that pressure. That generally translates
to ~ 500PSI depending on the enclosure size and the type of fuel. We
have holes in the ceiling tiles in the lab from heads of weakened M8 hex
cap bolts popping.
>
>> You should see the new testing requirements for opto-isolators
>> in the latest edition of 60079-11.
>
> I'm afraid to look. Is vacuum potting the opto-isolator in epoxy
> acceptable?
That's the old way of doing it, these days 'approved' opto-isolators are
simply hermetically sealed. One of our older products uses an older
style with epoxy sealing - Those are the ones we have to get tested to
the new standards, we have made the decision to redesign via a
'daughter' pca to carry newer, cheaper, more available parts.
>
> I'm seeing PCB's that no longer "air gap" an opto-isolator by milling
> a groove in the PCB. Instead, they're using optical emitters and
> receivers with a fiber optic cable in between. I'm not sure what spec
> change inspired this change.
I don't think that was a specification change. Fiber coupling ensures
less susceptibility to ambient noise at significantly reduced power, and
allows higher speed (smaller light/dark transient). However, I don't
think I've seen any discreet solutions like that in quite a while.
Everything used for data/signalling for the most part is monolithic
these days.
>
> The real fun is the focused power density at the end of an optical
> fiber with a few watts of IR behind it. I used one to destroy the
> "protective" caps that cover the transceiver ports. It seems like the
> perfect igniter for an explosive atmosphere and might work as a "laser
> welder". There was quite a bit of activity in about 2015 trying to
> conjure a fix. The choices were (1)reflect the light and destroy the
> laser, (2)use a "protective" cap and watch it either explode or burn,
> (3)let the light pass and turn it into a rack mounted death ray, or
> (4)change the design somehow so that it's reasonably safe. I have no
> idea what they did.
Lasers are a special case in HazLoc world. I only know since there are
specific chapters in the standards, but we don't have any laser
applications.
>
>> They went from one paragraph to three pages of assessment. The last quote we got to assess to the new standard showed a $5000 price increase - solely- to the additional testing required. The alternative is to specify pre-approved devices, which would require a rather substantial design change to the existing circuit. It would save us a month or two of testing and that $5K, but the cost of the devices is an order of magnitude higher - I'm leaving that decision up to the business management.
>
> Yep. Standards grow like weeds. In many ways, standards are a good
> thing. Unfortunately, they don't expire. When I helped design the
> AN/SRD-22 radio direction finder, one of the specs that the USCG (US
> coast guard) required was that all power wiring comply with some
> obscure spec that nobody could find. Filing an exception during the
> bidding process was a guaranteed way of not winning the bid. So, we
> played dumb and hoped for the best. We eventually discovered that it
> was for "rope lay electrical cables" left over from building cruisers
> and battleships during WW2. Nobody knows how it snuck into the specs
> for a modern (at the time) direction finder.
HazLoc hasn't been updated to exclude hi-pot testing for battery
operated equipment. The only out we've found is an 'end-run' one agency
gave us on one of our products because there were no isolation
transformers, but the standard doesn't specifically allow for that
exclusion. It was a judgment call from the reviewers - something they
have latitude to do. One reviewer looking over one of our accessory
components stated 'we really shouldn't have to do this, it makes no
sense'. It's basically something that got left over and someone has to
petition for a change.
>
>
--
Add xx to reply