Powering up puzzle

[MonkeeRench]

With the power switch off at the controller, when I plug in the USB power supply, all the LEDs seem to go thru test-blinking, then the Power and Mode LEDs remain on, just as they are when I turn on the power switch.  The red LED on the connected E-Field pre-amp board is also always lit.
Confused -- shouldn't the power switch shut off everything?

[kevinmcc]

That should not happen if the switch is correctly connected.

The jumper should be removed from pins 5&6 and the switch will then connect them.

Note the pins with thicker red lines in lower left corner of the board layout.

http://www.blitzortung.org/Compendium/Ha..._3_brd.png

[MonkeeRench]

That should not happen if the switch is correctly connected.

The jumper should be removed from pins 5&6 and the switch will then connect them.

Note the pins with thicker red lines in lower left corner of the board layout.

http://www.blitzortung.org/Compendium/Ha..._3_brd.png

Thanks, Kevin, that worked like a charm!  I really appreciate this forum in filling in the blanks of documentation that somehow eludes me.  Another puzzle is on the layout of 19.3b which does not show the "Power" thru-holes where I connected the external switch discussed.

[kevinmcc]

Are these the instructions you were using?

http://en.blitzortung.org/Compendium/Doc...B_20_1.pdf

Not sure what you mean by power thru-holes.

[MonkeeRench]

On my PCB 19.3b board (Sep. 2016), next to the pre-installed USB socket are two thru-hole pads labeled "Power", to which I connected my external power switch. On the layout I see no such thru-holes, only a single unlabeled symbol in the vicinity of the USB and the 5-6 header terminals.

[kevinmcc]

Must be an undocumented revision. Maybe on of the Devs can let us know?

[DelandeC]

On 19.3b, you have to close jumpers 5 and 6 to power up the all system.
Jumpers 1 and 2 have to be closed also to power the pre amplifier boards.

http://fr.blitzortung.org/Compendium/Har..._3_brd.png
http://fr.blitzortung.org/Compendium/Har..._3_sch.pdf

[allsorts]

Close. B-)

Short pins 5 and 6 of the S1 jumper block to apply power to the system.
Short pins 1 and 2 to provide +5 V to the E-field preamp.
Short pins 3 and 4 to provide +3.3 V to the E-field preamp.
Leave 1 and 2, 3 and 4 open to not power the E-field preamp.

Bad things are likely to happen is you short 1 and 2 at the same time as 3 and 4 or short pins 1 and 3 as doing so connects the +5 V rail to the +3.3 V rail...

The H-field pre-amp is powered directly from the + 5 V rail so is on when ever the system is powered up.

[kevinmcc]

I still think they should add a protection diode to the 3.3 V rail.

[BobW]

I still think they should add a protection diode to the 3.3 V rail.

You are right. Nice shot nearby and the 3.3 volt rail is going high. There's a choke and a cap that will help keep it down -spike smoothed by choke, and eaten by cap. Same deal with 8 pin to amp except it's direct connected to +5, so what comes in gets on +5

Coax  - Blitzortung or otherwise (CATV, Satellite, etc) - should go to ground block bonded to building electrical ground, and a surge suppressor can't hurt. Like this:

https://www.3starinc.com/dual_port_coaxi...26fd6%3D32
https://www.3starinc.com/f_male_to_f_fem...ector.html
 
Same deal with the 4 pair to the preamp. Somthing like this. Be aware I've ordered one, and will be looking inside it to see if all 4 pairs have diodes and what kind. Nasty post on Amazon if it's crap. 

https://www.amazon.com/gp/product/B00UYX...UTF8&psc=1

[cutty]

If this makes someone feel secure, why not????

However, for argument:
I spent my 35 year career before retirement working with, among others, devices damaged by Lightning Impulses.. either direct or on power line. 
With all respect, I believe you're wasting your time and money.

First,  a lightnng impulse isn't a single spurt of electricity... its leaders, discharge, secondaries, broad frequency bandwith, scary as heck.
... scads of volts, scads of current potential, who knows how many gauss M... etc... and un-predictable in manifestation

Second,  remember that the system, because of the external PS, should be completely isolated from AC / Earth ground...
... and if I remember correctly, that "ground" would be a 'perfect earth return'... ha, ha....
Now, external EMF would be looking for a ground.... which it can't theoretically find through the system,
unless you've added one....

What happens on the secondary discharges if the initial discharge 'blows' a so called protect diode? 
What happens if the 'discharge current' of the diode then flows along the tiny ground paths of the controller, for example... ?
These are in large part 3.3 - 5 volt devices. The circuit traces are fractions of millimeters apart.

Protect devices would likely only "prove a negative" that is to say, "well, that stroke didn't fry me... must be working"...

Various scenarios  were discussed at various depths during development... even possible 'fusing'....
the consensus generally was "In event of HIGH Emf or direct strike you'll still get toasted... "..

No ground is likely to be perfect, and some "protect" devices require that treacherous
'2nd connection' to ground... which 'bypasses' or 'offers' an alternative to the 'isolated' design condition...   
..."protects" often result in "circuit board path arcs" rather than simple component damage... sometimes because of a second 'ground escape'.... 

Properly installed and mounted, the E amp interface is about as secure as any device exposed to atmosphere.
If you have 'grounded' your controller, to minimise noise, or just because you thought it a good idea,
you may have actually opened up the possibility of arcing by providing
a higher' resistance path to ground' through a "coax surge protector"  which may direct the leader surges into
 a controller's "lower resistance"... multiple paths... when the 'controllers' path to earth is actually a fraction of an ohm
because of adding a 'ground' as mentioned above... it could easily be
lower than the 'air' to ground impedance... when otherwise the 'air might have completed the initial discharge directly to earth,
if it had not been offered an alternate path because of the protect!
(dang, that's a screwy paragraph... whatever... )

Conversely, a "standard' ungrounded controller can only connect to earth ground through a shielded RG45 cable
assuming the routers are not 'isolated' from AC ground... which is an incidental reason why we
suggest unshielded RG45 to any router... no multiple grounds. In fact, shielded RG45 to the router can result in 'ground loop' noise.

 Now, suppose the device does fire...  is it a one-time protect-then-replace technology, ? 
How do you know if has 'worked' and is now defective?
A 'reset type device? Is its reset response time in the nanoseconds?  ...
otherwise secondary discharge blows you away...
If this were an issue, obviously more kits would be shipped to 'dead stations' then are shipped to 'new installs'...
We've had 2 or three in Americas Region (3) since system Green rolled out, maybe... (I could be mistaken on numbers)
and they seemed (from my poor memory) related to power line over-voltage due to Lightning on AC line.
or direct strikes... which few devices survive regardless of 'protects'.
Remember the "FIRST" from above... direct hits are just too dang bad. Sorry. That's Lightning,

However, If it makes you feel more secure, as above, why not... ? 
Then check your E-field reception and delays to make sure it hasn't been affected by your modifications.

Cheers!

[BobW]

Not going to have a big discussion here about this. While there are much better devices for the purpose than what I referenced that do survive better, what I suggested doing is covered by article 810 of the NEC. It's tough to survive a nearby flash, but pretty much anything that picks of current at point of entry helps.

[allsorts]

I still think they should add a protection diode to the 3.3 V rail.

Can't use a series diode, it would drop to much voltage in forward mode.

Parrallel zener? Maybe but it has to have a high enough voltage such that an "enthusiastic" nominal 3.3 V supply doesn't cause it to conduct but low enough that it'll conduct before the nominal 3.3 V components get fried. The +5 V regulator is a nominal 250 mA device but has a typical short circuit current of 500 mA, so this 3.5 V has to dissipate 1.75 W or more...

Best not to fiddle with the power jumpers with the unit powered up.  B-)

[allsorts]

Not going to have a big discussion here about this.

Wise, lightning protection is up with politics and religion.

what I suggested doing is covered by article 810 of the NEC.

An American thing, this side of the pond finds it all a bit weird.  The UK and probably Europe as whole don't "ground" feeders, phone lines(*), etc where they enter the property. Exposed conductive parts (pipework, equipment chassis (unless double insulated, fittings etc) within a building are bonded to the "Main Earth Terminal" (MET) and that is connected to any conductive service pipes at the point of entry as well as the supply earth. But this is to ensure a low earth loop impedance for fault protection and to ensure that all exposed conductive parts are at the same potential.

Lightning "protection" in the UK is done with lightning rods connected to low impedance earth spikes via 1 x 3/16" ish copper strip. These leak the charge building up between cloud and ground, hopefully keeping the potentail below that required for a strike to occur. But you'll only find lightning rods on large commercial type buildings not on peoples homes.

It's tough to survive a nearby flash, but pretty much anything that picks of current at point of entry helps.

A nearby strike is going zap stuff or not. A few pence device (MOV, gas discharge tube, WHY) isn't going to make much difference, the energy levels are just too high. I'm with Cutty, if you have some "protection" device how do you know if that device is still functional after a nearby strikes or is that strike?

I don't know the details of that American electrical code but I doubt it is written with much attention paid to keeping noise out of sensitive RF receivers. With the E-field screen grounded when there is a storm in the vicinity it'll start to act like a lightning rod. There will be current flows and thus voltages in the screen. The E-field is an unbalanced input it'll "see" the screen potential moving around as a signal.

Found the spec of that inline coax supressor, triggers @ 65 V, 500 A, 2 x 10 us. If I've got the maths right [ (65 * 500) * 20^-6 ] = 6.5 joules, a tiny amount of energy. I like the claim "Does not degrade with repetitive strikes" decent strike and it'll be degraded all over you back yard! To be fair it doesn't make any claims about lightning.

Sorry BobW, as you can tell I'm of the view there is not a lot you can do about lightning, at least not without spending a small fortune on protection. Switching transients on your incoming electrical power is another matter but they don't have anything like the energy of a lightning strike.

(*) Just remembered phone lines have a gas discharge tube across the pair. No ground connection though.

[Jon Point]

"...as you can tell I'm of the view there is not a lot you can do about lightning, at least not without spending a small fortune on protection. Switching transients on your incoming electrical power is another matter but they don't have anything like the energy of a lightning strike."

Lightning protection is all about layout, not just fitting protective devices. I have worked on dozens of sites from MF to EHF that are constantly bombarded by lightning with few ill effects. Here, we are dealing with LF/VLF, and the only issue is that lightning 'frequencies' are in that realm. Any solution that blocks lightning will affect our receivers. Interestingly, the energies in lightning that cause the most damage (like blowing-out switchboards and the like) are 'long-tail events', which exist at much lower frequencies again 1kHz - 10kHz. If we can safely conduct these away from our systems, the remainder of energy at higher frequencies is not going to hurt (much).

My solution, being on the coast in an area that cops a lot of lightning in the warmer months, is to keep it inside the roof. If I mount it outside and an adjacent tree is struck, I would expect a side-strike to affect the system.

BTW, I am a ham radio operator and have a number of wire antennas covering HF, well grounded and protected with GDTs. Although we have experienced a number of strikes (recently repairing the roof gutter after the last tree strike sent timber everywhere), the only 'damage' we ever had was a cable modem that may have simply failed due to being 8 years old! It was replaced by the telco.

[kevinmcc]

I still think they should add a protection diode to the 3.3 V rail.

Can't use a series diode, it would drop to much voltage in forward mode.

Parrallel zener? Maybe but it has to have a high enough voltage such that an "enthusiastic" nominal 3.3 V supply doesn't cause it to conduct but low enough that it'll conduct before the nominal 3.3 V components get fried. The +5 V regulator is a nominal 250 mA device but has a typical short circuit current of 500 mA, so this 3.5 V has to dissipate 1.75 W or more...

Best not to fiddle with the power jumpers with the unit powered up.  B-)

Series diode would only drop voltage 0.2 Volts, and the power is for the AMPs which need a minimum of 2.5 Volts.

The voltage regulator needs a minimum of 2.0 Volts, so no problem there either.

I have not know if all the 3.3 V components can handle 5 V, but a protection diode would protect them.

As for fiddling with the jumpers, yes not to is best, but you know as well as I people will.

[allsorts]

0.2 V forward drop? Wanders off to google... ah schotty 0.2 to 0.4 V, rather lower than I thought, my bad.

Only really need the series diode in the +3.3 V feed to the S1 jumper block that powers the E-field preamp. That is where a mis-jumpering or tweezer slip is going cause the +5 V and +3.3 V rails to come together. If some one chooses to power a board just resting on a work bench along with nuts/bolts and wire off cuts that's their probelm. B-)

The Blue E-field pre-amp uses the ADA4891 op-amp, minimum supply of 2.7 V, a silicon diode would drop to much. Higher end schotty would bring the supply at the controller board down to 2.9 V. Assuming 10 mA pre-amp load, 100 m of coax will only drop 0.025 V, so plenty of head room. Assuming the pre-amp design will still perform properly with a supply below 2.9 V.