Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: Balanced line lightning protection and pass through  (Read 5482 times)
KK6RPX
Member

Posts: 113




Ignore
« on: September 04, 2017, 10:19:43 AM »

Hi All,

Just put up a 160 meter loop fed with 450 ladder line. I'm loving it, beats the pants off my dipole. So, now I need to deal with a winter worthy pass through and I'd like to add some lightning protection. Not for a direct strike by near hits or static build up. I see these out there:

https://www.arraysolutions.com/index.php?_route_=surge-and-rf-protection/as-309h

But YIKES! Over $100 with shipping and it appears to be two gas discharge tubes and a couple of caps. Doesn't seem like the impossible dream to build one. They specify the gas discharge tube voltage but not the cap size/voltage. I'm running a max of 600 watts, usually around 400. And ideas as to cap size? And I'm guessing 2kV or so would be ok for voltage?

I'm also going to add a 1:1 current balun for a choke and a knife switch to disconnect during possible lightning events. Should the knife switch go before the choke or after? Also going to have to figure out how to weatherize it to keep the contacts clean and yet be able to access it easily if need be.

As to the pass through, the shack is in a trailer so metal walls. I have a metal pass through, I removed a pane of glass in a louver window and have bulkhead fittings. It also has a metal hood shielding it from weather and bringing balanced line through the mess on the other side might be tough. I think it might be easier to just punch a hole through the trailer wall and use twin coax for the feed through the wall.  Thoughts on that?

Thanks!

Jonathan
Logged
KD8IIC
Member

Posts: 648




Ignore
« Reply #1 on: September 04, 2017, 10:50:33 AM »

 Wireman has the balanced line arrestor spark-plug assembly that mounts on a standard 8 foot ground rod.
Just make sure you hacksaw off the bottom caps to let the moisture out though.
As far as wire feed through goes, I'm running TV Twin-Lead through the window sill panel from MFJ. Roll ur own?
Dowel rods keep the burglars out doors if they are inclined to pry open the window.
Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #2 on: September 04, 2017, 11:42:11 AM »

You can build one and here are the electronic components.

Parts at DIGIKEY

BOURNES 2095-270-BLF, $1.41   2700V, 5000A gas discharge tube

OHMITE OY105KE, $2.51 1MEG resistor, 2W, 20kV pulse rating

AVX SV17HA802JAR, $11.14  8nF cap, 3kV, NPO

HAMMOND 1591LFLBK, $5.38 ABS box (you might want to select a different box)



« Last Edit: September 04, 2017, 12:02:32 PM by KH6AQ » Logged
N3QE
Member

Posts: 4880




Ignore
« Reply #3 on: September 04, 2017, 12:38:44 PM »

I use two spark plugs in a grounded metal plate right outside the house, where the ladder line enters.

The "out of box" spacing works fine for me, for powers up to legal limit.

Spark plugs get kinda rusty in this use, I have to replace every few years.
Logged
KK6RPX
Member

Posts: 113




Ignore
« Reply #4 on: September 04, 2017, 04:59:47 PM »

Hoe do the spark plug arrestors compare to the Array Solutions arrestor?

KH6AQ, thanks much for the part numbers! Looking at the picture it appears that the schematic is:

ant ---- cap--- transmitter
       |-- gas DT ---- gnd
       |-- resistor --- gnd

Also, just wondering if you know Bob, KH6AS. I ask since you are almost sequential call signs you might have tested at the same time and place.

Thanks for the help!
Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #5 on: September 04, 2017, 08:10:03 PM »

I don't know KH6AS and have had this call for only six months but you're probably right that KH6AS and the original KH6AQ tested right about the same time.

Your diagram does show the correct component connections.

GDT vs spark plug? This Littlefuse paper lists some of the GDT advantages, such as consistent firing voltage, lower arc voltage, and that the gas gap is protected from the elements. They mention an arc voltage of 15 volts. And the cost of a GDT is comparable to a spark plug.

http://www.littelfuse.com/~/media/electronics/product_catalogs/littelfuse_gdt_catalog.pdf.pdf

Arc Voltage is much lower for the GDT
This paper compares air gaps (similar to a spark plug to a GDT). They measure the arc voltage to be 200 volts for the spark plug and close to zero volts for the GDT.

https://www.orcadxcc.org/content/pdf/Notes_on_Coaxial_Suppressor_30sep2013.pdf

« Last Edit: September 04, 2017, 08:28:14 PM by KH6AQ » Logged
KK6RPX
Member

Posts: 113




Ignore
« Reply #6 on: September 04, 2017, 08:47:57 PM »

Great information. Thanks so much! OK, the gas discharge tube is a clear winner. One question, how does one tell when a GDT has been blown? One good thing about the spark gap is that one can see if the gap has been cooked.

Does the cap really need to be a NPO? Would variation in capacitance from temperature really matter? Not that I'm stingy, it's just that I'm stingy... :-)

And does anyone have thoughts about the pass through?

Thanks folks!
« Last Edit: September 04, 2017, 08:54:13 PM by KK6RPX » Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #7 on: September 04, 2017, 10:18:21 PM »

I selected the COG capacitor for low DF (Dissipation Factor) so it will handle higher RF current. The NPO cap has a DF of 0.1% while the Kemet C330C103KHR5TA X7R cap has a DF of 2.5%. I estimate the NPO is good for 10 amps.

The 8nF NPO cap has an Xc of 6 ohms at 3.5 MHz. With a DF of 0.1% its ESR is 0.001 x 6 ohms = 6 milliohms. At 10 amps the power dissipation is 0.6 watts. The 10nF X7R cap has an ESR of 0.1 ohms and at 2 amps it will be pretty hot.
« Last Edit: September 04, 2017, 10:23:00 PM by KH6AQ » Logged
N3QE
Member

Posts: 4880




Ignore
« Reply #8 on: September 05, 2017, 05:40:12 AM »

Arc Voltage is much lower for the GDT
This paper compares air gaps (similar to a spark plug to a GDT). They measure the arc voltage to be 200 volts for the spark plug and close to zero volts for the GDT.

https://www.orcadxcc.org/content/pdf/Notes_on_Coaxial_Suppressor_30sep2013.pdf

I don't think the voltage after arc occurs is awful relevant. The voltage before arc occurs is more relevant. I would say that a typical spark gap is kV or larger and the small coaxial GDT are 90V to a kV.

Also keep in mind they are specifically looking at coaxial (50 ohm nominal) surge suppression, and voltages on ladder line have to be many times larger because the impedance is many times larger.

If you are running legal limit through 450 ohm ladder line, and SWR is perfectly 1:1 (unlikely!), then the RMS voltage across the ladder line is given by:

P = E*E/R

1500 = E*E/450

675000 =  E*E

E = 820 Volts (RMS)

E = 1150 Volts (Peak)

Now if your SWR is not 1:1 you may have voltages several to many times larger than this.

Quote
Arc Voltage is much lower for the GDT

I hope you now see why lower arc over voltage GDT is not possible for my transmit feeder. (A couple times in past few years my link-coupled tuners have gotten too close to the cinder block wall - a good chunk of an inch, not a millimeter! - and drawn these incredible arcs.)

For my receive-only antennas I can see choosing a GDT.
« Last Edit: September 05, 2017, 05:43:10 AM by N3QE » Logged
KK6RPX
Member

Posts: 113




Ignore
« Reply #9 on: September 05, 2017, 07:13:19 AM »

OK, your cap choice makes perfect sense. Thanks for the detailed explanation, I learned something.

N3QE, the GDT is rated for 2700V. You don't think that is adequate?

Thanks!
Logged
N3QE
Member

Posts: 4880




Ignore
« Reply #10 on: September 05, 2017, 07:19:16 AM »

N3QE, the GDT is rated for 2700V. You don't think that is adequate?

If you are low-power (100W or so) only it may be fine. "May" depends on whether your SWR ever gets into the 10:1 range.

One dit from a legal limit amp at 2:1 SWR and those fingernail-sized GDT's are goners.
Logged
KK6RPX
Member

Posts: 113




Ignore
« Reply #11 on: September 05, 2017, 07:45:24 AM »

The loop antenna is new and I have not yet figured out how to attach my analyzer to balanced line so I can't tell you exactly what the SWR is. I run a max of 600 watts. I'm not sure what the maximum voltage available on the GDT is.

However, the manufacturer says:

* 4 kW CW / 8 kW PEP
* Ladder line or open line impedances: 100 to 800 ohms

That would be a max of 16:1, right?
Logged
N3QE
Member

Posts: 4880




Ignore
« Reply #12 on: September 05, 2017, 07:58:23 AM »

The loop antenna is new and I have not yet figured out how to attach my analyzer to balanced line so I can't tell you exactly what the SWR is. I run a max of 600 watts. I'm not sure what the maximum voltage available on the GDT is.

However, the manufacturer says:

* 4 kW CW / 8 kW PEP
* Ladder line or open line impedances: 100 to 800 ohms

That would be a max of 16:1, right?

There will be certain frequencies where you have a real hard time getting a good tune and the SWR at the shack is bigger than 16:1.

You kinda hope these will fall outside the ham bands but I don't know how to guarantee it. Modeling the antenna and feedline can be a good start but is not a guarantee.

By changing the length of the feedline, you can move around these horribly bad spots.

If you make your first tests at low power (20W, or at least less than 100W) you can be pretty sure you won't blow up a GDT even in the case of high SWR, as long as you are at low power.

Depending on how robust your tuner is, it is possible the GDT will be there mostly to protect your tuner from tune-up arcovers. There's some real value to that especially in the case of wimpy tuners that cost hundreds of dollars. Tuners can also fail due to overcurrent heating especially high circulating currents but with ladder line feeds to unknown antennas you will more likely see arcovers rather than overheating.
« Last Edit: September 05, 2017, 08:01:19 AM by N3QE » Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #13 on: September 05, 2017, 11:11:58 AM »

Arc Voltage is much lower for the GDT
This paper compares air gaps (similar to a spark plug to a GDT). They measure the arc voltage to be 200 volts for the spark plug and close to zero volts for the GDT.

https://www.orcadxcc.org/content/pdf/Notes_on_Coaxial_Suppressor_30sep2013.pdf

I don't think the voltage after arc occurs is awful relevant. The voltage before arc occurs is more relevant. I would say that a typical spark gap is kV or larger and the small coaxial GDT are 90V to a kV.

Also keep in mind they are specifically looking at coaxial (50 ohm nominal) surge suppression, and voltages on ladder line have to be many times larger because the impedance is many times larger.


It looks like you're correct that the arc voltage is not awfully relevant.

Let's explore the 50 ohm coaxial environment and work our way up to ladder line. I'll make some simplifying assumptions to get the model started.

The antenna is 20 meter tall monopole connected through the GDT circuit to a 50 ohm load.

The lightning E-field rises at 10kV/m for 8us then decays to zero in 20uS. The peak E-field is 80kV/m.

The voltage induced into the open-circuit monopole is 1/2 the E-field. Let's say the antenna source impedance is 50 ohms and the antenna model driving the GDT circuit is a voltage source rising at 5kV/us with a 50 ohm source impedance.

The GDT is a 750V device that fires at 1750V. The overvoltage required to fire depends on if any of the gas in the gap is ionized. It can fire as soon as 750V is reached or it could take an 1000V extra.

The GDT shunt is coupled to the 50 ohm load via a 10nF capacitor. When the GDT fires the arc voltage is 15V. The GDT fires at about 600ns and there is no protection, other than the 10nF cap, until then.

Running this in SPICE we see the voltage across the 50 ohm load rise to 1100V at which time the GDT fires and the voltage reverses to -550V. The voltage decays exponentially with a 50ns time constant. The energy dissipated in the 50 ohm load is 7 mJ.

Now we set the arc voltage to 200V to simulate the spark plug. Running this in SPICE we see the voltage across the 50 ohm load rise to 1100V at which time the GDT fires and the voltage reverses to -650V. The voltage decays exponentially with a 50ns time constant. The energy dissipated in the 50 ohm load is 6 mJ.

The frequency content is low except when the GDT fires. Its fall time can be <1ns and that will send a 550V fast rising pulse to the load.




« Last Edit: September 05, 2017, 11:19:47 AM by KH6AQ » Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #14 on: September 05, 2017, 12:18:01 PM »

With the 10nF cap, and GDT or spark plug, the let-thru energy is 7 mJ. Without the GDT or spark plug the let-thru energy is 1200 mJ. Without anything the let-thru energy is 75 J.

With the cap, and GDT, the peak current into the 50 ohm load is 21A before firing and 15A afterward. This current is low enough that a steering diode circuit with a two 3VDC biased rails offers protection for a receiver. Adding 50 ohms in series with the 10nF cuts the receiver peak current to only 1/4A for the cost of 6 dB higher noise figure.

The DC rails must be much high for a 1.5 kW TX and (~750V) and that makes it ineffective for a receiver unless a lower voltage steering diode circuit is inside the transceiver between the TX and RX stages.

I've designed "hybrid" arc protection circuits using a GDT followed by one or more steering diode stages. Between each stage there was a limiting impedance. So, combining the two devices (GDT and steering diodes) is my idea of the ultimate protection circuit.
« Last Edit: September 05, 2017, 12:28:09 PM by KH6AQ » Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!