Warning: Creating default object from empty value in /homepages/11/d132647312/htdocs/Amfone/mkportal/include/SMF/smf_out.php on line 47
Link Coupled Tuner Version Question




 
The AM Forum
April 14, 2021, 03:51:07 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: 1 [2]   Go Down
  Print  
Author Topic: Link Coupled Tuner Version Question  (Read 16835 times)
0 Members and 1 Guest are viewing this topic.
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2729



« Reply #25 on: February 03, 2014, 08:07:59 PM »

Usually the inductance of the transmitter side of the link is to be equal to 50 Ohms reactance. The series capacitor must resonate with the inductance at the lowest freq for the given band and provide any needed additional capacitance transferred from the antenna side of the link.

The voltage requirement is driven by whether tuning will be done at high power. If not, the spacing need be no larger than the typical loading cap for the given power. I've used a BC receiver tuning cap in this application with no problems at 1500 PEP. The assumption is that the impedance with be rather close to 50 Ohms once tuned.

Hope this helps.

Also an interesting way to do the coil tap connections at the link below.

http://www.pituch.net/Steve's%20Page/Radio/500%20Watt%20Link%20Coupled%20Tuner/500%20Watt%20tuner1.html

AMfone group,

  Have a question regarding capacitors in the K1JJ design as well as general link coupled tuners.  Specifically I'm concerned about the input capacitor.  I'm seeing 1700pF for input network capacitance on LB Cebik's write up and it appears 1kv is sufficient from what I understand.  However, I'm seeing 2kv in the K1JJ.  From what I understand the voltage rating depends on the Q and the type of tuner.  Looking at the MFJ sight I can see a plate load capacitor at 800pf rated at 1.1kv with a minimum capacitance of 25pf at $57.  
http://www.mfjenterprises.com/Product.php?productid=282-2573-1
  They also have an 800v unit at only $26 that goes 20-823pf.  The 1.5kw unit is considerable more expensive and not worth it.  
  So, I'm wondering if anyone is using 800 to 1.1kv caps to tune the input of an LCT with full legal limit?  
  On another note, I've recieved a quote to get the butterfly capacitors rotor/stator dies made along with production cost with tumbling/polishing.  For the output/balanced section this is going to be the way I go.  I'm going to invest in the dies and buy in quantity of 1,000 or more rotor/stator sets at a time and will be buying soon enough.  The higher value 1700pf units are not currently in the realm of something I can make as these seem to be soldered together like a broadcast capacitor.
  
  Just added a modification to this reply for something that is a bit puzzling on the input cap.  The LB Cebik (1960ARRL book) value for 80 meters is showing 900pf.  The modification page for the Johnson Matchbox to make it similar to the Annecke at 1kw output is a 1.5kv 330pf for 80mtrs.  I'm not getting the huge difference in values for C1 yet.   Any advice on C1 would be greatly appreciated.  

73,
Mike

Logged
WZ5Q
Member

Offline Offline

Posts: 109



WWW
« Reply #26 on: February 04, 2014, 07:00:07 AM »

Howdy Mike,
These are some of the things I learned (mainly from Mr. Cebik's grand site) when I researched and built my LCT...

 
  Just added a modification to this reply for something that is a bit puzzling on the input cap.  The LB Cebik (1960ARRL book) value for 80 meters is showing 900pf.  The modification page for the Johnson Matchbox to make it similar to the Annecke at 1kw output is a 1.5kv 330pf for 80mtrs.  I'm not getting the huge difference in values for C1 yet.   Any advice on C1 would be greatly appreciated. 


I would think that the difference in C1 series capacitor values is due to the difference in Inductance of the input link coils of the 2ea LCT designs. Up to a point, you can increase the Inductance value of the coil to then thereby reduce the amount of capacitance needed to resonate it at the lowest frequency of interest.


AMfone group,

  Have a question regarding capacitors in the K1JJ design as well as general link coupled tuners.  Specifically I'm concerned about the input capacitor.  I'm seeing 1700pF for input network capacitance on LB Cebik's write up and it appears 1kv is sufficient from what I understand.  However, I'm seeing 2kv in the K1JJ.  From what I understand the voltage rating depends on the Q and the type of tuner.


The voltage rating for the input series capacitor depends upon the capacitor reactance and the power level. For the power level, you have to go by the instantaneous peak voltages that will appear across the cap plates remembering that off resonance voltage peaks are considerably higher. If you use the recommended values of capacitance and inductance from LB Cebik's site, then the voltage seen across the input series cap plates will be around 400v at 1500 watts. You need to put a safety reserve into this and would generally use a cap rated at least double in voltage which would be around 1Kv. If you reduce the recommended input series capacitance by increasing the inductance of the Primary input coil, then the capacitor reactance increases as well as the peak voltage across the cap plates and a higher voltage cap will be needed.

Basically what all this means to is to use the biggest voltage rating cap you can find and then you won't have to worry about it.  Grin

I decided to go the brute force direction in my LCT build. I could not find the split stator or dual differential caps at the high voltage ratings I wanted, so I decided to use huge Vacuum Caps for the input and output circuits. I also opted to use plug-in Inductor Coil sets so I would not have to band switch and could optimize each coil set for my particular antenna system.

Take Care,
Logged

Mike
WZ5Q
From Deep Down in the Dark Dismal Swamps of the Calcasieu
http://www.wz5q.net
ke6cvh
Member

Offline Offline

Posts: 19


« Reply #27 on: February 05, 2014, 12:23:01 PM »

Hi Mike,

  I've seen your LCT web page many many times and it is awesome, thanks for such inspiring work like so many others who contribute to the hobby!!  I've got a quote on dies for the butterfly cap and will get another one from a different USA manufacturer as a sanity check before buying anything but will be producing my own butterfly caps soon enough.  Will be using 6061-T6 instead of that 7075 stuff because better electrical conductivity for a small bit of extra money.  What is the problem is making a butterfly cap with only 1kv voltage. The spacers would be more like washers and any slight bend in the rotor would be a disaster.  That said, I've been looking for a good source for the caps on the input of the network.  Was on Epay last night and saw there is a gent in Tawain selling caps real cheap and it looks to be the same as the MFJ units.  I'm thinking they are selling to MFJ and whoever else they can.  They had more than 10 pieces in two different styles, here are the links I saw:

http://www.ebay.com/itm/High-Voltage-Air-Variable-Capacitors-40-445pF-Amplifier-Tuner-REPAIR-DIY-/161086324757?pt=LH_DefaultDomain_0&hash=item25817e4015

http://www.ebay.com/itm/High-Voltage-Air-Variable-Capacitors-30-250pF-Amplifier-Turner-REPAIR-DIY-/160999183137?pt=LH_DefaultDomain_0&hash=item257c4c9321

http://www.ebay.com/itm/High-Voltage-Air-Variable-Capacitors-22-360pF-Amplifier-Turner-REPAIR-DIY-/151216034189?pt=LH_DefaultDomain_0&hash=item23352d918d

  That last link for the cap going for $20 is just asking to be modified.  It is 1kv rating and 360pF while only 1.56" long.  Not sure what size that all-thread is on it but if I bought 5 of them for a total of $100 then I could potentially get an 1800pF 1kv cap that is going to be under 8 inches long.  I cannot get plates/rotors stamped that cheap in USA so likely will only do the butterfly in the USA and this one overseas.  It will also need the last of the 5 sections switched in/out seperate from the other 4 sections like the Palstar does giving a minimum capacitance of 22 with a max of 1800pF.  The corners are a bit sharp on it but the voltage is not going to be high enough to make a difference I think.  Any ideas to make the 1800pF cap in a better way (in the USA) I'm open to.  I'll let you know what the seller says about a modification possibility on this cap and what size all-thread they are using for it.
  There are alot of different angles to this project that I'm working on still but thanks for the input on the capacitors. 

73,
Mike
Logged
ka4koe
Contributing
Member
*
Offline Offline

Posts: 1154


It's alive. IT'S ALIVE!!!


« Reply #28 on: February 05, 2014, 01:54:48 PM »

Reference

http://amfone.net/Amforum/index.php?topic=5931.0

Can I adapt the classic K1JJ design to match my inverted L? I realize the OUT POOT is balanced. If circulating currents are an issue, then I can just use a current balun on the feedline. What say you?

Tnx

Philip
KA4KOE
Logged

I'm outta control, plain and simple. Now I have a broadcast transmitter.
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2729



« Reply #29 on: February 05, 2014, 04:21:32 PM »

Does the cap on the transmitter side of the link need to be a butterfly type?
Logged
ke6cvh
Member

Offline Offline

Posts: 19


« Reply #30 on: February 08, 2014, 12:59:10 AM »

Hi Phillip and Steve,

  Sorry about the delayed response been crazy busy the last couple days.  I know the Johnson Matchbox could tune but to a different impedance range if I understand by strapping one of the outputs to ground making it unbalanced.  The reason why I'm looking into using a butterfly is because there will not be a wiper causing the Q to be very high and it to be less mechanically complicated.  That is why the magloop folks love the butterfly caps so much (when a vacuum variable cannot be found which is the best). 
  Learned a little about aluminum stamping.  They use sheets that come off a roll causing there to be sometimes 3 to 5 mils per inch of a bow.  Not sure how that can be alleviated but know that traditionally aluminum stamping was the only option for the stators/rotors before water jet and laser was around. I'll ask that question on a machinist thread I belong to for my CNC mill.  It would be nice to have one size in the 5 inch range and another in the 10 inch range but if the rotor/stator was 10 inches that would give a bow of 30-50 mils.  If the spacing between plate and stator is 1/8" (roughly) using 1/4" aluminum spacings and giving almost 7kv working voltage then I think there is enough to be able to accept that bow in the stamping.  This is likely the reason why I've only seen the giant capacitors in the Soviet transmitters like the one I see on Ebay right now.  I've seen some ex-Soviet variables that had plates that almost looked like they were cast aluminum.  Anyone seen some really large, 10-12 inch, plates on air variables that have the standard .040 inch thickness aluminum plates?

Best regards,
Mike
Logged
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2729



« Reply #31 on: February 08, 2014, 01:23:42 AM »

Non-butterfly type caps are used as loading caps in Pi-nets all the time without concerns of Q and loss. Why would it be any different in a link tuner?
Logged
ke6cvh
Member

Offline Offline

Posts: 19


« Reply #32 on: February 09, 2014, 01:32:29 AM »

Hi Steve,
 
  The matchbox uses a differential capacitor with a wiper so it definitely will work.  The butterfly capacitors are used in magloops due to the high current and low ohmic value so any resistance in a wiper becomes a significant amount of overall resistance.  The only thing achieved from a butterfly capacitor is less loss and mechanically a simpler design having no need for a wiper contact.  One other advantage, if a stepper motor was used to control the cap so the tuner could be placed directly below the feed point of the doublet it has virtually no turning resistance compared to a split stator with a wiper and only the inertia due to the weight of the rotors is of any concern with another trade-off is only 90 degrees from min-max vice 180 degrees less resolution per step of the motor but quicker time to the tuning solution. Also, if a LCT was used to tune an electrically short antenna the resistance of a wiper would start to become more of an issue than if it was tuning a doublet just over 1/2 wavelength.  Example, if I was trying to use an LCT with a 150ft doublet for 160-20 meters  due to space limitations of the property a house was on and was in series tune for the 160 portion I'd personally prefer a butterfly (or even better a vacuum variable) over a regular cap. The drawback it appears is a smaller spacing between rotor and stators that is 1/2 for a given voltage rating compared to a regular variable capacitor so tolerance on how flat the stator and rotor becomes more important.  This is what I did not take into account initially.  For a 5 inch butterfly cap at 400pf and 7kv (if the aluminum was perfectly flat) it seems that it is a good approach.  For a 10 inch butterfly air variable I'm not certain this is a good approach without a way to ensure the bow of the aluminum is a lot better than 3-5 mils per inch.  I suspect that all stamped rotor/stators have this same issue but could be wrong and sure would like to know how manufacturers controlled this.  Even if water jet or laser is used if the sheet came off a large roll and has any bow to it the same problem will be introduced.  I'll try to call around on Monday night but being overseas it is a real hassle.  Will also do some emails to water jet servicing companies to see how they ensure everything is nice and flat for their products.

73,
Mike
Logged
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2729



« Reply #33 on: February 09, 2014, 02:17:43 AM »

The matchbox does not have a cap on the transmitter side of the link. I thought you were planning on placing a cap on the transmitter side. But I think you are talking about the antenna side now.  Tongue Sorry for the confusion.
Logged
ke6cvh
Member

Offline Offline

Posts: 19


« Reply #34 on: February 09, 2014, 05:15:56 AM »

Hi Steve,

  You are absolutely correct, I'm planning on putting a cap on the transmitter side and have considered a split stator on the transmitter side and a butterfly on the antenna side.  Was hoping that the same stator/rotor set that is used for the antenna side could also be used for the transmitter side but not sure if that is possible as an 1800pF butterfly cap has the stator and rotor really close together for 2kv rating.  The old rule of "you don't get something for nothing" is definitely kicking me as the advantages of the butterfly are good but only in some applications (antenna side) and not necessarily in all (transmitter side).  What would be really nice is to be able to make all the parts using USA labor and make it really affordable.  Certainly is a hard puzzle to get all straightened out on.

73,
Mike
  Just found this site on allowable tolerances on bending of stock aluminum:
http://www.trident-metals.com/alumtolls.shtml
  And this web site on how to flatten sheet metal.  It seems to be near impossible and certainly the reason why capacitor rotor/stator plates just don't get very large:
http://bbs.homeshopmachinist.net/archive/index.php/t-40775.html
  Looks to me that around 5 inches is about as large as anyone would want to go with aluminum stator/rotors.  Has anyone seen ones that are larger than that?

73,
Mike
Logged
ka4koe
Contributing
Member
*
Offline Offline

Posts: 1154


It's alive. IT'S ALIVE!!!


« Reply #35 on: February 11, 2014, 07:49:01 AM »

This is from the Collins 32V manual. The last two diagrams indicate how to feed an unbalanced antenna with a LCT.

Philip


* LCT.gif (38.42 KB, 1116x1516 - viewed 604 times.)
Logged

I'm outta control, plain and simple. Now I have a broadcast transmitter.
ke6cvh
Member

Offline Offline

Posts: 19


« Reply #36 on: February 11, 2014, 12:08:33 PM »

Hi Philip,

  Thanks for the schematic!!!  It appears that Art Collins had a different way of coupling than the series/parallel way the K1JJ does it.  Lots to think about on this project and it is really getting interesting.

73,
Mike


 
Logged
W0GSQ
Member

Offline Offline

Posts: 6


« Reply #37 on: February 11, 2014, 03:17:52 PM »

.
Logged
ka4koe
Contributing
Member
*
Offline Offline

Posts: 1154


It's alive. IT'S ALIVE!!!


« Reply #38 on: February 11, 2014, 03:28:26 PM »

Here's MMANA GAL runs for my 150' inverted L.


* LC MMANA.JPG (71.85 KB, 654x793 - viewed 453 times.)

* Table MMANA.JPG (63.53 KB, 556x441 - viewed 363 times.)
Logged

I'm outta control, plain and simple. Now I have a broadcast transmitter.
ka4koe
Contributing
Member
*
Offline Offline

Posts: 1154


It's alive. IT'S ALIVE!!!


« Reply #39 on: February 11, 2014, 03:40:37 PM »

Top line should read 299 pF Cseries, 7 uH Lshunt.
Logged

I'm outta control, plain and simple. Now I have a broadcast transmitter.
ke6cvh
Member

Offline Offline

Posts: 19


« Reply #40 on: February 18, 2014, 09:05:09 AM »

AMfone crew,

  OK, as I'm working on the Link Coupled Tuner I've got a question for all the collective brain power and more importantly vast experience.  OK, from this first line you can tell I'm desperate to get some straight answers and/or debate on this question.
  As I prepare to wind the coil I want to do it the best way for Q.  W8JI states on his web pages that some of the best Q inductors he has seen are edge wound.  Not having any just sitting around to check I have to wonder "what about the distributed inter winding capacitance going up"?  So then the next thought is why not just use some 1/2" wide 20 mil copper strip?  I've got lots of copper tubing and a couple 100ft rolls of the copper strip to boot.  If it was edge wound it surely would be hard to bend but then I thought about all that counterfeit stuff showing up that is copper clad aluminum.  Actually, if advertised as such there really isn't anything wrong with it and it typically will have 15% or higher copper content.  One can buy everything from copper clad aluminum bus bar to flat wire and the aluminum would certainly make it easier (and cheaper) to wind edge wound. I've bent aluminum wire before in large gauges and it is ridiculously easy to bend so I'm thinking something with 15% or 20% copper cladding over aluminum won't be much harder.  
  But the nagging question still remains....what is better edge wound flat copper tape or copper tube for the ultimate Q on the ultimate link coupled tuner.  The matchbox has edge wound.  I've seen flat wound (or barrel wound I've even seen it called) in everything from a small inductor after the roller inductor of the tuner to inside amps.  Honestly, 20 mils is going to have a heck of a low inter winding capacitance and give a great (and easy to buff/clean) flat surface to any contact from a roller to something else.  
  Reading up on proximity effect makes me think the edge wound is the choice but then what would be the optimum width/depth?
  There has got to be an engineer out there that can shed light on the subject as it seems to be a complete mystery after scouring the net over and over and over.
  Oh, and then there is the flat wound strip in the Collins 180 couplers.  Certainly that was a mechanical choice but was it also a better Q to boot?
  I'm thinking with a 20 mil thick copper strip wound flat vice edge wound then the SRF will be extra-ordinarily high for a large link inductor (let's say one for 160 meters).  It would be the equivalent of 20mil wires side by side as far as the inter winding capacitance goes but of course much wider. 
  I've got some answers from that 3more company out of Taiwan who will sell individual rotor/stators or any other parts.  The butterfly capacitor dies and rotor stators from USA mfg are probably a "go" but am waiting for more answers from this gent in Taiwan regarding a 1700pF 2kv input capacitor for pricing with the butterfly on the antenna side of the LCT.  

Very best regards,
Mike
Logged
Pages: 1 [2]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.057 seconds with 19 queries.