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 3 4 Next   Go Down
  Print  
Author Topic: Length of coax from xcvr to amp affecting input swr, why ?  (Read 37333 times)
VE7RF
Member

Posts: 212




Ignore
« on: December 31, 2011, 04:44:27 AM »

With short length's of coax, [no more than aprx 3'], the input swr between
either my FT-1000-D  or my FT-1000-MK-V... and the input to my Drake L4B
is dead flat.   Once the coax is lengthened  to aprx 6', the input swr rises,
and can't be brought down to 1:1  No amount of slug tuning  will bring it back
down flat.  It will bring it down a tiny bit though.  What is  going on here ??
 
On my HB amp,  I use a variable tuned input, consisting of a pair of broadcast caps
+  tapped coil,  [pi- net].  The coax can be 25' long, and it can always be tweaked dead flat
from 160-10m.  The hb amp resides in the workshop next door.
 
Having to resort to changing all the C1 + C2 values in the drake amp is a pita. Then
even if I do that, and then change the coax length yet again, I will  have to go through
the same C1 + C2 rigmarole again.  I have 4 x L4B's..and this is getting to be a real pita
with this input swr.  Does anybody else notice this effect ??   Is their  some  'magic length'
of coax that will solve my problem ?   I don't understand the mechanism why the swr changes
with coax length.   I mean, coax length  to a dummy load has no effect on swr,  so why should
it, when feeding the tuned input of a GG triode amp ?
 
tnx.......... Jim   VE7RF 
Logged
KH6AQ
Member

Posts: 7757




Ignore
« Reply #1 on: December 31, 2011, 05:32:01 AM »

Why do you want a 1:1 VSWR? Is the VSWR low enough for the transceiver to operate without folding back?
Logged
W1QJ
Member

Posts: 2944




Ignore
« Reply #2 on: December 31, 2011, 05:36:45 AM »

Really Jim? 
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #3 on: December 31, 2011, 06:17:47 AM »

It is all explained here:

http://www.w8ji.com/tuned_input_circuit.htm

Anyone who understands transmission lines and SWR also understands SWR is constant with line length, so long as the SWR detector null and line have the same impedance and losses are low.

Any SWR change with line length obviously has to be a spurious or harmonic in the system.

Class AB grounded grid amplfiers have considerable harmonic energy at the cathode.

Use enough Q in a low pass at the cathode, and the problem will go away unless the meter or coax is bad.

Logged
VE7RF
Member

Posts: 212




Ignore
« Reply #4 on: December 31, 2011, 05:16:19 PM »

Why do you want a 1:1 VSWR? Is the VSWR low enough for the transceiver to operate without folding back?

##  Both my MK-V's  start folding back  with a 1.5:1  swr., but still have plenty of po to drive the amps.

Jim
Logged
VE7RF
Member

Posts: 212




Ignore
« Reply #5 on: December 31, 2011, 05:44:16 PM »

It is all explained here:

http://www.w8ji.com/tuned_input_circuit.htm

Anyone who understands transmission lines and SWR also understands SWR is constant with line length, so long as the SWR detector null and line have the same impedance and losses are low.

Any SWR change with line length obviously has to be a spurious or harmonic in the system.

Class AB grounded grid amplfiers have considerable harmonic energy at the cathode.

Use enough Q in a low pass at the cathode, and the problem will go away unless the meter or coax is bad.



##  yes, I realize that.  I also tried  inserting my coaxial dynamics wattmeter, with the low power slug in rvs..same results. [ cd meter installed on the output of the yaesu].   I also tried my rack mount bird 43..and also my array solutions  powermaster.  All the external meters  + the yaesu meter all  jive with each other dead on.

##  The C1 + C2 values on all the drake L4B tuned inputs  are almost double the values  used in the heath SB-220, so the Q  should be plenty high enough.  The only thing changed was the 6 x grid pins are now directly bonded to the chassis, and the 6 x grid caps + 2 x chokes have been removed.  But I have 4 x L4B's..and one of them still has the floating grids, and grid caps/chokes etc...and that amp too, has high  input swr with longer cable lengths.

##  The input swr on 10m, however, on the 3  x L4B's  with grids directly chassis grnded is now sky high, real high, and that's with short length of coax to xcvr.  Dunno why the 10m input swr is affected on 10m, but that's another issue itself.

##  On the L4B tuned input, there is NO coax  from the C2 cap to the cathode !  The double wafer small bandswitch is located right next to the cathodes..and all they use is a .01uf @ 2 kv disc ceramic, wired directly from tuned input bandswitch to cathode.

##  Bypass swr is also dead flat on the L4B's..and doesn't deviate, until u get up to aprx 50-60 mhz. [mfj-259b on input, and 50 ohm plug on output]

##  I'll try some new coax with new connectors.   I have 2 x MK-V's...and 2 x FT-1000D's...and  4 x L4B's, and any combo of any xcvr + any amp will produce identical [ high swr] results  with longer cable to xcvr.  So it's not a one off case of a screwball swr meter in one of the 4 x yaeus's....esp when I also installed external bird/cd/powermaster meters..and get the same results.

## It's baffling to say the least.  BTW, ur tuned input write up on ur site was very informative. I didn't realize the T type tuned input [ using 2 x coils and one cap] was a lo pass filter,..except it won't attenuate harmonics.  We tried that config on a 6m GG amp, and no amount of tweaking the variable cap or tweaking either/both coils would result in an swr below 1.5:1  We dumped it..and went to a PI net, with 2 x small adjustable caps and a fixed coil.  problem solved, flat input swr, loads of grid current/drive to the cathode, and eff was way up.

##  I can always use the built in auto tuner in the yaesu's if required, in conjunction with the L4B's. In the mean time, I will make up a 1/2 doz new jumper's, of varying length's, and run some more tests. perhaps a mess of large diam ferrite beads slid over some of the jumpers will do something. I have both type 77 and type 43 beads.

Later... Jim  VE7RF
Logged
VK4TUX
Member

Posts: 37




Ignore
« Reply #6 on: January 01, 2012, 03:38:49 AM »

It can be a weird thing. I replaced my patch leads a couple of weeks ago;
Between Txcr and amp I used a 6ft length (amp=sb-220) Input swr good on all bands bar 15m which was real high.
I then made a 9ft length and now input swr perfect on all bands.
I note 9 ft is 3 times what you have in 3 feet for which you have good results.
If you multiply an antenna length x 3 a similar matching effect occurs;
i.e; 1/4 wl to 3/4 wl is a 3x factor, Maybe a coincidence, however that worked for me.
I'm guessing as the termination impedances are not perfectly matched the coax acts similar to a tuning stub or matching transformer.


Adrian ... vk4tux
« Last Edit: January 01, 2012, 03:49:39 AM by VK4TUX » Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #7 on: January 01, 2012, 04:06:17 AM »

##  The C1 + C2 values on all the drake L4B tuned inputs  are almost double the values  used in the heath SB-220, so the Q  should be plenty high enough.  

The L4B, reverse engineering the network, looks like an effective Q of 2.1 on the input. Assuming a good layout that might be 15 dB or so attenuation on the second harmonic.

I'm not sure how the SWR meters respond to a second harmonic in the reverse direction, or how much power is in the 2nd harmonic at the cathode in the L4B, but Qeff of 2.1 might not be enough, or there could be layout problems limiting the attenuation through the network.  

If the coaxial cable matches the SWR meter null exactly, and loss in the cable is low, cable length can't affect SWR. Any length sensitivity has to be cable problems, instability in the radio, or spurious from the amp.

I've had to go up to Q~5 on some amps.
Logged
K6AER
Member

Posts: 5688




Ignore
« Reply #8 on: January 01, 2012, 08:32:38 PM »

Are you sure the transceiver output impedance is 50 ohms?
Logged
M0HCN
Member

Posts: 566




Ignore
« Reply #9 on: January 01, 2012, 09:10:07 PM »

Transceiver output impedance will make NO difference to the SWR measured on the line.....

If you have drive power to spare, a pad in series with the line will significantly reduce the reported VSWR, 6dB or so may help enough.

Regards, Dan.
Logged
VE7RF
Member

Posts: 212




Ignore
« Reply #10 on: January 01, 2012, 09:21:26 PM »

Transceiver output impedance will make NO difference to the SWR measured on the line.....

If you have drive power to spare, a pad in series with the line will significantly reduce the reported VSWR, 6dB or so may help enough.

Regards, Dan.


### They only do 200w  out...and the L4B  requires 110w  to drive it to 800 ma.  Even if a 3 db pad is used, then the xcvr is running at 200w.... so it's imd [xcvr] will be worse.  Plus the 3db will have to handle 100w.

Jim
Logged
VK4TUX
Member

Posts: 37




Ignore
« Reply #11 on: January 01, 2012, 11:37:10 PM »

Transceiver output impedance will make NO difference to the SWR measured on the line.....

If you have drive power to spare, a pad in series with the line will significantly reduce the reported VSWR, 6dB or so may help enough.

Regards, Dan.


The important measured swr here is that determined by the transceiver swr detection circuitry.
It's all about keeping that happy for best power output.

Adrian ... vk4tux
Logged
VK4TUX
Member

Posts: 37




Ignore
« Reply #12 on: January 01, 2012, 11:44:51 PM »

  Does anybody else notice this effect ??   Is their  some  'magic length'
of coax that will solve my problem ?   I don't understand the mechanism why the swr changes
with coax length.   I mean, coax length  to a dummy load has no effect on swr,  so why should
it, when feeding the tuned input of a GG triode amp ?
 
tnx.......... Jim   VE7RF 
[/quote]

The amp input circuit is not purely resistive like a non inductive dummy load.

Try 9' ?

Adrian ... vk4tux
Logged
VE7RF
Member

Posts: 212




Ignore
« Reply #13 on: January 02, 2012, 12:25:22 AM »

##  The C1 + C2 values on all the drake L4B tuned inputs  are almost double the values  used in the heath SB-220, so the Q  should be plenty high enough.  

The L4B, reverse engineering the network, looks like an effective Q of 2.1 on the input. Assuming a good layout that might be 15 dB or so attenuation on the second harmonic.

I'm not sure how the SWR meters respond to a second harmonic in the reverse direction, or how much power is in the 2nd harmonic at the cathode in the L4B, but Qeff of 2.1 might not be enough, or there could be layout problems limiting the attenuation through the network.  

If the coaxial cable matches the SWR meter null exactly, and loss in the cable is low, cable length can't affect SWR. Any length sensitivity has to be cable problems, instability in the radio, or spurious from the amp.

I've had to go up to Q~5 on some amps.

## Point well taken.  On a similar note, on my hb tuned input for the GG 3CX-3000/6000A7 tubes, I did not want to build 9 x bandswitched tuned inputs, so built just an adjustable pi net tuned input, with 2 x broadcast caps [padded on 160m only] and a tapped coil. I can make the Q anything I want, so tried a Q of 5-6 for an experiment.   To test the tuned input, I installed watt meters on both sides of it..then into a dummy load.  Low bands is ok, but on the high bands, with 200w going into it, I only got 160w out of it.  I increased the coil by 1/4 turn, and reduced the 2 x caps, and re-tuned for flat swr, then the Po rose to 195 watts.  The coil was 6 ga solid cu wire, and I could get it warm with 200watts. That equates to 2A  from the xcvr. Then with a Q of 5-6, it becomes 9-12 A of circulating current in the coil alone.  BW is less, but no big deal since it's fully adjustable anyway, via front panel. .  Once hooked to the cathode though, the C1 + C2 values are off by quite a bit, but a 1:1 swr could always be obtained.

##  Now imagine what happens with real small ga wires typ used on tuned inputs, and high Q. Inserting the watt meter between the tuned input and a dummy load , and varying the Q vs a real eye opener.

##  Other's have told me that when they went from 3'  to 6' of coax o their TL-922, that the swr rose as well. Some are suggesting that the added 3' of coax is like adding another 30 pf to the input of the amp.  I find this hard to believe, since the same 6' would have zero effect on a  dummy load.

##  I have also tried using arco trimmer caps   [padded on the low bands] and adjustable slugs or fixed coils.  The values used are dead on with the spread sheet, into a dummy load, but again, when hooked to the cathode, the caps are off by a bit, but always result in a  flat swr.

## OK, what happens if 6-20' of coax is used between xcvr and the input of an AL-1500/AL-1200 /alpha, dentron, sb-220, henry, etc, etc??  

## some have also suggested to use an electrical 1/2 wave of coax [cut for 3500 khz]..which then becomes 1 wave on 40m..and 2 waves  on 20m...and  3 x waves on 15M...and 4 x waves on 10m.

## some have suggested installing a T connector at the input to the amp, then inserting small amounts of C, to tune out the swr.

## Perhaps the answer is to replace the silver mica's used for C1..with arco's ?

Jim  VE7RF
Logged
VK4TUX
Member

Posts: 37




Ignore
« Reply #14 on: January 02, 2012, 03:42:42 AM »


## Perhaps the answer is to replace the silver mica's used for C1..with arco's ?

Jim  VE7RF
[/quote]

Jim, I guess you still have both the 3' & 6' patch coax pieces. In your massive amp resources kit can you find a 259 double female joiner to connect them together to make a 9' length?
If you could try that I would be interested to see the result if that's not too much trouble?

Adrian ... vk4tux
Logged
Pages: [1] 2 3 4 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!