The AM Forum
December 10, 2024, 04:11:09 AM *
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 3   Go Down
  Print  
Author Topic: grid circuit swr...  (Read 27261 times)
0 Members and 1 Guest are viewing this topic.
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« on: January 16, 2014, 11:41:20 AM »

I am playing around with driving the various homebrew rigs with a solid state exciter, and found something odd.
On my 2x 813 rig, I get no swr when I peak the grid circuit, works like normal.
On my 2x 4x150a rig, I get a very high swr, which decreases when the grid circuit is peaked, to a lower but not low value.

Both rigs use a link for the input, cold end twards the grid coil, and I started off with a 3 turn link on the 4x150a rig, then went to a 2 turn link, with only a slight improvement in the swr.

What drives the swr?
The coils in the 4x150 are made of thick enameled wire on a coil form, the 813 rig has thinner wire on pvc pipe pieces.

The antenna tuner in the icom 756 pro 2 will tune the swr out, but its not stable with drive power, changing the drive makes the tuner try and retune.

Very simple setup, very short piece of coax between the antenna input to the 2 turn grid coil.
I do not remember the tube side of the circuit, but it gets fixed and grid leak bias, has a cap across the coil to tune the grid, rf choke in the grid voltage feed.


The home brew vfo/exciter drives the rig fine, 6146, pie net output.

I was thinking of turning the vfo/exciter into a CW rig, plus making room in the rack for another power supply to run the modulator or RF decks.

Where does all the swr come from?
Logged
KA0HCP
Contributing
Member
*
Offline Offline

Posts: 1185



« Reply #1 on: January 16, 2014, 01:36:12 PM »

1.  SWR itself is unimportant.  It is merely an alternate way to measure impedance in a circuit, using ratios instead instead of actual figures, and the necessity to perform algebraic calculations.

2. The input impedance of the transmitter is determined by the interface circuit.  

3. Three approaches can be taken in feeding the transmitter
-Ignore impedance matches, accept loss, and with a SS exciter, accept power foldback from protection circuits.
-Match the impedance at the transmitter side
a. Use fixed tuned circuits for each band
b. Provide tuning circuits, either manual or automatic (LDG boards are used in some amplifiers)
-  Again, used fixed or tunable circuits on the exciter side

You have chosen to use the auto tuner on the 756.  That's fine.

Solution:  Let the auto tuner find a match, then turn OFF the "Tune on PTT" or "Tune on Transmit" function, in order to maintain the initial match.
Logged

New callsign KA0HCP, ex-KB4QAA.  Relocated to Kansas in April 2019.
W1RKW
Contributing
Member
*
Offline Offline

Posts: 4423



« Reply #2 on: January 16, 2014, 04:30:35 PM »

The SWR to some extent is not important as stated. What is key is the grid current. Don't exceed it. I have the same issue with GORT. It's a tuned input but I just watch the grid current and adjust accordingly with the drive of the exciter. I drive GORT with a Viking 2 and try to match for best power transfer with the V2.
Logged

Bob
W1RKW
Home of GORT.
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #3 on: January 16, 2014, 09:47:51 PM »

The issue is the solid state rig will not create enough drive.
I am not going to run it at higher power out into a high swr, that is asking for trouble.
Besides, other rigs have a very low (or no) swr with the grid circuit peaked.

The output of a 756 pro series rig in AM is 25 watts of carrier.
If the swr was zero, I might need 10 watts drive to get the right grid current, I have a lot of grid leak and a very high bias voltage, so it takes some drive.



Logged
KA2DZT
Member

Offline Offline

Posts: 2190


« Reply #4 on: January 17, 2014, 03:35:00 AM »

Why do you have the cold end of the link towards the grid coil??  I would put the hot end towards the grid coil.  You would have the link at the cold end of the grid coil.  This may have nothing to do the high SWR.

Fred
Logged
W2VW
Contributing
Member
*
Offline Offline

Posts: 3483


WWW
« Reply #5 on: January 17, 2014, 08:47:43 AM »

The grid in the 4CX250B takes nothing in the way of power until it pulls current so drive power changes load impedance to the plastic radio.

The auto tuner tuning at low power is designed to tune into a load that doesn't change with power level.

A resistor from grid to earth through a DC blocking capacitor to keep grid bias intact will help lower the load impedance. The value depends on the impedance stepup of the input tuning network.

A starting point might be 500 ohms. Be sure to rate the dissipation of the resistor for overdrive mistake situations.
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #6 on: January 17, 2014, 09:24:34 AM »

I think all rigs were designed that way.
I also think all my rigs were built that way.



Why do you have the cold end of the link towards the grid coil??  I would put the hot end towards the grid coil.  You would have the link at the cold end of the grid coil.  This may have nothing to do the high SWR.

Fred
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #7 on: January 17, 2014, 09:30:43 AM »

Ok, maybe you are on to something.
But if I turn the drive up and get say 5 ma of grid drive and tune the grid cap, the swr will go down to a low of 3 to 1, and not zero (at peak grid drive).

I have to look at the grid circuit, I had to work last night so I did not get a chance to play radio.
There might be some wierd stuff going on with the grid circuit values, I do not remember offhand how I did the bias hookup.

I also need to plug in the other RF deck (4D32) and see how that one acts.




The grid in the 4CX250B takes nothing in the way of power until it pulls current so drive power changes load impedance to the plastic radio.

The auto tuner tuning at low power is designed to tune into a load that doesn't change with power level.

A resistor from grid to earth through a DC blocking capacitor to keep grid bias intact will help lower the load impedance. The value depends on the impedance stepup of the input tuning network.

A starting point might be 500 ohms. Be sure to rate the dissipation of the resistor for overdrive mistake situations.
Logged
W2VW
Contributing
Member
*
Offline Offline

Posts: 3483


WWW
« Reply #8 on: January 17, 2014, 09:39:29 AM »

The load presented by the grid tuning network can be out of range of the ricebox tuner. If there's a way to adjust it try a few very small changes and careful SWR readings at typical drive levels. It might take an old fashioned SWR bridge and tube type driver substituted to do this.

OR cut and try. Edison method.

One unrelated question. How did you determine the operating grid current of the 250B in class C?

TIA,

Dave

Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #9 on: January 17, 2014, 09:58:16 AM »

I look at the tube data for the grid drive of the tube/s in class C plate modulated service.

I then increase the voltage (a lot) to make the tube run way into class C.

I then inject tones and study the intermod and harmonics and adjust things for the best display.


I posted the results here quite some time ago in the clean rig thread.

I could adjust things to make things dirty or quite clean...
Logged
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8301



WWW
« Reply #10 on: January 17, 2014, 11:45:23 PM »

My best SWR from the plastic radio driver is when the class C drive is adequate and rated grid current is flowing.
This tells me the power taken by the grid is probably lowering the impedance of the setup to a reasonable value.
It also says I got lucky with the match-up using an MB-40 tank. I did have to fiddle with the link's turns and position.
A 750 Ohm swamping resistor across the grid coil might make it perfect, but it works well so I am happy.
The behavior also means that I can not use something like a miliwatt sized MFJ analyzer to see the Z at the transmitter drive jack.
Logged

Radio Candelstein
KA2DZT
Member

Offline Offline

Posts: 2190


« Reply #11 on: January 18, 2014, 05:07:02 AM »

Still think your link is not correct as to which end is nearer the grid coil,  but I will check it further as you may be right.  In my HB xmtr I'm sure I have the hot end of the link nearer the grid coil.  Although, I don't think it has much to do with your SWR problem.

You should try loading the grid circuit down with some resistance.   As other have said, the grid circuit impedance probably swings over a wide range with tuning the grid circuit.

I'm not sure what your grid circuit looks like, but try a 100 ohms resistor to ground.  This will give half the 50 ohms you need for a match.  The remaining grid circuit will be in parallel with the 100 ohms getting you closer to 50 ohms.

Another thought,  make sure you have the correct by-pass caps in the grid circuit where needed.

Try to post a schematic of the grid circuit if you haven't already, I may have missed it in earlier post.

Fred
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #12 on: January 19, 2014, 09:00:53 PM »

I did a bunch of experiments, and nothing seems to help much.
I tried different numbers of link turns, thinner wire, rewound the grid coil, put an adjustable cap in series and parallel with the link, putting it across the link and closing it (maximum capacitance) lowers the swr a bit.

Maybe I should try putting the link over the grid coil.

I suppose there are grid driven amps designed for a 50 ohm input, what does the grid circuit look like on those?
I assume they are broad banded and show a low swr to the radio driving them...

Logged
KC2ZFA
Member

Offline Offline

Posts: 440



« Reply #13 on: January 19, 2014, 10:12:49 PM »

did you calculate the reactance of your link at the freqs you want to operate ? it's probably not 50 ohms. And the tuning cap you used was probably not large enough to resonate the link at the freqs you tested.

Peter
Logged
W2VW
Contributing
Member
*
Offline Offline

Posts: 3483


WWW
« Reply #14 on: January 20, 2014, 11:13:13 AM »



I suppose there are grid driven amps designed for a 50 ohm input, what does the grid circuit look like on those?




They look like a 1:4 or 1:9 xfmr with a resistor across the output.

 
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #15 on: January 20, 2014, 01:56:57 PM »

Well, since the icom does not seem to be selling, I will try some other grid coil setups.
Logged
Chuck...K1KW
Contributing
Member
*
Offline Offline

Posts: 163


« Reply #16 on: January 20, 2014, 07:45:21 PM »

You probably need to adjust the amount of link coupling and if the link isn't mechanically variable, try putting a variable cap in series with the link, as close to the link as possible.  A value of 300 - 500 pf variable should do the trick for 75M.  You may have to add a couple of turns to the link to make this work well. 
Logged

73, Chuck...K1KW
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #17 on: January 20, 2014, 10:04:17 PM »

I removed the grid coil and link and re did it with B+W coil stock, 3 turn link over slightly smaller coil stock.

I get a very low swr, no swr really, good drive with low power.

However, when I modulate the rig, the swr goes way up.

With the tube driver, I see no change at all in the 6146 plate current, or the RF deck grid current.
This is starting to be a pain in the butt.

I increased the grid drop resistance to get a very high bias voltage with more drive, and the situation got worse.
Logged
W3GMS
Contributing
Member
*
Offline Offline

Posts: 3048



« Reply #18 on: January 21, 2014, 07:37:16 AM »

Brett,

If you go back and ask yourself what the difference between a low power drive where things look good and that same signal with audio applied it may point you in some areas to check.  For 100% modulation source your drive power will  be running 4 times the carrier power on peaks.  Forget for just a minute the frequency domain points.  Take  your driver and increase  the steady state power to 4 times that of what  your unmodulated carrier.  Look at the SWR and make sure the amplifier is stable and the SWR is still low.   By doing that you will be determining how your amplifier is behaving as a steady state amplifier.  It may also help to look at the output on a spectrum analyzer to make sure its a stable as it needs to be.  Repeat the test just mentioned with the 4 times the unmodulated carrier test up and down say 5 KHz since that is what the amplifier will see under modulation peak power wise.    Every time you change something in the grid circuit, go back and check and make sure the amplifier is still neutralized since its a grid driven amplifier.   Trigger parasitic's can show under modulation, but I would not go down that path until you checked the steady state performance of the gain block itself while using the Icom as a driver. 

The output networks for your tube driver are quite different from the broad band SS driver and clearly their are compatibility issues based on the results your reporting.  My point in running the test I mentioned is to characterized the compatibility issues between the amplifier and that particular SS exciter.   

Joe, GMS           

   
Logged

Simplicity is the Elegance of Design---W3GMS
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #19 on: January 21, 2014, 08:05:00 AM »

I did turn the drive up quite a bit, not 4x normal, but maybe 2x, and still had the problem.
No sign of erratic operation or parasitics that I could see on the spectrum display, and the output swr stays low, so I do not think I have much power going out that is off frequency...

More tests tonight...
Logged
W3RSW
Contributing
Member
*
Offline Offline

Posts: 3307


Rick & "Roosevelt"


« Reply #20 on: January 21, 2014, 09:30:23 AM »

I didn't have such mismatch problems in both my single 813 rig and twin 813 larger rig.
Simply wound a few turns of 20 gauge hookup wire over the cold end on B&W mini stock with the windings having the right turn sense and both ground of link and RF ground (less neut. impedances) of grid coil connected together.  Are your coils wound with the same rotational sense?  Best impedance match sensitivity also is when the link is on the cold end of the resonant L/C grid coil.  Less variation in load- change sensitivity too.

In the dual 813 rig I made the ganged input band switch to select 5, 3, and 2 turns on the link along with full, half and quarter inductance ( approximately) for 160, 80, and 40 meters.

SWR is less than 2:1 in all cases (as roughly measured by the Drake TR 7 driver ) and power inputs of 10 to 20 watts is as expected.

In the single 813 rig I simply wound a 3 turns link good for all three bands and placed about 5k swamping across the grid to RF/neut. intermediate ground line.

I'll go to the downstairs computer and pull up a pix of the dual 813 grid input circuit to add here later.
Basically it's physically mounted on plexiglass to isolate a couple of variable caps from DC ground, etc.


* bottom wiring interior 8x6.jpg (219.67 KB, 1260x945 - viewed 401 times.)

* Link input circuit.JPG (2049.37 KB, 2592x1944 - viewed 416 times.)

* link input and corner angle.JPG (2114.6 KB, 2592x1944 - viewed 417 times.)
Logged

RICK  *W3RSW*
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #21 on: January 21, 2014, 11:24:09 AM »

That is some nice work.

In my 2x813 rig, I use seperate coils and links switched in for each band, 160 to 10, and do not have an issue with that rig.

There may be something different with the 2x4x150a rig, the 4x150 is different from something like the 813.

If I put a load resistor across the grid coil, that will just lower the Q, no?

I could put a resistor and cap from each grid to ground to load the grid...

I have the 3x4D32 rf deck on the bench, I plan on changing the grid input coils to the B+W coil stock with links over the coil like I did on the 4x150 rig, that got the swr down to zero (with no modulation).

Not sure why modulation would effect the grid so much, it might be a design issue, as the grid stuff is close to the high voltage feedthrough from the plate choke. Maybe some trace of RF gets into the grid circuit from that?

The deck is quite cramped, as it started out as a one tube deal on a chassis that was too small, then got a 2nd tube added...

http://n2dts.smugmug.com/Ham-radio#!/i-c57nxT7/A

Top:

http://n2dts.smugmug.com/Ham-radio#!/i-k3PMKRC/A

It works great as is with a tube driver, but maybe I should rebuild it on a bigger chassis and reuse the front panel.
I could then add 80 meters.
And maybe a 3rd tube....
Logged
W3RSW
Contributing
Member
*
Offline Offline

Posts: 3307


Rick & "Roosevelt"


« Reply #22 on: January 21, 2014, 02:40:06 PM »

Your stuff looks pretty darn neat too.  Yeah 5k does lower Q slightly, but this is a transmitter with tons of excess power compared with a receiver. Heath did this trick for years.  Broadens tuning, makes it less critical.  I found that I didn't really need it in the twin 813 rig. So both rigs take about 12 to 15 watts drive, the twin putting out over twice the power of course.  Is your link wound same direction as grid coil?

Grid current is measured but I find screen current is more critical when starting from a gross miss-tuned state.  When returning to previously recorded settings simply watching plate current dip is sufficient for touching up.  I haven't tried for minimum IP3 or some such.  I really should, given that I have the QS1R to watch everything.

For modulation to affect much of anything in indicated plate current usually indicates inadequate grid drive.  I found out early on that drive for CW in class C can be less that that required for AM.  Well that you knew.  So plate current shouldn't change much in the class C stage on a coil/mag. meter when going from 0 to 100% modulation. Since mine has a shared but very beefy power supply for mod. and class C final (3000v @ 1A P. Dahl, running 120 on 240 input) I see a small drop and sometimes increase in Ip as I swing the monkey.

Also you probably don't have complete neutralization if modulation is affecting your grid so much.  You need a big one of these piston or equivalent neut.caps.  Grin

The Raytheon fil. xfor is in honor of Derb.


* top view A 7x5.jpg (115.25 KB, 700x525 - viewed 466 times.)
Logged

RICK  *W3RSW*
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2303


« Reply #23 on: January 21, 2014, 03:16:11 PM »

The 813 and 4x150 rig are not neutralized.
The 813 rig does not seem to care.
The 3x4D32 is, but it does not seem to work right.

All the rigs work fine and do not do anything odd other then the solid state rig drive problem.

The 3x4D32 does 80 and 40 meters, the 4X150 is set up for 40 only.

I think I will rebuild the 4X150 rf deck and make it bigger, better, and neutralized.
80 meters would be easy to add.

That should be an easy fun winter project.

Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8886


"Let's go kayaking, Tommy!" - Yaz


« Reply #24 on: January 21, 2014, 04:24:35 PM »

That dual 813 rig is a real gem, Rick!

Let me know if ya ever decide to sell it... Wink   

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Pages: [1] 2 3   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 0.066 seconds with 18 queries.