[Central-Electronics] 600L Problem Diagnosis

[Tom Bruns]

You might want to deoxit the bandswitch contacts first and hope that's 
the issue as it's either that or the couplers themselves according to 
the schematic. If it's the coupler(s) your options are pretty much 
limited to liberating a good one from a parts chassis if you can find 
one or giving Nick Tusa K5EF at ce-multiphase.com 
<https://www.ce-multiphase.com/contact-us> a phone call to see if he is 
able and willing to take it in for repair.

At https://k9sul.fayoly.net/projects/ce600l (click on Initial repairs 
<https://k9sul.fayoly.net/projects/ce600l/ce600l_fix1.html>) you can 
find SWR plots of the output couplers using a 3900 ohm load resistor. 
The 10m plot is of a bad coupler.

While it is of course theoretically possible to fabricate a replacement 
coupler be aware that at the factory half of the couplers manufactured 
did not pass spec (and they knew what they were doing).

On my 600L on 80m I get 300W output with 6W of drive on CW and about 
350W peak output on SSB as measured on my LP-100A. Mine has solid state 
replacements for the 866AX rectifiers, one with the originals may have 
slightly less output due to slightly lower plate voltage.

Glad to hear from you directly that replacing a bad diode in the RF 
bridge circuit resolved the reflected power issue.

73 Tom KC5L


On 8/20/26 5:18 AM, Laurence M. Szendrei via Central-Electronics wrote:
> Hi Barry,
>
> If you have a Vector Network Analyzer (these days lots of folks own 
> the quite affordable nanoVNA), you can do a simple single-port 
> measurement indicating how well the couplers are matching the plate 
> circuit to 50 ohms at the amplifier output. Since the VNA ports are 50 
> ohms, you would connect the S11 port to the 600L RF output connector, 
> and temporarily connect a load resistor (small 1/4W or 1/2W carbon 
> comp or metal film are fine) from the plate of the tube to ground. THE 
> AMPLIFIER IS NOT POWERED UP FOR THIS TEST - IT IS STONE-COLD, AND 
> UNPLUGGED FROM THE MAINS FOR SAFETY's SAKE. ALSO MAKE SURE PLATE 
> SUPPLY FILTER CAPACITORS ARE COMPLETELY DISCHARGED. The value of the 
> resistor should be the specified plate voltage divided by [the 
> specified peak plate current X 1.6]. The factor of 1.6 is 
> approximately correct for class B/AB2 operation, and should be close 
> enough to reality for this test. Sweep the VNA over the appropriate 
> frequency range for all the couplers you want to look at, selected via 
> the band switch on the 600L. You can then look at the resulting graphs 
> of return loss or VSWR (which are different ways of looking at the 
> reflected power), and a Smith chart which will show the complex 
> impedance over the sweep. The ideal result would be a VSWR of 1:1, and 
> return loss at the noise floor, over the frequency range of interest, 
> indicating a match from the PA plate circuit to 50 ohms resistive (and 
> I guarantee you won't have that in the real world). You can compare 
> results for the couplers on the bands which are behaving themselves 
> with the couplers for the bands which aren't.
>
> Make sure you remove the load resistor and VNA from the 600L before 
> powering it up, or annoying things will happen! ;-)
>
> I haven't done this with my two 100V transmitters yet, although it's 
> on my bucket list. But I have used this procedure with a homebrew 
> high-power class C amplifier with a pi-network output to find the 
> optimum settings for the input and output capacitors, and the 
> inductor. For class C the fudge factor is 2 rather than 1.6.
>
> 73,
> -Larry/NE1S
>
> On 8/17/26 9:33 PM, Barry Graves via Central-Electronics wrote:
>> New member here.  I'm a CE fan and have a couple of 20As, 10Bs, 
>> slicers, MM, and most recently a 200V and 600L.
>> Recently a problem cropped up with the 600L on 80M band.  The amp has 
>> low output and more alarming, the grid current pegs on key up and the 
>> reflected power deflects the meter to the left (both conditions on 
>> the 600L meter).  A quick test on 40M was perfectly normal so this 
>> seems to be limited to 80M and not a meter problem.  I'm hoping this 
>> is not an output coupler problem.  Looking for advice on what to 
>> check.  Is there a test method and expected measurement guide for the 
>> broadband output couplers for example?Cheers,BarryKJ5GQM
>
> ______________________________________________________________
> Central-Electronics mailing list
> Home: http://mailman.qth.net/mailman/listinfo/central-electronics
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Central-Electronics at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html