[Central-Electronics] Ce 20A alignment questions
alancgut at gmail.com
alancgut at gmail.com
Mon May 18 11:12:14 EDT 2026
Hi Jon,
It looks like you're making progress.
In the manual, on page 21A it is mentioned that one possibility for reversed sidebands is an improperly wired modulation transformer. It says: "To reverse the sidebands, the secondary leads (brown and white) of either modulation transformer should be reversed."
Maybe someone replaced one of those transformers at one time. I've heard that they have a high failure rate and a lot of people use substitutes stuffed back into the original can.
Something to check out.
Regards,
Alan, W9CG
-----Original Message-----
From: central-electronics-bounces at mailman.qth.net <central-electronics-bounces at mailman.qth.net> On Behalf Of Jon Kannegaard via Central-Electronics
Sent: Monday, May 18, 2026 1:15 AM
To: central-electronics at mailman.qth.net
Cc: ranickels at gmail.com
Subject: Re: [Central-Electronics] Ce 20A alignment questions
I have interesting, upside down results and a couple of questions for the group.
I’ve achieved 50 dB carrier and opposite side band suppression on 80 meters. This 20A has a PS-2 not a PS-1. 50 dB was what they thought they could get with a PS-2 in a 20A but they never shipped one with PS-2. So this is a great result. Except it isn’t.
It’s on the wrong sideband. That’s 75M LSB on SB2 not on SB1. I wonder if I got the wrong RF phase shift output leading, and the wrong one lagging. I ended up with B (green wire, lower pot) leading A (red lead, upper pot) by 90 degrees.
Does anyone know if this is right or should A be leading B by 90 degrees?
When I put in the PS-2, I checked that the pin-outs were the same as a PS-1, that pin 4 was 90 degrees ahead of pin 8 in both units. I think they are. Am I right? Has anyone else done this? (I made both resistors around the AF balance pot the same because the PS-2 has equal level outputs while the PS-1 does not.)
Details:
I preset L1 to 9.02 MHz and L2 to 8.99 using a grid dip meter. I got my hands on some low capacitance (3.9 pF) scope probes and adjusted L1 and mainly L2 to 90 degree phase shift with voltages as close as I could achieve measured at the null pot wipers. I had to move L1 as far from L2 as possible and still had .3V P-P difference at 90 degrees. The position of the ferrite cores is completely opposite of the suggested presetting which is L1 1/2” up and L2 only 1/4” up. This seemed odd but I proceeded anyway
Painstakingly tweaking L2, L1, the null pots, and the audio level pots, I managed to get -50 dB for both the carrier and the upper sideband at 3.695 MHz with an audio frequency of 1230 Hz using the SB2 setting which is the wrong setting. Now L1 resonates at 9.027 and L2 at 8.96. I made no attempt to balance the suppression between the USB and LSB. USB looks terrible. I know it’s a trade-off but I think something fundamental must be wrong so I didn’t try to get USB under control.
I think I’ll leave these bizarre results for a couple of days so I can use it on the 75 meter West Coast Vintage SSB net on Tuesday, then I’ll take another run at it to see if I can straighten it out. If anyone knows about the A/B phase lead/lag or anything more about using the PS-2 in a 20A I’d appreciate some help.
Thanks,
Jon
The 9 Mhz oscillator was running 9,000,140 Hz — slightly high but within the +/- 1 kHz spec.
115 VAC. Out of cabinet but long warm-up.
The PS-2 was rebuilt using a recently calibrated Stanford Research SB 720 high precision LCR Meter but that was a many years ago when I was working on the 100V. I do remember every resistor was out of spec.
---------------
> On May 5, 2026, at 11:08 AM, Jim Candela via Central-Electronics <central-electronics at mailman.qth.net> wrote:
>
>
>
> Hi all,
>
> This topic sure peaked my interest. Most of us here on this board have been here for decades!
>
> I do have some comments to add. The coil position adjustability for the 9 Mhz coils is necessary for the first time setup, if it was done correctly originially. My memory is a little fuzzy, but there might be three holes for the two coils. If so, perhaps later on, CE just had two coils with one slotted for adjustment. I obtained a 20A once that had a coil in the wrong hole. The carrier still nulled, but at pretty extreme carrier balance pot positions. This unit would lose opposite SB rejection as the Mic gain was increased. It seems that the IN34 diodes were reaching a limit on the PIV rating, and may have been avalanched on signal peaks. Fixing the amplitude offset fixed the problem I was having.
>
> It is important to have the goal to get equal 9Mhz amplitude between the two coils, and the two pots centered when the carrier is at a null. Measuring the 9Mhz levels needs to be done with a very low capacitance probe. I once successfully used two identical 10x scope probes to a 100 mhx dual trace oscilloscope. These probes will each present a ~ 5pf load, which is a lot. I found some tiny 3pf ceramic disks, and one in series with each probe tip lowers the load to something like 1-2pf. Making sure each probe was compensated, and doing XY mode on the scope, the coil alignment is perfect when the Lissajous pattern is a perfect circle. If this is achieved, see where the carrier null pots are positioned when the carrier is in a deep null. They should be centered, but that is not always achievable. Then remove the probes, and fine tune, ever so carefully, the coils to get the null back. The exact sequence may vary.
>
> Troubles:
>
> ** Cannot balance the coil amplitudes. vary the coil position.
> ** The Null pots are near an extreme, check balance of coil amplitudes, and check the 1N34a diodes. Substituting 1N5711 hot carrier diodes is a worthwhile upgrade. Sort for the same forward voltage on a DVM diode mode.
>
> Hints: The pots can be wirewound since they are bypassed for RF anyway. I have one 20A using a pair of Bourne Knob Pots, 10 turn puts with a clock dial, and a nice knob.
> Hints: You can reduce the pot range by going to a lower resistance pot, and adding resistors either side to get to the original value. You can also stagger the two values to recenter the pot to fit your alignment while keeping the total resistance the same as stock.
>
> Jim
> Wd5JKO
> On Tuesday, May 5, 2026 at 12:27:54 AM CDT, Jon Kannegaard via Central-Electronics <central-electronics at mailman.qth.net> wrote:
>
> Thanks everyone for all the answers. It was actually the article that Ron posted that got me to take my 20A down off the shelf to work on it. Even with that article and the manual, I somehow hadn’t gotten it through my head that adjusting distance between L1 and L2 sets relative levels of the two RF phases. Alan explained that.
>
> The ZL2AMJ method is clever. I hadn’t heard of it. I have been using an SDR with an RF pickup off a dummy load which seems to give a pretty clear picture of the situation.
>
> Laurence mentioned the 100V. I have one of those. I brought that back to life many years ago. It does achieve better opposite sideband suppression with a more sophisticated audio phase shift network and a simpler (!) RF phase shift network. But the 20A is such a historic transmitter and lean by comparison, much easier to muscle around.
>
> I have to shelve the project for a week. When I get back to it, I’ll let you know how it goes or maybe ask more questions.
>
> I really appreciate the responses,
>
> Jon
>
>
>> On May 4, 2026, at 4:41 PM, Laurence M. Szendrei via Central-Electronics <central-electronics at mailman.qth.net> wrote:
>>
>> Bob - I like ZL2AMJ's method too - it's much better than using a very narrow RX. (Before I had a software-defined RX I used my R390 with the bandwidth in the 100Hz position to null out the undesired sideband, and tuning back and forth as one checks the suppression in the LSB, then USB, mode, back and forth.) I never thought of listening for a tone 2X the frequency of the signal at the audio input! It would be good to have the SDR listening in the AM mode with a bandwidth wide enough to capture both sidebands and get both the visual and audio "feedback" (not in the electronic sense!) while performing the nulling procedure. Thanks for getting me thinking!
>>
>> 73,
>> -Larry/NE1S
>>
>> On 5/4/26 11:43 AM, Robert Nickels via Central-Electronics wrote:
>>> On 5/4/2026 6:19 AM, Laurence Szendrei via Central-Electronics wrote:
>>>> Also, if you have a local SDR with a spectrum display use this to
>>>> do your sideband suppression procedure - it’s so much better at
>>>> telling you directly what is happening regarding sideband
>>>> suppression than using a scope or any receiver lacking a spectrum
>>>> display. CE didn’t have these when it developed the procedure in
>>>> the manual
>>>
>>> I'll just "second" what Larry is suggesting here. Even a cheap SDR is more than adequate - just make sure to loosely couple to the transmitter so you don't overload it, and choose SDR software that lets you zoom way in on the signal so you can clearly see both sidebands and the carrier. You'll see the interaction between controls but as described in the manual, it's a process of optimizing to get the best compromise when switching between sidebands.
>>>
>>> Another very useful method must be credited to Fred Johnson ZL2AMJ who first described it. Think of the spectrum produced by an SSB transmitter when modulated by a pure 1 kHz sine wave tone. With good carrier and opposite sideband suppression, only a single output will be produced (this is how many produce CW). If there is carrier leakage, it will be 1 kHz removed from the desired signa, thus a the beat frequency, a 1 kHz tone, will be heard in the output when listening with an AM detector. If there is opposite sideband leakage, it will be 1 kHz on the other size of the carrier thus a 2 kHz tone will be heard. The amplitude of the two tones heard with an unaligned phasing exciter will give an indication of the unwanted signal components that are present. Nulling the carrier should bring the 1 kHz tone down virtually into the noise. Then as you adjust the audio and RF phasing you'll be able to minimize the 2 kHz tone. It's an iterative process because the adjustments interact, and the optimal setting for one sideband won't be optimal for the other sideband. You just have to go back and forth til you get a good compromise and as Larry said, it's possible to get 30db or better suppression. The improved audio section of the 100/200V allows greater suppression than with the 10/20A design.
>>>
>>> Fred described a simple "crystal radio" type detector with headphones but all you need is an AM receiver but using an SDR in the AM mode gives the added benefit of a visual indication.
>>>
>>> 73, Bob W9RAN
>>> ______________________________________________________________
>>> 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
>>
>>
>> ______________________________________________________________
>> 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
>
> ______________________________________________________________
> 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
> ______________________________________________________________
> 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
______________________________________________________________
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
More information about the Central-Electronics
mailing list