EMRFD Message Archive 12195

Message Date From Subject
12195 2016-01-17 11:23:52 larry_n0sa High Output on Weber Tribander Clone

I am just finishing up on a transceiver I am building.

It is basically a copy of the Weber Trbander but I am controlling it with 

an Arduino Pro Mini and using a AD9850 DDS card.

I designe and etched printed circuit boards for it and set it up for

40, 30 and 20 meters. Everything is working just right and the output looks

clean on a spectrum analyzer. All except for 20 meters.

The 20 meter output is over 8 watts ( too high) and there are two spurs that are about 44dB down

that are 2 Mhz below and above the main signal.

Maybe I have a 2 Mhz parasitic oscillation going on.

Any ideas what the problem might be?

I know the AD9850 is not rated for a square wave above 1 Mhz but it does generate one

albeit an ugly one. It seems to be driving the buffer chip just fine. The square wave doesn't look any worse at 14 Mhz than at 10 Mhz so I don't think that is the issue, I could be wrong.

Anyway, any help would be appreciated.

Larry Naumann


12196 2016-01-17 18:19:44 wb8yyy_curt Re: High Output on Weber Tribander Clone

Suggestion is to watch how the spurs move with tuning the DDS.  Are they always 2 MHz below and above?  DDS spurs all except harmonics of the clock will generally move with changing the RF.  Of course the IF you rig uses will also be a factor.  A little detective work here could help ID the source of the spur.  You may need to generate a block diagram sketch along with the frequency math to get a handle what is going on. 

If 20m has the highest output, that is strange.  Do check your 20m output filter just to make sure it does not have a fault affecting the PA.

Very nice project! 

73 Curt
12197 2016-01-18 09:00:28 larry_n0sa Re: High Output on Weber Tribander Clone
I put the square wave output that feeds the final into the spectrum analyzer and sure enough there are the spurs 1Mhz above and below the main signal. They are much weaker on 30 and 40M. There is no IF on Xmit, the DDS drives a buffer chip then the three parallel BS170's. I now understand the spurs but not the high output.
 I have learned a lot from this project, Steve Weber has some interesting designs. I built and liked the Weber Tribander kit but wanted a full size display and thought it would be fun to clone his main circuit less the controller.
 I did check the 20M LP filter but I think it  needs improvement as the second harmonic is only down about 44dB. Not nearly as good as 40 and 30M.
12198 2016-01-18 11:13:40 Bill Carver Re: High Output on Weber Tribander Clone
What are the AD9850 clock frequency and output frequency?

12199 2016-01-18 15:23:56 larry_n0sa Re: High Output on Weber Tribander Clone
Clock is 125 Mhz output is 14 Mhz. The spurs come and go depending on frequency, just so happens that 14 Mhz is bad. I very carefully adjusted the trimpot that controls the duty cycle of the square wave while monitoring it on my spectrum analyzer, I was able to get a better signal to spur ratio. Now it all looks much better. I reworked the 20M LP filter as it wasn't doing a very good job and now everything seems to be just fine. The spurs are over 50 dB down now as well as the second harmonic.
12200 2016-01-18 17:22:18 iam74@rocketmail.... Re: High Output on Weber Tribander Clone

This appears to be a Class E amplifier, and if it is, all the old methods of fixing things (except layout) don't apply. I noticed in the assembly instructions for the Tribander several warnings about the switching mechanism -- You might be driving the wrong LPF (or more than one), or have a defective ground. There was also some comment about these in the e-Ham reviews.

Anyway, Class E has to be treated differently. You are right about the importance of the duty cycle and its adjustment. Perhaps yours is not quite fine enough. Also, it is probable that the DDS has some weirdness with the filtering. It might be judicious to put the whole thing through an amplifier/filter system before applying it to your rig. I am thinking that is the right thing to do with these DDS units and any rig, e.g., like the BITX.

So, I would clean up the DDS signal and look closely at the switching system.

Just a thought. I notice you got it working though. FB. 73

12201 2016-01-18 19:30:50 larry_n0sa Re: High Output on Weber Tribander Clone
 Yes, I fine tuned the duty cycle and made some minor adjustments to the LP filter to improve it's 28 Mhz performance. There is no filtering on the Chinese DDS cards for the square wave output. It looks good on 7 and 10 Mhz but at 14 the spurs kick up and can be reduced a bit by careful adjustment of the duty cycle pot.
  Now it is much better and acceptable performance. I do believe the high output was caused by the poor performance of the LP filter and a lot of second harmonic energy was being transmitted and showed up as power output as it obviously would. Just under half of the output power was at 28 Mhz. Amazing!!
 I sure learned a lot from building this project. I basically have a Weber Tribander with a nice display that I built myself. Now all I need to do is figure out how to program the RIT into it. That shouldn't be too hard.
12202 2016-01-18 21:14:17 Bill Carver Re: High Output on Weber Tribander Clone
12203 2016-01-18 21:23:42 Bill Carver Re: High Output on Weber Tribander Clone
Don't what why my question didn't appear on the reply to this email, but will try asking again.
Are these spurs occasionally running past the desired frequency, running much faster (sometimes much, much faster) than the tuning rate of the signal you want?

12207 2016-01-19 09:29:03 larry_n0sa Re: High Output on Weber Tribander Clone
 No, the spurs stay in place. I have determined that they are from the DDS card. I can reduce them by careful adjustment of the square wave duty cycle pot on the card. I can get them so they are over 50dB down so I believe that should be acceptable.
The high output was caused by  a faulty operating LP filter at 20M. I modified the filter and all is well.