Listen to Latest SolderSmoke Podcast

Friday, March 9, 2012

JBOT Thoughts

Posted today to the BITX-20 Group:

Good to see so many JBOTs under construction. I recently built two of them. Farhan has asked me to share with the group my experiences using -43 core toroids. Ironically, TV cores are a bit harder to find here, so I had to go with the fancier toroids!

In my first amp (used in a 17 meter DSB rig) I used the smaller FT-37-43 cores in all three transformers. T1 and T2 were wound and placed in the same way that Farhan did with the TV cores. For T3 I used superglue to stack 4 of the cores 2x2, then wrapped them with a bit of electrical tape. I wound T3 so the input wires were on one side and the output on the other. (On this amp, I had started out using some large binocular cores from the junk box, but I had a tough time getting the amp stable with these cores, possibly because using them resulted in longer lead lengths. So I went back to the smaller FT-37-43 cores).

In the second amp (used in a 17 meter SINGLE Sideband rig) I used FT-50-43 cores for T1 and T2. The bigger gores were a little easier to work with. I glued them vertically to the copper clad board. For T3, I again went with 4 FT-37-43 cores. (I tried using 4 FT-50-43 cores for T-3, but I found that output was low, so I went back to the smaller cores for T3). Note that on the second amp, I put some insulation between the heatsinks and the copper clad board (gorilla tape) -- I was worried about possible short to ground if the anodized layer on the heatsinks got breached.

In building my amps, I used the photo on Farhan's Phonestack page as my guide.

As is always the case with amps that I build, I found that a certain amount of "taming of the beast" had to take place before I got the devices stable. This is certainly the result of the kind of homebrew devices into which they are being inserted! Following advice from Doug DeMaw, I made liberal use of ferrite beads. I put one on the lead carrying 12 volts to the amp board. I put another on the lead from the 12 volt line to the final's RF choke. And I put one on the line carrying the .6 volt bias to the secondary center tap of T2.

I put a bit of shielding (copper-clad board) between the low pass filter and T3.

These steps allowed these amps to work nicely with my contraptions.

As long as we are talking about JBOTS, I have a question for the group:

On my second amp, I noticed that the output was a bit lower than it should have been -- only about 1.5 watts. I did a lot of troubleshooting, then I finally checked the bias currents in the first two stages. I found that BOTH were running at about 50 ma.

I looked at Farhan's JBOT notes and found that the second stage is supposed to be running at 100 ma. I found that the base bias voltage on the second stage was 1.13 volts, while the first stage had 1.9 volts on the base. That didn't seem right.

I then turned to the bias.exe program that comes with EMRFD. I plugged in the values from the JBOT schematic and, sure enough, the predicted collector bias current was 50 ma, not the 100ma the design was looking for.

Using that program, I made some changes to the base resistor in the second stage. I found that with 150 ohms instead of 100 ohms, the collector bias current goes up to the desired 100 ohms. This also seemed to bring power output up to the desired range of 3-4 watts (output should be a bit lower at 18 MHz, right?)

More info (and pictures) on my JBOT adventures can be found here:


73 Bill N2CQR

Our book: "SolderSmoke -- Global Adventures in Wireless Electronics" coffee mugs, T-Shirts, bumper stickers: Book Store:

No comments:

Post a Comment

Designer: Douglas Bowman | Dimodifikasi oleh Abdul Munir Original Posting Rounders 3 Column