Re: ATS3b Power Out
--- In AT_Sprint@yahoogroups.com, "Steven Pituch" <list2@...> wrote:
You've really done a lot of serious work on your rig! Sounds like
your method of optimizing is good, but you may have succeeded in
getting to much power out of the rig! I think an ampere going in,
after "matching", is too much.
Your dummy load error is significant. You may be able to correct it
close to 50 ohms with a parallel resistor. It would help a lot of
you could check it with an antenna analyzer at RF.
You RS meter may be fine. They key is to check it against
a "standard", such as a multimeter that has been recently calibrated
against real standards. If you can measure volts accurately, and you
can get 1% resistors, then you can check and measure current; you
also can find any error in your meter's current scale and correct it
The WM-1 may not read accurately when connected to a non-50-ohm load.
The cables will transform the non-50-ohm impedances into different
complex values also.
One warning is that once you get everything at 50 ohms, your ATS-3B
may deliver even more power than now, so the current may go too high
and the power out may be excessive. I hope not, but this might
happen. You may have to go through one more round of tweaking, since
the rig really needs to be set up for reasonable currents, good
efficiency, and a 5.0 watt maximum into 50 ohms.
I think you have more patience than I do.
At least we're pretty sure you're a victim of measurement errors
This is partly why corporations with many pieces of test equipment
spend huge sums doing scheduled calibrations and getting
certificates, even when the equipment seems to be just fine.
Hang in there!
> Hi Carey and All,
> Thanks Carey for your thoughtful and detailed reply. It is exactly
> of wisdom I was looking for.
> I think I purchased my multimeter after speaking with Wayne Burdick
> building my Sierra. I did purchase it to check RF power. That
would put it
> at over 10-12 years old. It is a Radio Shack 22-168A. It always
> be a nice tool. I hardly ever check current with it, mostly just
> and voltage. I don't have anything else to compare it to. I think
> to consider buying another one.
> The dummy load in my MFJ 949 E is reading 56.7 Ohms. There's 13
> right there! I can't back figure the math, but hoping this means
> Watt meter is reading low. I suppose you recommend the Norcal
> I was reading about the Accuprobe which is OP amp compensated, but
> Norcal calibration chart seems adequate enough. I had had built
> more accurate dummy loads years ago but discarded them after a few
> because of the leaking mineral oil.
> I would appreciate any advice on purchasing a decent multi meter.
> so many out there now. Should I stay away from auto ranging? It
> out there better than others under $100. Etc.
> The way I adjusted the output filters:
> I studied all the available documentation. I tried adjusting C3F
> had suggested and C4F as you had in a few cases but found that only
> C2F had a positive impact on the output power. I would first peak
> module with the stock LF1 and LF2 number of turns by compressing or
> expanding the turn spacing. If I got to a fully compressed or
> toroid, if I thought that adding or subtracting a turn would help I
> try that. Otherwise I would add about 20 uF to C2F and try again.
> after doing this I would have to add or subtract turns to get to an
> range. It seems that most of the time expanding turns on LF1
> current trip, while expanding turns on LF2 would increase the
> incrementally. I did this for hours at a time until I got the
apparent 5 W
> out. Of course some of the copper traces are gone at the toroids
> replaced them about a dozen times for each module.
> As I said earlier I was tripping the current limiter at about 4 W I
> I got 5 W out on 40 m after doing your mod of R47, but it was on
> edge when I hooked it up to an antenna. That's when I replaced R47
> 470 Ohm resistor. It wouldnake sense that maybe I am actuaally
> 5.5 W or more and thus the high current.
> I have the feeling that my power readings are one-half to one Watt
> 73 and thanks again,
> Steve, W2MY