Saturday, March 21, 2009

Enterprise



No I'm not going to write about Kirk, Picard or Janeway. I just needed a pic for this edition.

In Calibration I alluded to how cool it is that the entire F5K fleet of radios has a DDS signal internally that is within 0.1dBm of all the other F5k (3K's also fill this bill) It means you can download new stuff to the radio and not spend half the night recalibrating. With the SDR-1000 you had to do this. The SDR -1000 was not so automatic especially in its early days. All the changes were done in software to be sure but you had to sit there and bring everything in the software into alignment. For that you needed a dummy load and a calibrated signal generator and an external watt meter of good repute. To tune up the Power amp you still need an external wattmeter

With what we have today however as you have seen is you can add new software and if there has been major surgery to the radio's brain requiring realignment, you just plug in the dummy and in 11 minutes your radio is all aligned automatically, and your alignment will be very close to every other radio in the enterprise. Across the F5K universe things will be nearly identical for people running the same version of software no matter if your running a lap top or some fire breathing 16 core behemoth. So what's the big deal? Here is the big deal

This is a quote from a note I got from one of the software designers (n4hy):

"With the recent changes to the RX flowgraph (in the test branch now),
and the total and complete elimination of RX image (in the iqtest
branch), and the upcoming changes to the TX branch, it will hard to
see how anyone can say that even a partially conventional design can
hope to compete. The receiver now has the AGC [AT THE END] of the
processing. This means that none of the nonlinear processes ahead of
it will be modulated by the agc. That is simply NOT POSSIBLE with a
traditional design or hybrid design, not even one where the users,
producers, desperately wanting what we have provided them have figured
out some horrible set of ridiculously complicated and/or expensive
devices to try and get it. The QSD receivers and the high speed A/D
receivers are the only way to get there."

DING DING DING Here is a guy who decided to move the AGC loop to the end of the RX flow diagram after all the rigmarole has been done to the signal so there is NO AGC modulation of those preceding processes. The only thing that controls the AGC therefore is signal, and the ACG does not control the transfer curve of some "stage" ahead of it like the ANL or the NB. All those stages do their duty in a clean way, before the AGC. Why do I keep harping on this?

A friend of mine who is a low band DXer (an Orion owner) one night commented on how sometimes if there is a carrier close to the freq of a DX station you can magically hear the DX. This is due to the AGC becoming fixed and not pumping. It is a big deal. If the AGC pumps just at the instant you want the gain to be highest so you can hear the damn DX, the AGC decreases the signal strength. I've only talked to my friend on the radio, but I wonder if he is bald, because that AGC business is enough to make you pull your hair out! How does this fit into the enterprise idea?

The code was rewritten to eliminate this issue. In any other conventional radio where the AGC shows up in the flow diagram is fixed in stone. In any other hybrid radio like the Orion, even if the AGC is partially firmware, the tentacles of the AGC reaches back into the hardware, and the hardware is hard and can not be moved. So you can goof around at the margins making changes in timing and threshold, but you can't all of a sudden decide to move the AGC to the end of the line. This is true of all roofing filter designs, because you need hardware like amplifiers to send signals through IF's. Amplifiers are what holds up the roof. For the Orion this made a big deal. The Orion had the roofing filter and a subsequent second IF filter (you can never get enough of them filters can ya). The second filter had an amplifier and the gain in the IF when you turned on the second filter caused the dynamic range of the radio to degrade. Add a $200 filter and get a crapier radio. Yep sounds like boys at TT were smoking some of that TARP money when they designed this sucker. You get all the distortion and damn little benefit, but they get to sell you another filter. "Well, just modify the radio" you say? When was the last time you took a soldering iron and the dykes to a $4000 radio and expected to resell it for more than a tenth of what its worth? "Well have the factory send out an upgraded circuit board" Ya right... Why do you think we have the Orion II? There ain't no profit margin in fixing the Orion. Also TT put a microprocessor that ran out of memory in the Orion. No more room at the inn for fancy firmware upgrades. Every time you put in a new feature an old feature breaks. That's why the Orion owners over on the TT reflector are still waiting and waiting and waiting for new firmware... Real soon now real soon...

Here is something I saw over there:

"I am amazed with all these bugs ie Red Screen etc. Didn't TT see these problems during their months of testing? Their loyal customers spotted these problems within 48hrs. Lets hope it's not another 1 1/2 years for a correction. I have not downloaded this firmware as of yet. This reminds me of how the car companys let the buying public be the guinea pig." This quote is for the Orion II. The Orion boys are just sitting there wishing they HAD some new bugs to bitch about Is this any way to run an enterprise???? Remember this radio is billed as the last radio you will ever need. OPPS the update came out the same day you died!!

For the Flex radio however the entire fleet of radios just have to download new software and bingo the AGC is at the end of the flow chart where it belongs. But it is even more fundamental because the AGC has the potential to be written perfectly (in my opinion it is already there) every F5K will respond perfectly (or virtually so) where as the legacy radio will not, and the legacy manufacturer will just bring out a new model.

And this is the power of Enterprise. One change in code results in all the radios enjoying the benefit. The Flex scheme is the ONLY scheme in ham radio today that allows this. And because of this the rest of the designs will be left in the dust. The RX now adapts itself as far as image rejection goes. Recall I wrote this is a phasing radio. This adaption is key. This adaption means your radio is running at best image rejection all the time as you tune around the band, as you are on USB or LSB, or AM or what ever. It takes one part of calibration off the table. Soon the transmitter will be the same. Always best tuned. Always tuned in such a way as to minimize QRM to other stations for example. Always tuned in such a way as to eliminate buckshot. Always tuned in such a way that wide band CW and the damn keyclix are not possible.

Yaesu had a keyclick problem for 15 years in the FT-1000 family of radios, and it was never cured. The entire enterprise of FT-1000 radios D MP FIELD MARK 1 MARK 2 MARK 3 MARK 563 had this problem. To cure the problem took some surgery to the guts of the radio, AND if you cured the problem you screwed up the QSK. That was the solution fo the entire enterprise. Yaesu's solution? Screw 'em. It costs too much to redesign for "that" little problem, and we are not measured on that aspect of type acceptance. Imagine a radio where the fix gets sent out in a download and the entire enterprise of radios is cured. I have that radio on my desk.

Let me quote my friend Bob once again

"That is simply NOT POSSIBLE with a
traditional design or hybrid design, not even one where the users,
producers, desperately wanting what we have provided them have figured
out some horrible set of ridiculously complicated and/or expensive
devices to try and get it. The QSD receivers and the high speed A/D
receivers are the only way to get there."

The only way to get there? Where is he going? THE FUTURE is where he is going

What is a traditional design? Think TS 940 or Omni-6 What is a hybrid design? Think Orion or K-3 What lays in the future I think will simply amaze and no enterprise of traditional radios or hybrids will be left standing except as horribly grotesque kludges (
of course you can always stick in another filter) or an anachronism, (you know like those jokers who wear tights and bucklers and eat hanks of meet off the bone, swill gallons of ale and pretend to be knights of yesteryear and then get into their beemers and drive home, or those jokers who get out the HW-16 twice a year and "make a contact") If you think about it, the K-3 is an anachronism It's basically a 5th generation Omni-5. (green tights, bucklers, hanks of meat,..... imagine a bunch if hams in green tights, mawing hanks of meat, and swilling beer... WHEW what mental picture....... wait a minute..... that kind of sounds like field day)

73

I know this is a very Wayne Greeney thing to say but: Is it just me, or are you starting to also get a clue about how we have been sold a bill of goods by the other manufacturers? I was talking to an Alpha owner the other day and when I asked him about his 87 he said "Oh you mean my light show?" Flat out cracked me up. I know from where he comes. I bought the "other one" and I'm less than overwhelmed myself. Both my old AL-1500 (which I stupidly sold to buy the Acom whirly gig) and my 30 year old LK-800tny are better amps.