Using my SO2R station, I tested the 87A band switching by transmitting with
one rig (FT-1000D/87A) and listening on the other (FT-1000MP). Here's what I
1. When using a single dit or dah to change bands, the element is passed
through at low power.
2. Regardless of target frequency or keyer speed, I cannot hear any
shortening of the first element.
3. If the key is held down throughout the band change, the carrier is
transmitted at low power until tuning is completed. Then the amp ramps power
up to maximum over a period of about one second. No interruption of RF is
required for the amp to switch from low to high power.
4. There appears to be no loss of power when the signal is transmitted at
low power (i.e., peak S-meter reading is identical for a single dit that
causes a band change and for a single dit when the amp is in standby.)
I don't have a storage scope here, but I'd say this implies a sampling speed
that's alot faster than 50-100ms, and T/R circuitry that can hot switch the
PINs. I took a quick look at the schematic, and there does appear to be some
sort of frequency divider network made up of four CMOS 4520's and two TTL
74HC165's. Don't have time now to get the books down and figure it out.
73, Dick WC1M
> -----Original Message-----
> From: k8cc [mailto:firstname.lastname@example.org]
> Sent: Monday, January 22, 2001 10:29 PM
> To: email@example.com; firstname.lastname@example.org; AD6E@aol.com
> Subject: RE: [CQ-Contest] 87A question
> At 01:01 AM 1/22/01 -0500, Dick Green wrote:
> >I don't know how they do the RF sampling, but even if they had to wait an
> >entire cycle at 1.8MHz to determine the frequency, that would
> still be less
> >than a millisecond. My guess is that the detection takes place
> faster than
> >that (and certainly faster at higher frequencies.) So, my
> conclusion is that
> >no audible portion of the dit is lost.
> I suspect the sampling time is a lot longer than this. While I'm not
> certain which microcontrollers are used in the 87A, due to the age of the
> design I will bet that their clock oscillators run a lot slower than 10
> MHz. Therefore, the CPU cannot measure frequencies such as 10M or 15M
> (particularly with enough resolution to discriminate between) without a
> divider between the RF input signal and the CPU. While I've not done the
> required math, my educated guess is that the actual sampling time is more
> like 50-100 mS once these factors are taken into account.
> I also don't know if its a valid assumption that the PIN diode T/R
> circuitry is capable of withstanding hot switching, particularly on the
> output. I suspect the design will begin tuning on the dit, but will not
> switch into transmit until the RF input signal is absent.
CQ-Contest on WWW: http://lists.contesting.com/_cq-contest/
Administrative requests: cq-contest-REQUEST@contesting.com