Saturday, November 29, 2008

RF Power Issues with the 857D

Readers of this blog know that I love this little radio. It's the first "new" HF radio I've ever owned. I've always had older tube rigs so understandably I've been quite proud to own and operate this little gem of a solid state compact radio.

It has been fun to drill down into the menus and to occasionally discover a few of it's undocumented features. One interesting thing I noticed, early on, were momentary high power "spikes" when using the rig on low power - this may not be an issue when running the rig at 100 Watts output on CW and SSB.

The ARRL and other reviews of this rig didn't show a problem, yet speaking with other users and doing numerous Google searches suggests many FT857 and FT897 transceivers also suffer from this issue. In the case of my transceiver the effect seems most noticeable in the mid HF frequency bands and is present on CW, SSB and AM. For those who use AM, the spike can be horrific and exceeds 100 Watts out when the microphone PTT is first pressed. SSB doesn't fare any better with spikes occurring on the first syllable spoken after a pause. The spikes on SSB exceed 100 Watts when the power level is set to 10 or 20 Watts.

If you are a frequent reader of this blog you know that I lost the finals in my 857D a while back when transmitting thru a solid state amplifier on 2 meters. I have since been searching for a reason as to why that failure may have occured. I'm still afraid to use the amplifier again until I am more assured as to what happened.

A Gooogle search of my problem led me to a most excellent website, Dave's Astronomy Page http://www.astromag.co.uk/ . I don't know Dave and have not asked his permission to republish and otherwise plagiarize his excellent research but I trust he will not be offended that what follows is actually his work.

Dave's concerns over this "power spike" were for the welfare of his linear amplifier.

Running the FT857 "barefoot" doesn't give any indications on the front panel of a problem, and indeed in many cases there isn't an issue with doubling your power for the first 5 milliseconds on transmit. This can be a serious issue when using the FT857 to drive other equipment, such as a linear amplifier. Fortunately, Dave says, his Acom 1000 linear is rated to cope with spikes of up to 100 ms duration, and these are less than that. Users of other types of amplifier may damage their equipment.

Examples of the first CW character being sent are shown below, the power output setting was 20 Watts on the 7 MHz band.

The only cure found so far is to feed a fixed voltage of between -3.0 and -3.5 Volts into the ALC socket of the FT857. Using a fixed ALC voltage still allows the power output to be set correctly from the menu in the FT857, it is also beneficial in taming the SSB spikes without causing problems to normal speech peaks.

Not all bands produce the same results, the spikes were at their worst in the mid HF bands and almost non existent on 2m. Dave says his oscilloscope is not capable of working at 432 MHz, do he doesn't comment on that band.

From experimentation, feeding what appears to be a reasonable voltage level on one band can reduce the power output on another band to near zero. This indicates that setting the voltage could be quite critical. I haven't tried the transceiver at widely differing case temperatures, nor have I tried any other transceivers to see if using the voltage my transceiver works well with also has the same effect on a another 857.

Quoting from Dave's website: The test circuit I used is shown below. Using a 9 volt alkaline battery is not a good choice as the maximum voltage required is only -4 Volts. The output (tip and sleeve) are wired to a 3.5mm jack plug, which is inserted into the ALC input socket on the FT857D. The 1.5 K Ohm series resistor avoids shorting the battery out when inserting the jack plug into the transceiver. For longer term use 3 or 4 AA sized alkaline batteries would be a more sensible source of the voltage. Note the FT857D manual gives an ALC input range between 0 and -4 Volts in the specifications. The optimum ALC voltage feed on my transceiver was -3.1 Volts, this level still allowed full output on 2 meters. A setting of -3.6 volts worked well on HF and 6 metres, but killed the output completely on 2 meters.

A more elegant solution is to generate a negative voltage by the use of an NE555, as shown below (note the 47 uF capacitor across the zener was incorrectly shown in an earlier diagram). A 13.8 volt supply is available from the "linear" socket on the FT857. Be careful when wiring a mini DIN plug as they are quite small and tricky to solder without creating a short. Covering the rear of the pins with hot melt glue after soldering the wires helps keep everything secure and short free - the plastic glue can be peeled off for re-work without difficulty. The photo of the completed box also has a phono socket connected to the TX ground pin of the mini DIN plug to give a feed to my Acom linear.

Dave says he contacted Yaesu for their view on this problem and they apparently have a key click modification that is applied to the current FT857D series, from board 07 onwards.

Anyone wishing to modify their FT857 using the modification below does so at their own risk. While the modification was supplied by Yaesu, it doesn't make it easy to implement. The modification can be downloaded here .

The central part of the underside of the main board in an FT857D is shown below (not modified).

Tuesday, November 11, 2008

How to decode Yaesu serial numbers

If you need to know exactly when your Yaesu radio was manufactured here is the way to decode Yaesu serial numbers:

Position 1 = year made

Position 2 = month made

C = January
D = February
E = March
F = April
G = May
H = June
I = July
J = August
K = September
L = October
M = November
N = December

Pos 3 &/or 4 = lot number(s)

Pos 5<-8 sequence in that lot

Position 1: last digit of the year in which the radio was produced

Position 2: Determine the position of the letter in the alphabet (C=3,D=4,E=5, etc) and then subtract 2. The result gives the month in which the radio was manufactured. Note that the letters will range from C (January) throughN (December). Thus, a radio with a serial number of 5H221234 would have been manufactured in 2005 (5), during the month of June (H=8, 8 minus 2 = 6, June is the 6th month), was part of lot 22, and was the 1234th radio in that lot.