Sunday, 10 March 2019
The FT-991a has pc based ptt in all modes except CW! Note that I don't mean keying the radio through full or semi break-in. The reason for not using these break-in modes is as follows: 1) Full break-in is useless in contests as you never want to hear while you are sending. Also this mode thrashes the radio/amp TX/RX changeover relay, leading to an early demise of your radio/amplifier. 2) Semi-break-in incorporates a delay to prevent the radio from removing PTT between sent cw words and elements. This delay is also applied at the end of every transmission, meaning you miss the start of callsigns from fast return callers. Also if you vary the speed of sending or have a different manual sending speed than the PC, the delay value will need adjusting each time to avoid needless PTT drop-outs. To avoid these pitfalls it is more efficient for the PC just to apply PTT when it needs to. When the PC sends a cw message it engages PTT and then removes it immediately at the end of the message. Yaesu aren't interested in fixing this in their firmware, so here is a work-around for anyone who wants to use an application (N1MM+, Win-Test etc) to PTT the FT-991a in CW mode. The only way to PTT the FT-991 from a PC in cw mode is through pin 6 of the RJ45 mic socket on the front of the radio: (https://shop.technofix.uk/cable-for-pc-headsets-3-5mm-jack-8p8c-modular-rj45-for-yaesu-ft-817-ft-857-ft-897-ft-450-ft-900-ft-991-etc). The 1/4" PTT jack socket can be used to hook up the PTT output from your Win-Key or other device: (https://www.hamcrafters2.com/WKmini.html). FT-991 Menu items: 056 CW BK-IN TYPE (semi/full). 057 CW BK-IN DELAY (30msec to 3000msec). Yaesu have made a complete hash of this radios firmware in that 056 doesn't allow a PTT option (like every other radio on the planet). Even 057 doesn't allow a 0msec setting to get rid of needless delays, so here are my recommended settings for cw contesting using the above hardware: 056 CW BK-IN TYPE (semi). 057 CW BK-IN DELAY (30msec). Using the above hardware and firmware recommendations means that the FT-991 can now be used for cw contesting in a nearly efficient manner. Ok, a needless 30msec is tagged onto the end of every transmission, but its still better than using Yaesu's semi-bk-in with 300msec (28wpm approx) tagged onto the end of every transmission. It's a shame that Yaesu seem to have lost any in house experience in CW contest operation and are now producing sub-optimal equipment for the masses who don't care or know any better. Just imagine if they they did this this on data modes too, the FT8 crowd would be on them like flies. I can only hope that radio was an ooops moment for Yaesu and that the architect is demoted to coffee duties. Hopefully this article saves many ops selling up their FT991 in favour of a proper cw radio like an Elecraft K3.
Thursday, 3 January 2019
There are currently several problems with this unit. I won't go into the Dell tablet internal battery charge problem (if not connected to power for several months) as this has been admitted by Flex (not sorted) and is well documented elsewhere. The problem that concerns me is one that FlexRadio claims is the fault of common networking equipment e.g. Netgear GS105 gigabit ethernet switch. As a firmware engineer myself, I see this problem as being a Maestro 'power on' boot problem where if the negotiation between the Maestro PHY and the Gigabit switch PHY (auto-negotiation for ethernet link speed) is interrupted (for whatever reason), the two units connection is permanently corrupted until the next power cycle. This means either the Flex or Netgear implementation of this auto-negotiation protocol is incorrect. I tend to trust a networking specialist company over one or two peoples opinion in Texas, especially when everything else plugged into the switch works FB, however YMMV. Apparently if you buy a specific type of switch or mandate a 100Mbs link speed this problem goes away (I am less than impressed with this Flexradio solution? which should really have been a eureka moment for them). FlexRadio does hint at the real root of the problem being that most commercial networking equipment is meant for browsing the internet rather than permanently streaming spectrum and sound information and recommends moving to gaming network gear to achieve a better QOS. I have no problems with QOS, once the Maestro is successfully past the auto-negotiation sequence (which in my opinion is the real problem) everything works fine. The fault happens every 1 in 10 power ups and is realised with the Maestro showing high packet loss despite being wired ethernet connected to a gigabit capable switch. Many users have tried to convince FlexRadio of this fault, but due to the 'fanboys' applying pressure, and the ignorance of company, no reasonable discussion can take place on their community forum. If you are happy to pay £1300 for a Dell tablet with a 1 in 10 start up success rate then go right ahead and buy one, ps. all problems will be down to your own networking equipment!
(without the wire). If you are considering buying this antenna second hand, ropes and silicon tubes should be inspected for wear and tear as these are the only perishables on the antenna.
Sunday, 21 October 2018
Yaesu marketing are in full overdrive claiming the FTDX101 is a direct sampling SDR. Sure their display is driven using direct sampling SDR technology, but anything you hear out the radio (the bit that counts?) goes through a mixer to a 9MHz IF then through a roofing filter etc etc. The whole point of direct sampling is exactly what it says on the tin "direct sampling" of the RF from the antenna not of a 9MHz IF with analogue mixing spurii! Same goes for Icom 7300 and 7610. If you want real direct sampling technology "to listen to" get a Flexradio or an Annan. It is really that simple. I hope the market can come up with guidelines on how to properly describe the glut of new architectures coming onto the market. Something like: hybrid SDR - Anything that uses a mixer as part of the receiver chain that eventually produces audio i.e. Icom/Yaesu/Kenwood/Elecraft. SDR - Anything that directly samples (at more than double the input RF) from the antenna i.e. Flexradio/Annan.
All radios when sending cw with PTT enabled (no full or semi break-in delay) have cw sidetone alone coming out of the speaker to allow the user to hear what they are sending. Flexradio have decided that this is isn't enough and have filled the gaps between the sidetone with receiver hiss. No antenna is actually connected to the receiver during this time, but the engineers at Flexradio think it sane to decode the receiver open circuit anyway. Flexradio engineers fully engage with the full and semi break-in brigade (rag-chew, DXers, occasional contester) as this category covers 99% of the user base. These modes are useless for serious contesting, let me explain why: Flexradio Full break-in (really useless): 1) CW contesting is normally done with an amp over a 48 hour period of approx 50% duty cycle tx. This kind of usage will shorten the life of the tx/rx switching mechanism dramatically for both radio and amp. 2) When running a frequency I have no need to hear between my transmissions as those heard are mostly slow or poor operators. Flexradio Semi break-in (useless): 1) A delay is used to stop the radio unkeying between each cw element and character. This saves the tx/rx switching mechanism somewhat (if set correctly), but unfortunately this delay is also added to the end of the transmission where I want to be listening for fast return callers. 2) If I need to vary my cw speed the delay must also be adjusted. Not good for multi-op contests. Let me now explain why computer PTT or manual footswitch PTT is best: External Winkey/Footswitch PTT (brilliant): 1) When sending via my computer, it knows when the end of the stored message is coming and can remove PTT immediately allowing me to hear fast return callers. 2) When sending via my paddle (external K1EL winkey), its hang-time parameter is sufficient to save the tx/rx switching mechanism. 3) The hang-time delay parameter of my winkey changes value automatically with speed. With Flexradio implementation of what to do when PTT is applied, the operator has to listen to the following for a sent "CQ": long sidetone, false hiss, short sidetone, false hiss, long sidetone, false hiss, short sidetone, false hiss, long sidetone, false hiss, long sidetone, false hiss, short sidetone, false hiss, long sidetone, false hiss, real rx hiss. You can imagine how annoying this is going to get over a 48 hour period. Sure you can dial in some flexradio delay to get rid of the false hiss but then you get the above semi break-in disadvantages. Numerous users have reported this fault to Flexradio, but they think they know better...
Monday, 19 February 2018
I'm just back from doing the above contest. Using a K1EL winkey within the Microham DigikeyerII and Win-test contest logging software. I had the following problems: 1) When setting the Microham software to provide PTT with lead (20ms, to protect hot-switching amp) and constant tail (10ms, minimum allowable setting) I discovered that Win-Test drops the PTT on spaces when sending predetermined cw messages, even though it knows it hasn't sent the complete message. To counter this without lengthening the standard PTT tail, you need to replace spaces with several ^ characters (the halfspace character with 1/2 dot duration). 2) My laptop computer (Dell D630 single processor 2.4GHz processor speed) running cw skimmer, microham router software and Win-Test caused variable cw timing, pauses, burps and stalls when too many Win-Test windows were open (such as instant rate monitor, world map, statistics and other non-essential windows). This is puzzling as the K1EL is meant to be there to offload timing from the processor precisely to get rid of these timing problems. 3) CW sending speed would range every so often between the correct speed and about double that requested, normally at the "tu GM5A" end part of the exchange. 4) The winkey was set up to have independent speed control (alt-v for Win-Test messages and speed control on Microham for hand sent morse). In most cases the two speeds were the same, even when set differently!. A few taps of the escape key would return the hand sent speed to that suggested. 5) When joining a message with hand sent morse, there was always a stall, I think you have to tap the key then start sending with the key. All in all the net performance was far short of the required performance. Everyone hated whatever was causing all these faults, determining which part of the setup is responsible is difficult. The team recon that returning to a simple serial port transistor switched PTT and CW sending method would cure most problems. I am disappointed with the Microham and K1EL keyer for making such a bad job of a simple task!
Saturday, 30 December 2017
I bought one of these antennas at the Lincoln national hamfest from Martyn Lynch. I had bought a 40m OCFD from them before which seemed to work fine, so decided to go for something a bit more expensive. At £200 this antenna isn't cheap, but I bought it for some practical reasons: 1) I wanted an antenna with low noise 2) That would add height to my scaffold pole support (so the antenna would clear my roof line) 3) Multi-band capable as I have a small plot and only one support. 4) Needed something better than my OCFD for CQWW CW contest in 2017. The antenna arrived after an unplanned 3 month wait due to EAantennas not being able to keep up with orders. During the wait I complained to Martyn Lynch who offered a 10% reduction on price, then to EAantennas directly who responded, but failed to get the antenna to me for CQWW CW (It arrived the day after the contest). During assembly I noticed multiple problems: 1) The insulator clamps that hold the elements to the mounting plate are very odd. A clamp in my mind is something with 2 separate parts, these were one solid part with bolts through them. How does this clamp the element? (poor design). 2) The bolts that go through the insulator clamps are just long enough to catch the nut on the opposite side of the mounting plate. After much hassle I ended up flattening the spring washer with pliers to enable 1 turn on adjoining nut (poor design). 3) The balun is secured to the base plate and the element ends to the balun, if the element clamps fail, the elements can rotate and the connection bolts are long enough to short to the base plate (poor design). 4) The elements are too long and heavy for the surface area of the base plate, this means the element clamps are too close together, meaning the antenna is self-detonating in ANY sort of wind. With the antenna assembled I mounted it to the end of the scaffold pole (6m long) and raised the antenna in the sky. Inside I tried the antenna on 10 and sure enough the SWR was around a 2:1 in the cw portion. Jumping to 20m, the SWR was around 1.5:1. Not too shabby! I moved to 30m and used my LDG RM600 remote tuner (directly under the antenna) to bring the antenna into resonance. I was excited to try 30m as my OCFD never allowed me to operate there. I put the linear on and decide to start low at 250w to see how it played (the instructions say maximum potential 1Kw). I made 5x 1min qsos before the SWR started to rise, I could tell straight away it was the balun as after a while to allow it to cool down, the SWR would start behaving again. Maybe the antenna was so far out of resonance on 30m that I could only run 100w there. Oh well, I moved to 40m after an hour or so to let things cool and repeated the above experiment. Unfortunately the same thing occurred. At this point I realised I had bought a pup! The balun was completely useless! Luckily I had a proper 400w 4:1 balun made by GWhip. I went out and swapped out the EAantennas balun and tried the above experiments again and this time no problems (still fine at time of writing). The only downside was that I could no longer tune 30m (2.5:1 swr best case after tuning). The weather deteriorated after a week of having the antenna up and the forecast said 25mph winds. I had my doubts, but the specs said >85mph for the antenna, so I left it up. Big mistake, in what I would call light winds, the antenna started to bend at the base plate. Unfortunately I was at work, so by the time I got home, the antenna was at a 20 degree angle from vertical. I rotated the antenna round into the wind the other way and the wind straightened it up somewhat. What a pile of junk! Totally useless! This antenna is going in the skip (minus the after market gwhip balun, its actually worth keeping as its specifications are truthful and it performs to specification unlike the Delta 7B). Do not buy the Delta 7B unless you live on the moon.