Thursday, 25 May 2017

Meet the VHF Linear Amplifiers and general update

In reverse order (as is so often the case here).

Why reversed? Because I am "contrary".

By now (25 May) the Sporadic E ("Es") season is well underway, and every good VHF enthusiast has been filling their boots.

As an example yesterday the Es opening started at 08:44 and ended at 22:40 when I turned off and went to bed. I am not sure how long it went on after that. If you looked during late morning and late afternoon then you missed the best bits. There was a lull from 13:02 to 15:10. A classic two phase day.

I had 350 WSPR spots on 10m, as follows:-
24 hour map of WSPR spots at GM4FVM starting 09:00 24 May 2017
This usually provides a good guide as to where to point the VHF antennas, as well as telling me when it was the right time to go to Morrisons to buy bread (i.e. during the lull). The Seeded Batch is a good loaf and it will be nice with a boiled egg for lunch today. Just waiting for 10m WSPR to tell me when to put the pan of water on to boil. Wait for the lull to start. Don't want to miss ANYTHING.

There is a danger here that a guy could get obsessed. Not me of course.

I had very few contacts on 6m yesterday, and none at all apart from locals on 4m or 2m.
VHF stations worked at GM4FVM on 24 May 2017
I was somewhat hampered by local 6m QRM which blotted out several DX stations which I could hear in the clear. That tends to make me move on to JT9, which is a less popular mode.

The opening on 24 May was notable more for its length than anything remarkable. For example, I worked F1ABL on 6m at 22:00 and SP6IHE on 10m at 22:30. During the QSO, SP6IHE broke off to work a W station and then returned to finish with me. Some stations worked into the US on 6m but not me.

The few days which preceded this produced more variety, including a couple of German stations on 4m. Their 4m authorisation lasts until the end of August though so far I have not heard many, not even on meteor scatter. I am trying a different map background, just to prove to myself that not everybody speaks English.
VHF contacts at GM4FVM 18 to 23 May 2015.
As the Es season progresses I would expect more double hop and more openings on 4m plus possibly the odd one on 2m.

For most of yesterday, apart from calling CQ on 4m, I sat and watched the whole thing unfold. It is a remarkable aspect of nature. You might expect it to move West with the Sun - it doesn't (other than in a vague general way - it opened here due East last thing before I went QRT). You might expect many stations to seek out DX, whereas they seem very happy just to work the shorter distances.

Yesterday it started early and ended late - today it might do the reverse. In fact it probably will do the reverse just for badness, as the one constant feature seems to be that the same thing never happens two days in a row. These are the things about Es that amaze me and keep me watching for hours.

I guess I am just easily amused.

Better still, I spent some time trying to get the tone burst working on my £8 Baofeng 888 handheld. I never even knew it had a tone burst. I had a conversation with Ian, GM4UPX, who knows about them but so far it shows no sign of wanting to turn the tone burst on. It seems to be set by the software (just as well as that rig has no readout nor many controls). Ian says he has got it to work so more conversation there soon.

Then again it was good to hear from Chris, GM4ZJI. Chris has migrated towards 2m satellite operation which is a great idea but I am not there to hear him. Still, he has put up a 4m vertical and we had a "part text/ part phone call/ part QSO". A vertical is a good way to get involved on the Es as well as for FM contacts. Good to hear him too. I look forward to further chats.

Funny really. This is a hobby about communication, but we often sit alone in stuffy little rooms and ponder to ourselves about things. We set up computers or solder things together, and we often forget that conversation is very helpful. At least I have a few sensible people to chat to, and I am grateful for that. The little things matter - Richard, GI4DOH, must have seen me on 10m via the cluster at 23:06 local time last night and he sent a friendly e-mail.

Speak for yourself Jim, not everybody's shack is stuffy.
My linears are a motley bunch.

There was no strategy behind buying them, and in fact I am not keen on linears much at all.

For Es working, a linear is not that much use. Just like Es, auroral signals are often strong too. Tropo working here is limited by my site and if I cannot hear them there is no point having a linear. It is when it comes to meteor scatter that I felt the need for more power.

Meteor scatter signals can also be quite strong, but only for fractions of a second. They tends to start quite strong and fade away. By using more power you extend the time that the station at the other end can hear your signal. It is this time extension which makes the difference rather than the level of the peak "ping" at the start.

I do not have any 400 watt or larger amplifiers. I simply wanted to raise my power to be in the same order of magnitude as the stations I was working. For meteor scatter this turned out to be 200W on 6m and about 150W on 4m and 2m. I have not found any need to go further.
GM4FVM linear shelf as at April 2017
The photo shows the linears in a previous layout. Since then, as the shelf keeps sagging under the weight, I have moved the 6m linear onto a nearby shelf.

Just about the only thing which they all have in common is that they are all hard wired separately to the PTT line for their various dedicated rigs.

I ended up with this strange line-up mostly by chance. Back when I used transverters for 4m I could produce about 25W output. This was useful on meteor scatter, but not quite enough. I needed something a bit more muckle. So about 5 years ago I bought a TE Systems 0610G 70MHz linear rated at 130W, which happily ran at 100W if I added a pair of cooling fans.
My TE Systems 0610G for 70MHz now sadly sold.
This linear did great service for five years. I had no particular reason to change it, but it was replaced rather by chance.

My previous 6m linear was a Bremi valve CB linear converted for 50MHz. When I finally got fed up with the instability and unreliability of this vintage monster I decided to buy a Linear Amp Gemini 4 to replace it.

The Gemini claims to produce 270W on 6m, with me hoping to use about 200W for meteor scatter. While I was waiting for it to arrive I contacted a GM operator who had one, just to ask about it. He mentioned that he had a 375W 50MHz TE Systems 0552G which he was not using. I quickly arranged to buy that amplifier and it is the one on the left on the shelf. The Gemini then took up its place on the right as the 4m amplifier. So the Gemini replaced the 70MHz TE and its 300W theoretical maximum output allows me to raise my 4m power to 150W. In between is a Microset SR200 2m amplifier which I use at about 150W.

The Gemini comes complete with integral mains power supply and (noisy) fans. The TE linear, on the other hand, comes with a 12V input and no fans. The TE handbook states that with a high duty cycle mode you should fit fans controlled by a 12V TX output available from the socket on the back. It recommends using a timing circuit to keep the fans running for longer than just the TX periods. Being a good amateur who does what he is told, I tried to fit a timed relay only to find that the 12V output from the linear could not provide enough current to lift the relay I had.  At this point it seemed simpler to follow GM4JJJs suggestion and make the fans temperature controlled.

Cheap temperature controlled relays are available from eBay for about £4. These allow you to set a desired maximum temperature as measured by a remote wired sensor which I placed at the bottom of the linear heat sink near the output device fixings. The LED readout on the chip allows you to set your desired temperature and it then switches the fans to maintain the range to within 2 degrees C of the desired level. In my case I set it to 25C. At that temperature setting the fans come on within a minute of my first transmission and stay on continuously until after I have finished.
Cheap temperature control relay from eBay being fitted into an ABS box

It is a simple matter to fit the temperature board into a box, which you can equip with all the sockets you need. In my case, probably unwisely, I use phono plugs for DC supply leads, the fan leads and the TX 12V output from the linear. It might be best to use different plugs to avoid some wiring mix-up catastrophe but that seems to be my standard now. I make clear labels and hope that I read them! There are several sockets on the box for my fans and one which I use to illuminate the backlight on my 6m power meter. I cannot hear the fans so it is useful to have a visual indication that they are likely to be working.

When I comes to fans I really dislike the standard 80mm PC fans. I find them noisy and ineffective. For the TE Systems linear I use four 120mm fans. At 200W output this finds an equilibrium temperature of around 30C. At the quoted maximum for the TE 0552 is 375W, at which point it is drawing well over 50 amps from the power supply, it is running at about 33C thanks to the fans. Whilst the linear would be happy at 375W on SSB, on modes with 1 minute TX times it gets rather too warm internally.

I suppose if I went for screaming small diameter fans moving the air more quickly I could use more than 200W, but I prefer to under-run the linear at 200W. That way it stays nice and cool, (hopefully) linear and the power supply remains happy providing about 36 amps. The over-temperature switch fitted as standard inside the linear trips at 65C.

So that is the TE Systems 0552G. It runs happily and fairly cool at 200W. Apart from the switch panel supply failing as soon as I got it (the repair for which was described earlier in this blog) it seems to do well. Whatever caused the supply to fail to the switching circuit, I think that is also the cause of the lack of current available to switch the fans.  My bet is that despite what they say in the handbook, it is not capable of powering fans via the socket at the back for long before something fuses. Still, that is solved now.

It couldn't be that you put in too thin wire when you fixed the problem Jim? Aw no, that would be impossible.

TE might have used a socket type in more general use for PTT than the RJ45 socket they now use as standard. I built a break-out box for the PTT and fan supply. It might also have had "N-type" RF sockets. It might even have had the necessary fans already fitted (though a continuous duty rack version with fans is available). It had none of these things. I bought it second hand and so far to does the job nicely.

Moving on to the Linear Amp Gemini 4, it was bought initially for 50MHz use. As the TE linear turned up, I moved it to 70MHz. There it runs happily at 150W. It is self-contained, causes no heartache and runs away needing very little attention.

Drawbacks, well it is noisy, with the transformer, fans and the changeover relay all contributing. On SSB it is possible to trip the overdrive cut-out on speech peaks early in the transmission. I am pretty sure that this problem stems from an initial spike in the IC-7300 output. Reducing the power setting on the 7300 seems to function via the ALC. This arrangement can allow instantaneous spikes at the start of a tx period. This is a common issue with many rigs. If I set up the 7300 carefully enough it is not a snag, and it only affects SSB. I know another amateur who has the same issue with the Gemini, and others who do not. To my mind, this is a rig issue rather than a linear amplifier problem. That does not explain why two stations with the same rig and amplifier have different experiences, but then, am I am expected to explain all the mysteries of the Universe?

There is a simple reason why the spikes were never a problem with the TE Systems linears. TE Systems have no protection circuits apart from an over-temperature cut-out. You can never trip the overdrive limit by transmitting a spike into the TE Systems Linear because it has no overdrive limiter. Nor does it have a high SWR cut-out like the Gemini. You take a risk using linears with no protection, and I read that the TE Systems amplifiers have been prone to problems for this reason.

Dealing lastly with the linear in the middle, we turn to the Microset SR200. This is a 200W rated 2m linear. It has overdrive and high SWR protection. To produce the rated 200W output it needs about 50W drive. Like the TE Systems linears it has a built in GaAsFET preamplifier, but in both cases you can switch them out of circuit independently of the PA TX stage.

I bought the SR200 direct from Italy, and at the time I saved about 20% of the UK list price. In the four years or so since then it has worked without drawing any attention to itself. It never failed like the TE, nor tripped like the Gemini. It has always just worked away to itself.

The drawbacks for the Microset include that it needs to have fans fitted for high duty cycle operation. It never gets very hot, but then I never draw full rated power. It does not have a fan control output, so I will be installing another temperature control to its fans. For once I am going to try a variable speed control and we will see how that goes. You can buy fans with integral temperature control and remote sensors which I will try, though these have preset temperature ranges. I will be changing over to 120mm diameter fans at the same time. In the past I used the Microset with 2 small and 1 large fans with nothing more complex than a switch.

Strangely for a high-VHF item, the Mircoset is fitted with SO239 RF sockets rather than "N-type". These are easily changed.

I use the same 12V power supply for the Microset linear as  I use for the 6m one. This could be a problem if I accidentally transmitted on both at once, but that has never happened yet.

So back to the start. Why do I have linears? Because nobody can use a commercial rig at full output on meteor scatter QSOs and expect it to survive at full rated power. I drive the 6m linear with 8 W for 200W output, meaning that the rig only needs to run at less than 10% of its rated out (and the linear at just over 50% with added fans). On 4m it is 8W for 150W output, using just 16% of the rig's output (and again using only 50% of the linear's rated power). On 2m the rig provides about 25W. To try to get a lot of power from a rig on a regular basis is risking a very expensive item for little benefit.

Also, none of the linears is working near its full rated power. If, say, you took the manufacturers advice and drove the Gemini and TE with 25W and the Microset with 50W, you might get away without overheating them on SSB. However, you risk pushing them into non-linearity. Personally I set them up using a power meter, checking full power output and then backing the power off to well below the maximum possible, and ideally to 50 to 75% of the maximum.

To me anyway, linearity is very important. I do not wish to get a bad name for over-driving a linear amplifier. The drive levels quoted by a manufacturer will be maximum figures, and my meters are not sufficiently accurate to ensure that I do not overdrive the amplifier. So I tend to back everything off - and it saves the rig too.

There was a magazine article recently which tried to suggest that everyone should have a linear and run full legal power. It suggested that this was our duty as amateurs as otherwise our legal power limits would be cut. Well, if that is the case I am failing in my duty.

There is a school of thought that says that more power is just better in every situation. Sorry, but at 200W I have gone as far as I ever need or want to go.

I suppose therefore there is no logic in my choice of amplifier or the use I put them to. For now though I am fairly satisfied with them, and the results they deliver.



No comments:

Post a Comment