After many years of using
various
valve
amps for 50MHz, it was time to go solidstate but how to do this without
breaking the bank? An email from a friend in Perth alerted me
to "some modules that looked interesting" on ebay USA. When I
looked there were these ex TV transmitter amplifiers from a Canadian
company
called LARCAN.
As it turned out the USA had already turned off most
of
their low
band analouge TV transmitters and a large number of these modules had
hit the
HAM
market. In Australia there are a number of these LARCAN TV transmiters still in use, some on CH3, some on CH1 and at this time some still on our CH0 (46.250 Video) but I could not be sure that the amps would ever become available to HAM's so I took the plunge and sourced two of these via a very friendly club in the USA. As it turned out the freight was almost as much as the purchase prices, so AUS$1,300 later, I had two modules to start my journey with. |
ResourcesIf you are after some background on these, I would suggest you start at these sites: |
Power suppliesThis seemed to be a bit of a hurdle until I found that on the Oz ebay site there is a computer recycler that is selling brand new ex HP blade server 50V 3KW switch mode supplies for $45, what a steal, so I bought two. They are still listed on ebay Oz as at Sept 2011, look for these keyword details:HP 3000W HS Power Supply Proliant BL p-Class 226519-001 $45 If you are looking for a good source of info about these HP server power supplies, I would suggest you have a read of the pages from Lance Collister W7GJ here, this page is mostly about the HARRIS power amp modules but has some good info on the HP power supplies and noise filtering. When connecting up your PSU to the Larcan module, ensure you have a good earth from the case of the PSU to the earth of the PA module, otherwise you will have S9 sized white noise on Rx. |
My Journey with the Larcan 1KWMy modules were what is called a HI-LO, so they were designed for around 77 - 88MHz and above, as the LO-LO modules had all sold out. Okay so thats not too big of a step, as there are details on how to easily change a HI-LO to a LO-LO out on the GEMOTO web site here or the Minute Man Repeater group.So now I have my module converted to a LO-LO, by removing some links and then changing some chip caps. But wait, these are not just any sort of chip cap :-) these are high voltage multilayer ceramics from a US company called ATC because there capacitors are so popular for RF applications, it is common to see most of this style of capacitors referred to an "ATC Cap" So here I go, add some RF on 50.1 MHz and watch the needle on the bird wattmeter go twang on the end stop right? nope all I get is the PSU tripping off on over current but why? Okay the rig I was using to drive it (IC-575A) was spiking RF at the start of each test transmission (15W) when I used the front panel power control (ALC) to reduce the drive and causing the amp to draw excess current, bugger, need to solve that before I do some damage. Next step, pad down the input with some fixed attenuators, over current problem fixed, 3 watts in and I get 100W out on 50.1, gee that seems poor (insert favourite swear words and repeat often). Changed frequency on the driver transmitter up to 54.995 and now I get 350W out for the same drive level and more importantly a similar current draw from the PSU... Huh some thing is not right here. At this point I started walking in circles and double checking EVERY THING that I could think of. Why Me . . . Why ME ! surely I did not deserve this, I have been good santa ;-) Quite a few weeks go buy and I loose interest after lots of proding and poking as it seems that no matter what I do, this thing will not give a usable power level at 50.1 MHz and a simple calculation shows me I have around 20% efficiency for the DC going in (YUK) suddenly those valves are looking good again. Well according to some of the post's in the google groups for the Larcans, a few had been using them on 50 MHz but most of the published conversions were for FM repeater use, so I started to post my problems to the group, some of the group members suggested a few things to check but in essence no change to my problem was found but it did force me to go and work my way through a module alignment to set the FET bias levels. My next step was to do some low level sweeps of the module as a whole (HP tracking Gen and Spec-AN combination) which confirmed my suspicions from the higher power measurements, the modules have a crook old roll off below 55MHz. This led me to look at the output combiners performance and I could see that there was an increase in the combiner loss of more than 3db from 55 down to 50MHz... yikes more than half my potential output power would be lost in the output combiner. During this time I zap'd some of my FET's in my first module with each giving me the dreaded blue flash as they expired, so now I have three of my 4 FET's with dud halves... sob sob. Onto module two and start the conversion process all over again, remove shorting links, change input and output connectors and change caps maybe module number two will be better? Nope same outcome but at least I have four good FET's on this one. Then a posting in the google group from some one with similar problems to me, Dave Olean, K1WHS had run into the same problems I was seeing when trying his amp at 50.1 MHz and was working on a solution too. Now some might recognise him as being the man behind Directive Systems but it took me a while to put two and two together :-) thanks for your efforts Dave. Anyway Dave has spend quite a bit of time on this and come up with a set of mods that should be applied to ALL Larcan LO-LO modules (including conversions from HI-LO to LO-LO) if you want it to run on 50.1 MHz. It includes extra capacitors on each AMP module and extras caps in the output combiner. Quite simply the Larcan modules drop of dramaticially in performance as you go lower in freq and by the time you reach 50.1 MHz you are already 4 MHz below their bottom limit from the factory. See daves article here As Australia still has at least one active CH0 Larcan TV transmitter (due to go off in June 2012) that I know of, it seemed a natural leap that they would have details of how they got them to work at 46.250 MHz (video carrier) so stay tuned as we seek out what the factory did to get these modules to work on and below Six metres and maybe then we will see further improvements on the work done by K1WHS to get these kicking arse on 50MHz. Australia has quite a few Larcan transmitters that are now out of service, most are CH1 and above (LO-LO and HI-LO) so there should be an abundance of these modules available once the disposal process is worked out. March 2012: I am in the process
of applying Daves changes to
my LO-LO and will fill you in on how it goes.
Late March 2012: Have sat down and
made the changes to my now LO-LO module to fit in with the capacitor
values in
Dave Olean's article. WOW, the Bird watt meter (500W slug) now easily
goes full scale at 50.1 MHz with just 5W of drive from the FT-817.
I can finally say I am impressed and the power supply is hapy
too.
April 2012 time to do away wit
the Larcan DC power connector and use a standard Anderson power pole on
a flying lead from the PA module
May 2012I
have the go ahead to visit the last HIGH power CH0 LARCAN based
transmitter in Australia on service switch off day in early June 2012,
on this day about 30 analouge TV transmitters will be turned off
June 2012 I
have made the journey of 1,900KM (1,187Miles) from my home near
Adelaide to visit the CH0 site near Junee in
New South
Wales (VK2). I reality it is quite a long way away from Wagga Wagga,
NSW and have all the details of what makes a CH0 Larcan tick, I need to
corelate my findings and publish it here. I can tell you that
a few of the inductors on the PA board have an extra turn on
them
and some small but important changes to the cap values on the PA board
as well.
Jan 2013I
have inserted an in-line 6dB attenuator on the input combiner
boad to pad down my drive power, wired the antenna change over relays
to a RCA socket on the back panel and moved the AMP module to the main
operating desk. The Larcan 1KW is now in-line with the 6mx rig
and can be used at the drop of a hat. Measurements show I can easily
produce over 1.3KW os SSB (PEP). March 2013A
top lid is all but complete, just have to work out how to secure it to
the case without creating a new problem, amp has been used for many
local VK - VK contacts but nothing outside my home country yet. What is different between a CH1 and a CH0 (46.250 MHz) 1Kw moduleAll the PCB model numbers on a CH0 and a CH1 are the same.Input combiner, no changes On the input side of the PA board, these changes are evident: C3 62pF and not 63pF as seen on a LO-LO C4 160pF and not the 200pF seen on a LO-LO On the output side of the PA board, these changes are evident: C13 43pF and not the 82pF seen on a LO-LO C16 110pF and not the 51pF seen on a LO-LO C14 and C15 are divided up into a pair of chip caps and not a single of a larger value, reason unknown but I suspect it is to do with the distribution of RF currents. C14A 91pF C14B 91pF C15A 91pF C15B 91pF L4 and L5 the DC feed chokes to each FET half, each have 3.5 turns, compared to the 2.5 turns on a LO-LO. Output Combiner, no changes June 2012, vk5pj I have a LO-LO Do I need to change the capacitor values? YES you do!There seems to be some confusion as to whether the article by Dave Olean (K1WHS) applies only to a HI-LO conversion or to an existing LO-LO as well. The simple answer is BOTH need these changes to be made. The starting point for Daves investigations was that he had, like me, converted his HI-LO module to a LO-LO then found that the LO-LO configuration was not the complete answer. If you have a LO-LO module then you need to make the changes he outlines in his article.Following his suggested changes shown in the article has been the only way I have had any success with my Larcan PA on 50MHz. If you want to use it at 53MHz for FM then you can leave it as just a basic LO-LO with a some what reduced efficiency. I must thank Dave for his work and his contribution the the google group for Larcans. IPA-2For those looking for a slick 500W amplifier, I would HIGHLY recommend you skip the full PA module and grab what is called the IPA-2 module. It has two out of the four PA module on it and it is split/combined with commercial ARNEN hybrid's and has none of the problems with the high power combiners seen on the main 1Kw module, photos and more information here If you want to maximise its efficiency you should make the relevant changes to each PA module that is outlined in Daves article. |
Protecting from Over DriveOne of the many decisions that were needed was how to protect the AMP module from over drive, I have eventually settled on using a 6dB 100 watt block attenuator sourced from Henry Radio in the USA, these nice looking flange mounts look the same as one of the dump loads already used on the Larcan output combiner modules and fit right in. I purchased some 3dB, 6dB and 10dB attenuator modules as I had some other projects in mind and it would allow me to make changes as needed. At first I thought I would use a 10dB and run the driver rig flat out at 100W for drive and rely on the atten to disapate the un needed power, then I changed my mind and placed a 6dB into the input line. I used a mini grinder (Dremel tool) to make the needed cuts in the PCB, you can se in the photo that it sits just to the front of the middle split of the two heatsinks. |
DC connectionsIts time to retire the special larcan connector and replace it with some thing more common |
![]() |
Larcan DC connector removed, tapped out the bottom
right hole to 6mm (1/4") and roughened the surface to get a
good earth. The +ve is just another 6mm bolt and nut combination. This also allows me to get a RFI suppression ferrite close the DC input for the amp. |
![]() |
50A Anderson power pole connector, a bit BIG for the job but all I could find at the time |
![]() |
There is an un-used 4-40 threaded hole under the input connector, I have tapped to be 6mm (1/4") and it is used for the earth connection to the HP switch mode PSU. |
Pictures, in no special order
|
![]() |
Original
output connector.
|
![]() |
top view of
original output connector
|
![]() |
This is the
BIAS and control board on the rear pannel (output connector to the
right)
|
![]() |
Power
connector and BIAS / control board
|
![]() |
Input
connector after being changed to a BNC
|
![]() |
Original
output connector removed and a scheme to replace it is hatched
|
![]() |
A view from
the rear of the module looking along the fuses and FET's (input to the
left, output to the right)
|
![]() |
Top view of a
LARCAN PA, shorting links still on output strip lines
|
![]() |
Detail
of a single 250W section, showing the shorting strips on the output
lines to raise freq [HI-LO] (removed for [LO-LO])
|
![]() |
My version of a new
output connector for the P.A
|
![]() |
top view of the new
bracket and N type output connector
|
![]() |
Side view of the
new bracket and N type output connector
|
![]() |
rear view of the
new bracket and N type output connector
|
![]() |
After mods
to each 250W module, still need to improve the output combiner
(poor soldering on my first "test module" has since been cleaned up)
|
This web page was last edited on Saturday, 11th March 2013 by vk5pj |