TF3CY » 600W Mosfet Linear

600W – still going on

tf3cy — Mon, 01 Mar 2010 13:22:31 +0000

I had high hopes working on the project over the xmas time, but it turned out to be a very busy time, followed by a hectic work schedule, traveling etc… So after not being able to take part in the CQWW 160m contest due to lack of equipment I started the work again. I just finished making a filter board prototype and finishing the SWR metering bridge. Will post here late on. I’m now in the process of ordering a second EB104 with a splitter/combiner so the end result will be a 1200W linear.

600W – build prep

tf3cy — Thu, 03 Dec 2009 14:19:56 +0000

So, I have now designed the filter board and calculated the values of each of the 6 filters. I have some 1000V silver dipped mica capacitors that I’m going to use, and have been adjusting the toroids to match them 100%. I decided to use 5 element low pass filters in stead of the 7 element commonly used. I saw the design in the Ameritron ALS-600/1300 and according to Elsie filter modelling software they should work fine. PLEASE comment if you have something to say about this decision. I will post the filter calculation later when I finish transferring them from Elsie to some better form.

While @work, I played around making a block diagram of the amplifier + some front panel ideas.

[View with PicLens]

The complexity of the control board startled me, I see that I have to multiplex some of the input/output. The 16f877 has 8 AD inputs, 2x full 8 bit GPIO. I’m compiling a list of all IO and I will probably for simplicity use one of the 8 bit bus for controlling the LCD. I want to keep the serial UART unused for future expansions (remote control), then I have to figure out how to get the band data from the xmit, I have some old QST article showing a PIC based antenna switch so I will probably just use that design. Using a 74hc374 tristate latch to multiplex one of the data buses will probalby be the choice.

I plan to fire up the amplifier next week for testing and start the output module into a dummy load. I will probably use a manual bandswitch to control the filters. I’m NOT going to start operating the amplifier without the control circuits, but I have xmas to look forward to now.. I finally have something to do ;P

So the info I’m now starting to look into is the ALC circuit, and the band data input from the transceiver.

Spreader drilled and tapped

tf3cy — Sat, 21 Nov 2009 20:46:09 +0000

I had some issues drilling the spreader, I thought the drill would go through the copper like butter, that is partially true, but as this is a very thick piece I’m drilling, it grabbed the drill and twisted it. I had two 2.5mm drills twist to bits and I could not get them out of the hole. This actually happened on the first hole I drilled, so I ended up moving the whole PC board 2mm to the side. Now I marked the MRF150′s and I think it’s very important to mark them first, drill and solder afterwards so there will be no forcing of the pcb when they are clamped tight. Unfortunatly I broke on drill in one of the mrf150′s mounting hole, but I could drill the hole and tap it. I hope I did not damage the tapping drill, as a bit of the old drill was wedged in the hole!. secondary plan was to widen this one hole to M4.

Here is a picture, this was surprisingly badly drilled – I blame not having a proper tool to mark with ;P

here is a picture of the “handywork” – put some lines over to see how it really went, it’s still withing tolerance and the mrf’s are clear of the PCB, and no forcing when taped down. I used M3 screws – and for that I think you have some room for “mistakes”

600w update – pdf’s

tf3cy — Fri, 20 Nov 2009 22:13:22 +0000

Here are some excellent pdf’s from the father of the EB104

QST_Dec_1982_p13-16

QST_Jan_1983_p30-33

Motorola-600 Watts RF Power FETs EB104

And here are some of the best links I have found:

crazy italians (must stop this music!): http://i5uxj-2.cln.it/amp/amphf150/amp150.html

This is a good page, if not the best.. : http://www.g4apvweb.pwp.blueyonder.co.uk

One more: http://oh8jep.kotinet.com/eb104.html

This is perhaps not to par, don’t understand how this was posted in QST: http://www.w6kan.com/ssamp.html

Here are EXCELLENT pictures of the EB104 assembled: http://www.k6if.com/gallery2/main.php?g2_itemId=1768 This is the project page: http://www.k6if.com/ssamp/

Update Nov 20.

tf3cy — Fri, 20 Nov 2009 20:05:16 +0000

So, day7 in the H1N1 swine flue s***… just starting to gain consciousness

Got my way to expensive Mouser order today – happy that the FedEx charge was 0.00 usd… so they just smeared that on the rest of the components. Only reason I used Mouser was the selection of 1000v silver-dipped-mica capacitors, they seem to carry the best stock of these very expensive components.

So, now I have the following:

Amidon toroids + some nice high quality wire
PIC + LCD + some thermistors.
Filter capacitors
PSU filter/ripple capacitors
power transistors for the regulator
EB104 “kit” – now built – will blog specially about that
Copper spreader – Old friend works at workshop that produces switches and other high power stuff for the local power companies, they have ton’s of VERY nice copper stuff, was very happy to remeber this old friend.

here is a gallery link:

[View with PicLens]

As you can see I have started construction of the chassis. As I don’t have any good “sheet” bending machines I opted to build the chassis from 10mmX10mm aluminum square axles.. don’t have the English word here I tried using some aluminum “brazing” tech called durafix. My impression was this was like these products on TV market, look good but are really shitty…. and unfortunately the latter was true, so I ended up going to a shop with a TIG AC machine to point-weld this.

Now my brother, who happens to be the secretary of the Icelandic Radio Amateurs has asked me to give a short lecture on this amplifier business next Thursday, so that forces me to put something down on paper, and design this properly. So some Viso/power point drawings will be made. I was just thinking how good It would have been to have a proper block diagram with good checklists.

I started work on the PIC control circuitry, hooked up the LCD and testing out some A/D -> LCD. I asked my wife if I could select my xmas present, and she said YES… so I ordered an ICD3 PIC programmer. I have been wanting to build or buy a ICSP programmer so I can fully apply my trial and error method of programming

600W process report

tf3cy — Tue, 03 Nov 2009 22:46:38 +0000

This is based on the Motorola EB104 bulletin:

EB104 amplifier board ordered from Communications Concepts, including 4x MRF150
Omron 12V Relays ordered from Ebay – used for TR switching and filter selection
Toroids and accessories orderd from Amidon
Large heatsink gotten from my father, TF3T
1200VA toroid transformer was bought in my trip to Shanghai – loong story there!
Going to order some silver-dipped MICA capacitors and PIC MCU + filter capacitors from www.mouser.com

Still need:

Copper spreader
12v cooling fans
Hardware
Heatsink for the PSU – asking locally

Next thing is to build the case

Going to build it from Aluminium parts available locally, have made some sketches in AutoCad, but it’s hard to draw it exactly as I don’t have all the components in hand, and don’t know the exact dimensions of some of the components.

Project start – 600W 1.8 – 30 Mhz Linear Amplifier

tf3cy — Tue, 03 Nov 2009 22:45:43 +0000

I have been planning to build a 600w linear amplifier to use on the HF bands.

Why homebuilt??

I have been asking my that question many times. If I have to order a commercial built amplifier I have to give out at least $1500-2000 USD. That will put a dent in the home budget and my XYL will notice big spending’s like that. Homebuilt projects usually consists of many many small orders!

But I really like to undertake a project like this, there are many aspects to this project. Mechanical design, DC power supply design, RF deck, Filters and control circuitry . I have been playing around with programming PIC MCU’s and have decided to use one of those buggers for control and protection circuit.

Why use Transistors. Why not go for the ever so simple, reliable, tolerant Vaacum tube design?

I don’t like working with HV circuits. Don’t understand the theory behind the tube amplifiers. And the challenge is mostly mechanical !

Availability of parts here in Iceland is very poor, so I have to order most components online.

Design principles:

Single chassis for now – can be modified later on if I decide to double the power output
600W Output.
Manual and remote bandswitching – using the BCD interface from my Kenwood / Yaesu
Protection: Over voltage, Over current, Over drive, High reflected power. Over Temp on PSU and RF stage. ALC back to Exciter.
LCD display – showing vital amplifier stats. (just me playing around with the PIC)

Project timeframe is from now to the finished project. I really hope to test the amplifier around Xmas 2009… now the component collection phase starts.