JUMA PA1000 Linear Amplifier

General discussion

Re: JUMA PA1000 Linear Amplifier

Postby 4z4rb » 01 Apr 2012 06:59

Hello David and all
I have been making these modules for 8 month now and selling them either through eBay or on
direct trade basis.
I also make a lower power versions2x450W devices and 2x600W devices.
I also implanted feedback on them and got good results and better match. also good IMD results
The use of TLT had enabled me to get better results and superior band width as opposed to the regular tube and sleeve transformer.
as for marketing my boards please contact me direct with: 4z4rb@bezeqint.net
mail
I am also on 20 in the evenings around 14180 usb.
A9:1 transformer is a bit trickto understand but is very easy to make and use
a semi rigid or a regular 25 ohm coax can be used depending on power ratings
On this paticular board I used the semi rigid because Mr Duda wanted it but no big difference at 1kw level with regular Raychem coax was observed.
These devices represent a great step in RF linear rf power amp design.
best regards and 73's
Baruch 4z4rb
4z4rb
 
Posts: 10
Joined: 18 Nov 2011 12:25
Location: Israel

Re: JUMA PA1000 Linear Amplifier

Postby LB3HC » 05 Apr 2012 14:09

Could someone explain what kind of transformer configuration this is and how it works? --> http://www.nikkemedia.fi/juma-pa1000/ju ... view-2.jpg

I Have read the interesting information here --> http://ludens.cl/Electron/mosfetamps/amps.html and I think what is done in the Juma 1000 amp is that each transistor (two in a MRFE6VP61K25H package) pulls current thru its own separate transformer's primary winding. By doing with separate transformers there will be no flux linkages between the transformers. Therefore when one transistor conducts, the other will not see the voltage appear over its drain to source channel due to flux linkage common to both windings (the other transistor does not conduct other than the set bias idling current). It seems to me that each transistor conducts its own 180 degree portion of the waveform. To be able to swing both positive and negative on the output of the amp to reconstruct the input waveform, it is necessary to have a phase reversal arrangement in the output transformer. I think this is what is done in the Juma 1000 amp output transformer arrangement. In addition to the phase reversal, it is necessary with a impedance transformation to transfer the energy to the filters and then to the output load I expect. For example a 1:9 ratio into the filter. I see no balun on the output of the amp before the filter. so the output looks like it is unbalanced. I assume the filter is unbalanced to reduce the number of components necessary.

I have done a LTspice simulation to try to verify what this guy writes on his page http://ludens.cl/Electron/mosfetamps/amps.html and my simulations seems to match his findings. I can post the schematic and LTspice files if someone is interested.

Please explain how this output transformer arrangement is done, as it is quite interesting. Even better, is it possible to post a schematic of the output section?

73
LB3HC
LB3HC
 
Posts: 2
Joined: 18 Mar 2012 11:41
Location: Norway

Re: JUMA PA1000 Linear Amplifier

Postby G4ZOW » 28 Apr 2012 07:31

Matti,

Are there any further updates on this project?

David G4ZOW
G4ZOW
 
Posts: 3
Joined: 20 Dec 2011 19:43
Location: England

Re: JUMA PA1000 Linear Amplifier

Postby SP5G » 07 May 2012 15:59

Hi,
I would say that the amp works, but construction has been completely changed in terms of amplifier developed by Baruch, are listed transformers input and proposition because it is warming and has not worked properly on the 50MHz band, bad choice of materials and cause terrible harmonic distortion of the amplifier.

I'm not saying that the construction carried out by Baruch is bad, but required a thorough revision of construction and use of appropriate materials, besides the amplifier generates large amounts of unwanted signals, resulting in older should load components on the power and excessive heat.
used a 14dB attenuator that allows you to develop the full power of the control of up to 50W.
With the right load transistor output power> 2KW in the 1.8 to 50MHz for BLF578XR also tested (mrf6vp11kh, mrfe6vp61k25h, blf578), used a very stiff 50V PSU that even at the peak will give 120A. linearity of the amplifier at an acceptable level.
Measurements performed on artificial 50ohm load,
laboratory using a moderately power 10kw, by direct measurement of voltage at the fundamental frequency of all harmonic filtering for LPF. (3rd harmonic suppression of-60dB relative to the fundamental frequency).
which was also properly connected Tektronix Spectrum Analyzer for 1.8GHz.
ofc amp working in class AB and C. I used a AVR Mega2560 to control all events and save all measurments on SD card in development version.

At present working with the power of 1.8 kW during the pileup about 4 hours at an ambient temperature gave 26 melted inside the amplifier heat sink 29 degree and 30 degrees. supply temperature of 28 degrees Celsius.

Weight with power amplifier and filters to 20kg.

today tested the amplifier based on four transistors BLF578.

More info on sp5g.net, soon puts the results of measurements of the instrument and the linearity plots and shots of a spectrum analyzer.

sorry for the translation :), good luck.

atb, Robert SP5G
SP5G
 
Posts: 2
Joined: 07 May 2012 14:00
Location: Poland

Re: JUMA PA1000 Linear Amplifier

Postby 4z4rb » 12 May 2012 05:09

I am happy to hear that you got the beast tamed.
I have made many improvements to the board since the time you got your board
BTW your board was the third i ever made so there might have been issues with it.
indeed since then the output transformers had been completely reworked and redesigned using better materials that were not around when your board was made a year or more ago.
I wonder if the ALC circuit that you planned on building was ever built?
I see these advancements as a great step in SSPA design with mutual work of hams all over the globe.
SOUNDS GREAT!!!
4z4rb
 
Posts: 10
Joined: 18 Nov 2011 12:25
Location: Israel

Re: JUMA PA1000 Linear Amplifier

Postby SP5G » 13 May 2012 17:32

Hi,
It's the simple board to control ALC, overload the amplifier is back to trx and drive power will be reduced. In my amp I also use some new features ex. to correct a linearity.

Right now i'm testing a 4xldmos amplifier on my developing board :).

atb, Robert SP5G
SP5G
 
Posts: 2
Joined: 07 May 2012 14:00
Location: Poland

Re: JUMA PA1000 Linear Amplifier

Postby 4z4rb » 13 May 2012 19:47

Well that will be a real Monster
you go dual parallel push pull or dual with combiner
the 2nd option might suit better due to extremely low load lines
you will need a 5 -6 ohms coax and it is very hard to work with the inner conductor is almost as thick as the outside .
very hard to bend or fold.
A 12 ohms coax line might do the work in parallel
good approach on the ALC board
I have sent you in a PM the mods that I did too the board
Good work and good luck!!!!
4z4rb
 
Posts: 10
Joined: 18 Nov 2011 12:25
Location: Israel

Re: JUMA PA1000 Linear Amplifier

Postby wa1gfz » 12 Jun 2012 00:14

Hi All,
I finally got back on the Juma site after losing my password. Matti, glad you found some nice shielded wire. I like all the interesting transformer reading. I have 1 comment on the shunt feed DC. Remember both windings are not balanced flux in class AB so there will be lots of DC offset at peak power. Ever see how hot the core gets in the EB104? This shunt feed core must support peak current in one winding while the other is at zero current. I've increased the size of my shunt feed cores in my 300 watt MRI final. New project here is I picked up some 1200 watt MRI amps on ebay running 4 SD2932s. I have it working great 160 through 6 meters. The output transformers were wound on a single toroid 2 X 3 turns 25 ohm shielded wire (#24) with dc on the CT. I just stacked 4 cores to move it down to 160M. There is a balun after the transformer and again I stacked 4 cores (wound with RG188). (.87 inch OD type 61). The output transformer after the combiner 1.25 inch type 61 again 4 cores stacked. I wound this with home made coax. RG400 with the center removed then #14 teflon wire pulled into the center. (close to 25 ohms) No temperature rise on any transformers running 200 watts carrier output AM and well over 100% positive peaks. I'm working on a bigger heat exchanger to be able to run higher bias. MRI heat sinks not suited for ham use even with lots of air.
I would be interested in the Juma amp transformer design changes. I think you guys need to run 1:3 turns ratio so harder to do. Frank WA1GFZ
wa1gfz
 
Posts: 2
Joined: 29 Nov 2011 01:33
Location: USA

Re: JUMA PA1000 Linear Amplifier

Postby 4z4rb » 19 Jun 2012 15:09

Hi to All
I got lucky and bought a spool of 25 ohm silver plated RAYCHEM top quality space aviation approved cable.
it is 18wag inner conductor and cable dia. is about 4.5 mm appro. it is the superb ETFE (tefzel) inner part and the outside is white cross radiation polymer ( the stuff that replaces teflon) all around a low loss high power superb rf wire!!
i have 25' pieces that I can sell
if interested contact off forum to 4z4rb1@gmail.com
thanks
4z4rb
 
Posts: 10
Joined: 18 Nov 2011 12:25
Location: Israel

Re: JUMA PA1000 Linear Amplifier

Postby PA4MIC » 09 May 2013 13:04

I'm wondering how this project is going? Is it still running or is it stopped? I hope not.

Best regards,

Michael
PA4MIC
 
Posts: 1
Joined: 06 May 2013 21:32
Location: Netherlands

PreviousNext

Return to Juma Discussion

Who is online

Users browsing this forum: Yahoo [Bot] and 1 guest