3.4 GHz Transverter Conversion (3.5 GHz WiFi transceiver)

[ZL1TPH]

The PA's that many over here use:
http://www.ebay.com.au/itm/Stealth-Micr ... 35befc0a13
Inside here

Inside 3.4 GHz PA

I have two units: One puts out 20 watts and the other 28 watts

[VK3QI]

The Toshiba amps that I had for sale on Sunday at the ENDRC hamfest sold out within minutes - $200 each.

The source of Toshibas seems to have dried up (these are factory sealed with test results), in the last 12 months.

The last one to sell on Ebay, item 191787974276, was only an A model and went for US$250 (A$350) plus postage of about A$50.

As Stephen ZL1TPH says, your other alternative is the Stealth Microwave units from Israel, Stealth Microwave SM3437-43L 3400-3700 MHz 20 Watt Linear Power Amplifier, which go for about A$200 including postage, but you have to be careful that they are: (1) not used or damaged (2) that they have the correct options on them.

See: https://docs.google.com/file/d/0BzzdlPY ... ef=2&pli=1

And http://www.alldatasheet.com/datasheet-p ... 37-43.html

There is also a 10 watt version SM3436 that sometimes gets snuck in, pretending to be the larger unit, by unscrupulous sellers.

A number of VK hams are using theses Stealth Microwave amps with success - there have been a number of articles on how to interface them which you can google for.

Cheers

Peter VK3QI

[VK5HP]

Hi all,

I noticed there is a smd PCB around .

Are the PCB files been shared at all ?

Cheers Paul VK5HP

[VK2ZIW]

Will PLL2 go just a bit lower?
=========== TRI3500_transverter_0_2.ino ===============
//ONLY CHANGE THE FOLLOWING TO CHANGE THE TWO PLL FREQUENCIES
#define PLL1_FREQ 1477000 //Khz. VCO range is 1298 Mhz - 1487 Mhz (originally 1478 Mhz.)
#define PLL2_FREQ 0 //428000 //Khz. VCO range is 428 Mhz - 455 Mhz (originally 448 Mhz.)
//Alan #define PLL1_FREQ 1485000 //Khz. 3398.1 => 428.1
//Alan #define PLL1_FREQ 1485000 //Khz. 3398.1 => 428.1
//Alan #define PLL1_FREQ 1482000 //Khz. 3398.1 => 434.1 for Microwave Modules 70cm=>6m & FT-620
//Alan #define PLL2_FREQ 0 //425000 //Khz. x8 => 3400 MHz for Rx tuneup

//Possible tune up source
// PLL2 ref will not work below 200Khz. need to re-design/change loop filter
// VCO needs to be pushed up 30 Mhz. 7th harmonic of 485.6 => 3399.2 MHz or 485.4 => 3397.8

//DO NOT CHANGE ANYTHING BELOW THIS LINE
========== Alan VK2ZIW ===========

[VK5ZT]

Will PLL2 go just a bit lower??

In a word, no....at least not on a standard board. (These things are easy to check on my development system )

It locked at 428 MHz but appeared a little unstable...any lower and it was out of lock.

The LMX2306 can certainly handle more range so the pll on the board needs more investigation to go further down.

As Lee has suggested, there is always PLL1

Cheers

Tim VK5ZT

[VK2JDS]

Omron are no longer making the g6y 2.4Ghz relays we have been using for the panel antenna switching (RF output side).
Mark at minikits does have some stock of them and the associated pcb though.
A replacement appears to be the TE relay from RS components catalog number 616-8758 at $10.80 plus tax, freight free for online shopping.
Its rated to 3Ghz and looks similar to the omron, and its 12 volts. I ordered a couple to experiment with.
73 Dave vk2jds

[VK3TU]

I used the TE HF3 relays in my 2m and 70cm masthead pre-amps. The specifications are very similar and they work fine. They're also capable of switching a bit more power than the G6 relays as well. The footprint is different and I couldn't find a library file for Eagle for them so I ended up making my own.


Bert, VK3TU

[VK3MAT]

re: the HF3 relays

My Brother is doing an arts degree so i made use of him and he drew up the pcb footprint in a dxf file for me

(it wouldn't let me attach a .dxf, so if anyone is after it send me a message)

Matt.

[ADMIN]

(it wouldn't let me attach a .dxf, so if anyone is after it send me a message)

Matt, you could rename the file to a 'txt' extension and attach that.
(Don't forget to mention that downloaders need to change it back to a .dxf entension once downloaded!)

[VK3TU]

For those familiar with Eagle, here's the relay library with the HF3 included. Just copy it into the Eagle library folder and change the file extension from ".txt" to ".lbr" and remember to update the library (Click LIBRARY>UPDATE ALL) on the schematic page. The new library file "relay-edit" should be there and when you expand it, it should have the HF3 listed in there.

Bert
VK3TU

[VK1JA]

Has anyone experienced a panel with very low output?

We ran the first one I've been converting on the bench yesterday and while the rx is very good, the tx is only outputting 10dBm (10mW). A load has always been on the output when powered up, so maybe it's just a dead PA?

I have noticed that this particular panel has been opened before, and may be a Monday or Friday build, as the stub coax and one screw was missing from the old tx filter.

Never the less, I still have two other panels here ready to convert

One other question regarding the tune up - now that we're using the old rx filter as a tx filter, is there a need to 'tune' the now tx filter to peak tx output - similar to how the rx testing talks about adjusting the screws in the filter to peak rx? There's just no mention of adjusting the screws in the now tx filter to peak output, just that 'screw in all screws in both filters, except for those second from each end. Screw out the second screws from each end almost all the was out'.

If the now tx filter screws should be adjusted for peak tx output, it may explain the low tx output.....

Thanks

Jayson
VK1JA

[VK3ALB]

Hi Jayson,

Yep, my own panel is low in output - not as low as yours but it is less than the regular 1W that most people get. I expect that one of the PA devices is dead in my panel.

From what I have experienced and been told the PA failures are usually catastrophic with the HMC327 devices going short pulling the 5 volts to ground. You can measure this at the regulator next to the PA. Each device is fed by it's own inductor and in the case of a short you can pull these to find which chip is dead. That you can get any RF out probably means the devices have not shorted. You didn't say how much drive was required to get +10dBm. At the input to the attenuator in what was the 480MHz filter block you normally need around -4dBm for +30dBm output. In my case I need about +2dBm for a little over +26dBm output.

Regarding the tunable filters, they are now both used in the RX line only. The PA output goes straight to the output relay so adjusting the filters will have no effect on output power. If you check the label on the back cover you'll see the TX/RX split is 50MHz so it is important to adjust both filters for maximum RX performance.

[VK1JA]

Hi Lou,

Thanks for the quick reply.

Yep, my bad about the filters, I see now that both are in the rx line! The good thing is that means there may be a bit more to come in the rx

The 10dBm output was with -4dBm of drive, as outlined in the tx testing. To tell you the truth we didn't even try increasing the drive. What would be a safe max input drive to test at?

I may have mentioned before that the current draw on tx with no input (first stage of testing) was 580mA, rather than the ~800 suggested.

But all is fine if this one is down in output, I've still got 2 more panels to play with. I'm just very greatful that you (and your team) have made these panels available to us, let alone all the research and testing you've done to get them converted and fully documented.

3.4Ghz has gone from a 'no mans land' to one of the most active uW bands in the matter of 12 months!!!

[VK4CZ]

Started the mod process on my panel this weekend.... still a bit to go. But now considering the distances and terrain to be traversed in SE VK4 have started planning whether to just return it to the panel or further develop it with sequencer, PA, etc



Sent from my SM-G900I using Tapatalk

[VK1JA]

Hi All,

Are the switcher blank boards (http://vk3atl.org/3.4GHz%20switcher%20V1.1.pdf) available for purchase, or is there anyone making up a batch of them - if so, put me down for three of them!

EDIT - I just noticed that they are listed as 'out of stock' on the club website, so they do sell them, they are just out of stock I'll keep an eye on the web site and grab some when they have more stock.

[VK5HP]

Are the PCB files advaible for down load ? I would like to cut some for my self

[VK3ALB]

I hope to have more boards available in April.

[VK3ALB]

Hi All,

Today Chris VK3ACG and I had a bit of a play with a panel with a dead P.A. Chris had already removed the HMC327s and the first splitter block as well as putting a short coax tail at the output of the driver. We found the driver stages were quite linear and with this particular board we were able to get +13dBm for 0dBm input at the 480MHz filter block. This should be more than enough to drive any outboard PA. Of note was that the board draws 300 to 400mA and the TX enable line serves no purpose in this configuration as it only controls the supply to th PA chips.

This picture shows the test configuration we used. Chris plans to use this board as the heart of another 3.4GHz transverter. If you think about it, add a preamp, and external P.A. with suitable relays and perhaps a 10MHz OCXO and you have a very capable transverter. So those of you that may have a dead PA, take heart, there is still a way forward.

3.4GHz test setup

[VK2JDS]

If you burn a driver stage then they can be replaced by a GALI-39 by changing the dc supply resistor to 75 ohms. I have replaced 2 so far quite successfully.
A solution to the blown PA problem is that usually only one device fails, so carefully lift each of the inductors feeding the chips and see which chip is shorted.
Just run the panel with the inductor for the shorted device removed and it will happily put out a quarter of a watt.
Otherwise use a 2 watt amp from minikits as there is plenty of drive available as explained in the previous post from Lou.
http://www.minikits.com.au/electronic-k ... -Amplifier

73 Dave

[VK3ALB]

Thanks Dave,

You've reminded me that from the many panels we have seen that not all panels put out the same amount of power. We've concluded from this that not all the ceramic filters have the same characteristics and some have steeper skirts than others. Consider that the filters in the 2nd mixer and TX Driver are cut for 3429MHz and you can see that it's the luck of the draw as to how much TX power you ultimately get from your panel. The average is somewhat more than half a watt but I have seen some put out a little over 1W. For those that want to experiment, bring the IF input frequency up 20MHz or so to see the result.

TX close-up showing supply inductors