Radioberry 2.0

[PA3GSB]

Hi all,

I like to inform you all that i have the radioberry 2.0 (beta 2) running. 

The RB 2.0 contains an intell 10CL025YE144C8G FPGA. The PCB is also capable to support the cheaper 10CL016 version (not tested yet).

The PCB is a 4 layer board, designed using KICAD. (learned a lot from the HL 2.0 design). The RB can be clicked at top of a raspberry pi.

The firmware will be loaded into the FPGA via the raspberry pi.

The clock is an TCXO running at 38.4 and iam using the multiplier in the AD9866. Running at 76.8 MHz.

For RX iam using the same anti alias filter as used in the HL2.0

(thanks Steve; the cap did the trick)

Supply voltage is provided via the raspberry pi.

Having an  IO connector for i2c  and regular io, to control a bandfilter.

I have to clean folders before sharing it at github.

The assemblage was done by Makerfabs, the only component to assemble is the FPGA.

Will inform you when the github repo is updated.

73 Johan PA3GSB

[Alan Hopper]

Hi Johan,

congratulations, that is looking really neat. Do let me know if anything extra Is needed in SparkSDR to support it.

73 Alan M0NNB

[John Williams]

Johan,

Do you perhaps have a pdf of your latest schematic for the project? Does it have a open collector capability for band selection and PTT? If so, my extra boards for the v1.0 or v1.1 pre-V2 amp may be of use. How much power output for TX? The 1.0 board takes 10mw being a 2 stage amp. The 1.1 board only has the final and takes 10db more drive.

Attached are some data provided by Stew N8VET on the amp.

John - W9JSW

--
You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[PA3GSB]

Hi John

Hereby the PDF.

Can you provide me with a link to your amp board?

73 Johan PA3GSB

Op zondag 11 februari 2018 16:49:51 UTC+1 schreef John Williams:

[John Williams]

https://github.com/W9JSW/Hermes-Lite/tree/Amplifiers/frontend/v2-Amp

[Steve Haynal]

Hi Johan,

Congratulations on the next Radioberry. It is good to hear the Cyclone 10 is working for you. The Cyclone 10 is one of the best values on the market for this logic size in my opinion. For the same amount of logic and capabilities, the Cyclone 10 is $10 less ($25 vs $35) and lower power than the older Cylcone IV we are using on the HL2. I would like to move to the Cyclone 10 eventually, but there are 3 or 4 pin differences that make it nontrivial.  

It is also interesting to hear that you are using the frequency doubling on the AD9866. 

From your schematic, your are using the IAMP output of the AD9866 for TX. How are you dealing with the nonharmonic spurs that occur in the 12M and 10M range when using the IAMP output? We switched to the TxDAC output and clean op amp preamp to deal with this in the HL2. Do you have any RX or TX test results posted?

Did Makerfabs source your AD9866 within China? What is the date code on it? (Second line, right under AD9866 on package.) What temperature does it run at? What did Makerfabs charge for the AD9866 and for the entire assembly?

73,

Steve

KF7O

[PA3GSB]

Hey Steve,

Tnx. Iam very glad with the result.

I only have looked for a bug for almost 2 weeks; by deducting i found the error ; Quartus prime did an automatic update of the FIFOs and changed the data width from 48 tot 8..... 

As said the PCB is adapted for cyclone 10CL025 ($17 exVAT) and 10CL016 ($12 exVAT).

The temperature measured via Infra Red; AD9866 46 C and the Cyclone 55 C.

Makerfabs did also the procurement for batch of 10 pieces:  $30.46

The AD9866 are marked #1129; so i think produced in 2011 week 29. 

The assembly was $12.36 per piece. Stencil and start $28 .

I found the communication very good! and they did a very good job! Including the packaging.

As said the FPGA could not procured in china so i decided to order the cyclone 10 myself including the soldering.

The TX setup is the HL1.0; i wanted a cheap solution and i can live for now  with  some spurs in 12M and 10M .... 

I did a quick look at the RX performance... it looks the same as you posted on the forum.

Will look into it in more detail later.

Hope i did answer all your questions.

73 Johan PA3GSB

Op zondag 11 februari 2018 19:11:11 UTC+1 schreef Steve Haynal:

[Steve Haynal]

Hi Johann,

Thanks for the answers. Did all 10 units work? We had some AD9866 quality issues with the last run. Where are you buying the 10CL025 for $17?

Regarding the spurs, we found these serious enough that we couldn't legally transmit on 12 or 10 M with an amp. You may be able to reduce the spurs even when using the IAMP. Joe and Claudio found that attaching a trap to TxDAC helped. See Claudio's blog. I think lowering the power output by increasing Rset should also help. I did try this with no success as discussed in this thread but may have not lowered the power enough. There are also the IAMP bias control registers at 0x10, 0x11 and 0x12 which may be tweaked to help.

73,

Steve

KF7O

[PA3GSB]

Hi Steve,

I ordered the 10CL025 @mouser.

The result is now 2 out of 10. For the other 8 i have to assemble the FPGA.

73 Johan

Op zondag 11 februari 2018 23:33:08 UTC+1 schreef Steve Haynal: