Listen to Latest SolderSmoke Podcast

Sunday, June 9, 2013

OOPS! Why the AD9850 DDS Boards are So Inexpensive

N3ZI has a very plausible explanation for the low price (about 9 bucks!):

My guess is that this is a liquidation due to a design build error, since they are being sold for a price of about 1/2 the price of the DDS chip alone. The modules are assembled and tested. The design error I noticed is that the wrong output filter is used. These boards use the 9850 running at 125MHz. A 125MHz DDS should have a 50MHz LPF, but it seems that these modules have the 75MHz LPF the chip maker recommends for the AD9851 running at 180MHz. My guess is, someone just copied the wrong filter from the wrong data sheet, and it wasn't caught until they went into production.
But for amateur radio applications they work fine up to about 40MHz. You can push them to 50MHz by adding a correcting filter, which is included in my controller PCB, but the output level is low in the 40-50MHz range.

N3ZI continues to offer some really interesting microcontroller products.  He has a controller board that allows you to simultaneously control TWO AD9850 boards.  This might be exactly what we need when we have a separate receiver and transmitter with different intermediate frequencies: Set up one board with the VFO freq for the receiver and the other for the transmitter with the resulting operating freq displayed on the LCD.  Viola! No more "Spot" or "net" and zero-beat by ear! (But I may be one of the last people on the planet still doing this!)   

N3ZI's site:

Our book: "SolderSmoke -- Global Adventures in Wireless Electronics" Our coffee mugs, T-Shirts, bumper stickers: Our Book Store:


  1. I don't find this explanation plausible. There are too many cheap Chinese AD9850 board variants out there for this to be the reason. More likely they are chips which didn't meet AD production spec's at inspection, and sold for peanuts for Chinese home market. Also possible some may not be genuine AD devices.
    Chris M6CRD.

  2. I have two of these AD9850 boards that look identical to the board in your picture. After bringing up the boards with a Parallax Propeller micro-controller, one of the first things I did was test the output filter.

    The filter is is correct. It rolls off at the right frequency for the AD9850 with a 125MHz system clock (42MHz) and seems to follow what I expect is the response of a 7th-order Cauer (elliptical) low-pass filter. (The schematic provided by the ebay seller is not correct and they refuse to provide the correct schematic.)

    I believe the boards were made to accommodate EITHER the AD9850 or the AD9851.

    The AD9851 data sheet reference design with a system clock of 180MHz calls for a 70MHz 200 Ohm 7th-order elliptical low-pass filter. The AD9850 data sheet reference design calls for a 5th-order 200 Ohm elliptical LPF.

    The board can accommodate either chip, just change the value of the filter components. And that's what the manufacturer did. In my case at least.

    That's not to say there may be some AD9850 boards out there with the (incorrect) 70MHz filter. But the two boards I just received have the right filter. As with anything from China, YMMV.

    I would like to hear from more people who have tested the filter on their AD9850 ebay boards. Let's see if this is someone spreading fear and uncertainty - or not.

  3. M6CRD is just raising some good engineering caution. All those possibilities are real and happen every day in the industrial world. Boards this cheap were probably discarded for some reason.

    As for taking someone else's castoff and using it to build a radio, well that's just homebrewing, right?

    I just got two Chinese AD9850s and look forward to hitching them together into a VFO with perfect 90-degree I/Q outputs, for use with my KK7B receiver and whatever else.

  4. I meant to include N3ZI with M6CRD in my previous comment. They're possible and one or more are likely I'd say. So just keep your eyes open like any good SolderSmoker, that's the fun of it. 73 de KC7IT

  5. The lower 8 bits of the 32 bits of the freq control word are screwed up on my boards. The lowest 4 bits seem to be in reverse order. Both boards that I have behave the exact same way so my conclusion is that it's a design error in the silicon.

    Mr C.

  6. Please ignore my previous post. Something else is going on with my setup. I can't find anything definitively wrong at this point.


  7. Unfortunately N3ZI is sold out of the AD9850 motherboard kits as of 8/12/2013 as well as the blank boards and says he has no plans to produce any more. I am not aware of a comparable board at the same general price but if someone is, maybe they could chime in.
    Geoff WB6NVH


Designer: Douglas Bowman | Dimodifikasi oleh Abdul Munir Original Posting Rounders 3 Column