upgrading dave's AD9835 Sine/Square wave generator design for AD9850

[RoadRunner]

i am currently working on a project, AD9850 based DDS 40MHz Sine/Square wave generator inspired from dave's
http://alternatezone.com/electronics/dds.htm   

which is AD9835 based 10Mhz Sine/Square wave generator.

i picked up the AD9850 module from ebay ,which works quite well
i have done with the digital part of it firmware is almost done. i have used a tiny graphical lcd which display current frequency and few other stuff like amplitude , offset value and square wave duty. and i have used a rotary encoder and few switch for the interface.

now the problem i am facing is which opamp will be suitable for upgrading the analog section of dave's design to work with AD9850.i am an software and digital guy.i just want fixed gain of x3 or x4 and offset adjustment. i can adjust the output of AD9850 module by the R-Set pin using a digital pot. the output of AD9850 module goes upto 2Vpp for frequency of around 25Mhz after that it starts dropping (its more than good enough for my purposes).

please help me by suggest which circuit and opamp will be good for this job
 
i have attached few image of current prototype and the output of the AD9850 module wave form output.

[brabus]

You know what? I am approaching the exact same issue!  I bought an ebay module very similar to yours, just to build up and test the firmware stuff, but ASAP I will go for a proper PCB in the final stage. Absolutely need BNC connectors!

As an output, I want to go for a double output stage:

1) low impedance output for audio stuff: I will stick with a simple opamp, maybe something very hig-end from AD or TI, still to find the P/N.

2) 50 ohm output for RF. Need to use a very high bandwidth opamp, like e.g. AD8000 (Ghz range UGBW, 4 dollars x1pc on DGK), and design an adequate PCB for that. No ceramic stuff, just proper decoupling and impedance control.

The audio part will be switched off automatically after 100 kHz, details TBD.

Gain switching is somehow easy, just a bunch of switched resistors - maybe some PGAs exist as well, but I do not have any P/N yet.

For the second part, my only doubt is regarding AD9850 amplitude performance over the whole range. Can we investigate a little deeper on why the amplitude is drooping afret 25 MHz? I am very upset for the layout of your board, maybe that is the root cause of the pole we are seeing.

I'd like to propose you an affair: you can fix the firmware with controls, display, smooth operation & co., and I will design a proper PCB and prototype! It could fit well in a Hammond box, maybe someone else is interested...  Where are you from, precisely?

[RoadRunner]

i am build this project for testing simple opamp circuits and providing clock to digital circuits like mcu etc.

the pcb you see in the image is solely digital section of the project these is no analog output trace of the module on this pcb i will mount a saprate daughter board for the analog stuff , same as you i will use proper bnc connectors. i can design pcb but because this is currently a personal project so i did't go for it.
i am directly probing at the output of the module there is nothing to do with pcb.

i have tested few pga like MCP6S91 and TI's PGA112 but thy can't do anything beyond 2 Mhz.

i can do anything in the firmware, i am in Dehradun ,India.

[kripton2035]

found that on the net some times ago. hope this helps.