RF Explorer

This one has somewhat less bizarre menus

and seems to work pretty well. It's making a 6 GHz output at 400 mV p-p, not bad for $29.90.

No, it is set to 6G but only works to about 4.5.

Thanks. Good tip.

Is 6GHz a typo?

--
Science teaches us to trust. - sw

I have one those or very similar. Mine has a strange feature. If I increase frequency by 1MHz, there is no output at 60,70,80 or so MHz.

71,72,73 and so on work well. I have not tried every tenht Megahertz, but I tried those.

You can set it to 6, but it only really works to about 4.5.

Another interesting operator interface issue.

I don't know what synth chip it uses, but the one we use (TI LMX2571) is very complex to program. You have to do all sorts of numerical factoring and make a ton of decisions. I can see how that code might be buggy.

TI obviously has the code to drive the chip - it's part of their demo package - but wouldn't let us see it, so we had to do all the math ourselves.

Did they give a sufficient justification for not releasing the code?

From playing with the LMX2594EVM, there was a hell of a lot of special cases with the auxiliary control registers.

That wouldn't be too much of a problem if you just want preset the 2594 to a small number of discrete frequencies - just copy whatever magic the GUI shows.

But if you want to, say, sweep across a wide range, it would be a pain.

Page did not show any price; used "see all buying options" and found NEW $49.50 . So where did you get that $29.0 price?

No, they just refused. That makes no sense; making the c code available would help sell chips.

You have to select the right VCO, and then find the numbers for something like

Fvco * (N / M) = ref

then

Out = Fvco / K

and deal with a lot of other settings.

With mostly 32-bit integers. Something like that or maybe worse. The data sheet is 62 pages long and the user guide is another 32. It has about 60 registers inside.

We wanted any frequency, but we could add another divisor in our FPGA.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

I just searched for "rf signal generator" and there it was.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

You can truncate amazon urls after the "/ref" to make them much shorter:

--
Science teaches us to trust. - sw

Just so, but it is those "other settings" that are a pain, especially deducing the boundary conditions where you flip between settings.

Nothing insurmountable, but needs a lot of duplicated thought coding and testing. The /unnecessary/ duplication offends me.

One plausible excuse for not releasing the code is that it is crap, possibly wrong, and they don't want to answer questions like "why X, wouldn't Y be better?". Not inspiring, and possibly not showing their hardware off to good advantage.

I'm just through this for the ADF5356. Only 13 32 bit registers, but still 500 lines of C code when I remove the passages of the data sheet with the register descriptions.

It was a lot of ado to compare the values from the data sheet with those from the eval board software and work out the differences. But now I have at least a function set_adf5356(double f); that works over the range I need (no divided outputs).

There is still some doubt WRT programming the phase detector bleed current. I have opened a case in the AD EngineerZone, 76 views upto now, but no answer, let alone a answer from AD.

< >

The bleed current is for optimizing phase noise and spurious; it's hard to check the effects of it at all, let alone program a solution that fits to all other inputs.

cheers, Gerhard

Am 24.10.20 um 11:51 schrieb Gerhard Hoffmann:

... and how can you call a fractional value N in the data sheet, and the non-fractional part of it INT, which is a predefined type in most languages used to implement that.

:-(

I have a couple of demoboards for Analog devices version and bugs are probable, it is not a simple chip. Maxim has one similar if not same chip, too. Is its code secret too.

Hi, RF Explorer Signal Generator is not limited to 1GHz step increase. You can go to FREQUENCY MENU in the device using the MENU key (click it several times) then use the DOWN key to select "Freq Step", click ENTER and you can edit any step you want. You can also specify any specific frequency in the "CW Freq" field. More details in

In addition to that you can use the simple but powerful Windows/Linux/Mac tool which allows to select frequency with a slider or by precise value:

Regards arocholl