LTDZ devices: New firmware available!

Hello,

This post is dedicated to Domenico, who has been working hard on improving the performance of the LTDZ device, both in terms of firmware and hardware.

His efforts include the addition of new functionality to the firmware and modifications to the LTDZ hardware, to improve its performance by correcting design flaws.

The research started due to an odd fenomenon of his particular LTDZ, which causes some strange signal drops:

This is not a problem of my software, as the same happens in WinNWT, as well:

Recently, more and more LTDZ owners are in fact complaining about this problem!

Domenico took the D6 FW SDK (the D6 FW can be used on the LTDZ device, as it seems) and basically fixed this issue and added new functionality.

One of the new FW functionality is the possibility to select 4 different attenuation values for the RF output!

Another cool feature is an automatic noise level adjustment upon startup or through a specific command.

Domenico says:

Chinese firmware fools you with deliberately lowered noise level, to cover up the design errors in the hardware. That's why I put the Noise Floor Calibration command.

The latest release of my VMA Simple Spectrum Analyzer software supports these new features.

However, I am currently left with little time for my hobbies, so the implementation is very crude right now. Still, I did not want to withold the publication of this great work by Domenico.

You can download this beta release HERE.

You will need to flash the firmware included in the package to your LTDZ device. This requires a programmer. I have successfully used Segger J-Link v8 and a ST-Link (both are cheap chinese clones).

I will try to improve the software and publish more detailed instructions on how to flash the device as soon as possible.

Please note that the Domenico FW does work on the D6 device, as well.

After flashing, please select "Domencio" in the list of devices under the SETUP tab. A new "Domenico" tab will appear, where the desired attenuation can be set. Please note that no other function can be activated when changing the attenuation, as this will crash the software. Remember, this is a beta release!

Regarding the hardware modification of the LTDZ, Domencio has done a lot of research. I have not tried any of these modifications, because I will purchase a dedicated device just for this purpose (I need to keep one original LTDZ to check compatibility issues), but I will share them anyway.

Domenico says:

LTDZ, D6 etc. etc. they are all copies of the original German NWT4. Chinese copies, badly copied. Maybe I found the LTDZ copy problem. The capacity of Pin 1 in the AD8703 to ground must also be considered. The 120Khz filter works badly. This is the simulation:

Improvement 1:

The mounted capacitor is LNF. You have to put another 100nf in parallel, I changed it because I can work with SMD components, if you don't have this possibility just put a traditional 100n capacitor in parallel to the existing one, like this:

Improvement 2:

Here is another problem on LTDZ. There is a capacitor to remove, as per datasheets PIN 6 and PIN 8 must not be connected for operation above 50Mhz and in the fact the D6 does not have this capacitor!

I haven't checked the difference on the trace but I trust the datasheets.

Improvement 3:

The ADC value read by my LTDZ when everything is off is 190 * 0.003223 = 620mV. This is the real reason why LTDZ is deaf.

Here is the simple change available to everyone:

Disassemble the two inductors and reassemble them on the welding side, this shift brings the reading down to 100 * 0.003223 = 320mV, which is a great improvement.

After passing the two inductors to the welding side, the PCB mask must be scratched off as shown in the red square. Why? Because you have to screen the track that you see in that square.

That track is part of the ADC converter and it radiates 350KHz.

Then shield with a thin piece of copper. The copper must be placed on the red zone:

After this modification the read value is 77 * 0.003223 = 0.240mV.

I have almost reached 106mV.

I have to see if we can improve further.

Improvement 4:

This is the modification to improve the sensitivity for LTDZ. It replaces the first modification in a much more effective way. Raw level ADC without box 26 - with box 36 you lose 1db.

With my firmware you should have the noise floor level at about -45db if so, you need to make this change:

Please note that all credits go to Domenico, who wants to remain anonymous.

If you have any questions or suggestions, please post them in the comments and I will forward them to Domenico.

Disclaimer: All HW modifications are on your own risk. I have so far not tried them out myself.