Thread: FM/DAB digital radio - THAT WORKS!!

Originally Posted by craigbrass

nrgizerbunny (who initially posted this thread) isn't selling these units anymore. I am currently the only one selling a full unit.

As detailed on http://www.coolapplab.com/ultimate_c...ona/index.html, a standalone Windows application will exist as well as plugins for Ride / Road Runner and Centrafuse. We should be able to add keyboard shortcuts in no problem. I will add it to my notes.

Excellent with keyboard shortcuts! You see, I want to control my radio with this unit. If you think it says AM/FM, you are wrong. It says DAB/FM :-) For specific tasks I will make my own menu system controlled with this panel.



For the windows application, will it be skinable? Will you share code?

let me chime in here...
i just saw a picture of the board, and it had a LOT of LARGE capacitors
where is this capacitance being added to USB POWER?
i bet this is a VIOLATION of USB SPEC!

a better design would be use of external power
a better design would use a isolated DC-DC convertor

course i dont know the exact layout, but it looks fishy

one thing for sure USB power should be pretty clean tobegin with not much ac ripple

course if usb can power it < 500ma, then whats all that large caps for?

I have two of the boards that nrgizerbunny built. The original he built had smaller capacitors on it, it would not pick up 210-220Mhz.

The second board picked up the entire range and FM was much better. He said something about the M2 power supply upsetting the radio.

As Dave said, this board is designed for good support of both FM and DAB, especially with the Mini Box power supplies which caused problems.

Also, I would take a look at some of the Sony radios, specifically the £30 people are currently getting Venice 5 chips from. This uses large capacitors like this as well.

Originally Posted by bigdave

I have two of the boards that nrgizerbunny built. The original he built had smaller capacitors on it, it would not pick up 210-220Mhz.

I have seen the layout of nrgizerbunny's PCB, and must say it think the noise problems has much more to do with the very poor PCB routing to and from the capacitors than it's about too small capacitors. The way the PCB has been routed I doubt the capacitors has much effect at all in reducing noise, that's probably why he has used 3 huge capacitors, to compensate for poor PCB-layout. If the power and GND routing was improved I'm sure the noise problems would be reduced a lot, also with smaller capacitors. There's capacitors on the PCB where there's only one long trace going to the capacitor, carrying both the current to and from the capacitor in the same trace. This is not the right way to do decoupling and the effect of the decoupling capacitor is reduced a lot this way.
The LM31T voltage regulator he use for 3.3V is also run outside specifications. Better use a good LDO instead instead of a LM317T, which isn't designed to operate at such a low voltage drop.
I made some comments about improving the PCB-layout and choice of voltage regulator from in this topic: http://www.digital-car.co.uk/forum/s...329#post264329
This is not to take anything away from nrgizerbunny's efforts, he has done some nice work to get a power supply running for this module, but he's not an electronics engineer as he mentioned somewhere, so he's probably not too familiar with how to make good PCB-routing and how important it is for ripple reduction. This shows in the strange way the PCB traces has been routed to and from capacitors. The higher frequencies your circuit operate at, the more critical proper PCB-routing becomes, also of the power supply traces.
So for a commercial product like craigbrass' I really hope the PCB design has been improved a lot and that they have not just reused nrgizerbunny's schematic and PBC-layout as is.

Originally Posted by Borte

This is the antenna i have (Couldn't find any info in English so page is in Italian): http://www.calearoam.com/a_156_IT_132_1.html

It very easy to get the same page in English, you just have to click at the Union Jack flag at the top right corner.
http://www.calearoam.com/a_156_EN_132_1.html

Originally Posted by craigbrass

4. The aerial connector is SMB.

What impedance is the antenna input?
Looks like car DAB and FM/AM antennas use different impedance it seems.
For example this one: http://www.calearoam.com/a_156_EN_323_1.html

D.A.B. cable : L = 5 m / Z = 50 Ohm / C = 101 pF/m
AM/FM cable : L = 5,6 m / Z = 75 Ohm / C = 67 pF/m

Originally Posted by veletron

Hi

I have been looking for a solution to combine FM & DAB signals to pass into the venice 5/7. Looks like a masthead triplexer would do the job:

http://www.tvaerials.com/product.aspx?productid=2173

Would need to be removed from the case, The TV input can either be ignored or terminated at 75R I guess. There appears to be no such thing as a DAB/FM diplexer, they are all for TV/FM+DAB

Losses look 'reasonable', and the module can be removed from its waterproof case & shoved in the same box as your DAB module.

Main issue is the F-Type connectors. Re-terminating the DAB and FM down-leads with F-Types should not be a huge problem, but making up an F-Type <--> SMB cable might be more of an issue due to the totally different cable sizes.

I'll see how an F-Type fits my downleads tonight.

Nigel

A problem is as I mentioned above that car DAB + FM/AM antennas seem to use 75 ohm impedance for FM/AM, but only 50 ohm for DAB.
The triplexer you linked to most likely use 75 ohm for all in- and outputs. If you mate a 50 ohm antenna to a 75 ohm input you will loose some power which means lower signal sensitivity.
Here's the manufacturer's product website for Vision V24-310 like you linked to: http://vision-products.co.uk/product...ubcategoryid=8

have seen the layout of nrgizerbunny's PCB, and must say it think the noise problems has much more to do with the very poor PCB routing

It has changed a bit I think since the one you have seen but around 250 of these units (old and new style) are in the market and work fine with a larger amount of Car PC power supplies than the initial design posted on Digital Car. A couple of electrical engineers have seen this design and say it is fine and also, as I said before, take a look at the Sony design. This uses a similar method.

Hi

I've had some reasonable results with this, see:

http://www.veletron.com/php/gallery2...2_itemId=16827

Also see the posts on digital car:

http://www.digital-car.co.uk/forum/s...=8609&page=117

Its not ideal agreed - if an ideal solution comes about, i'll swap it out.

Nigel