DC Isolation - a bluetooth Roberts R200

[oxford33]

Hi everyone,

I'm posting this partially as a question and also in case it helps someone else. I have a Roberts R200 which I've added an aux input to based on some searching the forum. This works pretty well and the radio still works when nothing is plugged in.

My next idea which might be slightly controversial was to try to add bluetooth to the set, so I cobbled together a 9V>5V supply with a diode, voltage regulator and a couple of capacitors, and ordered a cheap bluetooth module called a "BLK-MD-SPK-B" for about £5 on ebay. This arrived today from china and turns out to be about the size of a postage stamp i.e clearly designed for surface-mount application.

I don't have the equipment to make a proper PCB but decided to give it a go anyway. I superglued the module to a bit of veroboard, then with some leads from a breadboard, more glue and a lot of swearing I managed to connect up the basic connections for power and audio (the module can also send remote control commands and even handle battery charging, but I wasn't planning on that) - image attached. Unbelievably enough this actually works!

The problem is it only works when the radio and the module are powered by different batteries: if I use the same battery for both, the radio just blasts out horrible noise as soon as the module powers on. I suspect that this is an isolation problem. Googling suggests some kind of isolated DC-DC converter, but I've never used one of those (and they're at least a few pounds each, so I don't want to buy the wrong thing). Does anyone know or have any better ideas how I can isolate the nasty digital circuit noise from the nice old analogue radio?

Thanks

[TrevorG3VLF]

The only capacitor that I see is a big electrolytic. This will not do much at radio frequencies.

[charambo]

There is an isolated and regulated power supply that might be of interest to you. The ADuM5000 from Analog Devices. Being small, if you need any shielding, it will not take too much space.

[McMurdo]

I suppose you need to ascertain where the noise is coming from. Is it coupled through the supply lines or induced harmonics via RF? Operating it successfully on a seperate battery suggests the former.

A typical component dc-dc converter will accept one voltage and convert it to the same or another but with galvanic isolation...a sort of baby switch-mode power supply as small as a cough sweet. Useful for providing an isolated supply with decoupling. Unfortunately they can also induce RF harmonics in close proximity to radios!

I'd try to decouple the various lines first with ferrite beads and/or ceramic capacitors before going down the floating supply route. It almost sounds like you have a similar problem to laptop charger noise when using one as an audio source for certain types of hifi setup. Try filtering the audio as well, you don't need the audio output of the bluetooth to have frequency response into the RF, and it's another way-in for noise.

[ronbryan]

Have you chosen to reference your audio ground to the R200 -9V rail (i.e. HT) when you are deriving the Bluetooth +5V supply from the R200 battery?

This approach could couple unwanted noise into the R200 audio amplifier, as it requires an audio ground referenced to the metal chassis, which is battery positive (or slightly negative of that if you have either of the later types of R200 set).

You could test this theory by isolating the Bluetooth audio with a small AF transformer and referencing the secondary to the R200 chassis.

Ron

[Martin G7MRV]

As already said, theres not much in the way of bypass there! Add a couple of 100nF and 10nF caps across your supply rails as well,

How are your audio lines connected? Have you put a blocking capacitor or a transformer in to isolate the device from any DC on the audio line, and vice versa?

[oxford33]

Thanks for the pointers. I'll add some more capacitors here and there but I think ronbryan may have hit on the real problem - I'm injecting the signal at the volume control, with a couple of stereo summing resistors followed by a 1uF film capacitor for isolation. However the ground was directly connected to the bottom of the volume control as I'd forgotten that chassis on the R200 is actually battery positive (which I suppose means I was also short-circuiting/saturating via the 5v voltage regulator, i.e. very bad!). This worked fine with separate battery-powered devices so I hadn't realised the error.

Seems to me it is the positve chassis vs. negative ground of the BT module, try a minus 5V regulator instead, then the grounds are referenced to each other.

[Scrapheap]

Hallo,

I have converted a couple of Roberts R200s to Bluetooth using a miniature matching 600 ohm transformer for the audio feed to the VC. Instead of fitting up a Bluetooth power supply, I simply use a Bluetooth module which has a rechargeable internal battery. Solves the interference problem completely!
Hope this helps.
Chris