Bluetooth Supply From Valve's Heater.

[PsychMan]

Like a lot of cheap chinese electronics there's no info provided on current draw. I'll have to check it on a multimeter. The thing is tiny and claims to have 3w power amplifiers on it, all for just under £3 delivered. I was curious how the absolute cheapest example would perform.

[paulsherwin]

I agree with the Biker that the current consumption needs to be under 10mA for the cathode trick to work without significant adverse effects. It's easy enough to hook up the module to three AA cells and measure the current.

[G6Tanuki]

In this situation I'd use a classic single-ended voltage-doubler (which works happily with one end of the transformer grounded, as you'd expect to find in a typical radio) then something like a 7805.

2 diodes, 2 electrolytics, a 7805, 2 small-value decoupling-caps for the 7805. The entire thing could fit inside a cubic-inch.

Job done.

[PsychMan]

Just an update on what I've been trying. I tried to measure the current draw from the Bluetooth module, but the device would simply not function with a dc ammeter in series.

I tried a 7805 regulator with the half rectified heater supply and a capacitor, and the device would also not power up. I experimented with different capacitor values (before and after regulator), but the result was the same. I suspect the power requirements may be too much for this particular module, and that perhaps the capacitor being charged every other cycle is not providing enough power and the regulator is not able to perform. I tried 4700uf at one point. Unloaded the voltage is nice and steady at around 5v, with the device connected it cycles between around 4.5v and 5v.

I have a spare 9v 100ma transformer and will experiment with a lash up supply from that. There is adequate space to fit this in the chassis where the mains lead comes in so may prove the better option. Its centre tapped so I think I should be able to make a supply with 2 diodes, 2 capacitor and 1 x 7805 module.

All good fun and a learning experience. Thanks all for your contributions

[PsychMan]

Been a while since I last looked at this. The Bluetooth module I had turned out to be faulty, and I found the transformer I had would not fit in the chassis so I shelved the idea for a while..

I've just revisited this and thought I'd try something slightly different and wanted some opinions on it. I've used 4 1N4007s to full wave rectify the 6.3v heater supply, smoothed with 1000uf capacitor, 7805 regulator, and smaller 2.2uf capacitor across the output.

The one issue with this of course, is if I now connect the Bluetooth modules output to the gram input of the set, I short my rectifier and burn up a diode and possibly my mains transformer for the set. To get around this I used a very small 1:1 audio isolation transformer to send the audio signal into the gram input, which works very well. I have a load of these and have been waiting for an excuse to use one. But before attempting to make this experiment a little more permanent, I thought I'd ask everyone's views on it?

While I cant see it happening, if for some reason there was a short between the primary and secondary winding of the small isolation transformer, it wouldn't be pleasant. Adding a small pico fuse could protect against this, but perhaps I might do well to simplify and go back to half wave rectification...

[paulsherwin]

It sounds as if the heater supply has one side connected to the chassis. You can't use a bridge in these circumstances, you need just a single diode connected to the hot side. Half wave rectification means the smoothing cap has to do more work, but that shouldn't be significant given the low currents involved.

You can keep the current arrangement by connecting the bluetooth earth to the chassis with a cap rather than a direct connection, but that's not really a good solution.

[PsychMan]

I wasn't sure if half wave rectification would provide sufficient voltage to use a regulator like a 7805, one way to find out

I'm seeing about 7.2v DC after my smoothing capacitor when rectifying all of the wave form, which I think is the minimum territory for the 7805 to work. If its low enough, perhaps I could drop it down with the spares diodes, its all on a big of tag strip so just a case of changing the wiring around

[paulsherwin]

Try it. In principle the voltage will be the same, it's just that the smoothing cap has to supply the voltage between alternate peaks rather than just between every peak. Use a Schottky to minimise voltage drop - you will find several in any scrap SMPSU.

As you say, 3 1N4007s or similar should take 7V down to 5V without bothering with the regulator.

[G6Tanuki]

Rather than use a stock 7805 (which is a nearly-50-year-old design!) which needs a rather high input voltage why not look at using the more modern 'low dropout' 5V regulators.

Some of these are happy with an input-voltage that's just a few hundred millivolts above the regulated output voltage.

http://www.linear.com/product/LT1086 series are cheap and readily available; will reliably regulate to 5V with only 6VDC input.

[PsychMan]

Good point G6Tanuki, I was just reading a datasheet the other night on the TI site about LDOs

[Ed_Dinning]

Hi As you are using an e-cap on the output of the 7805 make sure you put a diode "backwards" across the reg and have the usual 0.1uF from in and out pins to ground on the reg itself.
Your bridge rect circuit should work well. If you use telecomm 600R 1:1 transformers they have an input protection circuit in the very unlikely event of a pri/ sec short.

Ed

[PsychMan]

The small TXs should be 600 ohm from memory, but they were cheapo ones I bought online from China so not sure of the protection. I got a whole bag of them, maybe I'll open one up and see what's inside it. I primarily got them to deal with earth loops on some equipment I use, just thought Id try one with this as an experiment when I realised I would be shorting out my rectifier using the chassis ground on the bluetooth module.

Regarding the diode, would that be to protect against reverse voltage on the rectifier / capacitor when using half wave rectification?

I've seen various diagrams of 7805 circuits, some seem to feature 0.1uf on the input and output to ground, some not, some with different values instead. So I just used what I had around. Perhaps I'll have a read of the datasheets

[paulsherwin]

The protection diode is to prevent capacitors discharging through the regulator, nothing to do with half or full wave rectification. It's good practice to fit one, though lots of circuits omit them without ill effects. A 78xx datasheet will show example circuits.

I really wouldn't bother with a regulator though. The rectified heater voltage won't vary very much, and the BT module is likely to function as well with 4V as with 5, so you just need to drop the voltage to a reasonable level with a few diodes. It should be easy to get the right voltage by trial and error.

[vidjoman]

I have 3 BTooth units and each has a battery inside @ 3.6 volts. The 5 volt in goes to a charging circuit to keep the battery charged. Have not checked how long it runs without external power but guess it's several hours. I suspect the 5 volts is not very critical.

[PsychMan]

I've since had a newer module arrive which has 2 separate voltage inputs (1 labelled 3.2 - 5v), and another labelled (7-30v), which allows great flexibility. They must have a few different regulators on this board, though all tiny SMD stuff.

My lower voltage module works great with using half wave rectification, 1 capacitor and a few diodes to drop the voltage. The newer one needs just 1 diode and one capacitor to feed the 7-30v input and works great, very minimal indeed!

Initially I tried a 1000uf smoothing capacitor, but as soon as I sent audio to the module it reset itself due to low voltage. Adding another 820uf in parallel and it works fine, so off to Maplin later for some 2000uf capacitors so I can install it all properly.

I wont be modifying all of my radios like this, but for some it makes sense as my other half can also make good use of them without mucking about with a transmitter

The modules were very cheap (from China), happy to share the source via PM if anyone else is interested.

[paulsherwin]

Glad to hear you found a solution that worked.

The unit with two voltage inputs almost certainly has a 7805 type regulator built in.