French Bell Sets

[Adjuster]

I have recently got hold of a couple of French external bell sounders, which I think date from no later than the late 1940s. Both have a pair of coils in series, and measure approximately 1000 ohms total DC resistance, which is also their nominal resistance as marked on the units. They have no trembler contacts, and appear to be magnet - polarised, designed for AC drive.

These things look to be in good order and I am hoping to get them ringing by making up a suitable "AC" supply (probably just an H - bridge giving squareish waveforms). My problem is though that I don't know what voltage or frequency should be used - I suspect that this is not the same as whatever modern telephones use in France. Obviously I can try hunting to find the minimum voltage and the best frequency to give a decent ring, but I would prefer to know what was actually intended. Does anyone have any ideas about this?

[russell_w_b]

Originally Posted by Adjuster

I have recently got hold of a couple of French external bell sounders, which I think date from no later than the late 1940s. Both have a pair of coils in series, and measure approximately 1000 ohms total DC resistance...

...Does anyone have any ideas about this?

Sounds to me like they're the same as bells here in the UK - 70V RMS (but in reality 20V - 90V RMS) perhaps 16⅔ Hz, as ours used to be before being superseded by 25Hz. I would try them on UK telephone frequency of 25Hz, 50Hz 70V as well, for the hell of it, and, if you have a hand magneto from a field-telephone, you'll have 16⅔ - 17Hz.

Of course, if you have a frequency generator and suitable amplifier (100V line), you could 'sweep' them from 16Hz to 50Hz and see where the sweet spot is. Although the French seem to do things differently sometimes, I rather think telephone ringing frequency is (was) pretty universal. Remember, in the UK, the actual voltage across the bell-coils is increased slightly by part-resonance with the series bell capacitor. So stick a 1u8 capacitor in series with them. In practice, there's barely any perceptable difference to the ring whether a capacitor is used or not.

[Adjuster]

I suspect that these things are lower impedance than the ringers we see on modern UK standard telephones, at any rate I measured the coil in a BT approved ringer I have as more like 6kOhms DC resistance. I don't have much idea of how that translates to an REN value, which I guess is in terms of impedance at the ringing frequency - furthermore there is no way these old French things are approved by BT etc, so I shouldn't think they ought to be connected directly to any phone lines here, with or without a series capacitor. In any case, my requirement is for bells triggered off by a computer, to sound different from my existing landline ringers.

Having said all that, I agree that the "suck it and see" approach will probably be best after all - it will be interesting to see how well the "sweet spots" line up for the two old bells, especially as it now seems likely that one of them is pre-WW2, much older than I thought, while the other is more like 1950s. Perhaps a bit of tweaking will be needed to get them working together.

[russell_w_b]

Originally Posted by Adjuster

I suspect that these things are lower impedance than the ringers we see on modern UK standard telephones, at any rate I measured the coil in a BT approved ringer I have as more like 6kOhms DC resistance. I don't have much idea of how that translates to an REN value, which I guess is in terms of impedance at the ringing frequency - furthermore there is no way these old French things are approved by BT etc, so I shouldn't think they ought to be connected directly to any phone lines here, with or without a series capacitor.

The impedance of a bellset type 59A (of the type used in a 300, 700 series telephone) is typically 1550 + j2031 at 25Hz (Z = 2555) or 2324 + j4320 at 50Hz (Z = 4905), equating, with a series capacitor of 1u8 or 2uf, to a REN of 4.

You could try driving your french bell from 30V a.c. 50Hz, say, and measure the current through it and the voltage across it, and see if it approaches the impedance figure quoted. If it does, you'll be safe enough trying it on the BT network; you're unlikely to harm anything.

[Adjuster]

Thanks for all that useful information. Of course I should have taken into account that the little extension bell I measured is intended to be a strict REN = 1, designed to work in a modern setup with other things connected and still give an acceptable loading.

I did not realise that older telephone instruments had higher loadings than that, but thinking about it it seems fairly reasonable in order to get a good loud noise out of an electromechanical bell.

[russell_w_b]

The 'modern way' of decreasing the REN of an old BPO (or equivalent) telephone is to insert a 3k3 resistor in series with the bell-coils, a-la the conversion advice on offer everywhere. Doing so will drop about half the ringing voltage across the bell motor, with a resulting current (for a supply voltage of 75V RMS) of 14.8mA through the coils. Without the resistor (but still including the series capacitor), the current will sit at about 35mA. In practice, the ring hardly changes.

Of course, more oomph will ensure a constantly good ring, but I've never had a problem, and my telephones have rung with a current as low as 9mA.