GPO 746F Broken thermistor bulb

[AlexFitton]

Hi I hop someone can help. I just started to convert a GPO 746F, while doing this noticed the legs of one of what I believe is a thermistor bulb had corroded and come away from the base of the bulb. Could anyone suggest a suitable replacement?

[russell_w_b]

If you just want to get your telephone working, simply strap out the regulator of which the thermistor bulb is part of. This is an option on the 706, which is similar to the 746. If you have a look at diagrams N846 (746 Inst.) and N806 (706 Inst.) available here you'll see where to do it. Perhaps just use a couple of resistors in lieu of the thermistor lamp filaments of a value between, say, 10 and 30 Ohms (dependent on the length of your line and how it sounds).

But to have it right, you'll need to source a replacement off an old 706 regulator board or a scrap 746 chassis. I should have one kicking around somewhere. If I find it, you can have it. I'll PM you.

[AlexFitton]

Thanks Russell. This is my first time trying to do anything with these phones. What is the regulator for and when you say line length do you mean between the phone and the socket?
Sorry to sound like a complete novice.

[russell_w_b]

I'm on about distance to the exchange, Alex. The thermistor is part of the automatic regulator. When you're near the exchange and drawing a large-ish current 'off hook' - 70mA say, the thermistor resistance is about 36 Ohms, and when it is a long line drawing about 30mA, it falls to about 10 Ohms. This varying voltage biasses the back-to-back regulator rectifiers and shunts the speech accordingly. It's explained in full down the page about the 706.

On the 706 (the instrument that preceded the 746), the regulator is on its own PCB and can be reversed in the socket, with the connections shorted together. It's fixed on the 746.

I have a 746 thermistor assembly you can have. PM me your address and I'll send you it FOC.

[David G4EBT]

That looks like a wire-ended bulb rather than an actual thermistor as such, but will behave in a similar way, in that the resistance varies as the filament warms up, though of course you'd need to know what the voltage and current spec of the bulb was, and for obvious reasons that won't be marked on the bulb. Some years ago an LCR bridge in PW specified a wire-ended bulb and I've just built an audio sine wave generator based on a Wien Bridge oscillator design which used one for regulation. (28V 40mA).

[russell_w_b]

The thermistors in a 746 are actually two-wire lamp devices, two of which are connected externally in series and centre-tapped. So yes, they look like a lamp. The thermistor in a 706 is a three-legged device with the filament centre-tap built in. I expect they are just lamps but with the characteristics exploited for the job in hand.

If you look at the pic you'll see the placcy case lying about which holds them secure (and prevents the glow from un-nerving subscribers who think their telephone might be on fire).

[AlexFitton]

Thank you both, had tried to find a spec online for them but could not locate them anywhere. Thanks Russell ill PM you my address.
Cant wait to get it fixed I've always wanted one of these phones.

[russell_w_b]

From Sam Hallas' document repository: 'The characteristic of the lamp is such that when the line current is 76mA the lamp resistance is 36 Ohm, when the current falls to 30mA, however, the resistance is only 10 Ohm. Therefore a bias voltage change of approximately 9 to 1 is obtained with a current change of about 2.5:1.'

From my own experiments on a GEC Barretter lamp, a 12:1 voltage change gives a current change of very nearly 2.5:1. I daresay there are several small lamps that would exhibit similar characteristics.

You can ignore the mauve line on the graph. It's just a plot of inrush current from cold to hot.

[Nickthedentist]

I don't wish to be pessimistic, but do check the continuity of the ASTIC (transformer) and microswitch contacts before you do anything else. If it's been underwater (or whatever) for long enough for that to happen to the lamps, then who knows what else has been damaged.

Nick.

[ThePillenwerfer]

There is a story that when the 706 first came out an elderly subscriber saw the regulator glowing through the ventilation slots and thought the 'phone was on fire so dropped it in a bucket of water. Then the GPO started painting over them.

If you can get the part to fix it properly then great but if not I'd expect it to work fine by-passed as the pre-706 telephones didn't have them and line regulation is going to be better these days in any case.

If there are other faults on it you can get a complete new circuit board from Telephone Lines for a fiver including postage.

- Joe

[AlexFitton]

Thanks everyone. I've checked the continuity on the transformer and the switch and it is good. Strangely the legs on the other thermistor an not corroded at all and apart from a bit of surface rust on the terminal screws its in fairly good condition considering was saved from the skip.
I gave it a polish with some brass cleaner then silver cleaner and the case and handset have come up great.

[AlexFitton]

Ok I think I'm nearly there. Thanks to Russell I've got two new thermistor bulbs fitted. I have also replaced the old style cord with a new one, changed the links around and fitted a resistor between T4 & T5 and fitted the 205 rectifier between T1 & T2. Just for good measure I changed the curly cord and fitted the electronic microphone.

Almost everything works perfect. There is dial tone, it rings, answers, hangs up and the sound is perfect... The only thing is I can’t dial out.

I know the line will support pulse dialling as my normal phone has options for both pulse and tone and works perfectly on both.

When dialling the tone cuts out and you can hear the pulses but as soon as you finish dialling the tone comes back.

The dial takes about 1.5 seconds to go from 0 back to rest. I did notice a bit of soot on one of the contacts on the back of the dial so I cleaned it with a bit of electrical cleaner and a thin bit of card but this hasn't had any effect.

Any ideas?

[russell_w_b]

Originally Posted by AlexFitton

The dial takes about 1.5 seconds to go from 0 back to rest. I did notice a bit of soot on one of the contacts on the back of the dial so I cleaned it with a bit of electrical cleaner and a thin bit of card but this hasn't had any effect.

Your wiring looks OK on the face of it, the dial wires going where they're supposed to go.

How tolerant is your exchange to pulse duration and mark-space ratio? An electronic pulse-dialler will be spot-on in this respect, but dials can fall outwith the tolerances required by some exchanges (I'm told). Your dial should give ten pulses-per-second (1 second to return from zero) with 66.6 millisecond 'break' and 33.3 millisecond 'make'. I think the accuracy of the ten pulses-per-second might be an issue.

Do you have another dial telephone to compare it with? A known 'worker'? Can you tap out '1-2-3' on the handset (like 'tap - taptap - taptaptap') and get the speaking clock?

Can you try your phone on another exchange?

Have you stripped the dial, cleaned it and lubricated it? Have a look here...

Pleased the regulator thermistor works!

Just a thought: when your dial returns hard to rest (after impulsing satisfactorily in accordance with the number dialled), do the impulse contacts inadvertently break and remake ('bounce')?

[AlexFitton]

Ive checked the line with another GPO746 i was given at the weekend and it worked perfectly. Im in the process of stripping and lubricating the dial now, il let you know how it turns out.

[AlexFitton]

Finally got it all finished and working a treat.

Dial stripped down degreased and re-lubricated and all working perfectly. It was a bit more of a challenge than I first thought it got a bit fiddly but time and patience paid off in the end.

I followed the instructions carefully and took loads of photos of the dial before I started for reference. it also helped to have a working dial close at hand.

I found the best thing to degrease the dial was Servisol 40 as it is powerful enough to remove all the old oil but didn't damage the plastic parts. (Tested first on the plastic gears of an old CD player that was about to go in the bin) I also used it to give the contacts a very carful clean.

I re-lubricated with 3 in 1 oil on a cotton bud stick with the cotton ends removed.

Once I had reassembled I gave it a quick test and hey presto dials out perfectly!

But... later I discovered something wasn't quite right the bedroom phone wasn't working right. There ware not lights on the keypad and I couldn't dial out. When I unplugged the 746 it returned to working normally, so I plugged my other 746 in to check and it worked fine. I decided to just change the whole board and tinker with the old board later.

I'm really pleased with the finished result and have added a ringer on off button just to finish the whole thing off. I keep making excuses to make calls on the landline now!

Thanks to everyone for their help especially Russell for the thermistors (they wont go to waste I will get the old board working at some point)

My next project is a RTT 56B I'm converting / restoring.

[russell_w_b]

Originally Posted by AlexFitton

Finally got it all finished and working a treat.
Dial stripped down degreased and re-lubricated and all working perfectly...

Well done on the dial. Good fun, eh?

Originally Posted by AlexFitton

But... later I discovered something wasn't quite right the bedroom phone wasn't working right. There ware not lights on the keypad and I couldn't dial out. When I unplugged the 746 it returned to working normally, so I plugged my other 746 in to check and it worked fine. I decided to just change the whole board and tinker with the old board later

What readings do you get if you measure the resistance between pins 2&5, 2&4 and 4&5 of your BT plug with the handset on rest?

[AlexFitton]

Typically after switching the board back in it now works fine, which makes me think I was a bit hasty in buying a new one. Oh well I've got the new one as a spare.
It now looks and works great.
Servicing the dial was easier than I expected in the end just takes a bit of time and patience but rewarding when it's finished.
It's been good practice for for reparing the dial on the RTT half of which was missing and the rest was rolling about in the bottom of the case!

[russell_w_b]

You might've had a bit of swarf or 'whisker' or whatever lodged beneath your board, shorting things out.

You can't have enough spare boards! It'll come in.