Smiths car clock repair

[robbo_291]

I have a friend's Smiths car clock in for repair that has come out of an E type Jag. I'm trying to identify a transistor. The part number is E5032. I don't know whether this was a specific part number to Smiths production or not but I can't find a listing for the part number.

The Jag was registered in 1969. There were up to two stand-alone clocks made for this car of the same style; one of which used incorporated a mercury back-up cell and a later one that didn't that was used up to 1974 and I don't know when the switch over happened. I have the one that has no mercury cell. I'm assuming that the clock could well have been replaced at some point in the car as this is the later model of clock.

I did assume that the clock would have had a Germanium transistor however the package of the transistor is TO92. Were there any Germanium transistors made in To92 package or does To92 package definitely indicate Silicon? The transistor is of NPN type as the car has a negative earth and there is also a 4.7k carbon film dropper resistor inline with the 12v supply.

Anyhow, this clock is all riveted together I daren't strip it to bits. I'm assuming that the transistor energises two coils that are sat on a large wheel that swings the coils in and out of a magnet assembly which powers the clock. I'll see if I can get a decent photo or two up on here.

Any help would be great thanks. Rob.

[Herald1360]

I would think it's likely to be silicon, but just because the car's negative earth doesn't guarantee it's NPN or even bipolar- that will depend on how the circuit is configured.

Do you have access to the transistor to get meter probes onto the legs? If so, it should be easy enough to identify its polarity as long as the shunting effects of the rest of the circuit don't swamp measurements.

[AC/HL]

I think it's Texas Instruments, E5024 and E5036 were used in the Thorn 3000 chassis, but they are both PNP.

[Mike Phelan]

Rob,

It's PNP silicon. The base is turned on by one coil and the collector drives the other one to give an impulse to the balance. Some models just had a centre tap instead of two coils.

It was an update from the earlier model that used two contacts instead of a transistor.

Generally the only fault is due to gummed up lubricant on the balance pivot unless someone has reversed the supply however later ones have a diode in series to prevent this.

[robbo_291]

Wow! Thanks to all for your help you guys! Pleased I joined this site! I will be looking at this again over the weekend so will get a few pics up. In the meantime, what would be a more common transistor as a replacement that would do the job?

[julie_m]

BC327 is a general purpose PNP silicon transistor in a TO92 package, good for 50 V, 800 mA, with a hFE of 100; and should be cheap and common enough for a first test at least. (If I found I needed more gain, I'd try a BC557 next; same pinout, more hFE, less IC.) Note that it has an unusual pinout for a TO92, with the base, not the collector, in the middle. This may be an advantage when replacing metal can transistors, though.

[robbo_291]

Ok... Well I couldn't get a decent photo that would be of any use. However, I've found out the following.

Viewing the TO92 plastic cased transistor from the flat side with the E5032 on it and the legs pointing down.

The first pin goes to the coil. the second pin goes via a 4.7k resistor to +12v and the third pin goes to ground. I hope this helps..

[AC/HL]

If you PM your address, I'll send an E5024 and an E5036 to try. I've got a few gathering dust.

[robbo_291]

Sent PM Bill and many thanks for that. I take it both these parts are PNP? No house number by the way.

[AC/HL]

Yes they are, but that's all I know about them!
Should be in the post today.

[robbo_291]

Many thanks for the transistors Bill. They arrived yesterday. I shall give things a go on Sunday and let you know how I get on.

[zach1965]

I came across this thread while researching the Smiths clock in my 1969 Jaguar. As an English major in college, I have few technical skills and have been flailing about for literally months trying to figure out what has failed in my clock.

I have both the battery and the 12V clocks (and I am wearing a 1968-ish Smiths Astral 17 wristwatch as I type -- I am quite a fan of Smiths!).

The battery clock has a E5083 transistor.

The 12V clock has a E5032 transistor.

I am seeking the following:

(1) part numbers for replacement transistors that would commonly be available in the USA (for example, I ***think*** an equivalent is a2N3906, a PNP transistor) and

(2) the pin-outs for the ***original*** transistors.

In the course of my flailing, I have come to the conclusion that BCXXX transistors -- which I assume to be the closest analogue to the E50XX transistors --have the pin-out "C-B-E"; 2N39XX transistors have the pin-out "E-B-C".

Obviously I need to get the correct replacement transistor AND fit it correctly.

Based on the advice of somebody whom I thought was smarter than I am, I fitted a NPN transistor which was, I now believe based in large measure on this thread, incorrect. I also fear that I fitted it incorrectly as the pin-outs for the old and replacement transistors may be different.

(After installation of an NPN transistor, the balance wheel on the clock, which previously did not move when power was applied, now rotates a few degrees clockwise when powered until one of the balance wheelcoils is partially covered by the magnet . When the time adjuster/balance wheel starter is engaged and released, the balance wheel oscillates only briefly before returning the the state set forth in the preceding sentence; the balance wheel oscillates for a much longer period when the clock is unpowered. Thus, the electronic control is stopping the clock instead of running it. The problem could be the transistor or the wiring/pin-outs or both).

All help would be greatly appreciated.

[Mike Phelan]

Hi,

For a start, you probably need a PNP transistor and I'm not convinced that a silicon device would work on an earlier clock, but in fairness, I've not tried it. Later clocks were silicon PNP but the coil windings or magnet positions might be different. I think that coincides with the anti-reverse diode.

In the past, when someone has reversed the polartiy and destroyed the transistor I fitted an AC128 or AC188, the connections are E-B-C.

If the balance seems to be locking in a set position the transistor is probably SC.
Or the starting switch is ....

[zach1965]

Thank you kindly, Mr. Phelan, for your response.

I am fairly confident that the existing transistors are silicon -- the battery clock E5083 transistor is the first picture, the 12V clock E5032 transistor is the second, and both appear to be silicone. Further, I can find no evidence that germanium transistors were ever made with a TO2 case such as these.

Since the AC128 and AC188 transistors are both PNP and E-B-C, I am treating those parts of the mystery as being solved.

Those AC1X8 transistors are hard to come across, however, as it appears that they have not been commonly available in the US for decades.

So the penultimate question is "Which common-to-the-US transistor should be substituted?" I have purchased 2N2907, 2N3906, and 2N4403 from Radio Shack (in a multi-pack). The 2N2907 has the highest Ic at 600 mA and would thus seem to be the best choice (the AC128 and AC188 have Ic's of 1A and 2A respectively but much lower voltage ratings -- given that the voltage on the 12V clock after the 5.6K resistor is a little more than a volt and there is some relationship between voltage and current that I simply don't understand, hopefully the 2N2907 will be sufficient).

Can you opine on this point?

And the final question is what does "...the transistor is probably SC..." mean?

Regards and thanks in advance,


-- Kevin Williams
Nashville, USA

One more question: I have Googled in every way I can think of and can find no data sheets for specs on E5083 and E5032 transistors without any success what-so-ever.

Can any of you kind folks advise where I might find those specs?

-- Kevin Williams
Nashville, USA

[wave solder]

SC is short for short circuit.

[AC/HL]

Those referenced are in-house numbers, Texas Instruments as far as I know, and the package is the same. Rob never came back, so I don't know what the outcome was.

[Mike Phelan]

Originally Posted by zach1965

I am fairly confident that the existing transistors are silicon ...

In that case, Kevin, it's not a very early clock; the germanium ones were only made for a short while after they had mechanical switches.

[zach1965]

Thanks to all for their kind advice and counsel.

I will experiment over the next few days and advise on results and conclusions.

Regards,


-- Kevin

[Herald1360]

The resistor seen in the picture in post #14 is 56ohms 5% (green blue black gold), not 5.6K 5% (green blue red gold).

[zach1965]

Thank you for your response.

The resistor shown in the picture is for the battery clock. My posts concern the 12V clock (the battery clock was damaged after being knocked off my workbench and was not repairable). I have purchased another on eBay and will, probably, request the kind assistance of the forum as I attempt to get it working, too.

Thanks again!