[RichG]

These SCX xenon lights sound like a bit of a disaster.

Putting aside that they might be fragile? There are two possibilities.

1) Incorrect choice of series resistor, such that they are being overdriven, perhaps an attempt to get more brightness?

2) Lack of series protection diode to protect against reverse polarity.

Has anyone got a picture of the light board showing the components and how they are connected?

Rich

 

[RichG]

If you are saying that sometimes they mend themselves after a bump, then of course it may be reason 3. A loose connection, however I assumed that people had checked for that.

 

[RichG]

Thanks Rob. More complicated than I expected. However a couple of observations & questions to be going on with.

So I can see the LED's at each end, then working in R2 & R3, assume to be in series with the LED's?

The gold things are a couple of diodes, probably in series again, so look like diodes to protect against reverse polarity are there.

Thats all you get on a Scaley board so now it get's interesting.

There seems to be another resistor, a capacitor and a burnt out bit in the middle. I wonder if they have a voltage regulator for the LED's?

Anyway need a better picture / does anyone know what the burnt out part in the middle is.

That said it looks like your problem may be a one off. Other people do not seem to be reporting both lights failing & smoke.

Rich

 

[RichG]

Thanks Zipp. Yes more data needed if we are going to get to the bottom of this one. Another picture would be nice. Any more pictures / info from other peoples failed boards would also be good.

I have a feeling there may be more than one problem here. With the available data I would guess.

1) From some peoples descriptions it looks like SCX rely on a mechanical contact between the strips and the board, rather than solder. That said the picture looks like it is soldered?

2) Some evidence that perhaps SCX designed to run on a lower voltage. 14.8 Volts versus 16 Volts or whatever a Scaley Wallwart gives out?

However if they are fine on 14.8V & quickly fail on 16V that does not sound like much of a design tolerance to me. So I feel there is more to find out about this problem. Also SCX must know that these cars will be run on other tracks, so no excuse.

Rich

 

[RichG]

Ok I think, thanks to Zips pictures I have worked out the circuit. Am away from home so please excuse the picture.

Am not certain about the transistor, the marking looks like 6C and that seems to be a BC817, using it's pinout the circuit is unusual but makes sense.

So we have a series protection diode from the +ve, this feeds the collector of the transistor. I am assuming the other diode is a zener, unknown voltage but would guess 4 to 6 volts. The zener is fed through a 1k2 and will hold the base at a constant voltage. The voltage on the emitter of the transistor will be about 0.7V less than the base and is used to drive the two LED's each through a 68 ohm resistor.

So SCX seem to have put in a simple voltage regulator for the LED's. This should mean that the brightness remains constant over a range of voltages, so in analogue mode less variance than a Scalextric car?

So bear with me while I think out loud.

The transistor is rated at 250mW and 500mA. The current rating is fine. Am not sure how many mA each LED will take, but let's guess 50mA so 100mA in total.

So if the emitter voltage is say 5V and we have a 15V supply then the transistor will have 10V across it and is passing 100mA. So power dissipation is 0.1A * 10V which is 1W. Too much.

So perhaps I have assumed the wrong current for the diodes. So looking at it from the other end. The max current for 250mW is at 12.5mA each, 25mA total. Power dissipation now .025A * 10V = 250mW which would put the transistor at it's limit.

So if this is a BC817 it is not suprising that it fails, if there is a suprise it's that it survives even with 14.8 Volts. It could be I am still making a wrong assumption about the LED current. To be safe it would need to be a max of 6mA each.

Also possible that the device is not a BC817. Could someone make a couple of measurements on a good board. With the car at max throttle measure the voltage on the input +ve, then on the three connections of the transistor which is the black square in the middle of the board. Put the negative meter lead on the Negative (RH 0V) input strip.

Rich

 

[RichG]

Hi Kev

Thanks for the readings. They look a bit odd. Of course it is possible that I have made false assumptions about what 6C is or could just be that your board is faulty. In particular reading 1 looks wrong, and explains why the LED is very bright, but not why the other one is dim.

Will be easier for me if you could you do them again but this time put the meter negative, Black lead to the other input strip on the RHS.

Then make readings 0 to 3 with the Red lead.

If someone with a good board could make the same readings that would be very helpful.

EDIT... If making measurements could you do 2 more. Please measure the voltage on the outermost connection of each LED.

Rich

 

[RichG]

If the little black thingy is a BC817 then (3) is ok and is close to the voltage on the rails.

(2) also looks ok and will be the voltage of the zener. I had guessed 4 to 6 volts.

However if the base of the transistor (2) is 4.23 Volts then the emmiter (1) should be about 3.5 volts. It looks very high at 12.14 and explains the very bright LED.

So yes if it is a transistor, then it has failed. Also if this is some other device, say a voltage regulator then 12.14 volts is still very high for an LED being fed through a 68 ohm resistor.

Will be interesting to see the other two readings.

Rich

 

[RichG]

Very interesting Rob. Will be good if they are all like that, trouble is I have a feeling that if the "device" goes short circuit it will take out the LED's.

Need some measurements from a good board, but when I get back home Kev will mail me a car, or perhaps this is an excuse to buy an SCX car

and I will sort it out properly. You will have to wait a little while for a solution.

Device obviously needs replacing, but need to do some maths and fully understand why it is failing. May just need a beefier device or a change to the circuit.

Rich

 

[RichG]

Ok well this needs confirming by measurement on a good board, but here's some maths.

A typical High brightness LED takes 20mA.

We have 2 so 40 mA.

Taking even a relatively low voltage on the Rails, say 12.7V, to make the maths easy.

So if the circuit is right then 0.7V is dropped across the input diode leaving 12V at the collector.

The voltage at the emitter will be 4.23V - 0.7V = 3.5V (4.23V is from kev's measurement. This part of the circuit is probably fine and was about what I was expecting)

So the transistor will have 12V- 3.5 V across it & 40 mA flowing through.

So power dissipation will be 8.5 * .04 = 340mW
Of course on a Scaley track with say 16V this will increase to 12.5 *0.04 = 500mW

Transistor is rated ar 250mW. So this is where I started from. What I don't understand is why they work at all on any track.

It seems so far out that I feel I am doing something wrong.

Or SCX have screwed up big time.

So will reserve final judgement until I get a car to measure.

Rich

 

[RichG]

Thanks for that Col. Well the measurements seem to tally. Also you have confirmed that in practice this is a voltage regulator and has the benefit of giving full output on the lights at small throttle openings.

Unfortunately it seems SCX has got the maths wrong, for the dissipation in the BC817. To be honest I would expect them all to fail.

And the higher the voltage on the rails and the longer your straights, the quicker they will fail.

Solution will be to use a beefier device or do as Col suggests and replace with a simple series resistor.

 

[RichG]

So doing the same maths on Col's actual figures. Each LED is taking 17mA which gives power dissipated in the transistor of 247mW.

So right on the limit, and that with the relatively low supply voltage of 12.4 Volts.
In practice for reliable operation I would rather see a max of 150mW.

When the transistor fails. If it fails open circuit then the LED's will just stop working. If it fails short circuit then the LED's will go very bright but not last very long.

Have we confirmed what voltage a standard SCX set runs at?

So do all SCX cars have this style of light board or is it just some cars? Ooh and if I was to buy one, just in the interest of research, give me a couple of suggestions.

Actually make that a pair, can't have a single car running round the track.

Rich

 

[RichG]

The black device is an SOT23 case style BC817-40 (At least until proven otherwise) It's base is held at 4.9V by the zener diode. The emitter will track the base voltage but is 0.5V less (Col's measurements). This voltage is used to drive the LED's via 68 ohm resistors.

So what this means is as you press the throttle the brightness will increase but instead of full brightness being reached at full throttle it will get there at about 6 Volts, and then stay at that brightness.

Above 6 Volts power is dissipated in the transistor, and the rest as they say is history.

My suggested short term mod for those wanting to try it is to replace the two 68 ohm resistors with 120 ohm, or to be very safe or if you run greater than 15V 150 ohm. This will reduce the LED brightness, mabe too much, but will retain the function as described.

Another alternative would be to put in a beefier transistor. Will look into that. Final option is to remove the transistor from the circuit and replace with a resistor connected between points 1 & 3. Something like 680 ohms would be about right, but you would need to experiment.

Please treat all of the above as experimental. I would not normally make recommendations without testing first.

So yes the Audi & the Aston look nice.

Also like the look of the 360

But does anyone know how you know which cars have this type of light board?

Rich

 

[RichG]

QUOTE (col_de_turini @ 11 Aug 2006, 12:52) <{POST_SNAPBACK}>Just about any NPN transistor will do

Hmm I think thats how SCX chose theirs.

However 2N222 at 650mW dissipation, probably fine.

Don't forget you have to derate at 5mW/ degC above 25 degrees. Will propose some alternatives when I get a mo.

 

[RichG]

It's a bit tricky to check the LED's. However if they have glowed brightly at any point then they have probably failed, if not then they are probably ok. It all depends on how the transistor failed. If it failed short circuit then it will probably have taken out the LED's, if it failed open circuit they will probably be ok.

Sorry there is not a simple answer. No I did not buy a SCX as I have a bit too much on at the moment.

However if you want send me your car I will have a look, and replace the transistor, check if the diodes are ok and send back for testing.

Rich

 

[RichG]

Hi Steve

Just to let you know, car arrived safely this morning.

Have had a look at the front light board and the transistor is obviously blown, it has that telltale bump in the top.

That is however what we were expecting.

SCX have not exactly designed the board for easy removal as some of the plastic parts are assembled after the board is put in. However I have just managed to extract it without removing the plastic parts.

Have tested the LED's and the good news is that they are both fine.

To be continued....

Rich

 

[RichG]

Progress report. Removed the transistor and replaced with beefier one. Switched on but no lights.

Checked the other components and the zener diode is short circuit.

Am suprised that the zener has gone, had better do the maths on that as well. I guess if they can get one component wrong, they could get them all wrong.

Quick bit of maths shows worst case of about 50mW so should be fine assuming this was a 100mW zener, trouble is I don't understand the markings.

From the measurements that have been made looks like a 4.7V zener is needed, unfortunately it's about the only value I have not got.

I will be ordering some other components next week so will get one. However means that getting the board going will be a few days away.

Rich

 

[RichG]

Bits arrived from Farnell this morning so have completed the repair.

Fitted a BZV55 series zener, 4.7v 400mW (Farnell 305-0919)

I do not know why the zener had failed, worst case dissipation is only 50mW and it should see nothing like that. The transistor failing would not put additional load on the zener. Also the zener had failed short circuit so that would not have caused the transistor to blow. A mystery.

Fitted a ZXTN2031 (Farnell 108-4421) in place of the failed BC817

There is no mystery here why the original device failed. It is rated at 250mW and even with 12 V on the track will be dissipating 280mW at 16 volts the dissipation will be up to 600mW.

The replacement is rated at 1W however in practice will be less than that because of the very small heatsinking from the circuit board. (I would prefer a higher rating, but this was the best I had) Hopefully will be ok, I ran it for 30 mins on the bench at 16 Volts and it was Ok, but it needs Steve to thrash it round the track for awhile.

A question for anyone with the body off. The pictures earlier in this thread show the transistor as a 6C which translates to a BC817. The device I removed from Steve's car was burnt so I could not read the marking. Would be interested in what marking is on other transistors.

So final thought. I don't know if people from SCX read this forum? If they do then there is a major problem with your light board, I would expect you know this by the number of returns you get.

A comment would be nice, failing that if anyone has a suitable name / address I will drop them a line.

Final Final thought. The connections for the lights and the floating motor all just use copper to copper strip contacts. To my eye this looks like "Optimistic Technology". Do people have any problems.

Steve will put the car in the post tommorow.

Rich

 

[RichG]

Swiss it's as if you have ignored the whole thread.

 

[RichG]

QUOTE (PeteN @ 22 Sep 2006, 15:57) <{POST_SNAPBACK}>I must assume the Scaley power (16v?) is a touch higher than the SCX (14.8v?) and the headlight circuit is just built with very little safety margin?

Why dont you read the thread Pete?

Rich

 

[RichG]

Good question Rob.

There is a bit of me that says you should all send them back to SCX and get them to do something. The board is not fit for purpose and my calculations show that I would expect it to fail in most circuits, not just higher voltage ones.

Does someone have a name or a contact at SCX? If so I will drop them a line explaining the problem. Although I feel they must know of it already, surely some people must return cars?

To your question, a simple solution. Probably first thing to try would be a series resistor. My concern is that by the time you had put in a suitable value to protect the board, the lights would be a bit dim. I will need to do some maths. An alternative might be a series zener, as this would drop a constant voltage.

When I get home at the weekend I will give some thought to a simple mod.

Rich