Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Transistor for Flourescent Ballast

Status
Not open for further replies.

KipHansen

Newbie level 5
Newbie level 5
Joined
Sep 29, 2008
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,398
thin-lite transistor

Folks,

I have a number of 12 VDC flourescent lighting fixtures on my boat, which my wife and I use as a home in the Dominican Republic, where we are humanitarian relief missionaries.

After many years, these are failing one after the other. The fault is the "transistor" - looks like a power transistor, flat, black, bolted to the case, three leads. I have tried to have them repaired, but the local electronics wizes cannot identify the transistor.

It is marked

1 85
(a small circle with Malaysia 55 embossed in it)
THINLT
R9024

Seems a shame to fail to repair them. The CB only has eight parts! This unidentified transistor, a "transformer", two brown 'chicklet' looking things (capacitors or resistors - marked .01K 630 F and .22 K 250 U), a blue 'can' capacitor (52-8915), a heavy-duty ceramic 5W 1K resistor, and 2 diodes.

Can anyone help me identify the transistor?

It would seem that given the function - a flourescent 12 VDC 'ballast' - for running 2 F8T5 mini-tubes -- that some smart guy ought to be able to guess what this MUST be.

Thanks for any help you can be,

Kip Hansen
 

flourescent ballast kip hansen

Hi,

Unfortunately the marking on your transistor is unknown to me.

Assuming it is a bipolar transistor (and not a power MOSFET) you may simply substitute it with a hefty bipolar transistor for a test try.
You may be able to obtain this transistor from digi-keys:

**broken link removed**

its data sheet is here https://www.fairchildsemi.com/ds/KS/KSC5305D.pdf designed for light ballast circuits, it includes a diode between the collector and emitter.

If you cannot obtain it, then you may ask your local electric wizards for getting at least 200-300V collector breakdown voltage bipolar transistor with 3-4A current and in similar package like the faulty transistor.

As a first test you may try TIP3055 (flat, black) whether the circuit works at all with a bipolar transistor but this type is only with 60-70V collector so may not work on the long run, but probably it is easier to obtain locally near to you?

You connect the three legs in the same order you see at the faulty transistor legs.
(most probably from left to right it has base, collector, emitter)

If any try with bipolar transistor is a failure and the lamp remains dark then maybe you need to replace the transistor with a power MOSFET. At www.irf.com International Rectifier you may find easily types with 300-400V drain source voltage and also at least 3-4A drain current ratings (like IRF730 or IRF740 etc).

rgds, unkarc
 

kip hansen floursecent ballast

I found some cross referencs to IR parts.

It could be IRFR9024N.

http://search.digikey.com

Your local wiz-guy should be able to tell between a bipolar transistor and a FET with only a DC volt-meter. If he has an oscilloscope then even better.

If you Google: fluorescent 12 VDC ballast circuit

You can find schematics for both MOSFET and BiPolar devices. Example: http://www.aaroncake.net/CIRCUITS/flampdrv.asp

If you can send a good high resolution picture of the PCB (both sides) we may be able to figure it all out.

Good luck :)
 

world mosfet subsitute on net guide

I would suggest trying a low VCE NPN transistor. With a Vceo rating of at least 60V and a Veb rating of at least 5V.

The circuit is most likely a self oscillating circuit - most are, and most use npn transistors to operate.

It would be helpful to know the package size, otherwise you run the risk of incorrect pin configuration of the transistor.

Something like this may be worth trying.

https://www.fairchildsemi.com/ds/KS/KSD73.pdf

oddbudman
 

15watt cfl inverter using d882

Dear oddbudman,

Well, after year, I am back on this project.

Here are photos of the circuit and the components.

10Dec2008_002.jpg

and
10Dec2008_001.jpg


The black wire is (+) and the white is (-). The two red wires go the pins at one end of the tube and the blues to the pins at the other.

10Dec2008_003.jpg



I have more detailed images of the capacitors etc if you need them.

I found another non-functioning fixture, a Thinlite, whose transistor is unhelpfully marked THINLT 0130.

Thanks for your help and advice,

Kip Hansen
 

thinlt

Ah! "REC" is on the reverse side of the printed circuit board and "REC" is also on the transistor.

The maunfacture has private labeled all of the parts to their own code. I found Thin-Lite in California and this is their website where they offer new ballast: **broken link removed**

Most likely they're making a business out of replacing burned out parts, otherwise there would be no point to re-label the parts with their private part numbers.

Later, (I have to go right now) I'll reverse their circuit and select a new part for you.
 

patent us20070076386

Thanks, I sure appreciate it!

Kip
 

ballast transistor

The transistor in the photo is having a Ring washer over that. Isnt the number market below that? As per the size and the circuit, you may use D882 or D313 of the same size or D1061 or MJE3055 which are generally used in CFL and flourescent Tubes Good luck
 

thinlt r9024

Hi Kip,
Firstly I suggest you try to find out the transistor type NPN or PNP. Hopefully on one of the unit you have you can still find a good junction (BE or BC) on the transistor. Once you have determine the type then selection for substitute becomes easier.

At home I opened up a commercially made 12 VDC operated 20W fluorescent lamp ballast. Inside it I can find an NPN 2SC1061 transistor.

I hope this helps.
 

pspice model ballast

Thanks for the help.

I will try both suggested transistors and see what works.

If anyone can sketch out the circuit and fill in parts values, that would be great.

I have closer images of the parts if needed.

Kip

PS: Thinlite expressly refused to supply a circuit diagram when asked - and suggested I find a distributor and buy replacement ballasts with the following statement
"We do not recommend field repair. In most cases the transistors are not the only part that needs replacing.
You have several options:
1) Send fixtures without lamps or lens to Thin-Lite for repair and warranty renewal.
2) Purchase ballast kits from below listed Dist. Or your local RV store.
3) Purchase new fixtures from Distributor.
"
 

85 watt fluorescent ballast schematics

Kip:

Sorry to be delayed so long, I had to repair a friend's computer that had a bad hard drive controller.

I started reversing the PCB today and I need some help.

I'm creating a micro-Cap simulation scehmatic. Once we get all of the parts and connections entered we can confirm the function by simulation and determine why the inverters fail after only a few years.

First for your information, the T-220 transistor is mounted on the solder side of the PCB, if you wipe off the thermal grease there is a chance the true component number is hidden under the grease. Wipe of the grease carefully otherwise you may wipe off label's epoxy paint too. If it's blank, then Thin-Lite scrubbed the original part number.

On the PCB, I need to see the solder side more clearly, rosin is blocking a few places such that I can not be sure if there is a connection or a open area. Please use some hot water, soap and a brush to remove the rosin. Getting the PCB wet is not a problem, when finished just put it some where to dry and it'll be fine. Then photo the bottom again and post it here.

Also on the transformer, if you have a ohm meter and the skill to determine the internal wiring; that would be a great help. As it is now, I'm using hints of the wiring order of some coils I can determine from external connections and the wire guage. But I'm not completely certain of my guesses at this point.

In the attached PDF all the components are guesses, if you can determine two red colored capacitors' values, that would be a great help too!
 

how to you open a thin lite in a boat?

Tanuki,

This is great.

I am inserting the urls for a few close-ups of the caps. The values are

Big blue cap = 220 uF 35v
Big brown chicklet = .22k 250
Small brown chicklet = .01k 630

Grey ceramic resistor is marked = 5W 200 10%

The diodes are diodes :D

That's it besides the transistor. That sneaky old unidentified Q1!

Unfortunately, the circuit boards are on my boat, and I am in the mountains...give me a couple of days to retrieve one of them. I don't remember there being any other markings on these (I have several of these boards). I will check the other side and post the data, if any. My recollection is the heat grease side is blank. Once I have a copy of the board to hand, I can determine at least which wires in and out of the coil at the same loop!

I believe, contrary to Thinlite's rep's reply that the transistors are probably the culprits (eg: burnt out) ....but have only the opinions of the above experts to guide me to select replacements. I DO have a working copy of the board, but don't know how to check transistor types with a multi-meter.


caps_1.jpg

caps_2.jpg

caps_3.jpg

transistor.jpg


I will retrieve one of the PCBs and clean the rosen off to make that part clearer.

Stick with me here, I am thrilled to have someone knowledgeable helping me with this.

Kip
 

transistor role in cfl pcb

Found this patent issued to Thin-Lite for a switchable two lamp system which has the schematic for your PCB. There are several unloaded components on the PCB and most likely this schematic is the full monty. From what I've reversed so far, it looks nearly identical once the second lamp and it's associated circuits are logically removed. Patent US20070076386 is also interesting and related to other larger wattage lamps.
 

description of ballast transistor

You are absolutely awesome!

The unit I have been photographing is for one 15 watt tube, whereas the one in the patent is for two 8 watt tubes, but essentially the same.

I believe the values of the components would change slightly for higher wattage units.
See image of 30_watt_unit:


30_watt_unit.jpg


As for the one 15 watt tube PCB we have been looking at, as you requested, here are photos with the clear sealant and rosen removed. Some silhouette photos help clarify.

These may all be extraneous now, with the circuit from the patent. The Q1 is probably the same in all the units - an N-P-N Power Transistor such as 2N3055 ? with the differences being in the coil? Does the transistor have to be "stronger" to power higher wattage bulbs? I have units of 8 watts, 16 watts, and 30 watts.

Anyway, here are the photos. I will have a guy put in a 2N3055 or equiv. and see if it works!


wires_end.jpg

transistor_end.jpg

center_of_bd.jpg

component_side.jpg

PCB_foil_side.jpg

similar_unit.jpg


I never would have thought to search the patent office for the circuit diagram. Brilliant!

Kip
 

transistor marking kip

Great, the new photos take all of the guessing out of it. Backlighting worked really well. I've got the circuit completed now.

The NPN transistor pinout is Pin 1: Base, Pin 2: Collector, Pin 3: Emitter. The emitter is connected to white wire (i.g. ground). Collector to the transfomer primary coil. Base to the lamp heater and clamp diode.

The first photo above where you mention RF suppression, yes and more. The extra capacitor is a snubber to limit the transformer's primary leakage inductance fly back voltage, which reduces the high voltage stress on the transistor. If left unsnubbed there will be increased RF emissions too.

What is the value of the snubber capacitor?

In the micro-cap simulator there are many high frequency spurious oscillations that I had to suppress with a addition snubbers, otherwise the simulator aborts. In the real world the transistor aborted too!

As built, the inverter circuit doesn't work very well ... as evidenced by their rapid failure rate.

If you have a ohmmeter, what is the resistance across the lamp heater pins?

Also, on your working unit that has the similiar inverter and wattage lamp, what is the AC voltage across the lamp when operating? Take Care!

One last item, if you have a ampmeter, when the inverter is operating, what is the supply current to the inverter (black wire) and the supply voltage?

With these values and guessing that the inverter operates around 40KHz, I can tune the transformer parameters to a reasonable guess. Once that is done, then the micro-cap will tell us what's happening and why the units fail prematurely and what we can do to fix it! My guess is adding the snubber that your one unit already has!
 

lamp ballast power transistor

Tanuki,

Here's the best I can do:

1. What is the value of the snubber capacitor? .033 K 630

2. What is the resistance across the lamp heater pins? If we mean resistance across the pins of the lamp itself, when not connected to the circuit, it is between 20 and 30 ohms. Between the two red wires at the lamp socket, 10 ohms. Between the two blue wires at the lamp socket 150 K

3. Also, on your working unit that has the similar inverter and wattage lamp, what is the AC voltage across the lamp when operating? Take Care!
I do not have a working model of the same wattage here in the mountains (again). I plan to visit the boat later in the week again, and can bring a single tube working model up for these measurements.

4. One last item, if you have a amp meter, when the inverter is operating, what is the supply current to the inverter (black wire) and the supply voltage?
I do not have a working model of the same wattage here in the mountains (again). The supply voltage comes from the battery bank on the boat, and is ranges from 11 volts (batteries very low) to almost 14 volts (chargers powered by shore line, wind generator, or onboard gen set going full blast). As above, I will bring a working unit for this measurement later in the week.

5. "guessing that the inverter operates around 40KHz,"..Note that the patent specifies "between 15 and 25 kHz, preferably about 20kHz."

The snubber cap only exists on the 30 watt units (which use larger tubes), but again, neither of the 30 watt units failed, only the 8 and 16 watt units failed.

Should I go ahead and put in a new 2N3055 (or equiv.) on the single-tube failed unit that I have and see if it will work? Then I could do the other tests....:)

Kip
 

cfl schematic why they fail

Kip

Thanks for the pointer on the patent, I didn't read it just looked at the pictures ... 20KHz will be my target frequency then.

One other thing, when the lamps fail are they operating or do they fail to start.

If they were operating then it was most likely overheating of the power transistor. If the lamp failed to start then it's most likley a over voltage on the transistor's collector-to-emitter voltage during the high flyback voltage during resonance mode start up.

I'm busy for a few days so it will be the weekend before I can continue on this.
 

code model dll 2n3055

Tanuki,

Thanks, I have taken the one unit to the electronics shop and asked him to throw in an NPN power transistor and see if it goes.

Truthfully, I can't remember if they fail when ON or, more likely, when TURNED ON for the millionth time.

I'd be interested if your program can suggest an NPN transistor by number.

I should hear from the shop on Monday, and I'll let you know the result.

Kip
 

snubber self oscillating ballast

Try D313 or D882 or MJE3055. Good luck
 

d882 ballast

Kip:

"I'd be interested if your program can suggest an NPN transistor by number"?

No, Micro-Cap is circuit simulator (SPICE); the user has to select the specific components from a limited library or make their own SPICE model of a part.

I think many NPN transistors will work, at least for some time, as the original OEM transistor did.

I want to solve the initial design problem that causes the failure of most of the units, as you described initially: "After many years, these are failing one after the other". In a properly designed circuit this should not happen; it should go to the landfill still working!
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top