Redesign of Floursent Light
Redesign of Floursent Light
I am trying to figure out a way to operate a floursent light that uses 2 AA batteries to run off of a 4 volt battery. I hooked up the 4 volt battery to the light and the only transistor on the board began to smoke. I looked all over the web for information on the transistor, but did not find much. What I did find was it is a audio amp transistor, somewhere around 2 watts. Thanks for any help.
Re: Redesign of Floursent Light
How can I operate a flourscent light built to run off of 2 AA batteries, using a 4 volt battery? This is just a little project I am working on using a 4 volt rechargable battery. I have a small solar panel that came with the battery for recharging. Thanks.
Re: Redesign of Floursent Light
j611:
Seems highly logical that if that unit was designed to run from 2 AA's.....3VDC that it would certainly tolerate 4VDC operation. Whats the chance that you miswired polarity when you connnected in the 4VDC into the circuit or came in at the incorrect points .This considering that you were not making connection to the old battery holders terminals ? Also on the xstr can you pass on all of its symbolization on to us, considering that overheating or fracturization did not leave it in a legible state.<p>73's de Edd
[email protected] ........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
Seems highly logical that if that unit was designed to run from 2 AA's.....3VDC that it would certainly tolerate 4VDC operation. Whats the chance that you miswired polarity when you connnected in the 4VDC into the circuit or came in at the incorrect points .This considering that you were not making connection to the old battery holders terminals ? Also on the xstr can you pass on all of its symbolization on to us, considering that overheating or fracturization did not leave it in a legible state.<p>73's de Edd
[email protected] ........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
Re: Redesign of Floursent Light
<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>Originally posted by Edd Whatley:
j611:
Seems highly logical that if that unit was designed to run from 2 AA's.....3VDC that it would certainly tolerate 4VDC operation. Whats the chance that you miswired polarity when you connnected in the 4VDC into the circuit or came in at the incorrect points .This considering that you were not making connection to the old battery holders terminals ? Also on the xstr can you pass on all of its symbolization on to us, considering that overheating or fracturization did not leave it in an illegible state.<p>73's de Edd
[email protected] ........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
<hr></blockquote>
j611:
Seems highly logical that if that unit was designed to run from 2 AA's.....3VDC that it would certainly tolerate 4VDC operation. Whats the chance that you miswired polarity when you connnected in the 4VDC into the circuit or came in at the incorrect points .This considering that you were not making connection to the old battery holders terminals ? Also on the xstr can you pass on all of its symbolization on to us, considering that overheating or fracturization did not leave it in an illegible state.<p>73's de Edd
[email protected] ........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
<hr></blockquote>Re: Redesign of Floursent Light
Hi Edd,
The wires were connected properly. The light was on and the transistor started smoking. The markings on the transistor are C2328A and below that Y-932. Thanks for your help.
The wires were connected properly. The light was on and the transistor started smoking. The markings on the transistor are C2328A and below that Y-932. Thanks for your help.
Re: Redesign of Floursent Light
Well looks like that xstr is a 2SC2328A , a NPN sil epitaxial unit in a stretch TO-92 case. 30VDC Collector voltage max. 2A max collector current. Possibly, 932 is, the date code of the device, being the 32nd week of 1999 for the xstr propers manuf. The Y is indicative of a selected unit in the 160-320 Beta (DC gain) range. A bit higher than usual for this power output level of device. Unfortunatelly though, the case configuration precludes a good thermal heatsinking of the semiconductor devices die. A equivalent replacement device in a TO-5 or TO-39 metal casing would permit infinitely better heatsinking (cooler running) of the unit, as well, so would the tab on a TO-220 encased unit.
Since this is divided between 2 posts. I believe your final situation was the device flaming, thus precluding the acquisition of operational current consumption. Also check for damaged/drifted low value associated resistors in its support circuitry.
Back initially, on an operational unit, my choice would have been the same as Jolly Rogers, the incorporation of a properly polarized, in series pair of diodes, but my preference would have been the 3A rated 1N5400 series of diodes and then cut back to only one diode if it proves to be permissible on current consumption.
BTW any chance that your unit is the small palm sized unit that Radio Shack sells. I got one for a buck at an electronics flea market but it was working…didn’t have to fix it except for cancerous FeCuSo4’ed batt connector springs…so I’ve never had reason to explore it’s innards. Makes a great reading light.
The main manuf of that xstr is Unisonic but looking at the Coll curr spec /Device wattage/Beta at those levels looks like a NTE 2363 might work, staying in that casing. This potentially also being an easier to acquire device from common electronic suppliers.
Also, if the unit is that RS unit I might be able to assist U more by researching my unit.<p>73's de Edd
[email protected] ........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
Since this is divided between 2 posts. I believe your final situation was the device flaming, thus precluding the acquisition of operational current consumption. Also check for damaged/drifted low value associated resistors in its support circuitry.
Back initially, on an operational unit, my choice would have been the same as Jolly Rogers, the incorporation of a properly polarized, in series pair of diodes, but my preference would have been the 3A rated 1N5400 series of diodes and then cut back to only one diode if it proves to be permissible on current consumption.
BTW any chance that your unit is the small palm sized unit that Radio Shack sells. I got one for a buck at an electronics flea market but it was working…didn’t have to fix it except for cancerous FeCuSo4’ed batt connector springs…so I’ve never had reason to explore it’s innards. Makes a great reading light.
The main manuf of that xstr is Unisonic but looking at the Coll curr spec /Device wattage/Beta at those levels looks like a NTE 2363 might work, staying in that casing. This potentially also being an easier to acquire device from common electronic suppliers.
Also, if the unit is that RS unit I might be able to assist U more by researching my unit.<p>73's de Edd
[email protected] ........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
Re: Redesign of Floursent Light
Hi Edd,
Yes, this is the Radio Shack small hand held flourescent light. I bought all I could find in my area, especially at such a cheap price of $ 1.98 each. Any help getting this 3 volt unit to work using 4 volts will be greatly appreciated. Again thank you Edd for your help.
Yes, this is the Radio Shack small hand held flourescent light. I bought all I could find in my area, especially at such a cheap price of $ 1.98 each. Any help getting this 3 volt unit to work using 4 volts will be greatly appreciated. Again thank you Edd for your help.
Re: Redesign of Floursent Light
J611:
Sorry so long on the come back, couldn’t find where I last left that unit of mine,
that is the same as your now defunct unit. Hidden in bedside night table drawer!
Exposing the units electronics, there is NO spare space in that unit ! A look at the circuit components and a trace of foil circuitry paths reveals it to be just a single xstr pwr osc ckt. With the ¾ in sq dub’l E ferrite xformer on one end feeding the U bent fluorescent lamp. Measurement of its pri was 120uh ctr tapped with a 75mh secondary, single winding for the tube ionizing and drive voltage. You don’t have to worry about any low value res being fried as I mentioned, without knowing what U had. Seems a 22 ohm is the lowest value on the board, hardly a ½ w at a constant/full 3v dump into it. A check on current consumption with a solid 3v on the unit via a DC ammeter in series with the batteries revealed a current consumption 1 ¼ a. Getting a base drive figure, this computed the Beta of the xstr as being 125…..nothing slouchy for the low voltage at that level of coll current. Yet shy of that manufacturers code Y designated Beta spread. They must have expected and planned for it, as at the far end of the PCB there is VR1, a 500 ohm trim pot. It is configured as a 2 term rheostat, vice divided potentiometer. I took a light output measurement and found that the light output diminished only 6% when I adjusted the current consumption down to 1a with VR1. Further dropping revealed at least a 600ma power consumption to successfully fire the tube on turn on. I’d definitely cut back and run it at that 1a (with 3v batteries) level just from the reliability aspect of that 2SC2328A xstr. Thinking of your smokimus maximus xstr, I did try at least 10 different TO-202, TO-212, TO-220 cased npn darlington types of xstrs with one of the TO-202’s running the coolest…but, casing too big. Just for curiosity I also tried using IRF540 and IRFZ44 FET’s but since the feedback winding on the xfmr was design oriented to a current mode bipolar utilization that drive level was inadequate. I was able to successfully skirt that deficiency by utilizing a 2222 npn and 2907 pnp connected in a super darlington configuration for FET gate drive , but NO way would that cluster go into the allotted/avail space. Strangely an old design ECG/NTE 48 in the jumbo TO-92 casing (same as 2sc) ran close to the same temp at 1a, that the 2SC did at 1 ¼ a. So that’s an available 99 cent number that one could use. The last thing I tried was some xstrs used on some batt pwrd neuro-surgical medical electronics equipment I have worked with. They were in SOT223 pwr casing with about an 1/8 in heat sink tab exposed. Pig tail leaded and put into circuit reveals the unit having a Beta of 190, required VR1 at its max res to get down to 1a…the case temp on the tab felt about the same as the 2SC..with its thermally insulated case around its die. Then I soldered a 3/8 in wide by 3 in long .005 in Cu foil strip to the tab…that totally cools it ! Wrapped the foil with Kynar tape both sides and it now nestles behind the dual parabolic hot pressed silver mylar reflector of the tube. The final xstr choice was a Zetex brand FZT651TC , seems they have a forte in their low voltage bipolar high Beta darlingtons with an all important low Vce(sat). Yes I did test the unit with the 2SC2328A in circuit at your requested 4v power level but natch U need to adj VR1 to then have only 750 ma as the current pulled, for a total 3w consumption. I even scaled the current down further and made short runs at 5 and 6 v on the supply. Overall on your solar fed 4 battery, I’m not to enthusiastic on your run time of the lamp at a 3w drain level on the batt. Unless you have one heftier SC & batt than I expected. BTW, got one good used 2SC2328A ,radial leads perfectly laterally preformed for your downed units specific application… if you want it…gratis. Just hit me at the Comcast addee with ur snail mail addee.
( Micro soldering expertise desirable) But its access is easier if one swings that riveted brass lug on the mylar ground 90 deg and temp lift that black ground wire at the emitters pad for easier access to the 2SC…. <p>
73's de Edd
[email protected] .........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
Sorry so long on the come back, couldn’t find where I last left that unit of mine,
that is the same as your now defunct unit. Hidden in bedside night table drawer!
Exposing the units electronics, there is NO spare space in that unit ! A look at the circuit components and a trace of foil circuitry paths reveals it to be just a single xstr pwr osc ckt. With the ¾ in sq dub’l E ferrite xformer on one end feeding the U bent fluorescent lamp. Measurement of its pri was 120uh ctr tapped with a 75mh secondary, single winding for the tube ionizing and drive voltage. You don’t have to worry about any low value res being fried as I mentioned, without knowing what U had. Seems a 22 ohm is the lowest value on the board, hardly a ½ w at a constant/full 3v dump into it. A check on current consumption with a solid 3v on the unit via a DC ammeter in series with the batteries revealed a current consumption 1 ¼ a. Getting a base drive figure, this computed the Beta of the xstr as being 125…..nothing slouchy for the low voltage at that level of coll current. Yet shy of that manufacturers code Y designated Beta spread. They must have expected and planned for it, as at the far end of the PCB there is VR1, a 500 ohm trim pot. It is configured as a 2 term rheostat, vice divided potentiometer. I took a light output measurement and found that the light output diminished only 6% when I adjusted the current consumption down to 1a with VR1. Further dropping revealed at least a 600ma power consumption to successfully fire the tube on turn on. I’d definitely cut back and run it at that 1a (with 3v batteries) level just from the reliability aspect of that 2SC2328A xstr. Thinking of your smokimus maximus xstr, I did try at least 10 different TO-202, TO-212, TO-220 cased npn darlington types of xstrs with one of the TO-202’s running the coolest…but, casing too big. Just for curiosity I also tried using IRF540 and IRFZ44 FET’s but since the feedback winding on the xfmr was design oriented to a current mode bipolar utilization that drive level was inadequate. I was able to successfully skirt that deficiency by utilizing a 2222 npn and 2907 pnp connected in a super darlington configuration for FET gate drive , but NO way would that cluster go into the allotted/avail space. Strangely an old design ECG/NTE 48 in the jumbo TO-92 casing (same as 2sc) ran close to the same temp at 1a, that the 2SC did at 1 ¼ a. So that’s an available 99 cent number that one could use. The last thing I tried was some xstrs used on some batt pwrd neuro-surgical medical electronics equipment I have worked with. They were in SOT223 pwr casing with about an 1/8 in heat sink tab exposed. Pig tail leaded and put into circuit reveals the unit having a Beta of 190, required VR1 at its max res to get down to 1a…the case temp on the tab felt about the same as the 2SC..with its thermally insulated case around its die. Then I soldered a 3/8 in wide by 3 in long .005 in Cu foil strip to the tab…that totally cools it ! Wrapped the foil with Kynar tape both sides and it now nestles behind the dual parabolic hot pressed silver mylar reflector of the tube. The final xstr choice was a Zetex brand FZT651TC , seems they have a forte in their low voltage bipolar high Beta darlingtons with an all important low Vce(sat). Yes I did test the unit with the 2SC2328A in circuit at your requested 4v power level but natch U need to adj VR1 to then have only 750 ma as the current pulled, for a total 3w consumption. I even scaled the current down further and made short runs at 5 and 6 v on the supply. Overall on your solar fed 4 battery, I’m not to enthusiastic on your run time of the lamp at a 3w drain level on the batt. Unless you have one heftier SC & batt than I expected. BTW, got one good used 2SC2328A ,radial leads perfectly laterally preformed for your downed units specific application… if you want it…gratis. Just hit me at the Comcast addee with ur snail mail addee.
( Micro soldering expertise desirable) But its access is easier if one swings that riveted brass lug on the mylar ground 90 deg and temp lift that black ground wire at the emitters pad for easier access to the 2SC…. <p>
73's de Edd
[email protected] .........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
Re: Redesign of Floursent Light
Why not just use a resistor in series with the supply to create a voltage drop?
Re: Redesign of Floursent Light
Extenuating circumstances…. 
He had opened the emitter of the units power xstr by exceeding the power/op temp spec of that unit with a 4V operation. Since the unit will have to be opened for xstr replacement , with the info I researched, might as well set the trimpot for 1 amp consumption at 4 V operation, thus precluding the lossy series pwr resistor dissipative loss ( this being fed from a solar pack and storage batt).
However, starting with an operational unit, a series .8 ohm/cool 2 watt metal film for a 1.25A feedthru or a 1 ohm for a more conservative 1 A pass.<p>73's de Edd
[email protected] .........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
He had opened the emitter of the units power xstr by exceeding the power/op temp spec of that unit with a 4V operation. Since the unit will have to be opened for xstr replacement , with the info I researched, might as well set the trimpot for 1 amp consumption at 4 V operation, thus precluding the lossy series pwr resistor dissipative loss ( this being fed from a solar pack and storage batt).
However, starting with an operational unit, a series .8 ohm/cool 2 watt metal film for a 1.25A feedthru or a 1 ohm for a more conservative 1 A pass.<p>73's de Edd
[email protected] .........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
Who is online
Users browsing this forum: No registered users and 45 guests