If the circuit is OK, than the operator is wrong. What exactly are you doing and what is the load that you use?

I use a motorcycle tail light bulb. Maybe a couple amps at 13.8 volts? The LM317 (TO-220 w/heat sink) will light it fully when the 4.7K resistor is lifted.

 

hi Jack,
I guess you know a 'cold' light bulb can draw upto 5 times its normal operating current, until the filament fully heats.?

E

 

hi Jack,
I guess you know a 'cold' light bulb can draw upto 5 times its normal operating current, until the filament fully heats.?

E

I do not think he knows that. And it makes sense, higher current with lower voltage.

 

hi Jack,
I guess you know a 'cold' light bulb can draw upto 5 times its normal operating current, until the filament fully heats.?

E
I do not think he knows that. And it makes sense, higher current with lower voltage.

Refer to my post #21. The LM317 can light the bulb by itself.

I never said, but this is a 20 amp supply.

 

A 20Amp supply with Lm317 seems so wrong from so many ways.

 

A 20Amp supply with Lm317 seems so wrong from so many ways.

Huh? Why is that? Did you even look at my schematic?

 

When you were checking and found the 0.6 V across the 470 ohm resistor, was there any voltage at all across either of the 0.1 ohm resistors?

Power transistors usually fail short-circuit, at least at first. If there is lots of power available, a shorted transistor will sometime then go open-circuit. It does seem rather odd that you should measure 0.6 V across the 470 ohm resistor - it suggests that the base-emitter junctions of the 3055s are OK but there is no current through them. Iif one or both were shorted, the voltage would be lower, if both were open it would be volts, and if there were any current through a 3055 it would be higher because of the drops across the 0.1 ohm resistors and the basic relationship between Vbe and collector current. Is there any chance there could be something as simple as a connection problem?

What voltages do you get if you hook up your light bulb and measure at each terminal of the each 3055? Try to measure right on the pins for base and emitter and on the can for the collector.

Be careful with the 2N2905 as the driver. It really isn't beefy enough to supply continuous drive for the 2N3055s if you want 20 amps at the output. The MJE104 is much better suited. It is obsolete (probably mostly because the package is obsolete), but finding a sub shouldn't be difficult. I can't find a full datasheet for the 104 on the web, but I likely have one in an old Motorola book.

The tail light bulb probably runs at something in the half to one ampere range when hot, and ten times that when cold. The LM317 would go into current limit when the bulb was first connected, but recover to normal operation as the filament heated up.

 

Hello,

In an old motorola book the replacement of the MJE104 would be an 2N5976.

Bertus

 

You will find by practice!

A 20Amp supply with Lm317 seems so wrong from so many ways.

Huh? Why is that? Did you even look at my schematic?

 

Hello,

In an old motorola book the replacement of the MJE104 would be an 2N5976.

Bertus

Thanks

 

Here's another schematic with voltage readings.
The top voltage is with no load on the output.
The voltage below it is with the tail light bulb connected.

 

First, I don't have data for the MJ104. It isn't even listed in the master index of my Motorola discrete semi databooks from mid 1970s. There is an MJE105.

I'm not sure I'm interpreting the schematic with the voltages correctly. Are the 0.125 & 13.8 values associated with the MJE104 measured between the emitter and the base? If they are, something is wrong there. The maximum voltage between them should be no more than about a volt, with the emitter positive with respect to the base.

Is the driver actually in the circuit the MJE104 or the 2N2905? If the 2905, I suspect it may be connected incorrectly. The emitter is the lead near the tab on the rim of the case. The collector is the lead diametrically opposite, and if the bottom of the case is metal there will be no insulating ring around it. The base is the lead between them. The pinout of the MJE should be ECB, left to right, pins pointing down, marked side up (some chance it differs, but this was the common pinout for that package).

 

First, I don't have data for the MJ104. It isn't even listed in the master index of my Motorola discrete semi databooks from mid 1970s. There is an MJE105.

I'm not sure I'm interpreting the schematic with the voltages correctly. Are the 0.125 & 13.8 values associated with the MJE104 measured between the emitter and the base? If they are, something is wrong there. The maximum voltage between them should be no more than about a volt, with the emitter positive with respect to the base.

Is the driver actually in the circuit the MJE104 or the 2N2905? If the 2905, I suspect it may be connected incorrectly. The emitter is the lead near the tab on the rim of the case. The collector is the lead diametrically opposite, and if the bottom of the case is metal there will be no insulating ring around it. The base is the lead between them. The pinout of the MJE should be ECB, left to right, pins pointing down, marked side up (some chance it differs, but this was the common pinout for that package).

The MJE105 is the NPN. The 104 is the PNP.
Yes. Those are voltages E to B of the 104.
That pinout for the 104, as you stated, is what I used.

I just double checked it. When I substituted the 2905, and then replaced it with the 104, I soldered the 104 in backwards. DUH!

It works again. How embarrassing?
Thanks everyone.
Jack