Need to get MG Electronics Power Supply repaired...or NOT!?

7812 voltage regulator test results:

w/ Power off and writing facing me

Black lead on the center pin, read lead on the left pin = 1.2mV and it trickled down to 0

Black lead on the center pin, read lead on the right pin = .5mV and it trickled down to a negative number

w/ Power On and writing facing me

Black lead on the center pin, read lead on the left pin = 30V

Black lead on the center pin, read lead on the right pin = 12.04v

I re-tested the D313 for continuity (after the unit has been turned on and off)

There is NO continuity between the base and the emitter or collector

There IS continuity between the emitter and the collector

I retested the corroded diode and it is ok.  The corrosion was giving me a bad reading previously but I scraped it off and the diode shows a voltage reading of .519v in one direction and nothing in the other.

All sounds good and normal except the D313. We need to test that one out of circuit and verify if it is good or bad.

At some point I mistyped Read when I meant to type Red. I mention it so that we don't start a new convention for what is the correct nomenclature.  Sorry about that. I always try to proofread but at 1:30 AM I may miss things.

John

John,

I have been following your steps in helping to solve this issue with great interest.  I wish that I could be of more help, but my days of being a technician and doing any real heavy device debugging are sadly for behind me.  I tend to fix broken equipment by buying a replacement.  To that end, my search skills have have found a couple of replacement power supplies (same brand - one for 99.00 and free shipping).  Let me know if you need any details.

I am (and always have been) in absolute awe of people who can fix broken gear and save it from the trash.  I seem to have a phobic thing that prevents me from opening up anything.  With a schematic and plenty of time, I have a fairly good chance of figuring it out.  But my running 'blind' skills are severely lacking.

Best of luck guys!  I will continue to follow along, and who knows, maybe I will come up with some way to help.

Hi Matt,

I tried following the thread and it's hard to follow, but it got me wondering, does slot car racing need such a supply?

It seems overkill, and I can see that it could fail through accidental misuse, or even damage the slot-cars if set accidentally to too high a voltage.

Probably a few amps is enough, if it is a 'scalextric' style slot car racing system.

HI Shabaz,

My problem isn't my purpose for having the supply.  My problem is that it doesn't work and I'd like to fix it.  I do not do slot car racing.

Thank you for the thought, though.

Matt

Hi Gene,

I'm just trying to fix this one for fun, I realize I can buy other power supplies.  I also wouldn't purchase another MG Electronics power supply out of principle now, they have neglected to return emails, neglected to provide contact information on their website, and flat out refused to honor warranties to their distributers.

Matt

John,

As you well know this isn't my area of expertise.  Too bad there is no schematic.  Some thoughts on things I didn't see but may have missed:

• have all of the electrolytic caps been checked
• can the transformer be isolated and checked
• if the caps, diodes, transistors, assorted ICs, and transformer are good then is it worthwhile to go through the circuit with an iron and touch up all the joints to make sure all the connections are good

Frank

Thanks for the input Frank, Shabaz, and Gene. I will keep your suggestions in mind.

John

Hi Matt,

Let's Continue:

Test the output transistors with the Analog meter by choosing one. Since I believe they are all in parallel we will look at all of them at the same time. With the unit unplugged and turned off. Put a meter probe on base of transistor I believe it is the yellow wire and test to the other two. Note meter readings. Switch probes and test in the other direction. Now leave out the base (Yellow) lead that was common in the first test and test both polarities between the other two ( Emitter and Collector).

Next test. Still using the Analog meter with the unit off. Test to see if there is connection between the negative of the big capacitors and the black (negative) banana plug jack on the front of the unit. Test to see if there is a connection between the negative of the large capacitor and the #4 pin of each of the LM741 ICs.

Now with the unit still off test to see if there is a connection between the Blue wires of the power transistors and the RED positive Banana JAck on the front of the unit. If not is there a connection between the yellow wires and the Red Banana Jack?

With the unit turned on and using your digital meter set on DC volts measure the voltage between the Black binding post on the front of the unit and the Collectors of the output transistors (Orange? RED? Wire I can tell the color clearly) Try turning the range selector on the front. We should see the same voltages that we saw when we tested the large capacitors.

With the unit turned on and the black lead of your digital meter still on the black banana jack check the voltages on the yellow and blue wires to the power transistors.

Let me know what you find.

John

You are getting some really good practice in using your meters.

Hi Matt,

I just wanted to let you know that I feel that we are making real progress. I have a preliminary concept of how they are making this circuit work. We hopefully are only a couple steps away from zeroing in on the source of the problem. You are doing an excellent job of interpreting my measurement requests. I understand the #3 pin reading on the lower Right LM741 and I believe we are good on the circuit through this 741.

John

If I'm measuring Resistance, then it's 15 Ohms between Yellow (base) wire on different transistors in both polarities (at X1)

Between Red to Reds it's also 15 Ohms

Between Blue to Blues it's also 15 Ohms

Between Red to Blues there is no reading

Affirmative continuity between negative on both big capacitors and negative banana plug

The continuity between the big capacitors and the LM741 is screwy.  It seems to change...

Negative continuity between negative on both big capacitors and #4 pin on each of the LM741 ICs, except that there was continuity for a second

Positive continuity between negative on both big capacitors and #2 pin on each of the LM741 ICs, and then there was no continuity

Intermittent continuity between the positive on both capacitors and some of the pins of each of the LM741 ICs

Positive continuity between the top row of rectifier diodes on the non-line side and the #2 pin on each of the LM741 ICs, then negative continuity

Negative continuity between the bottom row of rectifier diodes on either side of diode and the #2 pin on each of the LM741 ICs

Positive continuity between each of the blue wires on the power transistors and the Red Banana Jack

Negative continuity between each of the yellow wires on the power transistors and the Red Banana Jack

16.5V between Red wire of power transistor and black banana plug, then turning the front range selector

26.14V

36.9V

Black banana plug to yellow transistor wire (all were the same)

36.7V

Black banana plug to blue transistor wire (all were the same)

.6V

I wanted to thank you for your patience and thoughtful explanations!   I really appreciate it!! 

I wanted to thank you for your patience and thoughtful explanations!   I really appreciate it!! 

Hi Matt,

The voltages that you provided tell me that the Output transistors are likely failed open. I just need to confirm that the voltage on the blue wire was 0.6 Volts?

I recommend that you bite the bullet and remove one of the output transistors so we can test it and verify that it is bad. Don't order replacements without talking to me first. Once it is out use the analog meter to run the same test we just ran on them when they were in the circuit.

With 36 volts on the collector and 36 volts on the base the emitter should also be at approx 36 volts not 0.6 volts as this is an NPN transistor.

John

Confirm, I'm reading .6V between the blue wires and the negative banana plug

36V between Red and the negative banana plug

36V between Yellow and the negative banana plug

Actually, it's -.6V

Hi Matt,

Here are some thoughts on the Output Transistors. If the wires are long enough you might try pulling the wires through the holes in the heat sink and then desoldering them or just clip them off and make new connections to the new transistors. I believe that the coils of wire on the circuit board are emitter resistors. One blue wire should run to one end of each coil and the other ends of the coils should be connected to the red banana jack on the front panel. Confirm to me that this is the case. Otherwise we may need to install our own emitter resistors to help the outputs avoid thermal runaway. When you remount the new output transistors you should either use a teflon insulator or if you are going to use the original mica insulators silicone heat sink compound needs to be used. We want the heat from the 2N3055s to easily transfer to the heat sinks. I saw that Newark has some reasonably priced 2N3055s

https://www.newark.com/multicomp/2n3055h/bipolar-transistor-npn-60v/dp/78M7081?st=2n3055

Get yourself at least one extra just in case. Also you should be able to find the teflon heat sink insulators on Newark. I still want to see the results of your out of circuit test of the 2N3055 before we go ahead but I had some time so I jumped ahead to these ideas on replacing the 2N3055s.

John

Hi John,

The wires are not very long and there is no slack. It would be a delicate operation without much room for error to desolder and solder.  I would probably need surgical tools.

The blue wires go to the bottom of the circuit board then up to I believe the curly q resistance coils (I don't what they are called) then on to the resistor next to it.  From there it's a maze.

Working from the red banana jack back, it's tied to the front Amp Meter and then back to the circuit board (it's the thick red cable coming out of the middle of the board)

In my research, bad power supplies seem to have usual suspects:

If not blown fuses, loose wires, corrosion, broken traces, or cold solder joints then -

Typically goes out:

Transistors

Electrolytic Capacitors

Integrated Circuits

Diodes

Sometimes goes out:

Mechanical switches and potentiometers

Resistors

Non-polarized capacitors

Usually never goes out:

Transformers

Variable capacitors

Inductors

Also testing components in circuit could yield bad readings.

So I'm willing to order replacements for the ICs, transistors not yet replaced, the power transistors and whatever else you'd recommend.  The parts are cheap but shipping adds up.  I'll give replacing the components a go and if it works great, if not, I'm not sure the unit is worth any more trouble. 

Thanks,

Matt

Hi Matt,

At this point based on your voltage readings the Output transistors are definitely a problem. I would start by replacing them. If that doesn't solve the problem we can go back and look at secondary causes. There are two wires on each output transistor that need to be replaced. The collector is connected through the screws that mount the transistor. You could even cut the pins off the old transistors and then solder on a short piece of wire and seal the connection with heat shrink. This would give extra wire to come through the heat sink and solder on to the new transistor. This extra wire could then be pushed back through the heat sink when the transistor is remounted to the sink. Hold off on replacing anything else until we have done the outputs. If you want to take more detailed pictures of the difficulty replacing the outputs I will see if I can come up with more ideas to make it easier.

John

Hi John,

I have a thought.  If the output transistors (or is it pass or power transistors?) NTE130 or 2N3055H are giving a bad voltage reading in the voltage test, could it be that any component between the blue terminal and the red banana jack could be bad, (and not the output transistor) thus creating the low voltage?

For example, the blue terminal on the output transistors tested positive for continuity with pin 2 on one of the LM741 IC (the lower right one) but gave a weird electronic sound when testing for continuity with pin 2 on the top left LM741 IC.  If the LM741's function is the same on this linear power supply as the ICs on switching power supplies, as detailed HERE in the 6th paragraph down, (which regulates the output voltage of the transistors) then a bad IC could result in low voltage.

Thoughts?

Matt

Hi Matt,

Sorry but due to the physics of a functional transistor the voltages that you have measured are not possible. Therefore we must conclude that these are not functional transistors. there is an energy barrier of about 0.7 Volts between the Base and the Emitter. Any time that energy barrier is exceeded current flow in the BE circuit. This also causes the CE current to flow. I think the BE junction can handle a few volts for a short period of time but anything more than that would cause it to burn out. Your measurements indicate that there is over 36 volts across this junction and therefore it is apparently already burned out. Until these transistor are replaced and functional we can't further trouble shoot the circuit easily.

John

Hi John

So what do I need to order?

NTE130 or 2N3055H

Do I need thermal paste?

What type of LM741CN IC?  This one?  If so, it's backordered, is there a suitable replacement you'd recommend?

Ceramic capacitors? Diodes?  Resistors?

It cost $9.99 to ship a .10 cent component.   I'm already in for about $35 with the analog multimeter and the transistor.  I could buy a new unit with free shipping for $83.

It doesn't make sense to buy one component at time to see if it's a fix.

Thanks,
Matt

mpulliam  wrote:

...

It cost $9.99 to ship a .10 cent component.

...

For companies that have free delivery above an amount, I add consumables like solder, flux, wick, spare cutter blades ...

Your repair will never be economical. If you do it for the fun of learning, you'll have a good time. Even if it fails.

But it 'll never be worth it in $$.