Replacing Voltage Regulators

[mtl777]

Hi Smell (Can I just call you that? )

How's it going with your PSU build? I'm really interested in it, even though I'm still not competent enough to do it. It will be a very good project for the future when I am skilled enough.

BTW, would you have time to answer my question on whether it's alright to replace the caps C14, C17, C22, C60, C18, and C12 with Panasonic FM or FC?

Thanks!

 

[Groundloops]

Moin (as we say here)! Call me smell if you like, I don´t bother. It´s absolutely alright to replace the caps as you´ve written.

PSU build is proceeding, but it takes time. I´m no electronics pope, doing my layouts by hand (no CAD skills). The PSU will fit into the case when any of the original parts is desoldered. Due to limited space it is not trivial. The regulators will be fitted on a aluminum bracket and this will be bolted to the chassis. Thermal grease between bracket and chassis will give good heat dissipation.

The transformer will be replaced by a one with dual 18 V secondaries. I´m not shure yet about how to connect it, but I will find a way.

If you have some patience and are willing to pay the parts and postage, I will build another one for you.


Cheers,

Gerd

 

[mtl777]

Hey guys, I need your help, all the help I can get! Something has gone wrong with my 1010. A few days ago, when I turned the system on, I noticed some static noise on output #4 even though nothing was playing through it. I took voltage measurements on the output and got erratic readings that varied anywhere from 0 to 10VAC! I turned the system off, hoping that the problem would later go away. Then yesterday, when I turned the system on again, another channel, output #5, started having the same problem. I immediately turned the system off. I am now afraid to turn this thing on, for fear that still another output might get damaged!

All I had done so far was to replace the four 1N4001 rectifier diodes in the PSU with MBR1100 Schottky type. That's all. I just wanted to see what the input voltage to the +/-15V regs would be after that change. I found that the input voltage increased from 23V to 24V (a 1V increase). So far so good. I tested all the inputs and outputs, played some music, made some recordings, and everything seemed to be working fine. But days later, when I turned the system on again, it started manifesting this problem. I thought it was already working great, but now, why the sudden deterioration?

Is it possible for the simple change of diodes to destroy the +/-15V regs? Could there be an incompatibility between the MBR1100 diodes and the regs? If the regs were damaged, what other components are likely to be the first to go? I hope not the opamps or DAC chips!

The brand of MBR1100 that I chose for the replacement was the On Semiconductor. It has the following characterstics:

Replacement MBR1100 - On Semiconductor

Maximum Ratings
Peak Repetitive Reverse Voltage ... 100V
Average Rectified Forward Current ... 1A
Non-Repetitive Peak Surge Current ... 50A
Junction Temperature Range ... -65 to 175C

Electrical Characteristics
Max. Instantaneous Forward Voltage ... 0.69V @1A 100C, 0.79V @1A 25C
Max. Reverse Leakage Current ... 0.5mA @25C
Typical Junction Capacitance ... 77pF
Max. Voltage Rate of Change ... 10V/nS
Typical Forward Voltage ... 0.47V @1A 150C
Typical Reverse Current ... 0.02mA @20V 125C

Do you see anything in these data that would make the On Semi a bad choice for the application? Here's the data of the original 1N4001's for comparison:

Original 1N4001 - General Semiconductor

Maximum Ratings
Peak Repetitive Reverse Voltage ... 50V
Average Rectified Forward Current ... 1A
Non-Repetitive Peak Surge Current ... 30A
Junction Temperature Range ... -50 to 175C

Electrical Characteristics
Max. Instantaneous Forward Voltage ... 1.1V @1A
Max. Reverse Leakage Current ... 0.005mA @25C
Typical Junction Capacitance ... 15pF

What should I do now? Please help me with ideas on how to fix this problem!

Your help is greatly appreciated. Thanks!

 

[AndrewT]

Hi,
is there any possibility you have inserted one/some of the diodes back to front?

 

[mtl777]

I was very careful to follow the correct orientation of the diodes. I checked and rechecked many times, so I'm absolutely sure they were put in correctly. Is there any possibility that some diode's cathode band is marked incorrectly or in reverse due to an error in the manufacturing process?

If one of the diodes I installed was marked incorrectly, wouldn't my unit just not work at all, and I would notice a problem immedately? But my unit worked fine initially, the input and output voltage readings were fine, the input voltage ripple even got reduced to 2V or less.

Some other possibilities: Are the diodes causing too much heat in the PSU due to the increase of 1V in the input voltage to the regs? Are the diodes inducing a strong surge of current whenever I turn the system on that is damaging some components?

Thank you so much for replying!

 

[Groundloops]

Extremely unlikely that the diodes are marked incorrectly; never heard about such a thing. If one of the diodes was reversed, the voltage doubeler simply would not work and give no voltage at all.

The voltage increase does no harm as the maximum input voltage of LM7815 is rated to 35 V when output is and you are still well below. Too much heat would damage the capacitors, but not immediately.

Is it possible that you shorted a psu cap to something when swapping the dodes?

 

[mtl777]

I checked for shorts and didn't see any. I'll check again just to make sure.

Comparing the data of the original 1N4001 and the On Semi MBR1100 that I used as replacement, do you see anything in the data of the On Semi MBR1100 that indicates possible incompatibility or other problems?

Theoretically, are Schottky diodes supposed to help reduce ripple and make for cleaner power in this application? I already had 23V input voltage before I put the Schottky's and that seems to be enough voltage already, right? So my only purpose for using the Schottky's is to improve the quality of the power, if they indeed can help with that. What do you think?

Thanks!

 

[CBRworm]

Maybe put the original diodes back and see what happens.

I also have a Delta 1010 and have not had any problems with it. I guess it is a little late now to suggest not modifying it - that M-audio probably knew what they were doing when they made it.

Maybe it is a glitch not related to your mod. I would put everything back to stock and try it again. Worst case - start looking on e-bay for another breakout box.

 

[mtl777]

Hey guys, I'm back! I put back the old 1N4001 diodes. It didn't make the static noise go away. I guess the damage had already been done to the outputs. Well, at least the damage has stopped and there are no additional channels that are getting messed up.

I tried replacing all the non-SMD caps in the damaged outputs #4 and #5, as well as in outputs #3 and #6 as they are part of the stereo pairs 3/4 and 5/6. Result: The noise is gone on output #4, but the output level of this channel is now low, like 7 dBFS lower than normal (compared to the level of output #3). No luck with output #5, though, as the noise is still present there.

Any idea what components could have been damaged in these outputs? I don't think it's the DAC because it's still playing audio. I hope it's not the opamps either. How likely would opamps rated for 36V maximum supply voltage be damaged in this scenario?

I hope it's just one of those SMD caps and/or resistors. I think I'm gonna do a "shotgun approach" and replace all of them SMD caps and resistors in the path of outputs #4 and #5, and see what happens.

Hi Gerd, since you have a 1010, I hope you could help me with the specs and what I need to know to be able to order the correct replacement parts for those SMD caps and resistors. Or anybody who would like to help, please give me at least an idea of the key words to Google based on the SMD caps and resistors that you see in the following picture:

An externally hosted image should be here but it was not working when we last tested it.

Notice that the SMD caps are unmarked but they all look exactly the same. Are they all the same value, type, etc.? I hope they are all identical as that would greatly simplify things, for then I would only need to know the details of one, and the same info would apply to all.

The SMD resistors have numbers. What do the numbers mean? For example, what does the number 1910 on resistor R107 mean?

Gosh, I'm so clueless I wish someone could help me! A link to an article or tutorial on SMD components would be of great help. Any info you can share would be greatly appreciated!

Thanks in advance!

 

[AndrewT]

Hi,
I think the codes on SMD are just that.
It allows the manufacturer and builder/assembler, if he sticks with one supplier, to differentiate between components but gives us little clue to what is being used.
If you can find the manufacturer, I'm fairly sure they will give you a listing of the component codes.

 

[CBRworm]

Is it possible that heat from soldering around the regulator area could have caused a cold solder joint or break at a nearby component? I would look for dull solder, or possibly a solder bridge somewhere.

You could also try and figure out the specs for the D/A converters and see if upping the voltage could have caused something to go outside their SOA.

As unlikely as it is, it could be totally unrelated. It could even be the cable between the PCI card and the breakout box.

 

[mtl777]

As unlikely as it is, it could be totally unrelated. It could even be the cable between the PCI card and the breakout box.

Thanks for mentioning this! I actually had a bad breakout cable during the time when the problem started. How I knew it was bad is because it would affect my mouse and make it malfunction (seems unrelated but it's true). When I replaced the cable with another one from my other Delta 1010, the mouse would work fine. And now it has dawned upon me that the cable may have very well been the culprit! I have checked carefully for cold solders and shorts and didn't find any, so the cable must have been the one.

Thanks!

 

[mtl777]

Hi Gerd or anyone,

Do you have any suggestions for modding the 1010 PCI card? I'm just looking to do simple cap replacements. Like, how do you clean up the supply rails to the crystal clock oscillator? Can this be achieved with a simple cap change? If so, which capacitor(s) should be changed and with what value?

Thanks!