Soekris Dam Dac

[rickmcinnis]

I can finally say I have finished the wiring. I have turned it on and there was no smoke or flame.

The downside is the + part of the mravica reg is not working properly. I hope I screwed up the voltage adjusting solder bridges.

One strange thing - my ground switch - one pole is connected to the input power grounds the other pole has ALL ground wires for batteries and regulators connected. YET, the green light is on no matter if the switch is off or on. I cannot figure out how this can be. Is anyone else seeing this?

The switch is not damaged - I verified with the ohmmeter.

So if the regulator is not destroyed I should have it working/playing tonight.

And, of course, there will be pictures. It is not a thing of beauty but it is small.

[randytsuch]

I can finally say I have finished the wiring. I have turned it on and there was no smoke or flame.

The downside is the + part of the mravica reg is not working properly. I hope I screwed up the voltage adjusting solder bridges.

One strange thing - my ground switch - one pole is connected to the input power grounds the other pole has ALL ground wires for batteries and regulators connected. YET, the green light is on no matter if the switch is off or on. I cannot figure out how this can be. Is anyone else seeing this?

The switch is not damaged - I verified with the ohmmeter.

So if the regulator is not destroyed I should have it working/playing tonight.

And, of course, there will be pictures. It is not a thing of beauty but it is small.

Rick
What is "Input power ground"? Ground of the DAC?

With the switch off, does it buzz out between the battery ground and the dac ground? I'm guessing it does, and you need to break that connection for the switch to do it's intended function.

The led will only come on if you're supplying power to the DAC, so there has to be a path from battery ground to dac ground.

One other thing. Are you going to leave the regulators on all the time? If you turn them off, I suspect it will drain the batteries.

Randy

[rickmcinnis]

Yes, I am going to leave the "chargers" on all of the time.

The drain when off seems to be minimal.

The ground I have one pole connected to is to the two holes marked PWR GND at hifiarduino - between the two PWR IN pins at the J1 header. Only thing attached to that pole and everything else is attached to the other.

I replaced the rectifier with two lengths of wire but that should not make any difference, I would not think. I wonder if I should remove those wires? What do you think?

Told myself one hundred times to check those regs before wiring them in. Some people never learn. With everything tight it will be proper penance having to troubleshoot, NOW.

Thanks for the help,

[randytsuch]

The ground I have one pole connected to is to the two holes marked PWR GND at hifiarduino - between the two PWR IN pins at the J1 header. Only thing attached to that pole and everything else is attached to the other.

I replaced the rectifier with two lengths of wire but that should not make any difference, I would not think. I wonder if I should remove those wires? What do you think?

In thinking about it, I would remove the rectifier wires, and wire to a ground point on J2. Those should be the real ground pins of the DAC, versus the pwr gnd pins in J1. The J2 pin closest to J1 is ground, so this would be easy to do.

And after you remove the wire from J1, before reconnecting it, see if the led goes off. If the led is still on, there is another path from battery ground to dac ground you need to find.

I am not switching ground, I switch the positive and negative voltages. Battery grounds are tied the various ground points on my dac, and this works fine for me.

Randy

[rickmcinnis]

I will give J2 a try but SOEKRIS said these were test points and the traces are pretty skimpy. I am using them for the 1.2 but only because we have no other choice.

First will remove the shorts across the rectifier pads - did that before Nigel got the batteries to work for Vref.

Wish SOEKRIS would release a schematic!

Thanks,

[randytsuch]

I will give J2 a try but SOEKRIS said these were test points and the traces are pretty skimpy. I am using them for the 1.2 but only because we have no other choice.

First will remove the shorts across the rectifier pads - did that before Nigel got the batteries to work for Vref.

Wish SOEKRIS would release a schematic!

Thanks,

I think Soekris was talking about the power traces. Since there is a ground plane, I would expect the J2 ground points to tie directly into the ground plane.

Another option is, if you've removed the electrolytic caps above J2, find a ground pad and wire to that. I connected one battery to a pad up there.

[rickmcinnis]

The caveat, for me, is - are those pins connected to the ground plane or to a trace that is connected to the ground plane?

I have the board positioned so I can no longer look underneath it, at this point.

All of the can caps are removed so one of those pads could be used - I would think they are directly connected to the ground I am using. Only thing I can think of is if there is some way the rectifier bypasses are making the connection but now that I draw it out that is impossible - unless there is some cap that is allowing the leakage. Hoping for the ridiculous. If the rectifiers were not bypassed I would HAVE to use the "big" cap pads. AS I think about it that would be superior, tying in there as opposed to where I am - two less solder joints!. My ground wire from the switch will be longer but that is moot.

Only caps remaining are SMDs around the FPGA, the clock, generally that stuff on the left side of the board and those caps that Nigel replaced with polyester caps after the FPGA, all others have been removed.

Maybe magic will occur and moving the ground point will solve the problem.

Thanks for you thinking about this. Much appreciated.

[randytsuch]

The caveat, for me, is - are those pins connected to the ground plane or to a trace that is connected to the ground plane?

I have the board positioned so I can no longer look underneath it, at this point.

All of the can caps are removed so one of those pads could be used - I would think they are directly connected to the ground I am using. Only thing I can think of is if there is some way the rectifier bypasses are making the connection but now that I draw it out that is impossible - unless there is some cap that is allowing the leakage. Hoping for the ridiculous. If the rectifiers were not bypassed I would HAVE to use the "big" cap pads. AS I think about it that would be superior, tying in there as opposed to where I am - two less solder joints!. My ground wire from the switch will be longer but that is moot.

Much appreciated.

Standard practice (at least that I'm aware of) would be to have a small clearance around the hole pad, and then have "spokes" between the ground plane and the hole pad. These spokes are VERY short traces. I expect that is how the gnd pins of j2 are connected to ground, but without artwork there is not way to verify. But, it would make NO sense to run traces from the ground pins to ground.

The capacitor ground pads have vias that connect the ground pad to the ground plane.

The post you're thinking of is very specific for connecting to the J2 A+ and A- pads, Soekris said not to do it because they were thin traces not meant to carry power.

And yes, there are traces between pwr gnd and the dac ground which will add impedance, better to connect directly to the dac ground somewhere else.

[rickmcinnis]

It seems I chose the wrong place for the main ground. - I would have thought this was the best place but I based this on nothing more than assumptions since this is where the transformers were connected.

So at this point what do you think is superior? Using all four points on J2 or to use the pads from the removed capacitors?

I am tending to think the more central location of J2 might make it the place and I have the feeling you would, too.

Thanks, again, Randy for your consideration of my problem.

I posted a question at DIYAudio begging SOEKRIS to give his opinion on the best spot. 50/50 chance of him answering it!

[nige2000]

pics please rick

gnd should be gnd anywhere at least from a functional point of view
what way have you wired the switches

id leave gnd out of the switch connections and have it permanently connected
you don't want differentials

whats powering the 1.2 and 3.3v for the fpga etc?