Case grounding (need assistance)

[Echopraxia]

Hi everybody,

Just finished my MBFM but when I would make contact with the front panel and the bottom panel I would get little sparks. The whole case is metal. When I have the panels together My MBFM won't work at all. I read from an old thread about a metal front panel can become a antennae and you should ground it. I have always been testing the MBFM with the front plate away and when I put them together I was hearing a buzz sound which was the sparking.

So my question is do I take an unused ground wire from one of the DIN or DOUT modules (mounted on bottom panel) and solder the wire to the front panel directly? Or should I be worried about something else.

I have a metal power jack on the metal bottom case with a two prong 9v DC adapter going in.

[Shuriken]

I don't understand. As your whole case is metal, there shouldn't be any potential difference between back and front. But perhaps the adapter has reverse polarisation and the V+ is connected to the case?

Edited September 13, 2010 by Shuriken

[Echopraxia]

i tried swapping the v+ and negative and my box still powers up but acts the same way with the sparks. I am using the ultracore and I don't know if there is anything that I might be doing wrong with it. I don't have any of the IIC parts of it stuffed. I used the ultra core for the banksticks.

Anything else I can try? I checked all my solder pads and they are not touching any of the case bottom.

[nILS]

If you have a metal case it should be properly grounded. A metal connector is bad, because there is going to be a potential between the PSU ground and your MIDIbox ground after the rectifier.

0) Remove power.

1) Remove all PCBs and CS

2) Multimeter in beep-mode :)

3) Are all parts of the case connected? If not, connect them.

4) Put the parts back in

5) Isolate the power connector :excl:

6) Is anything else connected to the case? +5V, +9V, anything on the CS, any data line? If so, find out where and eliminate the short.

7) properly connect MIDIbox ground to the case

8) It works now :D

edit: To visualize how the potential difference happens import the following into http://falstad.com/circuit/

$ 1 5.0E-6 10.20027730826997 50 5.0 50

v 176 208 176 144 0 0 40.0 5.0 0.0 0.0 0.5

w 176 144 304 144 2

w 176 208 304 208 2

d 304 144 336 176 1 0.805904783

d 304 208 336 176 1 0.805904783

d 272 176 304 144 1 0.805904783

d 272 176 304 208 1 0.805904783

w 272 176 272 288 0

w 272 304 448 304 2

w 336 176 448 176 0

r 448 176 448 288 0 2000.0

g 448 304 448 320 0

x 472 241 525 247 0 24 Load

x 130 286 224 292 0 24 Potential

w 192 256 192 240 0

w 192 224 176 240 0

w 192 224 208 240 0

w 256 304 240 320 0

w 256 304 240 288 0

w 256 304 192 304 0

w 272 288 272 304 0

w 448 288 448 304 0

w 192 240 192 224 0

[/code]

[TK.]

edit: To visualize how the potential difference happens import the following into http://falstad.com/circuit/

nice tool!

Best Regards, Thorsten.

[Echopraxia]

Thank you for the detailed respone nILS,

Before I begin doing eveything wrong :) Could you clarify these steps for me?

5) Isolate the power connector:

Do I just put something around it so the metal jack doesn't touch the metal case? Any recommendations for a fix? Plastic washer maybe or alternative connector? I am using a two prong 9V AC 400ma supply.

6) Is anything else connected to the case? +5V, +9V, anything on the CS, any data line? If so, find out where and eliminate the short.:

I have just DIN and DOUT cables ran to the CS. I have a +12v and -12v power supply ran to the opl3 board. The MBFM works great when the two panels and not connected. Do you think I still may have a short somewhere?

7) properly connect MIDIbox ground to the case:

Do I just run any of the Vs connections and screw it to a nut on the case? Do I do the same for the CS?

This is a very important subject for me as I have curious young children around the house. I just read about grounding metal cases, and if by not doing so, say my little header pins came loose which are running Vd and touch the case I could get a harmful shock correct?. I'm now even more worried about doing this properly. I also should repeat this for my MBSEQ which I just put in a metal case.

Thanks again,

Edited September 14, 2010 by Echopraxia

[nILS]

Before I begin doing eveything wrong :) Could you clarify these steps for me?

Most certainly ;)

5) Isolate the power connector:

Do I just put something around it so the metal jack doesn't touch the metal case? Any recommendations for a fix? Plastic washer maybe or alternative connector? I am using a two prong 9V AC 400ma supply.

Oof. Is it one of the those?

If so: Oof²!

Typically those little funny DC connectors are called DC connectors for a reason :) All the metal ones I've come across so far have one of connections being the socket-casing - typically the outer one which (again) typically is ground and hence grounds the case. In your case: Congrats, you have AC on the case :frantics: So whatever you do, don't plug it in again until it's fixed :) 9VAC won't hurt, but it can easily kill all the electronics inside.

After careful consideration I'd like to change "isolate" to: "Throw it away and use a plastic one."

6) Is anything else connected to the case? +5V, +9V, anything on the CS, any data line? If so, find out where and eliminate the short.:

I have just DIN and DOUT cables ran to the CS. I have a +12v and -12v power supply ran to the opl3 board. The MBFM works great when the two panels and not connected. Do you think I still may have a short somewhere?

Unlikely. Yet: At some point attaching the front panel gives you sparks, which means that *something* on the CS is shorted to the panel - if it's ground, all good, if not you need to find the short. Multimeter > Beepmode > find out what's connected to the panel (with wires to the core unconnected)

7) properly connect MIDIbox ground to the case:

Do I just run any of the Vs connections and screw it to a nut on the case? Do I do the same for the CS?

Star wiring seems pretty popular. Run a connection from the core module (J2 or sth that has ground) to the base of the case and fix it to that with a nut, check with the multimeter if the connection is good. Usually you'd now run wires to all parts of the case from that spot. In your case a single wire to the front panel should do, since the rest seems to be connected (to be on the safe side, test with the multimeter and add wires if necessary).

This is a very important subject for me as I have curious young children around the house.

Good thing you're working with small voltages then. Before you wire up a tube amp, make sure to ask again ;)

[Echopraxia]

"Throw it away and use a plastic one."

I don't seem to have a local store for these. What would you reccomened? I am using Radio shack panel mount type N for my plugs and it fits perfect. Or should I just cut the connector off of the power supply and wire it to some headers? I'd rather not have a blocky power supply always attached to the unit ya know.

Edited September 14, 2010 by Echopraxia

[Echopraxia]

next question. I drilled a hole in the bottom case and strapped a screw and nut to it with another nut on the underside of the bottom case and tightened a wire down between the top side nut and case. I then ran the wire to J2 Vs of the core and checked for voltages. I did get the 4.9v when I checked random Vd spots with the lug nut as ground.

Did I do this correctly?

Do I now just need to run another wire from the same nut to the front panel and tighten that wire down with another nut on the front panel correct?

Edited September 15, 2010 by Echopraxia

[Echopraxia]

I put the plastic jack on there and no sparks! Running well.

next question. I drilled a hole in the bottom case and strapped a screw and nut to it with another nut on the underside of the bottom case and tightened a wire down between the top side nut and case. I then ran the wire to J2 Vs of the core and checked for voltages. I did get the 4.9v when I checked random Vd spots using the lug nut as the ground point.

Did I do this correctly?

I should point out that when the top panel is put on it is screwed down tight to the base panel. It was an older roland mixer that I gutted. So would I really need to still run a ground wire to the front panel?

Edited September 15, 2010 by Echopraxia

[nILS]

The wire to the front panel won't be necessary if you only run the unit with the panel on and it actually makes connection to the rest of the case.

Glad you got it working now!

[Echopraxia]

I booted it up everything was working , audio was good yada yada.. BUT when I scroll to the far right (I.E USER) of menus I get weird glitch noises. When I press any of the 8 general purpose buttons and use the menu encoder to scroll to the very right this happens and as long as I stay on the far right icon the noise is constant. :turned: Not sure if this is a grounding thing. please advise.

[nILS]

I am not sure either. I lost my magic crystal ball.

6) Is anything else connected to the case? +5V, +9V, anything on the CS, any data line? If so, find out where and eliminate the short.

[Echopraxia]

After playing with my new glitch feature I think I may not fix it :thumbsup:

For the record it is probably a pin barely touching the case because the unit does stop producing audio sometimes and I have to reboot.

[Echopraxia]

I scared myself this morning. I thought I blew up my FM. It turned out the voltage regulator pins on the underside were too close to the case and caused the regulator to get real hot and power down the synth when I gave it a thud on the table because the audio dropped out again. I found out it was hot by caveman exploration. :o I'm no expert but I trimmed the leads, let it cool down and now no glitching and no audio drops. Yay!

[strophlex]

:flowers: