|
Hello, I wrote a program for communication between the 6711 DSK's MCBSP's and some RF transceivers. The code by itself was working fine, I was testing it on a logic analyzer. I designed a buffer circuit to attenuate the 3.3V serial outputs of the MCBSP and add some DC bias to it. I was testing the setup when I noticed that I was no longer getting any valid clock or serial data signals out of the MCBSP0 any more. This is using the same program that I had verified the other night. The clock output of the MCBSP1 is only 1V peak to peak. I can't get a clock out of MCBSP0. I checked the 3.3V and 5V signals on the daughtercard connector, they are fine. The CheckDSK tool in the GEL menu of CCS tells me that the DSK is fine. Is there a posibility that I may have blown something on the board? I don't see what I could have blown since I wasn't feeding anything into the daughtercard connector. I did connect the ground of the connector to the ground of my DC power supply though, but I don't see how that could have done anything harmful. Does anyone know of any ways to pinpoint the problem? Fraidun |
|
|
Blown MCBSP ?
Started by ●September 7, 2003
Reply by ●September 7, 20032003-09-07
|
Fraidun,
"Current flows in mysterious ways, until you truly understands all of the
ways that current flows"
Speaking as someone who has blow some parts over the last few decades,
let me play back your description.
1. Everything worked
2. Connected some circuitry and probes to a few pins
3. Everything still works except for some of the pins that were
connected to my probes and circuitry
4. "Do you think I could have caused the problem in #3??"
The cause could have been ESD, shorted probe, incorrect design, or ??
The McBSP signals are pretty much unprotected on the 6711. Break out
the schematics and see if you lucked out and fried the analog switch instead of
the DSP...
BTW, the CheckDSK on the GEL menu only checks CCS - DSK
communication.
mikedunn
akhifra <a...@Eng.Auburn.EDU> wrote: Hello, |
Reply by ●September 7, 20032003-09-07
|
I had the same problem, my level shifters were damaged, dunno about yours. so i had to order the level shifters which are between the pins and cpu, and swap them out. they are hard to remove and resolder by hand, but possible. i used expensive polymide tape to shield all other components, then applied heat to remove the level shifter chip. applied some liquid flux on the footprints of the old chip. then aligned the new shifter chip in place and held it there with pressure from the top. took a fine solder tip and soldered pins on all four corners, then remaining pins. anyways, difficult to do if you have not seen it done before. Best is to find someone whom has done this kind of work before to help you, IF this is your prob. chiko --- akhifra <> wrote: > Hello, > > I wrote a program for communication between the 6711 > DSK's MCBSP's > and some RF transceivers. The code by itself was working > fine, I > was testing it on a logic analyzer. I designed a buffer > circuit to > attenuate the 3.3V serial outputs of the MCBSP and add > some DC bias > to it. I was testing the setup when I noticed that I was > no longer > getting any valid clock or serial data signals out of the > MCBSP0 any > more. This is using the same program that I had verified > the other > night. The clock output of the MCBSP1 is only 1V peak to > peak. I > can't get a clock out of MCBSP0. > > I checked the 3.3V and 5V signals on the daughtercard > connector, > they are fine. The CheckDSK tool in the GEL menu of CCS > tells me > that the DSK is fine. > > Is there a posibility that I may have blown something on > the board? > I don't see what I could have blown since I wasn't > feeding anything > into the daughtercard connector. I did connect the > ground of the > connector to the ground of my DC power supply though, but > I don't > see how that could have done anything harmful. > > Does anyone know of any ways to pinpoint the problem? > > Fraidun > > _____________________________________ > Note: If you do a simple "reply" with your email client, > only the author of this message will receive your answer. > You need to do a "reply all" if you want your answer to > be distributed to the entire group. > > _____________________________________ > About this discussion group: > > To Join: Send an email to > > To Post: Send an email to > > To Leave: Send an email to > Archives: http://www.yahoogroups.com/group/c6x > > Other Groups: http://www.dsprelated.com > ">http://docs.yahoo.com/info/terms/ __________________________________ |
Reply by ●September 8, 20032003-09-08
|
Chiko- > I had the same problem, > my level shifters were damaged, dunno about yours. > > so i had to order the level shifters which are between > the pins and cpu, and swap them out. they are hard to > remove and resolder by hand, but possible. > > i used expensive polymide tape to shield all other > components, then applied heat to remove the level > shifter chip. Hey too much work. Another option is to use a SHARP exacto knife to GENTLY, SLOWLY cut the pins. Then you can easily clear the pads with a fine-tip soldering iron. Be patient when cutting, use a slow gentle "sawing" motion until the blade goes through the pins, otherwise you can damage a pad. Jeff Brower system engineer Signalogic > applied some liquid > flux on the footprints of the old chip. > > then aligned the new shifter chip in place and held > it there with pressure from the top. > > took a fine solder tip and soldered pins on all > four corners, then remaining pins. > > anyways, difficult to do if you have not seen it done > before. Best is to find someone whom has done this > kind of work before to help you, IF this is your > prob. > > chiko > > --- akhifra <> wrote: > > Hello, > > > > I wrote a program for communication between the 6711 > > DSK's MCBSP's > > and some RF transceivers. The code by itself was working > > fine, I > > was testing it on a logic analyzer. I designed a buffer > > circuit to > > attenuate the 3.3V serial outputs of the MCBSP and add > > some DC bias > > to it. I was testing the setup when I noticed that I was > > no longer > > getting any valid clock or serial data signals out of the > > MCBSP0 any > > more. This is using the same program that I had verified > > the other > > night. The clock output of the MCBSP1 is only 1V peak to > > peak. I > > can't get a clock out of MCBSP0. > > > > I checked the 3.3V and 5V signals on the daughtercard > > connector, > > they are fine. The CheckDSK tool in the GEL menu of CCS > > tells me > > that the DSK is fine. > > > > Is there a posibility that I may have blown something on > > the board? > > I don't see what I could have blown since I wasn't > > feeding anything > > into the daughtercard connector. I did connect the > > ground of the > > connector to the ground of my DC power supply though, but > > I don't > > see how that could have done anything harmful. > > > > Does anyone know of any ways to pinpoint the problem? > > > > Fraidun |
|
|
Reply by ●September 8, 20032003-09-08
|
look at the schematics for you board. mine was a dsk6711 and there were two types of voltage levels shifters from 3.3V to 5V 1)SN74CBT3257PW for Mcbsp0 2)SN74CBTD3384PW for Mcbsp1 LOOK below for comments: --- wrote: > Chiko, > > Here is some more information about this problem: > > Lets say I want to activate the CNTL0 bit and an LED in > the IO port: The LED > goes on, but I don't sense anything on the CNTL0 line. > > Also, I tried to run the MCBSP's in loopback mode, and it > was working. So I > doubt that the DSP is fried. Yes it should still work in DLB > > If I output a clock out of MCBSP1, I get the right > frequency, but the voltage > level is wrong, around 1.3V or so and the clock doesn't > look as square as it > used to. Yes i had lower voltages > > Someone mentioned a melfunctioning "analog switch" may be > the cause > as well. Could this be the "level shifter" that you are > talking about? What > is the part number of the level shifter that I should > order from TI? you can order from digi-key > > Did you also have these problems? > > I really appreciate your help on this. > > Fraidun > > > Quoting Chiko <>: > > > I had the same problem, > > my level shifters were damaged, dunno about yours. > > > > so i had to order the level shifters which are between > > the pins and cpu, and swap them out. they are hard to > > remove and resolder by hand, but possible. > > > > i used expensive polymide tape to shield all other > > components, then applied heat to remove the level > > shifter chip. > > > > applied some liquid > > flux on the footprints of the old chip. > > > > then aligned the new shifter chip in place and held > > it there with pressure from the top. > > > > took a fine solder tip and soldered pins on all > > four corners, then remaining pins. > > > > anyways, difficult to do if you have not seen it done > > before. Best is to find someone whom has done this > > kind of work before to help you, IF this is your > > prob. > > > > chiko > > > > > > --- akhifra <> wrote: > > > Hello, > > > > > > I wrote a program for communication between the 6711 > > > DSK's MCBSP's > > > and some RF transceivers. The code by itself was > working > > > fine, I > > > was testing it on a logic analyzer. I designed a > buffer > > > circuit to > > > attenuate the 3.3V serial outputs of the MCBSP and > add > > > some DC bias > > > to it. I was testing the setup when I noticed that I > was > > > no longer > > > getting any valid clock or serial data signals out of > the > > > MCBSP0 any > > > more. This is using the same program that I had > verified > > > the other > > > night. The clock output of the MCBSP1 is only 1V > peak to > > > peak. I > > > can't get a clock out of MCBSP0. > > > > > > I checked the 3.3V and 5V signals on the daughtercard > > > connector, > > > they are fine. The CheckDSK tool in the GEL menu of > CCS > > > tells me > > > that the DSK is fine. > > > > > > Is there a posibility that I may have blown something > on > > > the board? > > > I don't see what I could have blown since I wasn't > > > feeding anything > > > into the daughtercard connector. I did connect the > > > ground of the > > > connector to the ground of my DC power supply though, > but > > > I don't > > > see how that could have done anything harmful. > > > > > > Does anyone know of any ways to pinpoint the problem? > > > > > > > Fraidun > > > > > > > > > > > > _____________________________________ > > > Note: If you do a simple "reply" with your email > client, > > > only the author of this message will receive your > answer. > > > You need to do a "reply all" if you want your answer > to > > > be distributed to the entire group. > > > > > > _____________________________________ > > > About this discussion group: > > > > > > To Join: Send an email to > > > > > > > To Post: Send an email to > > > > > > To Leave: Send an email to > > > > > > > > > Archives: http://www.yahoogroups.com/group/c6x > > > > > > Other Groups: http://www.dsprelated.com > > > > > > > > > ">http://docs.yahoo.com/info/terms/ > > > > > > > > > > > > __________________________________ > > > __________________________________ |
Reply by ●September 8, 20032003-09-08
|
Yes, you can cut the chip out too. thnx, jeff. To be clear, 'applied heat' means using a high temp heat gun. I really do not recommend anyone performing this repair unless he/she has seen/done surface mount device soldering by hand. good luck! chiko PS my board was the dsk6711 --- Jeff Brower <> wrote: > Chiko- > > > I had the same problem, > > my level shifters were damaged, dunno about yours. > > > > so i had to order the level shifters which are between > > the pins and cpu, and swap them out. they are hard to > > remove and resolder by hand, but possible. > > > > i used expensive polymide tape to shield all other > > components, then applied heat to remove the level > > shifter chip. > > Hey too much work. Another option is to use a SHARP > exacto knife to GENTLY, SLOWLY > cut the pins. Then you can easily clear the pads with a > fine-tip soldering iron. Be > patient when cutting, use a slow gentle "sawing" motion > until the blade goes through > the pins, otherwise you can damage a pad. > > Jeff Brower > system engineer > Signalogic > > applied some liquid > > flux on the footprints of the old chip. > > > > then aligned the new shifter chip in place and held > > it there with pressure from the top. > > > > took a fine solder tip and soldered pins on all > > four corners, then remaining pins. > > > > anyways, difficult to do if you have not seen it done > > before. Best is to find someone whom has done this > > kind of work before to help you, IF this is your > > prob. > > > > chiko > > > > --- akhifra <> wrote: > > > Hello, > > > > > > I wrote a program for communication between the 6711 > > > DSK's MCBSP's > > > and some RF transceivers. The code by itself was > working > > > fine, I > > > was testing it on a logic analyzer. I designed a > buffer > > > circuit to > > > attenuate the 3.3V serial outputs of the MCBSP and > add > > > some DC bias > > > to it. I was testing the setup when I noticed that I > was > > > no longer > > > getting any valid clock or serial data signals out of > the > > > MCBSP0 any > > > more. This is using the same program that I had > verified > > > the other > > > night. The clock output of the MCBSP1 is only 1V > peak to > > > peak. I > > > can't get a clock out of MCBSP0. > > > > > > I checked the 3.3V and 5V signals on the daughtercard > > > connector, > > > they are fine. The CheckDSK tool in the GEL menu of > CCS > > > tells me > > > that the DSK is fine. > > > > > > Is there a posibility that I may have blown something > on > > > the board? > > > I don't see what I could have blown since I wasn't > > > feeding anything > > > into the daughtercard connector. I did connect the > > > ground of the > > > connector to the ground of my DC power supply though, > but > > > I don't > > > see how that could have done anything harmful. > > > > > > Does anyone know of any ways to pinpoint the problem? > > > > > > Fraidun __________________________________ |
