DAC AD1862: Almost THT, I2S input, NOS, R-2R

I experienced 2 capacitor explosions. One small electrolytic flew over my head due to the reversed polarity (from that moment I started wearing safety glasses) - it is really like projectile ... and the second explosion was THT tantalum (the yellow thing in ceramic package), not sure why it happened because it worked for a few months - it was like a grenade with shrapnels :cool:
Always have great respect when diagnosing or testing something. Rochester does have the original AD1862, but doesn't offer spare eye :fight:
This "hot projectile" can fly quite fast, slow mo guys measured 60 m/s, it can do a serious injury to the eye:
Funny is that a large capacitor with the weakening cross on top is most likely safer than a tiny one without cross :scratch1:
 
  • Like
Reactions: 1 user
Hey guys,
I made some measurements on the AD1862 DAC's outputs without installed chips and oamps.

  1. Impedance when it's turned off.
    It shows 2MOhm 🤔
    tempImagedmmnyn.gif
    Another channel shows 4MOhm

  2. Impedance when it turned on.
    It shows 0L.
    tempImagezsmGNM.gif
    Another channel shows 100KOhm

  3. Voltage when it turned on.
    it was showing around 4V and going down slowly.
    tempImageZwLMBU.gif
    Another channel shows around 500mV.

    This is quite strange, right?
    Do you guys have approximate correct values for DAC the outputs?
 
Where did you buy ldo's from? Did you check all the values of passive components? Check solder joints as well.

Btw, if you use jlspunds board, my advice is to change the heatsink on the regulator (xmos supply). It is piping hot even with lower supplied voltage (i supply it with 4.6V). It is lukewarm to the touch now, while it was fingerprint erasing hot with default heatsink.
 

Attachments

  • 20221218_215114.jpg
    20221218_215114.jpg
    298.3 KB · Views: 106
  • 20221218_214431.jpg
    20221218_214431.jpg
    297.3 KB · Views: 108
  • Like
Reactions: 1 users
Where did you buy ldo's from? Did you check all the values of passive components? Check solder joints as well.

Btw, if you use jlspunds board, my advice is to change the heatsink on the regulator (xmos supply). It is piping hot even with lower supplied voltage (i supply it with 4.6V). It is lukewarm to the touch now, while it was fingerprint erasing hot with default heatsink.
how do you mount the heatsink? heatsink glue?
 
Member
Joined 2014
Paid Member
I agree that the regulator does indeed, get hot with the supplied heatsink, but I surely would not use a heatsink of that type on this application. There’s no way to securely mount it to the pcb and could lead to shorts if it was bent and touched surrounding components.
 
I agree that the regulator does indeed, get hot with the supplied heatsink, but I surely would not use a heatsink of that type on this application. There’s no way to securely mount it to the pcb and could lead to shorts if it was bent and touched surrounding components.
hmmm, maybe i will desolder and mount it under the board, mounted to the chassis base. any info on this transistor?
 
Yeah those bigger heatsinks usually have a couple differently spaced hokes. You can use a bit smaller heatsink though, factory one is really bad.
@Vunce yeah, it needs isolator below. I'll be mounting different one when i pack it in box. Nixie will use a smaller to220 one, and if it does the job, i'll mount the ones i mounted on psu ldo's, 17c/w.
 
http://www.kelco.rs/katalog/detalji.php?ID=9675

I ordered these from a local shop, I think they are around 20k/W. Before that, I reduced the heating by putting a 1N4001 diode on the output of the +5V regulator, the voltage drop on the diode is almost 1V at 400mA (dissipation is 400mW on that diode). So now the XMOS is powered with about +4V, which is not a problem because that regulator is for 3.3V and there is another one behind it for 1V. Now it is hot but not too hot as before.