If your opa1632 doesn't have soldering pad on the underside,it has poor power dissipation .Your estimation(336mW) probably means super hot.
Your board has 2 pcs opa1632 per channel. It makes thing worse. I think super hot situation is normal.It's better to use opa1632 with power pad,which decreases thermal resistance.
Another choice is to remove one opa1632.The second opa1632 is not necessarily needed because the 3rd opamp is also differencial from Twisted Pear's IVY design.I'm not sure but you can remove the 2nd opa1632 with a little modification.It also decrease heat problem.Good luck.🙂
Not only Icq, I/V OPAMP are sinking 5.6mA*2 from DAC output. To Simplify calculation, assuming the OPA1632's output stage are class B and no Icq variation relative to load current, 12V*5.6mA*2=134mW is also dissipated at zero signal...I see it must be super hot.
About Vocm issue, a correction where I wrote sink with CCS, you can use just resistors for the purpose. Read the AD1955 datasheet. A caveat, pay attention the +- input voltage not to be below 0V, under the DAC supply rail voltage. DC offset at the IV opamp output is no problem, even the application of the datasheet have it.It is canceled out in latter stage's CMR. But DC current to the 2nd OPA1632 also produce the heat in both OPAMP. It may not wise choice.
About Vocm issue, a correction where I wrote sink with CCS, you can use just resistors for the purpose. Read the AD1955 datasheet. A caveat, pay attention the +- input voltage not to be below 0V, under the DAC supply rail voltage. DC offset at the IV opamp output is no problem, even the application of the datasheet have it.It is canceled out in latter stage's CMR. But DC current to the 2nd OPA1632 also produce the heat in both OPAMP. It may not wise choice.
Yeah, next revision would be MSOP with thermal pad or through hole...
Still debating what to do with the Vocm issue. DC offset is undesirable both due to the 2nd OPA1632 as well as not wanting to feed DC to whatever sits next in the chain. Supposedly it's a balanced line receiver in which case it'll be fine, but perhaps at some point it gets connected to a device that isn't actually handling balanced signals properly and lacks a DC blocking cap. Is that a reasonable concern?
Still debating what to do with the Vocm issue. DC offset is undesirable both due to the 2nd OPA1632 as well as not wanting to feed DC to whatever sits next in the chain. Supposedly it's a balanced line receiver in which case it'll be fine, but perhaps at some point it gets connected to a device that isn't actually handling balanced signals properly and lacks a DC blocking cap. Is that a reasonable concern?
hello,
i am also using those dac's (pcm1794a) with opa1642 (http://www.diyaudio.com/forums/digi...dsp-audio-preamplifier-crossover-project.html)
i must say that the soic8 opamps after the dac's runs a quite hot even at 12v symmetrical power supply. I am using this design for a while and i don't have any problem with it. I also checked for oscillations and there is no any. So it seems to be normal for those small soic packages.
i am also using those dac's (pcm1794a) with opa1642 (http://www.diyaudio.com/forums/digi...dsp-audio-preamplifier-crossover-project.html)
i must say that the soic8 opamps after the dac's runs a quite hot even at 12v symmetrical power supply. I am using this design for a while and i don't have any problem with it. I also checked for oscillations and there is no any. So it seems to be normal for those small soic packages.
- Status
- Not open for further replies.