Using the AD844 as an I/V

[jcx]

really "Iout" and "R2R Ladder" are pretty definitive in describing the circuitry connected to the Iout pin - metal film resistors and CMOS switches

the data sheet should call out any pins with special V, I limits in the "Absolute Maximum Rating" table - no special pin limits are mentioned in the PCM1702 datasheet

so we really don't expect anything to break with any V, impedance up to the supply V being attached to the Iout pin

[georgehifi]

After reading a member here (you'll have to search for it) who tried different a progressivly lower I/V resistors, he got down to I think it was 10 or 20ohm and then damaged his one of his PCM1702/4's, it still played but had the spits noise, and he did say it got warm to the touch when he did it.

Cheers George

[jcx]

may have static damage - a problem with tweaking and CMOS - I would definitely be using ESD control measures working on a EOL $75 CMOS chip


what exactly do think "virtual gnd" looks like to the chip, how is it different from a wire to gnd - the data sheet doesn't include any verbage on how close to ideal the op amp may be == how much like "real gnd" the virtual gnd input looks

you can easily use op amps with output current capability >> the 10 mA often seen as substrate diode forward bias recommendtion
the datasheet shows a feedback C - instantaneous current is only limited by the op amp slewrate, output current limit

[abraxalito]

Seems to me more likely George you could damage the PCM1704 with a higher resistor than lower as above 100R or so there are protection diodes that start conducting, limiting the output compliance. But since the max Iout is a measly 1mA or so, really I can't believe that was the reason the chip fried.

[dirkwright]

Originally Posted by AudioLapDance

PS Now if we could just figure out how to do all the fancy sims and tests, we'd really harness the power of TINA!

Running tests in Tina is pretty straight forward. There are Bode plots, FFT's, noise, etc. that you can all run on the circuit. You can even use a custom waveform, but I have never done that. It would be cool if someone had the "staircase" sine waves that were the actual output from a DAC. It has to be in WAV file format. Anyone?

[AudioLapDance]

Well, I'm new to TINA but I've been meaning to learn ...

Which would be good tests to show the performance of the ad844 CB IV?

[dirkwright]

Originally Posted by AudioLapDance

Well, I'm new to TINA but I've been meaning to learn ...

Which would be good tests to show the performance of the ad844 CB IV?

Well, obviously you want to know things like noise and distortion. So, the noise test is important. When you run the FFT, it starts from the frequency of the generator in your schematic. If you have filtering in the circuit, you should delay the FFT by 10->50usec to give the circuit time to settle. You can learn a lot about a circuit by changing values and re-running the tests and comparing the results. Good luck.

[georgehifi]

Ok this is by far the best so far, super dynamic, stunning deep tight bass, great mids with real body, and extended crystalline top end.
A little stair case noise on the 20k sine wave on the output but not intruding into the extended and delicate top end this has, maybe when I get the courage I will put the 100,000pf to ground after the dac and then 10ohm in series to pin 2 of the 844 to give a second filter, that should kick any stair case noise right out then.

BTW: I have a left over relay muting circuit that I can still use, that shorts the output to ground, for a couple of seconds during switch on and switch off, can anyone say if the output of the discrete buffer will handle that as is or do I need a series say 50-100ohm resistor? I'd rather not use a resistor. The dc offset thump at switch on and off is around 1v but after it settles it is steady at around 1mV+&-

Cheers George

[AudioLapDance]

Hmm ... I should add your current source to the sim. It will probably affect the input a little. Any idea where on 20k trim you are (measure)?

[georgehifi]

Originally Posted by AudioLapDance

Hmm ... I should add your current source to the sim. It will probably affect the input a little. Any idea where on 20k trim you are (measure)?

A touch to the - for the left from middle, and I would say half way between mid and - for the right, as the right dac had about +30mV dc offset and the left about +10mV from memory.

Cheers George