You're quite correct - I'm an idiot. There's no sign of a MAX232-like charge pump inside that chip ... and there's a specific Output Swing characteristic that says you can only get to Vdd-0.04.
Very simple DAC - 2 solutions here
- Thread starter KTarke
- Start date
Yes, the original purpose of this thread was to find some alternatives to a external dac-chip...
In last two weeks I have also done a little something on the matter:
-Took a industrial valve-adjuster (input 0-10V) and tested both those circuits, that were shown in the beginning of this thread. In that application BOTH worked just like expected. But as "the court" said, one can not draw conclusion about that ; the circuits still are not suitable to any other app...
I will later re-engineer the valve-adjuster and try to see, what kind of signal conditioning (if any) is used in it. Will report on that.
-Have also searched the Internet for a suitable ONE_CHIP external dac (output 0-10V, single supply)
Reading those datasheets and app-notes have been a good lesson.
In this, again, I search for simplicity:
-single supply
-no extra op-amp needed to gain 0-10V
Findings: most single-supply ones does NOT reach rails... closest is Vss+0,3V. On the Vdd side that is no problem, it can fixed by feeding 10,3V...
0 and 10V can be reached by adding a negative feed, and that can be "extracted" from single supply with one chip, but that makes two chips together
(as does dac+op-amp)
Conclusion: no VERY simple solutions available there... (or just I haven't found the right chip since)
BTW: I was surpriced to find several "commercial" serial-driven 0-10V dac-modules ,that use simple op-amp as outputs. All of those were guaranteed to be accurate and fast...
So, the op-amp -way seems not to be any worse than a "real" dac-chip.
(all those seemed to have a uP to handle the serial side, and pwm:ing the op-amps)
In last two weeks I have also done a little something on the matter:
-Took a industrial valve-adjuster (input 0-10V) and tested both those circuits, that were shown in the beginning of this thread. In that application BOTH worked just like expected. But as "the court" said, one can not draw conclusion about that ; the circuits still are not suitable to any other app...
I will later re-engineer the valve-adjuster and try to see, what kind of signal conditioning (if any) is used in it. Will report on that.
-Have also searched the Internet for a suitable ONE_CHIP external dac (output 0-10V, single supply)
Reading those datasheets and app-notes have been a good lesson.
In this, again, I search for simplicity:
-single supply
-no extra op-amp needed to gain 0-10V
Findings: most single-supply ones does NOT reach rails... closest is Vss+0,3V. On the Vdd side that is no problem, it can fixed by feeding 10,3V...
0 and 10V can be reached by adding a negative feed, and that can be "extracted" from single supply with one chip, but that makes two chips together
(as does dac+op-amp)Conclusion: no VERY simple solutions available there... (or just I haven't found the right chip since)
BTW: I was surpriced to find several "commercial" serial-driven 0-10V dac-modules ,that use simple op-amp as outputs. All of those were guaranteed to be accurate and fast...
So, the op-amp -way seems not to be any worse than a "real" dac-chip.
(all those seemed to have a uP to handle the serial side, and pwm:ing the op-amps)
hippy
Ex-Staff (retired)
Bear in mid that some things labelled "0V-10V" input may actually not need the full range of voltage. For a valve, 0.5V and below may be fully closed, 9.5V and above may be fully open. This provides extra safety and caters for where people haven't designed perfect controllers and can't output swings to the power rails.
Not everything has top and bottom dead-band but it's common to design them in or have them naturally occur, and it becomes pointless to provide a perfect signal when the thing controlled doesn't need that. Don't forget that manufacturers usually make both ends of a system and will likely be making it as simple for themselves as possible.
In the stage lighting industry when using 0V-10V control, bottom dead-band works the other way round, no matter how hard you try to got 0% (off) the 'valve' will always be forced slightly open to pre-heat the bulb. That stops huge in-rush currents and exploding lamps when they're suddenly cranked-up to full. That's also why it's highly dangerous to connect pyrotechnics into any lighting rig socket. There's usually some top dead-band as well to cater for the fact equipment may have diode mixing internally or that may be added by the user.
Not everything has top and bottom dead-band but it's common to design them in or have them naturally occur, and it becomes pointless to provide a perfect signal when the thing controlled doesn't need that. Don't forget that manufacturers usually make both ends of a system and will likely be making it as simple for themselves as possible.
In the stage lighting industry when using 0V-10V control, bottom dead-band works the other way round, no matter how hard you try to got 0% (off) the 'valve' will always be forced slightly open to pre-heat the bulb. That stops huge in-rush currents and exploding lamps when they're suddenly cranked-up to full. That's also why it's highly dangerous to connect pyrotechnics into any lighting rig socket. There's usually some top dead-band as well to cater for the fact equipment may have diode mixing internally or that may be added by the user.
Again, You are right, Hippy.
That is what I have noticed too, that the "last" 0,5V on both ends are not used, in valvemotors, dimmers, etc.
Though, valvemotors have built-inside switches to recognize end-positions, it is ensured that the valve will close, by "twisting the rules" in input voltages.
I just consider this: usually "ends" are not used, but once I might bump into a thing that needs really the whole range! And then I am in trouble... if my circuit does not handle it.
(actually, I know some: atleast one "lab" powersource and a robot-weldingmachine. Those two needed really the whole range to operate correctly)
I am designing these gadgets so, that they could be used with a wide variety of devices, and I am going to sell some of my "deeds" ,too.
I am on the safe side, when I can say, that it is the users device, that does not meet the standard...
That is what I have noticed too, that the "last" 0,5V on both ends are not used, in valvemotors, dimmers, etc.
Though, valvemotors have built-inside switches to recognize end-positions, it is ensured that the valve will close, by "twisting the rules" in input voltages.
I just consider this: usually "ends" are not used, but once I might bump into a thing that needs really the whole range! And then I am in trouble... if my circuit does not handle it.
(actually, I know some: atleast one "lab" powersource and a robot-weldingmachine. Those two needed really the whole range to operate correctly)
I am designing these gadgets so, that they could be used with a wide variety of devices, and I am going to sell some of my "deeds" ,too.
I am on the safe side, when I can say, that it is the users device, that does not meet the standard...