One of the CD 63 mods is to replace generic 7805 5V regulator by LM340T5.
I looked at datasheets for a couple of devices and someone correct me if I misunderstand them.
The STM is given as 10 uV/VO, so for a VO of 5 V, the noise is 50 uV right?
So that is marginally higher than the TI part that has 40 uV at 5V output (noise quoted for 3 different voltages.
generic 7805 STMicroelectronics L7805ACV
eN Output noise voltage TA = 25C, B =10 Hz to 100 kHz 10 uV/VO
LM340T5
Texas Instruments LM340T-5.0/NOPB
VN Output Noise TA = 25C, 10 Hz <= f <= 100 40 75 90 uV (5V 12V 15 V)
.... whilst on the subject, there is Texas Instrument TPS7A4700 variable voltage devise with noise of 4.17-μVRMS, an example:
http://uk.farnell.com/texas-instrum...Id=&categoryId=700000005152&iscrfnonsku=false
or
http://tekdevice.com/chapter2/index.php?route=product/product&product_id=87
I looked at datasheets for a couple of devices and someone correct me if I misunderstand them.
The STM is given as 10 uV/VO, so for a VO of 5 V, the noise is 50 uV right?
So that is marginally higher than the TI part that has 40 uV at 5V output (noise quoted for 3 different voltages.
generic 7805 STMicroelectronics L7805ACV
eN Output noise voltage TA = 25C, B =10 Hz to 100 kHz 10 uV/VO
LM340T5
Texas Instruments LM340T-5.0/NOPB
VN Output Noise TA = 25C, 10 Hz <= f <= 100 40 75 90 uV (5V 12V 15 V)
.... whilst on the subject, there is Texas Instrument TPS7A4700 variable voltage devise with noise of 4.17-μVRMS, an example:
http://uk.farnell.com/texas-instrum...Id=&categoryId=700000005152&iscrfnonsku=false
or
http://tekdevice.com/chapter2/index.php?route=product/product&product_id=87
Last edited:
The noise voltage of all these regulators is measured under ideal test conditions (nothing wrong with that, and its the only way to do it).
What you have to consider is that the introduction of PCB print from the regulator output to its destination (such as DAC etc) absolutely destroys that noise figure when measured at the destination.
So sweating the fine details is all very well, just remember that it all gets swamped in the real world when measuring a few cm from the reg.
What you have to consider is that the introduction of PCB print from the regulator output to its destination (such as DAC etc) absolutely destroys that noise figure when measured at the destination.
So sweating the fine details is all very well, just remember that it all gets swamped in the real world when measuring a few cm from the reg.
It depends what the regulator is feeding really. If its pure analogue stuff like the output filters and so on then any deterioration is normally minimal.
If its feeding something like a 100 leg smd all-in-one servo/dac/system control processor then the noise and hash generated on rails by the device the regulator is feeding can be significant. The overall impedance of the line to the reg then becomes very important. Those are two extreme cases, in practice there is every scenario in between those two.
When you are talking a basic noise figure of only a few tens of micro volts, then its very very easy to degrade that in applying the device.
If its feeding something like a 100 leg smd all-in-one servo/dac/system control processor then the noise and hash generated on rails by the device the regulator is feeding can be significant. The overall impedance of the line to the reg then becomes very important. Those are two extreme cases, in practice there is every scenario in between those two.
When you are talking a basic noise figure of only a few tens of micro volts, then its very very easy to degrade that in applying the device.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.