Thank You guys.
I will keep the lower noise one in, it is an On-Semi brand 7915 C type.
What do you guys think the more important the ripple reject number or the output noise voltage number?
I am thinking of changing the 7815 for a better part. The one I have in there is a LM7815A.
Thank you again. Reading the specs sheets after a while makes one lose perspective.
Robert
I will keep the lower noise one in, it is an On-Semi brand 7915 C type.
What do you guys think the more important the ripple reject number or the output noise voltage number?
I am thinking of changing the 7815 for a better part. The one I have in there is a LM7815A.
Thank you again. Reading the specs sheets after a while makes one lose perspective.
Robert
I have found a problem using these regulators.
I found I had to decouple the output of the regulator with an electrolytic very close to the regulator or I got instability on the reg output.
Perhaps I just bought cheap versions.
It used to be standard practice to put a 1uF tantalum capacitor on both input and output of the 78xx and 79xx regulators.
The design I am upgrading uses 10uF on the in and outs of both the 7815 and the 7915.
Anyone have any thoughts on specs of what is more important. Either the ripple rejection number or the output noise voltage number?
I can use the old 7815 in the circuit that has a 80db rejection and 90uV/VO noise or a new 7815 with 70db and 90uV noise total. The new one is made by ON-Semi, (I hope I read the specs right) it is the On-Semi NCP7815TG and is says its good for appliances, but no mention of audio use. I hope I am not chasing my tail with all the specs.
I am thinking the lower noise would be better (per Jitter's comment), but I am uncertain on the rejection of ripple. (side note, I think the new 7915 with 90uV noise and 60db rejection sounds cleaner than the old one with 320uV of noise, could be my imagination...I do not have proper measuring equipment to prove this)
Thank You,
Robert
Anyone have any thoughts on specs of what is more important. Either the ripple rejection number or the output noise voltage number?
I can use the old 7815 in the circuit that has a 80db rejection and 90uV/VO noise or a new 7815 with 70db and 90uV noise total. The new one is made by ON-Semi, (I hope I read the specs right) it is the On-Semi NCP7815TG and is says its good for appliances, but no mention of audio use. I hope I am not chasing my tail with all the specs.
I am thinking the lower noise would be better (per Jitter's comment), but I am uncertain on the rejection of ripple. (side note, I think the new 7915 with 90uV noise and 60db rejection sounds cleaner than the old one with 320uV of noise, could be my imagination...I do not have proper measuring equipment to prove this)
Thank You,
Robert
Last edited:
The datasheet says you dont need a decoupling capacitor on the output.
I found I did need a capacitor on the output, my class D amp was making very high frequency noise on the 12 volt rail and causing the amp to howl.
The 10uF caps are after the big 6800uF filter caps but before the regulators, so maybe they are part of some bypass for the big filter caps.
The design is mur860's into 6800uF the into 10uF caps and the into 7815 and 7915, then out to 10uF, the to 47uF then to op-amp power rails. (btw this is on Behringer Truth B2031 powered monitors,with LM3886 amps)
out,
Robert
The design is mur860's into 6800uF the into 10uF caps and the into 7815 and 7915, then out to 10uF, the to 47uF then to op-amp power rails. (btw this is on Behringer Truth B2031 powered monitors,with LM3886 amps)
out,
Robert
Maybe so, but a board full of TTL with a 2 MHz clock and 10 ns rise times is not something I would hang off a slow regulator with no decoupling. The other advantage is the decoupling acts as a low pass filter reducing regulator noise.
Yes, I think that you are putting too much faith in spec. sheets. Manufacturing techniques change with time, mfg. seldom update the spec. sheets.
I wouldn't be too sure of that. Recently I downloaded some datasheets fom ST and others of some very standard components (78xx, LM317, etc). to answer a question in another thread (http://www.diyaudio.com/forums/parts/82638-my-transistors-original-copy-29.html) and actually found them to be very up to date. All of them had a revision date in 2010.
These were downloaded directly from the manufacturers and not at third party, though...
Jitter,
it looks like you are correct, I found new spec sheets at On-Semi. The newer 78XX and 79XX have different specs (and generally better) than some older spec sheets I found on the internet.
The newer 7915 form On-Semi has better specs than the older ones.
I guess the question I still have is the On-Semi NCP7815TG with 6.8uV/VO output noise voltage and 70db of ripple rejection better than the old LM7815A I have in there? The LM7815A is over 10 years old, the newer spec sheet I found on the LM7815A has a better noise spec of 90uV. but I do not know if this is the total noise or if it is measured in uV/VO.
Thank You,
Robert
it looks like you are correct, I found new spec sheets at On-Semi. The newer 78XX and 79XX have different specs (and generally better) than some older spec sheets I found on the internet.
The newer 7915 form On-Semi has better specs than the older ones.
I guess the question I still have is the On-Semi NCP7815TG with 6.8uV/VO output noise voltage and 70db of ripple rejection better than the old LM7815A I have in there? The LM7815A is over 10 years old, the newer spec sheet I found on the LM7815A has a better noise spec of 90uV. but I do not know if this is the total noise or if it is measured in uV/VO.
Thank You,
Robert
The units tell you if it's total noise or noise per volt output.
total noise = uV
noise per volt output = uV/VO.
If the manufacturer used the units correctly, that means that 90 uV is total noise.
A bit of mathematics shows this is most likely true:
6.8 uV/VO * 15 VO = 102 uV (total noise for the NCP7815TG)
90 uV/15 VO = 6 uV/VO (noise per volt output for the LM7815A)
total noise = uV
noise per volt output = uV/VO.
If the manufacturer used the units correctly, that means that 90 uV is total noise.
A bit of mathematics shows this is most likely true:
6.8 uV/VO * 15 VO = 102 uV (total noise for the NCP7815TG)
90 uV/15 VO = 6 uV/VO (noise per volt output for the LM7815A)
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.