OK...
Heres a few ideas and questions...
1 - Feeding the LME from a "clean" separate higher voltage supply than rail voltage from output. Feasible ???
2 - wouldn't these LME's accept balanced signals ?? wouldnt that be more desirable input signal on clean amp like this ???
will have more but off the top of my head this is what i got...
Using Panson's schematic i'm making my own PCB layout that will allow me to run 4 pairs or run 2 boards in series that will net 8 pairs per channel !!
Already have a nice Sturdy enclosure ( 4U ) and most of the power supply Including a Hammond 182V40 ( 1.5 KVA - 80VCT 18.8A ) and 10 x 10000uF 100V Panasonic TUP Series caps...
Last piece of the puzzle for the power supply is diodes ...
I have these picked out, but not sure which ones to settle for ...
RURG5060, RUHG5060,
FFH50US60S
Heat Sink would be M-Sink and would basicly cover all 4
Input please
Heres a few ideas and questions...
1 - Feeding the LME from a "clean" separate higher voltage supply than rail voltage from output. Feasible ???
2 - wouldn't these LME's accept balanced signals ?? wouldnt that be more desirable input signal on clean amp like this ???
will have more but off the top of my head this is what i got...
Using Panson's schematic i'm making my own PCB layout that will allow me to run 4 pairs or run 2 boards in series that will net 8 pairs per channel !!
Already have a nice Sturdy enclosure ( 4U ) and most of the power supply Including a Hammond 182V40 ( 1.5 KVA - 80VCT 18.8A ) and 10 x 10000uF 100V Panasonic TUP Series caps...
Last piece of the puzzle for the power supply is diodes ...
I have these picked out, but not sure which ones to settle for ...
RURG5060, RUHG5060,
FFH50US60S
Heat Sink would be M-Sink and would basicly cover all 4
Input please
Last edited:
OK...
Heres a few ideas and questions...
1 - Feeding the LME from a "clean" separate higher voltage supply than rail voltage from output. Feasible ???
2 - wouldn't these LME's accept balanced signals ?? wouldnt that be more desirable input signal on clean amp like this ???
will have more but off the top of my head this is what i got...
Using Panson's schematic i'm making my own PCB layout that will allow me to run 4 pairs or run 2 boards in series that will net 8 pairs per channel !!
Already have a nice Sturdy enclosure ( 4U ) and most of the power supply Including a Hammond 182V40 ( 1.5 KVA - 80VCT 18.8A ) and 10 x 10000uF 100V Panasonic TUP Series caps...
Last piece of the puzzle for the power supply is diodes ...
I have these picked out, but not sure which ones to settle for ...
RURG5060, RUHG5060,
FFH50US60S
Heat Sink would be M-Sink and would basicly cover all 4
Input please
Thank you for your connector comment.
I have run LME with a higher rails without any issue.
A linear regulator for the LME is absolutely feasible. LME chips have decent PSRR. I wonder how much improvement (measurable and audible) we can get.
Sounds like you aiming high power. The regulator is likely a discrete one.
IMO, balanced input does not mean a cleaner or better signal by nature. For home use, unbalanced interconnect is just fine.
Yes, according to the datasheet the PSRR is very fine. You should be able to get great performance with just by-pass capacitors.
I still think we can reduce the amount of THD+N by supplying the LME with a clean and properly designed RCL supply and possibly get even try a regulated version...
I know 16W is a lot of heat to dissipate, but whats a lil bit more when your running 8 x 250w output transistors ??
I agree.
The front end of power amp can usually be separately powered and most seem to benefit.
The decision is between series or shunt regulation.
Hi, let met add my observations on the LME 49830 (not the LME49810) but as they are close to eachother....
Well - originally I'have had a +60/-60V Regulated PSU for the LME and a 50V/-50V unregulated for the power stage.
In fact the 60V was done using the 50V of the output stage, and adding on top a 15V PSU ...
in clear ... this PSU :Imageshack - alimampli.gif
So you need to imagine a regulation after the 65V PSU. this regulation is done with a LM317.
SO ... What I observe .... removing the the regulation improves the THD+N of about 2-3 dB What ever the output power level (From 1W up to 150W).
Note that the THD also improves !!!! Conclusion: my regulation sucks !!!!
BUT, I've removed the regulation and added a Resistor in the PSU in order to have a nice CRC. C= 22000µF R=47Ohms C= 2x4700µF, for the LME.
The results are much better in term of THD+N.
So the LME is sensible to PSU, even if its PSSR looks good.
Now... I cannot tell you yet if there is any change on the sonic aspect as this Amplifier only heat a Resistor at that time. But Next week, i will listen and will tell you.
here are my 2 cents.
EDIT : My measurement system is far away from Panson's system. I use ARTA + a Computer+ an Audigy2ZS Sound card. The best THD+N I got is 0.01% and THD = 0.0025%
The Noise I've is linked to the measurement methodology, I will try to improve this later.
Well - originally I'have had a +60/-60V Regulated PSU for the LME and a 50V/-50V unregulated for the power stage.
In fact the 60V was done using the 50V of the output stage, and adding on top a 15V PSU ...
in clear ... this PSU :Imageshack - alimampli.gif
So you need to imagine a regulation after the 65V PSU. this regulation is done with a LM317.
SO ... What I observe .... removing the the regulation improves the THD+N of about 2-3 dB What ever the output power level (From 1W up to 150W).
Note that the THD also improves !!!! Conclusion: my regulation sucks !!!!
BUT, I've removed the regulation and added a Resistor in the PSU in order to have a nice CRC. C= 22000µF R=47Ohms C= 2x4700µF, for the LME.
The results are much better in term of THD+N.
So the LME is sensible to PSU, even if its PSSR looks good.
Now... I cannot tell you yet if there is any change on the sonic aspect as this Amplifier only heat a Resistor at that time. But Next week, i will listen and will tell you.
here are my 2 cents.
EDIT : My measurement system is far away from Panson's system. I use ARTA + a Computer+ an Audigy2ZS Sound card. The best THD+N I got is 0.01% and THD = 0.0025%
The Noise I've is linked to the measurement methodology, I will try to improve this later.
Last edited:
What i meant is to use a discrete ( completely separate ) supply on its own with no ties between then two as by "boosting" with a second trafo you still have the noise from the main supply. Basicly what end up happening is now you have 2x the amount of noise.
CLRC works better, but finding s 5uH choke for 10-20A is hard and expensive !
Even better a regulated then fed through a CLRC in my mind ( twisted i know ) would still be the cleanest solution along with decoupling the rectifier bridge with caps as well
Last edited:
An electrical engineer at work likes building amps and he has just built an amp around this chip. He has not used regulators and it works really good! He said National Semiconductor know what is best for LME49810! He used just by-pass capacitors as per the data sheet.
I heard this amp with a Sony SACD player and ATC speakers and it was very lovely. I know this is just subjective opinion. But this experience teaches me that the chip is capable and does not need regulators.
Good luck with it.
: )
I heard this amp with a Sony SACD player and ATC speakers and it was very lovely. I know this is just subjective opinion. But this experience teaches me that the chip is capable and does not need regulators.
Good luck with it.
: )
Yuli this is what I'm going to be using...
Rail supply has 10 x 100V/10000uF caps for "storage" and uses 3 x 470uF caps for filtering, 6 x 1uF for decoupling, 2 x 5uH chokes and 2 0R33-10W non-inductive resistors.
Still searching for a suitable pair or regulators , but the filtering arrangement will basicly stay the same. C21 & C22 is a pair of 1500uF 200V caps with a 1uF decoupler on each one.
An externally hosted image should be here but it was not working when we last tested it.
Rail supply has 10 x 100V/10000uF caps for "storage" and uses 3 x 470uF caps for filtering, 6 x 1uF for decoupling, 2 x 5uH chokes and 2 0R33-10W non-inductive resistors.
Still searching for a suitable pair or regulators , but the filtering arrangement will basicly stay the same. C21 & C22 is a pair of 1500uF 200V caps with a 1uF decoupler on each one.
I can't measure the HF effect.
Many designers say that the HF spikes and all and HF output impedance of the reg are very important to how the amplifier behaves and how that unsuppressed noise and interference affects the low level resolution.
I can hear the difference between a 317 and a Salas Shunt. The LF performance of the regs is similar. The HF performance is very different. What I hear (or rather don't) seems to be the much improved HF performance of the shunt reg.
It's the same with a snubbered or unsnubbered rectifier in the PSU.
The snubbering attenuates the HF. If done wrong the caps can increase the rectifier noise.
Many designers say that the HF spikes and all and HF output impedance of the reg are very important to how the amplifier behaves and how that unsuppressed noise and interference affects the low level resolution.
I can hear the difference between a 317 and a Salas Shunt. The LF performance of the regs is similar. The HF performance is very different. What I hear (or rather don't) seems to be the much improved HF performance of the shunt reg.
It's the same with a snubbered or unsnubbered rectifier in the PSU.
The snubbering attenuates the HF. If done wrong the caps can increase the rectifier noise.
I always have used .5uF 50V Tantalum caps and makes a world of a difference specially when used with a 220uF or 470uF cap for noise suppression.
When i switched over my supply in my Yamaha ( my test mule ) i could not believe how much of a difference it made over what was in there, but then again, new rectifier and new caps and topped it off with a new transformer ( one at a time of course ) but the biggest change was not the caps ( which i have tripled the original capacitance ), but the snubered rectifier and the common node choke i used ( ripped the ferrite ring out of an old car amp i had laying around from the mid 90's and wound the 5 turns on it ) and also used a 220uF cap across the output of the rectifier bridge for HF filtering
The amp still sounds like
and its loud enough that neighbours wanna 
Last edited: