Power supply application note
new app note!
In the app note there is also reference to app notes AN-1625, AN-1850 Reference designs for 49810 and 49830 , so I guess these will show up shortly.
new app note!
In the app note there is also reference to app notes AN-1625, AN-1850 Reference designs for 49810 and 49830 , so I guess these will show up shortly.
Yap wow ! Thanks for info rikkitikkitavi
... would be nice to build the amp with some evaluations boards ... 🙂
Anyhow nice infos in the AN ... (not only in due to the LME)
... would be nice to build the amp with some evaluations boards ... 🙂
Anyhow nice infos in the AN ... (not only in due to the LME)
I can't wait to see which output transistors uses national in the AN-1625. Any idea about what to expect? The LM4702 "reference" design was based on impossible to find sanken devices, hope the thing does not go that way again.
differences between LME49810 and LME49811
I have spent some time looking at the datasheets for the LME49810 and the LME49811 and it seems hard to understand why national is selling these two.
From what i see the LME49810 has two extra features (keeping bias while muted and baker clamp-clipping indicator) and a much higher slew rate, while the LME49811 has lower distortion at some points (but not in the whole frequency/power spectrum) and a much flatter closed-loop frequency response (Its hard to tell why the LME49810 isn't that flat).
Any idea about why national is selling/developing these two? Is the LME49811 simply a lesser LM49810?
I have spent some time looking at the datasheets for the LME49810 and the LME49811 and it seems hard to understand why national is selling these two.
From what i see the LME49810 has two extra features (keeping bias while muted and baker clamp-clipping indicator) and a much higher slew rate, while the LME49811 has lower distortion at some points (but not in the whole frequency/power spectrum) and a much flatter closed-loop frequency response (Its hard to tell why the LME49810 isn't that flat).
Any idea about why national is selling/developing these two? Is the LME49811 simply a lesser LM49810?
Re: differences between LME49810 and LME49811
None of the two is less. It is a matter of choices. For some commercial designs the baker clamp can be very useful. The LME49810 have much higher slewrate and higher current capacity then the LME49811. Personally I prefer hose two things in the LME49810.
Unless u work with darlingtons or mosfets the LME49811 will nog have enough current without an additional driver stage to feed an output stage.
With kind regards,
Bas
ionomolo said:
None of the two is less. It is a matter of choices. For some commercial designs the baker clamp can be very useful. The LME49810 have much higher slewrate and higher current capacity then the LME49811. Personally I prefer hose two things in the LME49810.
Unless u work with darlingtons or mosfets the LME49811 will nog have enough current without an additional driver stage to feed an output stage.
With kind regards,
Bas
Two pole compensation
Will any of these chips work with two-pole compensation schemes? Basically, can I stick a C-R-C where a single compensation C is being used currently?
If this works, it should give us the ability to take the open-loop gain further out into the HF range and therefore increase GNFB and reduce distortion in the HF area.
Will any of these chips work with two-pole compensation schemes? Basically, can I stick a C-R-C where a single compensation C is being used currently?
If this works, it should give us the ability to take the open-loop gain further out into the HF range and therefore increase GNFB and reduce distortion in the HF area.
Re: Two pole compensation
Two-pole compensation will of course improve the distortion but it has to be used carefully. It is not a generic compensation method like one pole. It must be tailored to the layout ( parasitics) and load. It does'nt tolerate capacitive loading or parasitic capacitors with the feedback network. So layout ( avoiding ground close to the input and feedback node) is important. One of the bad points is reaction to overload. Due to the very low phase margin in a frequency band, the recovery becomes slow from non linear operation
Conclusion, use it if layout, load and dynamics of signal is well controlled and in the right range. In this way, it is a DIY solution.
JPV
tcpip said:
Two-pole compensation will of course improve the distortion but it has to be used carefully. It is not a generic compensation method like one pole. It must be tailored to the layout ( parasitics) and load. It does'nt tolerate capacitive loading or parasitic capacitors with the feedback network. So layout ( avoiding ground close to the input and feedback node) is important. One of the bad points is reaction to overload. Due to the very low phase margin in a frequency band, the recovery becomes slow from non linear operation
Conclusion, use it if layout, load and dynamics of signal is well controlled and in the right range. In this way, it is a DIY solution.
JPV
Re: Re: Two pole compensation
Okay, thanks. I hadn't realised it's difficult to use. I have relatively little practical experience with amp designing, but I saw Randy Slone uses it in his designs quite liberally, so I thought it would be a good idea.JPV said:
Re: Re: Re: Two pole compensation
It is a good idea but it will require some testing with real load real layout and overload. Tuning componants value will help.
JPV
tcpip said:
It is a good idea but it will require some testing with real load real layout and overload. Tuning componants value will help.
JPV
AN-1850 is now on National's web site, for those interested.
http://www.national.com/an/AN/AN-1850.pdf
There are also files for the boards used in AN-1849 and AN-1850 if anyone wanted to build them. Get all the info here:
http://www.national.com/analog/audio/pwramps_evaluation_boards
-SL
http://www.national.com/an/AN/AN-1850.pdf
There are also files for the boards used in AN-1849 and AN-1850 if anyone wanted to build them. Get all the info here:
http://www.national.com/analog/audio/pwramps_evaluation_boards
-SL
Hi Spittin Llama,
Thanks for sharing the link. Would you recommend this amp for subwoofer application? Seems like a good sounding amp based on the posted parameters.
Thanks for sharing the link. Would you recommend this amp for subwoofer application? Seems like a good sounding amp based on the posted parameters.
Output device OK ?
Hi,
I'm planning on buying a set (2) of LME49810 from e-bay for $15 including shipping but first I'd like to know if my output device are OK. These are original (15 years old roughly) Toshiba 2sc3281 / 2sa1302.
How many device per rail would I need to get let's say 200W@8ohm ?
I already received as a sample a set of LME49830 and I'm planning to use some original (15 years old roughly) TO-3 Mosfet model 2SK134 and 2SJ49. I'm waiting to build this amp since I think the bipolar transistor will sound better than the mosfet...I only have one power source at +- 85Vdc salvaged froma Crown CE-2000.
Based on your experience would the bipolar output device be OK or should I go with the mosfet amplifier ?
TIA,
Eric
Hi,
I'm planning on buying a set (2) of LME49810 from e-bay for $15 including shipping but first I'd like to know if my output device are OK. These are original (15 years old roughly) Toshiba 2sc3281 / 2sa1302.
How many device per rail would I need to get let's say 200W@8ohm ?
I already received as a sample a set of LME49830 and I'm planning to use some original (15 years old roughly) TO-3 Mosfet model 2SK134 and 2SJ49. I'm waiting to build this amp since I think the bipolar transistor will sound better than the mosfet...I only have one power source at +- 85Vdc salvaged froma Crown CE-2000.
Based on your experience would the bipolar output device be OK or should I go with the mosfet amplifier ?
TIA,
Eric
Fredlock, it would be fine for a subwoofer. Typically you just want plenty of headroom on the power amp for the sub for good dynamic range along with a high current power supply. So an amplifier with the power you need with a correct power supply will work great for a sub, no matter if it is an LME49830 amp or something else.
Alexcd, just read the note, it shows 8 ohms from a +/-60V supply. If you want more power, increase the supply voltage. You have to make sure that you stay within the SOA of the output devices which is what is going to limit the supply voltage and hence the output power.
e_fortier, I don't know if that question is for me but even if it is I don't know what you are actually asking. Sorry.
If none of the questions were meant for me then sorry to jump into other's discussions. If you haven't read the note do that first then ask for help from those that know more, which may or may not be me. I am not an expert about anything.
-SL
Alexcd, just read the note, it shows 8 ohms from a +/-60V supply. If you want more power, increase the supply voltage. You have to make sure that you stay within the SOA of the output devices which is what is going to limit the supply voltage and hence the output power.
e_fortier, I don't know if that question is for me but even if it is I don't know what you are actually asking. Sorry.
If none of the questions were meant for me then sorry to jump into other's discussions. If you haven't read the note do that first then ask for help from those that know more, which may or may not be me. I am not an expert about anything.
-SL
Hi SpittinLama,
The question was a general one using the LME49810 and not specifically aimed at you per say. My question was simply asking for help or advice in choosing between a MOSFET amp vs a bipolar one using the output devices I mentionned in my previous post and based on these devices and a rail voltage of +- 85Vdc how many device would I need to maintain a good SOA.
TIA
Eric
The question was a general one using the LME49810 and not specifically aimed at you per say. My question was simply asking for help or advice in choosing between a MOSFET amp vs a bipolar one using the output devices I mentionned in my previous post and based on these devices and a rail voltage of +- 85Vdc how many device would I need to maintain a good SOA.
TIA
Eric
Re: Output device OK ?
Not just devices per rail, but the thermal impedance of the heat sink. I think that the recommendation for the 125W version was 4 inches of Aavid Thermalloy 65605. If you read that article you can work backwards the supply voltage to the thermal impedance you need.
Aavid has revamped their website -- well worth a visit. For the moment Arrow North America's (arrow is a distributor of Aavid's stuff) website is down over the 4th of July Weekend.
Maybe I will have to write a little TCL program for amplifier heat sink calculations.
e_fortier said:
Not just devices per rail, but the thermal impedance of the heat sink. I think that the recommendation for the 125W version was 4 inches of Aavid Thermalloy 65605. If you read that article you can work backwards the supply voltage to the thermal impedance you need.
Aavid has revamped their website -- well worth a visit. For the moment Arrow North America's (arrow is a distributor of Aavid's stuff) website is down over the 4th of July Weekend.
Maybe I will have to write a little TCL program for amplifier heat sink calculations.
Just finished a mosfet (IR) LME49810 amp with a couple of ccs's (a sink to the source and a source to the sink) at about 10 ma's to prevent the driver from leaving class A at high slew rates and found them to give a better sound.
Definitely something to try, given the price of the ccs's (0.20€ each).
Definitely something to try, given the price of the ccs's (0.20€ each).