| glennb |
Hi chip heads,
LME49810 is a new audio power amp driver chip which appears in National Semiconductor's latest Analog Product Guide (2Q2007), available at http://www.national.com/whatsnew/fi...ct_selguide.pdf (88 pages, 1.7MB)
The product folder is at http://www.national.com/pf/LM/LME49810.html and the PDF data sheet (22 pages) can be downloaded from there.
Its been briefly mentioned in the LM4702 thread http://www.diyaudio.com/forums/show...&postid=1222529 but I thought a new thread should be started.
The LME49810 is a 'single' cousin of the 'dual' LM4702, although its output current is about ten times that of the LM4702 and new features have been added like soft clamping and dedicated bias pins for the Vbe multiplier. The power supply rails are up to +/- 100V, slew up to 50 V/uS.
My samples have just gone into 'shipped' status from Singapore, so the first prototype amp is not far away.... |
|
|
| glennb |
About 10 minutes after that post the nice UPS guy knocked on the front door. Ripping into the package and pulling out the camera....
The LME49810 is in the middle, with a ruler for scale. It is flanked by a LM4702 and LM4780 also for scale. It has tiny pins like the LM4780, but at least they don't have to pass the full speaker output current! 2mm pin spacing on the left and 4mm pin spacing on the right. |
|
|
| sayang001 |
My samples are due to arrive today as well. Will start some testing with TIP142 & TIP147 first, mainly following the shematic in the datasheet.
Next step would be a Sziklai Output like the C200 from ampslab here:
http://www.ampslab.com/c200cfp.htm
Of course I'm open for other suggestions. |
|
|
| Nordic |
| National wanyed $56 to ship the samples... madness!, I'll just wait for a local retailer to stock these. |
|
|
| Russ White |
Yes, I too have taken notice of this chip. :) I would say it is much better suited for a FET amp than the 4702 is. And I have my samples and will be testing that scenario with IRFP240 and IRFP9240 very soon.
The bias arrangement and Baker clamp, which basically helps keep the amp stable into clipping if I understand correctly, make this chip much more robust than the 4702. I just want to see how it sounds first, and the MOSFET arrangement here could not be simpler.
Here is a brief explanation of a Baker Clamp from on old TI doc:| quote: |
What is the function of a Baker Clamp?
The collector current of the power transistor in a switching regulator is proportional to variable output load current. Normally, with bipolar transistors the fixed base drive current is optimized for a maximum output load current. This can result in unacceptably large storage time at light load, because the transistor will be driven into deep saturation. The baker clamp prevents deep transistor saturation by providing a path for excessive base drive current. Many applications, such as flyback and forward converters, are utilizing this technique. The baker clamp diode must have a fast reverse recovery time.
|
Cheers!
Russ |
|
|
| alexcd |
| I expect big things from you guys with this chip. :) |
|
|
| dfdye |
| I'm not ready to ditch the 4702 yet, but it certainly seems this chip will do everything the 4702 would and then some. I do wonder what the price point will be. |
|
|
| glennb |
I've put up a web page for this chip at http://home.pacific.net.au/~gnb/audio/lme49810.html, to sit along-side my pages for the LM4702 and LM4780. More information will be added as experimentation with the chip proceeds. Feel free to send me comments, corrections, ideas.
The LM4702 would still be preferred for lower power stereo amps, but for high power mono blocks the LME49810 looks better. |
|
|
| Russ White |
| I am not sure I can find a good reason to prefer the LM4702 for any application..... Honestly. |
|
|
| dfdye |
| price?? do we know what they are charging for this thing yet? |
|
|
| jackinnj |
| quote: | Originally posted by Russ White
I am not sure I can find a good reason to prefer the LM4702 for any application..... Honestly. |
It's two channels, so less real estate consumption. |
|
|
| Russ White |
| quote: | Originally posted by jackinnj
It's two channels, so less real estate consumption. |
Sure, but that does not carry much weight...
Also to get the same capability you would need to add more parts to the LM4702 (clamp and drivers etc) which would probably take more space than two 49810s.
Not saying the 4702 is bad, but I think the 49810 is better (on paper as I still have to hear it) in my book.
Cheers!
Russ |
|
|
| jackinnj |
for DIY it's a small consideration -- for a manufacturer of high end products heeding National's guidelines there's the need for localized decoupling and storage at the power pins of either chip --
a hint for both chips -- mount the decoupling caps on the underside of the PCB -- and if possible use polypropylene devices -- you want the legs on the decoupling caps to the as short as possible. |
|
|
| Nordic |
Despite the extra cost and size, I do still prefer mono chips to their stereo competition... i.e. i'd rather use opa134 than 2134....
As seen from a sonic quality viewpoint... not to mention some of the other distortion mechanics involved in a shared chip... |
|
|
| Russ White |
| quote: | Originally posted by jackinnj
for DIY it's a small consideration -- for a manufacturer of high end products heeding National's guidelines there's the need for localized decoupling and storage at the power pins of either chip -- |
Hmm I thought this was DIY aduio.... :)
But just to play along, I am sure other factors important to commercial applications are:
- Overdrive stability
- Flexible drive capability
- Wider range of power device choices
I don't see how it would be any more difficult to produce a great stereo amp with the mono chip. In fact it would probably be less expensive to use it if you matched feature for feature.
It may be an apples/oranges comparison.
The 4702 simply does not have the same capabilities on its own. Thats the main point. |
|
|
| ozonek |
I am thinking about the difference between LM4702 and LME49810. Is there possibility that these two chips are made using same die and configured (enable/ disable certain circuits such as Baker clamp) according to their own specs. For memory devices, it is common practice, not sure if it is the same in the analog world.
Has anyone opened these two chips to see the difference of the die itself? |
|
|
| glennb |
| quote: | Originally posted by ozonek
I am thinking about the difference between LM4702 and LME49810. Is there possibility that these two chips are made using same die and configured (enable/ disable certain circuits such as Baker clamp) according to their own specs. For memory devices, it is common practice, not sure if it is the same in the analog world.
Has anyone opened these two chips to see the difference of the die itself? |
The chip design of the 49810 is probably based on half a 4702, with improved slew rate, added baker clamp, clip flag, higher current output transistors. Neither are listed on http://www.national.com/packaging/parts/DIE.html
I have yet to have the misfortune of letting the blue smoke out of a 4702 or 49810, but if I do, it would be worth trying to expose the die. |
|
|
| minicar |
| quote: | Originally posted by glennb
The chip design of the 49810 is probably based on half a 4702, with improved slew rate, added baker clamp, clip flag, higher current output transistors. Neither are listed on http://www.national.com/packaging/parts/DIE.html
I have yet to have the misfortune of letting the blue smoke out of a 4702 or 49810, but if I do, it would be worth trying to expose the die. |
I think you are right. It is base on the 4702 but not the same die. Their slew rate and some other spec are different, should be another design. |
|
|
| Russ White |
All you have to do is ask....
I called National. It is a different die. It is in fact a significantly different bias scheme. Though National will not give away the farm. ;)
According to the nice gentleman I spoke with there are more new devices on the way as well... Could it be a new dual part coming based on the 49810? Hmmmmm that would be sweet. :)
Cheers!
Russ |
|
|
| Eva |
| I really miss a current limiting feature in those chips :bawling: :bawling: |
|
|
| jackinnj |
 |
|
|
| Russ White |
More important than the distortion of the test circuit is the distortion of the final amp. And current drive and slew rate will have a lot to do with that. Also you won't have as much XTALK with a properly designed LME49810 circuit(or crcuits, dual mono).
LM4702
XTALK Channel Separation (Note 11) f = 1kHz at Av = 30dB 85dB
SR Slew Rate VIN = 1.2VP-P, f = 10kHz square Wave, Outputs shorted 15 V/μs
LME49810
SR Slew Rate VIN = 1VP-P, f = 10kHz square Wave 50 V/μs(min)
Sure the distortion of the LM4702 looks great in the test configuration (no load, especially not a capacitive one). Driving a realistic power stage I would think the LME49810 would have a clear advantage, especially with MOSFETs.
So while thats an interesting graph, I would say its not very representative of the results you will get with a real power amplifier, even using National's excellent guidelines.
I would say the LME49810 is in part the result of lessons learned on the LM4702.
Cheers!
Russ |
|
|
| Russ White |
Also notice this graph from the LME49810 DS,
The standard (there is only one part) LM49810 easily surpasses the 4902C and equals (or betters under a real load) the 4902B.
When reading these data sheets always keep in mind, we are talking no load here. Add a load and the story will change dictated by the capability of the driver to adequately meet the demands of the load. |
|
|
| jackinnj |
the load on the source and sink is different on the two test boards --
with respect to slew rate -- (among other things) this is a function of the compensation capacitor value -- there is a forthcoming application note on compensating all of these driver chips. i have played around with this quite a bit and found that a couple tweaks were necessary to max the bandwidth without having the amplifier self destruct -- found that 30 pF was about right. i also found that the feedback network had to be compensated a bit -- i have run the amp thusly at 140 kHz without catastrophe!
in a new application note, National shows BD139 and BD140 transistors to drive a variety of MOSFETs (Renesas, Magnatec and Vishay) -- this is a good idea.
here's a comparison of the two driver chips -- they probably used 30kHz bw on their distortion analyzer.
 |
|
|
| Russ White |
| quote: | Originally posted by jackinnj
with respect to slew rate -- (among other things) this is a function of the compensation capacitor value -- there is a forthcoming application note on compensating all of these driver chips. |
Yes that will be a good app note, but I doubt it will change the maximum slew rate of the LM4702 which is stated in the data sheet. and I have read the FET app note, but you would not need any additional driver to use the LME49810 with many FETs. |
|
|
| Russ White |
| quote: | Originally posted by jackinnj
|
You should note that that is the expensive LM4702B and not the more common LM4702C. |
|
|
| jackinnj |
| quote: | Originally posted by Russ White
You should note that that is the expensive LM4702B and not the more common LM4702C. |
I was never able to get my Lateral MOSFET amplifier to attain the THD+N% levels of the reference design -- but it was (is) still very, very good -- the "B" has just become available in quantity and is a little more than twice the price of the "C".
I have an LM4702 amp sitting on the bench as we speak -- oscillating at 8.123 MHz. :bigeyes: |
|
|
| Russ White |
Yes I bet the FET LM4702 amp sounds great. But even in the app note I think in most cases your are running the LM4702 very close to its margins with respect to the bias voltage. I guess what I am saying is I would doubt very much that with a real amp the LM4702 performs any better than a LME49810. You can look at the LME49810 as having a built in pre-driver for FETs etc.
Lets say you started with an LM4702 plus BD139 and BD140 transistors to drive FETs. Naturally your THD+N would be much worse than the LME49810 or LM4702 alone.
Also, I just don't see being able to compensate the LM4702 so it could match the LME49810's slew rate. Though I will anxiously await the app note to see what National says. |
|
|
| ozonek |
| Will NS continue their development on LM4702A (the top grade) while they now have a better design like LME49810? (dual version is coming soon?) |
|
|
| glennb |
| quote: | Originally posted by ozonek
Will NS continue their development on LM4702A (the top grade) while they now have a better design like LME49810? (dual version is coming soon?) |
49810 is the evolution of the 4702 technology. A dual 49810 in a 4780-like package would make the 4702 obsolete. After all, they stuffed two 3886 dice in the 4780, so why not use the same trick again. |
|
|
| Gcollier |
No question about it...this is an exciting new part :D It is very clear from the datasheets and application notes that the engineers at National used what they learned from the LM4702 to build an improved version.
For those of you who are just as excited as me I have created library for this part, and put the datasheet schematic into Eagle. I will be coming up with a board design in the next few days (if the wife gives me enough free time that is :whazzat: ) I am intending on using BJT's and following the app note for now. I'll give MOSFETs a try later if it warrants it.
Have fun with the files (and yes I did misname the osense pin onsense).
G. |
|
|
| Gcollier |
| Here is the schematic from the datasheet in eagle format. Nothing like waitng 60 seconds to post again ;) |
|
|
| gmikol |
| quote: | Originally posted by Gcollier
I will be coming up with a board design in the next few days (if the wife gives me enough free time that is :whazzat: ) I am intending on using BJT's and following the app note for now. |
I hate to steal gcollier's thunder, but I was killing time and I came up with this board design (used my own symbol). It fits in the standard free Eagle 80x100mm footprint. What do y'all think? Be gentle on me, I'm pretty new at laying out boards. But I want to get better, so fire away.
I know there's one issue already, and that's the output sense line (the thin red line closest to KK2) runs parallel to the bias lines for quite some time, but I can't figure out how to fix that.
--Greg |
|
|
| gmikol |
| ... and the schematic. |
|
|
| Gcollier |
| Nice Work! I was laying out a board last night...but your design looks pretty darn good at first glance. Would you mind sharing the files so we can all have a go at it? |
|
|
| gmikol |
You hardly need the driver for this configuration. Since 50mA * 75 hFE(Min) = 3.75A, for MJL3281, good for 30V rails into 8 Ohms. This is about as hard as you'd want to push a single device anyway. But I laid it out as a demonstrator for doing larger output stages. If and when I get around to building any of this, I may try a driver-less single device output stage just for the heck of it.
I haven't shared any Eagle files before. Here's a .zip with the BRD and SCH files. Let me know if you need anything else.
--Greg |
|
|
| Gcollier |
Thanks for the files they work perfectly fine for me :)
I hope to get around to experimenting with this soon. It will likely become the replacement in my guitar amp build which is currently using an LM4780. I'm going to be pushing 60 volts on each rail so I will have to make some adjustments, mostly to the output stage...but I really like your tidy layout.
Thanks Again!
G. |
|
|
| BWRX |
Hi Greg. According to your schematic you don't have AGND connected to PGND. You need to have them connected, preferably at one point.
Also, your 10uF input coupling caps with the 243ohm input resistors set the -3dB high pass frequency at 65Hz... That's pretty high and will be very audible. I would raise the values of the input and feedback resistors to get the -3dB frequency down around 10Hz or below so you don't have to increase the value of the coupling cap. |
|
|
| Russ White |
| quote: | Originally posted by BWRX
Hi Greg. According to your schematic you don't have AGND connected to PGND. You need to have them connected, preferably at one point.
Also, your 10uF input coupling caps with the 243ohm input resistors set the -3dB high pass frequency at 65Hz... That's pretty high and will be very audible. I would raise the values of the input and feedback resistors to get the -3dB frequency down around 10Hz or below so you don't have to increase the value of the coupling cap. |
The 10uf cap sees a resistance of around 7K to GND. A 10uf cap should be sufficient for that. But the Power GND and Signal GND does seem to be disconnected.
Cheers!
Russ |
|
|
| glennb |
| quote: | Originally posted by Russ White
The 10uf cap sees a resistance of around 7K to GND. A 10uf cap should be sufficient for that. |
Yep, agree. There is no reason why RS=6.81K can't be increased to 22K. C7=10u can then be reduced to 0.47u - 1u, which is a more reasonable size in a film cap. That will give f-3 of 14 Hz or 7 Hz.
| quote: | | But the Power GND and Signal GND does seem to be disconnected. |
They do need to be connected. The AGND should be a 'star' network and the PE (PGND) should be 'star' network, and the two star points should be linked by one thin conductor. This prevents high currents in the PE ground circulating into the AGND connections, and degrading THD.
RIN=243 Ohm is not actually required for the amp to work. However, it can be used to limit the HF bandwidth to prevent RF getting into the amp, provided a small cap is added across RS, say about 10nF. f-3 with then be 65 KHz. It can be scaled, ie. 2430 Ohm and 1nF to get it in the range of a polystyrene or mica cap. |
|
|
| BWRX |
Quite right Russ. I looked at his schematic and thought I saw 10uF in both places, but the schematic clearly shows 220uF in series with the feedback resistor. My mistake.
As glenn also mentions, I also think it would be a good idea to increase the resistance values so smaller value caps may be used. |
|
|
| gmikol |
Thanks for all the replies everyone. My initial goal was to lay things out according to the datasheet, and to provide the option to short the input and feedback caps. I think there's a typo in the National datasheet that describes the input pole for Rin/Cin. I agree on increasing the resistance.
Re: The grounding. I know that PE and AGND are separate on the board. I was planning on having the star on the PSU board. Any reason not to?
Aside from the grounding issues, it sounds like the layout is generally sound. Power amp layout is a lot different than submicron SRAM layout.
--Greg |
|
|
| Triodas |
I don't think that increasing the resistance values would be a good idea. I have got experince with LM3875: lower resistance values - better sound. But I don't know how to explain this phenomena.
Pls excuse my english. |
|
|
| gmikol |
| quote: | Originally posted by Gcollier
I'm going to be pushing 60 volts on each rail so I will have to make some adjustments, mostly to the output stage |
Yeah...some big adjustments. I've been trying to learn about SOA and thermal calculations, and by the spreadsheet I put together, it looks like you'd need 4 pairs and a pretty beefy heatsink (< 1 C/W). Of course, there's no guarantee that my spreadsheet is right.
Good luck...
--Greg |
|
|
| BWRX |
We mean use slightly higher resistance values, not outrageously higher values. For example, increase the 243ohm resistor to 2k and the 6.81k to 56k. Now a 10uF cap will give you a -3dB frequency of 8Hz in the feedback loop, and a 1uF cap will give you a -3dB frequency of 3Hz at the input. Now you can use very good quality caps in the signal and feedback paths. The tradeoff is of course slightly higher noise because of the higher value resistors.
The next step up is to get rid of the caps and DC couple the signal. |
|
|
| Avenida |
| there will be a demo some time next month in London to show the design of using these high voltage driver |
|
|
| AndrewT |
| quote: | Originally posted by BWRX
........ Now a 10uF cap will give you a -3dB frequency of 8Hz in the feedback loop, and a 1uF cap will give you a -3dB frequency of 3Hz at the input. ......... |
I recommend that the filter frequencies be reversed so that the input filter provides the effective bandwidth limit for the amplifier.
I think that using the NFB roll-off as the bandwidth limiter is not a good way for good sound.
I would go even further.
Keep the F-3db=3Hz at the input and change the NFB F-3db to about 2Hz (needs an RC about 4times higher than the BWRX proposal). |
|
|
| glennb |
| quote: | Originally posted by gmikol
You hardly need the driver for this configuration. Since 50mA * 75 hFE(Min) = 3.75A, for MJL3281, good for 30V rails into 8 Ohms. This is about as hard as you'd want to push a single device anyway. But I laid it out as a demonstrator for doing larger output stages. If and when I get around to building any of this, I may try a driver-less single device output stage just for the heck of it.
--Greg |
Greg as you probably know, hFEs on the spec sheet are meant to be a guide only. A real-life batch of transistors from a particular manufacturer tested at a load current of a few Amps on a heatsink might show hFE anywhere from 50 - 120.
Also, your example was good for an 8 Ohm resistive load, but for a speaker system load you should allow for impedance dips and V-I phase angle changes around the speaker resonances. The output transistors' peak current and average power dissipation should be calculated as if a 3-4 Ohm load existed, for an 8 Ohm nominal speaker system. There is a fair bit of leeway here because the low impedance loads only exist at some frequencies and the average long term power in music at these frequencies may not be significant.
Even with a 50mA drive capability from the '48910, its still not much good to directly drive a typical output transistor to achieve over about 25 Wrms output into 8 Ohms.
MOSFETs and Darlingtons are a different situation... |
|
|
| monkey29 |
:smash: The LME49810 as the core, one to three pairs of Toshiba output stages.
heatsink size: 230mm x 110 mm
PCB size: 210mm x 60mm
Further experiments and then published the results .
Display samples. |
|
|
| monkey29 |
| 49810 Module show!(2/3) |
|
|
| monkey29 |
| 49810 Module show(3/3) |
|
|
| AndrewT |
| quote: | Originally posted by glennb
Greg as you probably know, hFEs on the spec sheet are meant to be a guide only. A real-life batch of transistors from a particular manufacturer tested at a load current of a few Amps on a heatsink might show hFE anywhere from 50 - 120.
Also, your example was good for an 8 Ohm resistive load, but for a speaker system load you should allow for impedance dips and V-I phase angle changes around the speaker resonances. The output transistors' peak current and average power dissipation should be calculated as if a 3-4 Ohm load existed, for an 8 Ohm nominal speaker system. There is a fair bit of leeway here because the low impedance loads only exist at some frequencies and the average long term power in music at these frequencies may not be significant.
Even with a 50mA drive capability from the '48910, its still not much good to directly drive a typical output transistor to achieve over about 25 Wrms output into 8 Ohms.
MOSFETs and Darlingtons are a different situation... |
Hi,
the peak current into a speaker can approach twice what is predicted using nominal impedance. I tend to assume that the minimum impedance for short term transient signals is half the Re of the driver.
On that basis an 8ohm speaker could present 3r0 to the amp and should not damage it nor activate the protection.
4ohm speakers could present 1r5 to 2r0 to the amplifier.
That is why I keep repeating that chipamps are not suitable for 4ohm speakers if you expect to get near maximum output. |
|
|
| monkey29 |
| The LME49810 Amplifiers test pictures. |
|
|
| monkey29 |
No.3
Passed the test and the actual hearing, the overall performance has improved over the LM4702. This benefit from the Slew Rate and mono design.
Testing the use of 2SC5200/2SA1930 (Fairchild Ver.), Hearing on the flu fluent, dynamic.
The overall impression that the LME49810 is a good chip. |
|
|
| Triodas |
Hello Monkey29,
My LME49810 samples will be delayed, so I can't do my tests.
What is sounding difference betwen LME49810 and LM4702?
Thanks. |
|
|
| VLSI |
Hi monkey29
Great to see you are developing a PCB for the LME49810 chip so soon.
Tony and I are very impressed with your work with the LM4702 and are very excited to try your LME49810
Board as soon as it is available.
I have built two LM4702 amps with my own design boards and Tony and I have built one each with yours.
Although they basically all sound the same, There are slight differences in the way they sound and it is hard to say exactly why.
I prefer not to say which ones sound the best until I have determined why.
I have a question! In the NS applications note AN-1490 for the LM4702 NS have specified a 15nf/1250v polypropylene capacitor across source and sink. I note you have used 100nf/100v “orange drop” polyester in your design and it is across the trimmer! I have moved the capacitor to the original NS position and replaced it with 15nf/300v mica (this is a big cap for a mica but I just happened to have some) and I am sure this improves the sound.
Why have NS specified such a high quality capacitor for this application?
Also would you consider developing a compact LME49810 board for just one pair of Darlington or MOSFET transistors without drivers and DC coupled throughout.
I know I should try and develop it myself but your boards, as so neat it’s hard to compete.
Keep up the good work.
Greg |
|
|
| jackinnj |
| quote: | Originally posted by VLSI
Hi monkey29
I have a question! In the NS applications note AN-1490 for the LM4702 NS have specified a 15nf/1250v polypropylene capacitor across source and sink. I note you have used 100nf/100v “orange drop” polyester in your design and it is across the trimmer! I have moved the capacitor to the original NS position and replaced it with 15nf/300v mica (this is a big cap for a mica but I just happened to have some) and I am sure this improves the sound. |
Leave the mica caps to the compensation circuitry and RF bypass on the input -- a high quality PP or PE cap across the VBE multiplier couples source and sink at audio frequencies -- The 15nF/1250V ECG cap is very, very low inductance -- you can also use a WIMA PP cap here. Digikey seems to be a preferred vendor for the Nat Semi folks and DK doesn't carry WIMA. Mica is not correct here -- fwiw, I found that 100nF across the VBE multiplier works just fine...
I have built the AN1490 amplifier with and without the 30pF mica from Source to Base of the VBE multiplier and can neither detect or measure a difference in performance.
With the LM4702 and LME49810 you can play around with the compensation capacitors to trade off stability vs slew rate...more to follow...stay tuned. |
|
|
| monkey29 |
| quote: | Originally posted by Triodas
Hello Monkey29,
My LME49810 samples will be delayed, so I can't do my tests.
What is sounding difference betwen LME49810 and LM4702?
Thanks. |
The LME49810 and My M-4702-F module, all have a good sense of hearing. Sound of performance , LME49810 slightly better, more details.
I think the voices of the description is more difficult. After listening to more, there will be further awareness. |
|
|
| monkey29 |
| quote: | Originally posted by VLSI
Hi monkey29
Great to see you are developing a PCB for the LME49810 chip so soon.
Tony and I are very impressed with your work with the LM4702 and are very excited to try your LME49810
Board as soon as it is available.
I have built two LM4702 amps with my own design boards and Tony and I have built one each with yours.
Although they basically all sound the same, There are slight differences in the way they sound and it is hard to say exactly why.
I prefer not to say which ones sound the best until I have determined why.
I have a question! In the NS applications note AN-1490 for the LM4702 NS have specified a 15nf/1250v polypropylene capacitor across source and sink. I note you have used 100nf/100v “orange drop?polyester in your design and it is across the trimmer! I have moved the capacitor to the original NS position and replaced it with 15nf/300v mica (this is a big cap for a mica but I just happened to have some) and I am sure this improves the sound.
Why have NS specified such a high quality capacitor for this application?
Also would you consider developing a compact LME49810 board for just one pair of Darlington or MOSFET transistors without drivers and DC coupled throughout.
I know I should try and develop it myself but your boards, as so neat it’s hard to compete.
Keep up the good work.
Greg |
Thank you for your support !
The existence of the capacitance, high-frequency distortion rate will be improved. Value increases, the transistor base-base of the lower impedance.
NS will be published LME49830, High Fidelity Driver
for MOSFETS |
|
|
| monkey29 |
PCB processing completed, is expected to release next week, with matching heat sink.
PCB Size: 210mm x 60mm (2mm/75um) |
|
|
| AndrewT |
Hi,
that assembly looks pretty and neat.:)
You should sell copies. |
|
|
| monkey29 |
| Yes. on my websit ... |
|
|
| sayang001 |
| Will there be a version with 10 (2 x 5) output transistor ? Would be really nice for heavy PA use. |
|
|
| glennb |
Yay, NS finally did a news release for the LME49810 on 23 July 2007, see http://www.national.com/news/item/0,1735,1269,00.html
The news release for the new LME49860 dual +/-22V opamp is linked from that page. Its an uprated LME49720 (aka LM4562).
Its spreading like wild-fire around the electronics news web sites. |
|
|
| panson_hk |
This is the schematic of my design under development. The output BJT is OnSemi ThermalTrak, NJL3281D and NJL1302D. The feedback network and protection circuit is a copy of that of the Leach Amp.
I will use prototyping board to try the design before going to fab the PCB. |
|
|
| panson_hk |
| How about using the test circuit/demo circuit in the data sheet as headphone amp? |
|
|
| megajocke |
| Will the protection work? The datasheet doesn't say the LME49810 is current limited. Do you know if it is? |
|
|
| artquake |
Hi monkey29,
what a chance. I wrote you today an email to your office, in the hope that they will understand english because the homepage i found was all in chinese ... and now i find you here ... Please contact. I am interested in a pair of LME49810 Amps. Are they finshed allready with heatsink or do i need to solder ?
Wich Power Supply does this set need ?
Thanks in advance for your efforts
Michael |
|
|
| panson_hk |
| quote: | Originally posted by megajocke
Will the protection work? The datasheet doesn't say the LME49810 is current limited. Do you know if it is? |
The protection is for the output stage. |
|
|
| AndrewT |
| quote: | Originally posted by panson_hk
.........The feedback network and protection circuit is a copy of that of the Leach Amp........ |
Hi,
the Leach protection has omitted a resistor to rail.
The Leach protection locus has a constant current characteristic when Vce>~Vrail, i.e.single slope. This allows overload of the output transistors into reactive loads.
Go look at the Leach clone thread for the location of the extra resistor for true two slope protection.
Better still ask Mikeks for a copy of his paper on protection schemes and how to calculate the component values. He also shows a better method of allowing transient signals through without triggering the protection. |
|
|
| mag001 |
Hi Dear Monkey29 & Panson_HK:
I am editor for a HiFi & DIY Magazine in Beijing, I wish you can write a artical about your LME49810 jobs for my magazine. could you please send a email to me?
my email is: radio@radio.com.cn
Thank you all! |
|
|
| megajocke |
| quote: | Originally posted by panson_hk
The protection is for the output stage. |
You are missing the point. If the output stage protection activates it works by shunting away drive current from the VAS in the case of the Leach or the chip in this amplifier. The VAS in the Leach is current limited so it won't blow up when the output stage protection activates and shorts its output to ground. So I ask again, is the chip current limited? |
|
|
| glennb |
| quote: | Originally posted by megajocke
You are missing the point. If the output stage protection activates it works by shunting away drive current from the VAS in the case of the Leach or the chip in this amplifier. The VAS in the Leach is current limited so it won't blow up when the output stage protection activates and shorts its output to ground.
So I ask again, is the chip current limited? |
Yes, LME49810 spec is 50mA output current capability, which is much more than typical VAS's. |
|
|
| panson_hk |
| quote: | Originally posted by megajocke
You are missing the point. If the output stage protection activates it works by shunting away drive current from the VAS in the case of the Leach or the chip in this amplifier. The VAS in the Leach is current limited so it won't blow up when the output stage protection activates and shorts its output to ground. So I ask again, is the chip current limited? |
Thanks for pointing this potential problem. My current idea is put the chip in mute mode and move D4/D3 after R6/R7 (base of Q1/Q2). |
|
|
| megajocke |
| Does anyone know if it is current limited or not? It kind of sucks if it isn't, really. If it is current limited at 50mA, it wouldn't matter much that this current is higher than what a typical VAS can source/sink. The internal diagram does not show any current limiting, but it is a simplified diagram and these often omit stuff like current limiting. So it might still have it. |
|
|
| Russ White |
| quote: | Originally posted by megajocke
Does anyone know if it is current limited or not? It kind of sucks if it isn't, really. If it is current limited at 50mA, it wouldn't matter much that this current is higher than what a typical VAS can source/sink. The internal diagram does not show any current limiting, but it is a simplified diagram and these often omit stuff like current limiting. So it might still have it. |
No it is not current limited. At least according to the design engineer I called to ask. |
|
|
| Russ White |
| I suppose the only way to know for certain would be to short the output with an ammeter. |
|
|
| panson_hk |
| quote: | Originally posted by Russ White
No it is not current limited. At least according to the design engineer I called to ask. |
How about a current-limited voltage regulator for the chip? |
|
|
| Russ White |
| quote: | Originally posted by panson_hk
How about a current-limited voltage regulator for the chip? |
That is a very interesting idea.... |
|
|
| glennb |
| quote: | Originally posted by Russ White
I suppose the only way to know for certain would be to short the output with an ammeter. |
Or to be slightly nicer, hook up the Fig. 3 test circuit in the datasheet with +/-50V rails and load the output with a 470 Ohm 5W resistor. Drive the amp with say a 1Khz sine wave and gradually turn up the wick until the output starts to clip on a CRO. The maximum output current capability is then simply Vpeak/R. Hopefully the chip won't expire.
The slew rate limit can also be measured by increasing the frequency until the output is a triangle wave, then measure the slope in V per microsecond. Ccomp influences the slew rate.
I just haven't had time to try this with my samples.... |
|
|
| AndrewT |
| quote: | Originally posted by panson_hk
How about a current-limited voltage regulator for the chip? |
yes a CCS set to quiescent +20% and then shunt regulate that extra 20% to ground.
The chip is then protected at quiescent current level until the shunt starts to reduce it's demand to try to keep the supply voltage up, plus peak/transient demand from any supply caps after the CCS.
Simple and short circuit proof. |
|
|
| panson_hk |
| I built the data sheet test circuit (Fig. 3) on a prototyping board as shown in the image. The circuit is powered by an unregulated +/- 22 VDC. The scope shows the output of a 1 kHz sinewave. |
|
|
| panson_hk |
| The Cc in the prototype is 22 pF. I then measured some curves using AP S1. The output residual noise is about 45 uVrms (BW = 22 kHz). Several THD vs output level curves are shown in the pdf file (note the input impedance of S1 is 100 kHz). My next step is to load the circuit similar to that of headphone amp. |
|
|
| panson_hk |
| Quiescent current Icc = 12 mA and Iee = 8 mA were measured. Does anyone have these data? Are they supposed to be not identical? Data sheet says Icc=11 vs Iee=13 mA. Output DC is 0.4 mV with input grounded. |
|
|
| AndrewT |
Hi,
what connections have you got to ground (0V)?
The output should be open circuit for checking output offset.
Will this chip operate correctly with both outputs unconnected?
If you know the ground currents then you can calculate the difference between Icc & Iee.
If there is zero ground current, then Icc=Iee. |
|
|
| panson_hk |
| quote: | Originally posted by AndrewT
Hi,
what connections have you got to ground (0V)?
The output should be open circuit for checking output offset.
Will this chip operate correctly with both outputs unconnected?
If you know the ground currents then you can calculate the difference between Icc & Iee.
If there is zero ground current, then Icc=Iee. |
Input (IN+) is grounded via Rin. Am I answer your question?
Yes, offset was measured with output open circuit. The ground current is 4 mA, exactly the difference between Icc and Iee. Is it normal? |
|
|
| AndrewT |
Hi,
you are getting 4mA flowing into the ground connection from Vcc.
Why do National show 2mA flowing out of the ground to Vee?
Your input pin ground current <1uA and is swamped by others.
Why is your circuit flowing in the opposite direction?
Is your 4mA the mute current? |
|
|
| panson_hk |
I added a 100 ohm resistor to output as load (test circuit of Fig 3 in data sheet). The THD degrades substantially!!! :xeye:
THD vs output level with and w/o are given in the pdf. |
|
|
| panson_hk |
| quote: | Originally posted by AndrewT
Hi,
you are getting 4mA flowing into the ground connection from Vcc.
Why do National show 2mA flowing out of the ground to Vee?
Your input pin ground current <1uA and is swamped by others.
Why is your circuit flowing in the opposite direction?
Is your 4mA the mute current? |
The mute current measured is 100 uA. |
|
|
| panson_hk |
| The different is from the 7805! |
|
|
| panson_hk |
| I didn't not measure the exact Icc and Iee, but +I and -I. Now is okay! |
|
|
| AndrewT |
| quote: | Originally posted by panson_hk
Quiescent current Icc = 12 mA and Iee = 8 mA were measured. Does anyone have these data? Are they supposed to be not identical? Data sheet says Icc=11 vs Iee=13 mA. Output DC is 0.4 mV with input grounded. |
| quote: | Originally posted by panson_hk
I didn't not measure the exact Icc and Iee, but +I and -I. Now is okay! |
Should I ignore what you type in future?
What are you on? |
|
|
| panson_hk |
| quote: | Originally posted by AndrewT
Should I ignore what you type in future?
What are you on? |
Sorry Andrew. I should not do typing while measuring. |
|
|
| Tolu |
Has anyone listened to the the LME49810, except of Monkey29?
How does it sound? Or are the last posts just theoretic discussions about data sheet items?
Regards
Thomas |
|
|
| panson_hk |
| quote: | Originally posted by glennb
Or to be slightly nicer, hook up the Fig. 3 test circuit in the datasheet with +/-50V rails and load the output with a 470 Ohm 5W resistor. Drive the amp with say a 1Khz sine wave and gradually turn up the wick until the output starts to clip on a CRO. The maximum output current capability is then simply Vpeak/R. Hopefully the chip won't expire.
The slew rate limit can also be measured by increasing the frequency until the output is a triangle wave, then measure the slope in V per microsecond. Ccomp influences the slew rate.
I just haven't had time to try this with my samples.... |
Why use 470 Ohm?
I think the test I recently did is similar to what you described. I tried different resistor at the output of Fig 3 with 1 kHz sine wave and +/- 60 V rails. With 600 Ohm load, the chip clips at about Vpeak = 28V. The current at this point is 47 mA. However, the chip clips at a much lower output for 100 Ohm. |
|
|
| AndrewT |
Hi Panson,
starting from the left, I think you are showing us loadings of
100r, 600r, 1k0, 10k & open superimposed.
There is a big jump from 100r to 600r and again another big jump from 600r to 1k0.
It appears that the chip is much happier with the 1k0 load.
This would indicate a current gain of 300 is required to drive 8ohm speakers and about 600 to drive 4ohm speakers.
Neither of these gains are available from single BJT devices so it looks like a driver and output EF stage must follow the chip.
Alternatively a FET output stage could be implimented. |
|
|
| panson_hk |
| quote: | Originally posted by AndrewT
yes a CCS set to quiescent +20% and then shunt regulate that extra 20% to ground.
The chip is then protected at quiescent current level until the shunt starts to reduce it's demand to try to keep the supply voltage up, plus peak/transient demand from any supply caps after the CCS.
Simple and short circuit proof. |
Hi Andrew,
Thanks for you comment on output drive capability!
For current-limited regulator, you proposed a shunt regulator? The chip's Iq is 11/13 mA. If the CCS is set at 50 mA, a shunt transistor will consume about 39 mA x 60 V = 2.34 W in idle mode. Will it be too much? How about a series regulator with current limit?
Panson |
|
|
|
