| MikeLewis |
| holy ****..... I found my next project when I'm done with my 10 channel 3886 NIGC. |
|
|
| MikeLewis |
| cr@p is a blocked word? |
|
|
| jaudio |
| "I found my next project" Me too. I was thinking of using mosfet instead of darington. |
|
|
| Gcollier |
There a bit of a lead time on these little puppies. I have had a few on order for a couple of weeks now. Anyone have suggestions as to what to use for the darlingtons on the source/sink pins. This will give me an excuse to use those XFO's that I have sitting around that put out too much voltage for a GC. Should be F.U.N!
G. |
|
|
| pinkmouse |
That does look very interesting.
However, this looks much more fun! ;) |
|
|
| MikeLewis |
From the Spec Sheet:
POWER DISSIPATION AND HEAT SINKING
When in “play” mode, the LM4702 draws a constant amount
of current, regardless of the input signal amplitude. Consequently,
the power dissipation is constant for a given supply
voltage and can be computed with the equation PDMAX = Icc
* (Vcc – Vee). For a quick calculation of PDMAX, approximate
the current to be 25mA and multiply it by the total supply
voltage (the current varies slightly from this value over the
operating range).
:bigeyes: Constant power draw? This sounds like it's going to be a real hog. Is this Class A? |
|
|
| Gcollier |
| quote: | Originally posted by MikeLewis
From the Spec Sheet:
POWER DISSIPATION AND HEAT SINKING
When in “play” mode, the LM4702 draws a constant amount
of current, regardless of the input signal amplitude. Consequently,
the power dissipation is constant for a given supply
voltage and can be computed with the equation PDMAX = Icc
* (Vcc – Vee). For a quick calculation of PDMAX, approximate
the current to be 25mA and multiply it by the total supply
voltage (the current varies slightly from this value over the
operating range).
:bigeyes: Constant power draw? This sounds like it's going to be a real hog. Is this Class A? |
Sounds that way. can't wait until I get my samples. I could use a little help with some suggestion for the Darlingtons.
G. |
|
|
| aletheian |
| HOLY COW! I'm ordering a pair right now...and maybe bye bye bridged 4780... |
|
|
| Easyamp |
| Very exciting, I can't wait to get my sample's. I've moved away from chip amps but these seem worth the time. |
|
|
| aletheian |
| Has anyone done any "back of an envelope" math on current consumption and voltage requirements on one of these? |
|
|
| tiltedhalo |
| This is the same package as the 3875,76,86 ect..... is a power darlington complimentary output pair with much higher power handling up to +- 70V to as high as 100 +- dissipation is still only 125W quiescent current is 30 ma distortion is .004 at 14RMS out, it actually is not as good overall as its conterparts just handles higher power supply voltages................ |
|
|
| UrSv |
| quote: | Originally posted by MikeLewis
From the Spec Sheet:
POWER DISSIPATION AND HEAT SINKING
When in “play” mode, the LM4702 draws a constant amount
of current, regardless of the input signal amplitude. Consequently,
the power dissipation is constant for a given supply
voltage and can be computed with the equation PDMAX = Icc
* (Vcc – Vee). For a quick calculation of PDMAX, approximate
the current to be 25mA and multiply it by the total supply
voltage (the current varies slightly from this value over the
operating range).
:bigeyes: Constant power draw? This sounds like it's going to be a real hog. Is this Class A? |
| quote: | Originally posted by tiltedhalo
This is the same package as the 3875,76,86 ect..... is a power darlington complimentary output pair with much higher power handling up to +- 70V to as high as 100 +- dissipation is still only 125W quiescent current is 30 ma distortion is .004 at 14RMS out, it actually is not as good overall as its conterparts just handles higher power supply voltages................ |
Guys do the maths...
Sure it's Class A, but it's a DRIVER not a power amplifier. To build a complete amplifier you need to add external components for the output stage and that's where the big power disspation will be. The current for the driver IC is 25 mA and the current available model in production is the A grade which means maximum +/- 75 VDC supply. This gives Pmax=(75-(-75))*0.025 W = 3.75 W of dissipation. Yes, that's less than 4 W. No need to worry... |
|
|
| youyoung21147 |
Nice thing ! :D
The point is to find the suitable power follower for this awesome chip, and have the money for the heatsink invloved :D
Gainclones may still remain the most popular chip amps :) |
|
|
| metal |
Hi
Cool stuff in deed for the power hunger guys...
Who can recommend the power output transistors to be used with this chip LM4702...
Thanks |
|
|
| elementx |
| quote: | Originally posted by metal
Who can recommend the power output transistors to be used with this chip LM4702...
Thanks |
MJ11032 from Onsemit should work fine, samples are available. I know I've had a sample request on the 4702 for over a month... it's on backorder... anyone actually play with them yet?
| quote: | Originally posted by pinkmouse
That does look very interesting.
However, this looks much more fun! ;) |
I agree! I actually got that chip, but it is surface mount and incredible small! I can't handle soldering it... anyone else upto the challenge?? I'd be willing to work out some sort of trade/purchase if you could just glue the thing to a piece of cardboard and solder some leads on it! |
|
|
| wes-ninja250 |
ElementX -- can you make a PCB good enough for it?
You can use solder paste and a toaster oven to actually solder it on. If that's the only component on the board, I expect you would get it right after only a few tries. :)
http://www.seattlerobotics.org/enco...06/oven_art.htm
Wes |
|
|
| demogorgon |
***** ********* ******, this is awsome..
I have got to get myself some o'these..
how's nationals sample program these days? just as tight as before i presume? |
|
|
| RCBandwidth |
For ElementX: I have successfully soldered the LM4970 device and
tried it out with a picmicro,it's an interesting device if you are interested in adding some "Bling" to your next project...
BTW: I have a stereo microscope that I used while soldering....Those little buggers look like 18 wheelers under the scope!!
Regards Bob C. |
|
|
| metal |
Hello
I might get some of these chips !!
Upon this & that, I will start designing a PCB for this nice thingy, I will post the design of the PCB asking for your opinions and suggestions to make it as blameless PCB design as possible for the LM4702.
Just wish me good luck... |
|
|
| Easyamp |
I was just thinking, if these are going to be like $5 or $6 it really wouldn't be worth it.
IMO, if it's only replacing the first and second stage of a typical lin amplifier, then it's only replacing a few resistors and maybe 5 transistors, which would cost you ,mmm, maybe a dollar.
I could be wrong but I don't think I'd pay for something that simple to make.
The're like a 200v opamp with no output stage, and while .005% d is nice it's not hard to attain with discrete componets.
Just a thought. |
|
|
| Tony |
@metal,
are you designing just the chip part pc or is it going to have the output devices as well?
if there will be a group buy for pcb, please count me in! |
|
|
| javven |
I'd definately be interested in this. I'll have to be the tail-end-charlie, though. I'm certainly no developer.
Leaders of the pack, please let us know just how much heatsink you need for a given application.... |
|
|
| metal |
Hello
I am designing the whole thingy in deed, so the PCB will contain one LM4702, two driver transistors and four output power transistors, also the "diode" transistor...
I 've concluded that it will be a better idea to be able to change the driver transistors, that will yeild the flexibility of having as many output transistors as the designer wants. For now, I will only include four output power trnsistors.
I have finished designing the library of LM4702, I will also design a library for the output power transistors, I will have MJL series.
For now, here is the library for LM4702, guys who use Eagle and are interested in this project, shall see this library and make sure its absolutely mistake free...
I also might omit the driver transistors and use power darlington instead, I am still confused on which one to go with...
Good LM4702 day ! |
|
|
| metal |
Hello
I have found these complementary darlington output power transistors to be a really good choice. I don't know if they are available in my country and other countries, but I will try to find out tomorrow.
MJ11031/2 complementary darlington output power transistors are rated at much higher power, but they are intended for general purpose audio power amplifiers, I don't assume they will be as good as those transistors that have transition frequency of 30MHz....
2SD2449
2SB1594
Let me know if there are any other suggestions for the darlington output power transistors that are as excellent as these and have transition frequency of 25MHz or higher, but rated at higher operating voltage and power..... |
|
|
| Tony |
for pa duties, hte mj11032/33 complements will suffice imho!
nice to have ouptu trannies expandable in groupd of 4's
how about dicrete drivers and outputs?
thnaks,
tony |
|
|
| johngalt47 |
Bob,
What make and model of stereo microscope do you have? I need to start using one myself. |
|
|
| RCBandwidth |
Hi John:The scope I'm using is quite old and probably not available,it has 1x to 3x manification and is more than adequate for soldering SMD devices.
Regards Bob C. |
|
|
| RCBandwidth |
John:I use the following technique for soldering SMD chips to an IC socket.I glue a small piece of wood to the base of an IC socket,I then use a small dab of hot melt glue under the chip and then secure it on top.I then solder 30awg wires to the IC socket,I then carefully align the wires to the chip and use solder paste to solder.I have successfully done several dozen chips with this method and have not lost one yet!....
Bob C. |
|
|
| metal |
Hi
I have already finished the PCB design, it looks complicated and unfortunately its double sided !!
Tony:
I agree with you, but we are not making a PA thingy, can't you see man, we need to cover the whole range of audio spectrum :) |
|
|
| demogorgon |
after all the chip is a "high end" driver, as national claims.
concerning the argument that this chip only replaces 4-5 transistors and resistors, it offers much lower space consumption, i envision myselv a very compact high power amp using this driver. |
|
|
| wes-ninja250 |
Hey, Bob, do you do all the wires at once, or one at a time?
How do you melt the solder paste?
Wes |
|
|
| metal |
Hello
I don't think its a matter of 4 to 5 transistors in deed, this IC, has current mirrors for the constant current consumption, which lowers the THD very much, also a thermal shut down feature, which also need more transistors, you start counting.....
The most important thing is the neat looking of the amplifier, and the high performance at the same time, that can be obtained using this IC...
The PCB I posted is just a prototype, I am still working on it, and will make it single sided if I can, I will also place LM4702 on a separate PCB, and the output stage on another PCB, the output stage PCB will have many verions, I mean there will be one PCB for LM4702, and two or three PCBs for the output satge from which you can select one that satisfies your transistors....according to the power transistors used....TO-3, TO3P...Darlington.....with drivers......etc... |
|
|
| RCBandwidth |
Hello Wes-Ninja:I first solder all of the wires into the IC socket,I use ic sockets with gold or silver plated round holes.I solder one wire at a time,a small amount of solder paste is applied to each pin followed by soldering with the 30 watt soldering iron.I do hold the wire to the pin with a fine point tweezer.
I can complete a 16 pin device in less than 1/2 hour.
Regards Bob C. |
|
|
| enzedone |
| Metal, that looks very nice. Of cource i'm no expert, just a pleb....... But looks very good, you are very talented, i only wish i had the ability to do this. I have ordered some chips as samples ,but i wonder how long they will take to arrive. That is a very good idea to enable people to decide on outputs as some of us might not be able to get what we really want and this gives us a few more options.... good thinking |
|
|
| metal |
Hello enzedone
Thanks man !
Glad you like it.. :) |
|
|
| enzedone |
| I just hope i can keep up with this thread. It looks very interesting, but being married and completely under the thumb...hehehe. I have completed just one gainclone, but started another..... man if my mrs found out i was looking at more projects instead of working on the house.....blood will be spilt....hehe. I think I'll have to start bargaining with her, she can get her hair done this month if i can buy another transformer......sounds good to me |
|
|
| wes-ninja250 |
Bob -- thanks!
I'll have to add that one to my book of tricks.
Once upon a time, I had to do something similar, but on a larger scale -- replace a wide DIP with a skinny DIP. I was replacing a defective SRAM chip and couldn't find a wide DIP of the right type.
I ground down an old EPROM, removed the die, then bent the pins straight on the skinny DIP and finished the connections to the EPROMs pins with 22awg solid copper wire. I also had to add an extra decoupling capacitor.
Wes |
|
|
| demogorgon |
metal, the output transistors can easily be mounted off board as you say, just holes for the wires, and possibly make room for putting in a small transistor with a resistor on the board for yer very own darlingtons.
i like the sound of that, compact.. :)
conserning the pcb you have allready posted, it's nice, i like the symetri, but for diyers making dublesided is perhaps a but to hard?
dunna, only make singlesided pcb's at school, and never god any solder to stick to em either.
just my 2 cents.
-Marius |
|
|
| metal |
Hello demogorgon
When I first started drawing the PCB, I only put all passive components and LM4702 in the schematic.....I was able to make it singlesided for LM4702....cheer up guys :)
As I have already said, I will make it this way, one singlesided PCB for LM4702 which is the main PCB module here, and many optional singlesided PCBs for power output transistors stage. My self, I prefer to be able to put the transistors on a separate PCB, thats much more flexible, and helps the guys who can't obtain a specific transistor.
Any way, it seems we got three choices here:
1. Using MJ11031/2 darlingtons pairs, I will design one for them, to make the whole thingy looks neaty and nice, also the small diode transistor will be included in the power transistors PCB, any way, don't expect the PCB for TO-3 footprint to be small. This choice will be the most powerful, but not the best performing one in the high ends of the audio spectrum.
2. Using 2SB1594/2SD2449 darlingtons pairs, I still don't know if they are available, but this one will have the best perfomance in the whole range of audio spectrum, and the small diode transistor is also included in the power transistors PCB
3. Using MJL4281/4302 pairs, these are not darlingtons, and a PCB for them is a must, so we have to use MJE15035/4 pairs as drivers too, and the small diode transistor is also included in the power transistors PCB. This one will be very nice, in case we can find out how to calculate this resistor between the emmiters of 2SB1594 and MJE15035, the same applies to 2SD2449 and MJE15034.
Note that I have maintained the pure complementary structure for the last option of power transistors trying to obtain a nice design, and finally semi solid, semi chip audio power amplifier GC solution.
Let me know if there are any other suggestions :)
Please note that I will not post the new PCB designs till I receive the required components... |
|
|
| AmA |
Hi Metal
what do you think about the new MJW0281A (NPN) MJW0302A (PNP) ON semiconductor ? |
|
|
| metal |
Hello AmA
Cool choice too ! |
|
|
| Russ White |
Hey guys,
Just a note, I ordered some samples of these a while back and have a PCB complete for the 4702 and SAP16N and SAP16P which are temp compensated and thus eleminate two transistors from the circuit. Those devices are also very easy to parallel, although they do need a 200ohm pot/rheostat between the pairs to set the current.
Just waiting for the SAPs to get here.
Cheers!
Russ |
|
|
| AmA |
very interesting Russ
Can you please post your schematic for us ?
many thanx |
|
|
| Russ White |
| As soon as I make it presentable I will post it. :) |
|
|
| Mike Gergen |
The LM4702 reminds me of the Sanken STK chips. I have had pretty good luck with the STKs making amps. Very satisfied.
This chip has me interested. It's cheaper then the STKs and appears to be available. One of the issues with the STKs was getting my hands on them.
Mike |
|
|
| metal |
Oh yea Russ
Here we go :xeye:
Do we really need schematics as much as we need PCB designs !? |
|
|
| Russ White |
| quote: | Originally posted by metal
Oh yea Russ
Here we go :xeye: |
Errrrmmmm?? What do you mean by that? |
|
|
| Russ White |
| My circuit is slightly different from the datasheet, but only slightly. The compensated output devices help a lot. |
|
|
| metal |
Hi Russ
Are you gonna share the PCB design ? |
|
|
| Russ White |
| quote: | Originally posted by metal
Hi Russ
Are you gonna share the PCB design ? |
Sure. I always do. |
|
|
| metal |
When, before, or after selling the PCBs...
I will be interested to have these transistors as well...but not interested in buying PCBs from any one... |
|
|
| Russ White |
| Well I guess you don't want it then. ;) Seriously, I have given away nearly all of my PCB layouts, not sure where your angst comes from. Sorry if I stepped on your toes. Forget it if you like. |
|
|
| metal |
| quote: | | Well I guess you don't want it then. Seriously, I have given away nearly all of my PCB layouts, not sure where your angst comes from. Sorry if I stepped on your toes. Forget it if you like. |
Well, I have asked you for the Sch & PCB files in eagle format many days ago for the design of the preamp/selector, till now, I have not seen any thingy from you....I thought you will never post any thing....
Any way, I am not angry, but I had the some doubts about you posting Sch & PCB files on the forum, which are clean now, we will be waiting for your work Russ ;)
I will also post my work on the PCB, so we can improve the final PCB for this design, thats what I really want... |
|
|
| AmA |
| this great site is for sharing information and help is n t it ? |
|
|
| Russ White |
| quote: | Originally posted by metal
which are clean now |
Hmm what is clean?
The other (IO selector) design you requested of me is not my work alone, and I would not presume to give it away. You should not think any worse of me for that. That is very petty.
You should do some research. PDF and brd files of my designs are available on the forum.
Your attitude is only going to make people less prone to share. I suggest you treat contributers with a little more consideration.
Cheers!
Russ |
|
|
| metal |
Hello AmA
Thats absolutely the case here, I hope other members also feel the same about what AmA said...No offence, ok... |
|
|
| metal |
russ,
Why making things more difficult here, while its much more simple. I asked you for the final eagle files, not for the PDFs nor the Brd file alone, your answer was not what you have already written here....but a little explaination at that time, just like the one you have already written, would have removed the vague many days ago...
Any way, I am always used to "consideration" for the members on the forum. What already happened here, shows that you are used to deal with sharing matters that way :) anyway, a member who is used to share, won't just be proned because of me, I don't think so russ....I always receive private emails from many members thanking me for the sharings I did...any way, thanks for the sharing and contribution you did, and might do.. |
|
|
| carlosfm |
| quote: | Originally posted by metal
russ,
Why making things more difficult here, while its much more simple. I asked you for the final eagle files, not for the PDFs nor the Brd file alone, your answer was not what you have already written here.... |
Wait a minute...
Who designs a PCB is free to share (or not) PDFs, BRDs and/or Eagle (or any other software) files.
Give the man some time.
What if he only shares PDF files, is it bad?
You are not paying for the files, so...
Go and make your own layout, don't be lazy.:dodgy:
:smash: |
|
|
| metal |
| quote: | | Who designs a PCB is free to share (or not) PDFs, BRDs and/or Eagle (or any other software) files |
Why ain't he just say that...
| quote: | | You are not paying for the files, so... |
Whoi mentioned paying here ? |
|
|
| carlosfm |
| quote: | Originally posted by metal
Whoi mentioned paying here ? |
Nobody.
But you can't be in such a hurry.
When they come, they come.
Hope for the best. |
|
|
| metal |
| quote: | Originally posted by carlosfm
But you can't be in such a hurry.
When they come, they come.
Hope for the best. |
Hey man, I always hope for the best...:), I will be waiting...why hurry, here I am, waiting till they come :) |
|
|
| Gcollier |
I see sooooo many petty little arguments on this forum...90% of them are communication issues. If we go back and actually READ what somone is saying...and not jump to the assumption that they are hurling insults and trying to start a fight we would get a LOT farther in advancing our designs. We have to understand that not everyone on this forum communicates in perfect english, which often leads to missinterpretation. God knows I am hardly comprehensible half of the time...and the other half no one can understand me ;)
Lets all make a concious effort to be better communicators!
Thats it...by the way glad to see we are getting some designs here, anyone got any real parts from National yet? :D |
|
|
| Russ White |
Hi G, Yes I have my samples they arrived today, now if I can just get my hands on those SAP16s :) BTW its very nice to hear from you! Its been too long for me. :)
This is going to be a fun project.
I apologize for snipping at metal a bit, but he did not seem to realize that contributions are freely given, but they should not be demanded. I probably misunderstood his intent.
I plan on sharing all my data and findings very openly including PCB design.
Carlos, just one word. Thanks.
:drink:
Cheers and no ill feelings on my part. |
|
|
| Gcollier |
Hi Russ!
Well glad to see somone has a few of these chips...at least I know they really do exist...I have some hope that mine will arive soon!
Anyway, I wasn't really reffering to anything or anyone in particular with my previous rant...just trying to keep things on track. It was more of a general comment to point out that we sometimes need to re-read things before we let out fingers start hitting the keys. :D
And yes it has been a while Russ...I've been hanging around the digital forum a bit (I have a TDA1543 design there...who doesn't) and I've been looking in on things here too. This is really the first thing that has caught my eye as something new to try. Now all I need to do is find the time to actually work on it.
G. |
|
|
| Russ White |
Here is my single sided prototype layout.
Keep in mind this is not optimized in any way it is just meant to give me something to test with.
I will post the output board later.
First here is the layout: |
|
|
| Russ White |
| Here is the schematic |
|
|
| metal |
| quote: | Originally posted by Russ White
Hi G, Yes I have my samples they arrived today, now if I can just get my hands on those SAP16s :) BTW its very nice to hear from you! Its been too long for me. :)
This is going to be a fun project.
I apologize for snipping at metal a bit, but he did not seem to realize that contributions are freely given, but they should not be demanded. I probably misunderstood his intent.
I plan on sharing all my data and findings very openly including PCB design.
Carlos, just one word. Thanks.
:drink:
Cheers and no ill feelings on my part. |
Hey Russ,
No problem, this is what I usually call a rage hour, after that we all start to apologize to each other, any way, no hard feelings man, and no offence was intended, so I want you to acept my apologies too...any way, there is no problem with demanding a design from some one here on the forum :)
Here is what some guys asked for, the PCB for LM4702 alone, with no output transistors, I think its fully optimized now and single sided, I spent the rest of yesterday night on it, hope you like it guys... |
|
|
| Russ White |
| Notice that the comp caps are not on the board yet, that is because I was initially going to put them right on the pins. I changed my mind and I will add them soon. So look for an updated PCB soon. |
|
|
| metal |
russ,
How much does it cost for two pairs of SAPs you have already ordered, just need an idea of how much they cost here.. |
|
|
| Russ White |
| quote: | Originally posted by metal
russ,
How much does it cost for two pairs of SAPs you have already ordered, just need an idea of how much they cost here.. |
About $6 each so 2 pair is roughly $24. |
|
|
| Russ White |
Metal, what where you planning on using for Ci and Cin, it will be tough to find caps in the size you have on your PCB.
Also you should bypass the PS pin on the chip as close to the chip as possible. |
|
|
| metal |
Russ,
Thanks for the advice, I will take that into consideration in the next PCB post.... |
|
|
| demogorgon |
good to see no hard feelings remain, i was getting worried an argument might get in the place of constructive audio discussion.
well, Russ, the "close-as-possible" bypassing for a chip that draws a constant current of 25ma is not needed. the overbypassing does not help the sound, it helps the sound you hear. and thats the sugarpill effect you know.
making the board for a specific pair of transistors serverly limits it usefullness, if this is something you make for yourself, only sharing the work for fun, then ok, but for most people, like me, your pcb is useless.
just a notion.
-Marius |
|
|
| metal |
Hey demogorgon
Glad to hear from you, I do really care about your opinion, and other member opinions too, as you have a nice perspective for such matters Marius.
Any way, I hope you find what you asked for in the PCB design I 've already posted :)
See ya ;) |
|
|
| Russ White |
| quote: | Originally posted by demogorgon
well, Russ, the "close-as-possible" bypassing for a chip that draws a constant current of 25ma is not needed.
making the board for a specific pair of transistors serverly limits it usefullness, if this is something you make for yourself, only sharing the work for fun, then ok, but for most people, like me, your pcb is useless.
-Marius |
Sorry to say so, but your first point is patently false, some LM3886 circuit can draw 15ma to 25ma at idle, yet good bypassing can make all the difference in the world, it is not just sound I am after either, it is also stability. Bypassing as optimally as possible is always the best choice, not that compromises don't sometimes have to be made.
As for you second point, any suitable power transistor you choose could be used. I don't make any assumptions on that. |
|
|
| metal |
Hello All
I will post a new modified PCB design for LM4702, I will modify the PS bypassing a little bit, so all guys be satisfied, and feel good about the PCB design, no problem.
Also I will work on the other PCBs for the power transistors to which I pointed out yesterday, (2SB1564/2SD1449), (MJL4281/MJL4302-MJE15034/MJE15035), (MJE11031/MJE11032).
I have read the datasheets for SAPs new transistors, and they have very good features, but I found this to be frustrating a little bit:
5. Destruction capacity of the built-in emitter resistor
The built-in resistor is fabricated with polysilicone on the chip for the SAP08P/N and a thick-film resistor is
used for the SAP10P/N and SAP15P/N. The latter, the thick-film resistor, has weaker destruction point in the Pc area (especially for large current flowing area) rather than that of the transistor chip itself.
This is subject to the area beyond Safe Operating Area (S.O.A).
However, under the evaluation like a short circuit test in which the current exceeds the guaranteed value, it
may cause the emitter resistor to be destroyed before the transistor itself is destroyed.
Consequently, the current value (or time) that operates the protection circuit is to be set at lower than that of discrete device configurations. In the application of car audio amplifiers, the same manners as the above need to be considered because the large current is flowed at low impedance.
In addition, once the transistor falls into thermal runaway due to a soldering failure to the external VR added between diodes or other failure manners, as the worst case, there may cause a resin crack or smoke emissions by flare up.
Flame retardant molding resin is used, and the material of the product is conformed to the most sever standard UL94V0.
However it is recommended that the careful consideration be given to a protection circuit, and the protection circuits should be provided appropriately in due course.
If the operating conditions are not to be matched to the ratings, it is also recommended that the E (Emitter resistor) terminal should be opened and the external emitter resistor should be added to the S (Sensing) terminal shown as below. (However this is not applicable to the SAP08P/N because a thin inner lead is used for S terminal.)
Looking from a cost-effective perspective, I don't think these transistors are best suited to this application in deed. What do you think guys ? |
|
|
| Russ White |
I am not sure what concerns you. Shorting just about output resistor beyond its rated power dissipation will kill it. Just don't short it. :) Or you could build an output protection circuit.
Still there are plent of good parts to choose. |
|
|
| metal |
| quote: | Originally posted by Russ White
I am not sure what concerns you. Shorting just about output resistor beyond its rated power dissipation will kill it. Just don't short it. :) Or you could build an output protection circuit.
Still there are plent of good parts to choose. |
Any audio diyer can easily short the output power transistors by mistake in deed, but I will try not to short them ;)
Do you have any short protection circuits on mind russ... |
|
|
| Russ White |
Yes, but I will not use for me first prototype.
You will have a hard time finding good devices that can take a full on short for long. |
|
|
| metal |
russ,
you might post some kind of schematic for the protection circuit later, I have the intention to add them to the power transistors PCBs.
Ok guys, I have to go now, see you the next day ;) |
|
|
| Tony |
| quote: | Originally posted by metal
Hello All
I will post a new modified PCB design for LM4702, I will modify the PS bypassing a little bit, so all guys be satisfied, and feel good about the PCB design, no problem.
Also I will work on the other PCBs for the power transistors to which I pointed out yesterday, (2SB1564/2SD1449), (MJL4281/MJL4302-MJE15034/MJE15035), (MJE11031/MJE11032).
I have read the datasheets for SAPs new transistors, and they have very good features, but I found this to be frustrating a little bit:
5. Destruction capacity of the built-in emitter resistor
The built-in resistor is fabricated with polysilicone on the chip for the SAP08P/N and a thick-film resistor is
used for the SAP10P/N and SAP15P/N. The latter, the thick-film resistor, has weaker destruction point in the Pc area (especially for large current flowing area) rather than that of the transistor chip itself.
This is subject to the area beyond Safe Operating Area (S.O.A).
However, under the evaluation like a short circuit test in which the current exceeds the guaranteed value, it
may cause the emitter resistor to be destroyed before the transistor itself is destroyed.
Consequently, the current value (or time) that operates the protection circuit is to be set at lower than that of discrete device configurations. In the application of car audio amplifiers, the same manners as the above need to be considered because the large current is flowed at low impedance.
In addition, once the transistor falls into thermal runaway due to a soldering failure to the external VR added between diodes or other failure manners, as the worst case, there may cause a resin crack or smoke emissions by flare up.
Flame retardant molding resin is used, and the material of the product is conformed to the most sever standard UL94V0.
However it is recommended that the careful consideration be given to a protection circuit, and the protection circuits should be provided appropriately in due course.
If the operating conditions are not to be matched to the ratings, it is also recommended that the E (Emitter resistor) terminal should be opened and the external emitter resistor should be added to the S (Sensing) terminal shown as below. (However this is not applicable to the SAP08P/N because a thin inner lead is used for S terminal.)
Looking from a cost-effective perspective, I don't think these transistors are best suited to this application in deed. What do you think guys ? |
i wouldn't worry much about the built-in 0.22ohm resistors, what can be done is to add external resistors in series with the emiters, say another 0.22 then add more sap's in parallel to get a better SOA rating for the output stage! this is how i will do it when using the sap devices! |
|
|
| Tony |
two of these chips in two input boards will make a really high powered amp, with a balanced input, ballanced outputs. i.e. with rails of +/-85volts, 600 watts is possible!
national has come a long way since the lm391 days! about 20 years, i'd say well worth the wait! |
|
|
| demogorgon |
| quote: | Originally posted by metal
Hey demogorgon
Glad to hear from you, I do really care about your opinion, and other member opinions too, as you have a nice perspective for such matters Marius.
See ya ;) |
why, thanks, i'm almost blushing here i'm sitting ;)
| quote: | Originally posted by Russ White
Sorry to say so, but your first point is patently false, some LM3886 circuit can draw 15ma to 25ma at idle, yet good bypassing can make all the difference in the world, it is not just sound I am after either, it is also stability. Bypassing as optimally as possible is always the best choice, not that compromises don't sometimes have to be made.
As for you second point, any suitable power transistor you choose could be used. I don't make any assumptions on that. |
I think you need to read me better, either that or i need to practis my english, as we think each other wrong, but trying to say, basically, the same thing.
bypassing as close as practical, not possible, and as good as practically possible. thats my view.
point two, the chip dosent need darlingtons after all?
i was under the impression that it did after reading the datasheet.
what will the effect of removing qmulti1 on the datasheet resoult in?
I was thinking this, with the resistors, were there to help counter thermal runaway?
-Marius
edit: metal i just ah a slight idea on how to drop one of the links:
but it has been decided one-chip-darlingtons are to be used, and the biaskiller transistors gone?
they can offcource be hardwierd on, but then the using of a pcb looses much of its advantages\usefullness over a p2p or a bread board. |
|
|
| Russ White |
| quote: | Originally posted by demogorgon
point two, the chip dosent need darlingtons after all?
i was under the impression that it did after reading the datasheet.
what will the effect of removing qmulti1 on the datasheet resoult in?
I was thinking this, with the resistors, were there to help counter thermal runaway?
|
If you use theramally compensated devices like SAP16 you do not need qmulti,, otherwise they should be included on the power device PCB.
Also you can easily make your own darlington pairs and use a wide range of devices.
My idea is to make the driver PCB as generic as possible by taking any output device dependancies off the driver board and moving them onto the power device board.
Cheers!
Russ |
|
|
| demogorgon |
| quote: | | If you use theramally compensated devices like SAP16 you do not need qmulti,, otherwise they should be included on the power device PCB. |
Using thermally compensated devices require the money to buy such.
I have under a dollar a day to live for, at least for another 15 days, so naturally i cant get those, though i can get conventional power transistors.
edit: just read that line closer, power device pcb?
gahh, hate building stuff out of 3-6 different boards when it could be done better on one.
| quote: | Also you can easily make your own darlington pairs and use a wide range of devices.
|
not with the pcb as such i cant, then i'l be having to hardwire, and the pcb's usefullness is again limited.
besides, there isn't a very wide selection of thermally compensated BJT's around are there?
wich answers this:| quote: | | My idea is to make the driver PCB as generic as possible by taking any output device dependancies off the driver board and moving them onto the power device board. |
my view on thing
cheers
Marius
btw, forgot to say metal, nice pcb, i love the symetri :)
though it's not the one i'l need ;) |
|
|
| Russ White |
demogorgon,
If you have a set of requirements you would like to see in a PCB I would be glad to see if I can accomodate you.
I chose the SAP16 simply because it has a great reputation, and it removes something like 6 components from the schematic. :)
You could build this amp with 2 SAP16P 2 SAP16N and the LM4702 for around $30 + passive component costs which are quite variable.
That seems pretty darned reasonable to me. ;)
But I also understand living on a budget.
So give me your ideas.
Cheers! |
|
|
| demogorgon |
Russ, This dialog turned out better than i had thought, good thing you are a diplomatic man :D
well, i'm trying to point this\these design\designs into a more intigrated solutions, thats based on logic and knowledge rather than the latest fashion on the forum :)
like mauro's 3886 amp. no BS, just a good design. no snake oil. no matter how cheap it comes ;)
i'l have to sit down with pen and paper and make a scetch of what i envision, but not tonight. tomorrow.
oh, btw, something that cost you in the states 30bucks will cost me about 60 here up north. if not more.
there's a reason the wages up here are about twise as high as yours. :apathic:
edit, not as in you, but rather all americans offcource.
gonna go try out Quake4 now. good evening to you all. :)
-marius out. |
|
|
| Gcollier |
This discussion has led me to one clear and concise conclusion....I really need to finish my new speakers!
Seriously, both of these designs look pretty good so far! Personally I would like to see individual output boards as it allows for more flexibility where heatsinking is concerned (lets face it that will be the biggest cost). It also allows for more flexibility in the chassis design and layout. This will also allow you the flexibility to experiment with different output transistors etc.
I look forward to seeing what happens with this design. Has anyone given some thought to a PSU for these????
G. |
|
|
| metal |
Hello
demogorgon:
If I am not wrong, you need to see two transistors (MJE15034/5) on the LM4702 PCB in order to hardwire (MJL4281/MJL4302) directly, not forgetting the addition of the bias transistor....
I already said, I will make it versatle, and satisfying to all valuable members interseted in this project. I will try to do what is mentioned above. I will also try to keep it single sided....
In this one, I have made the small signal tracks shorter as much as possible, and moved the PS capacitors in order to allow using a connector for the PS. I added the mute resistor also which I completely forgot, now its on the PCB, along with another connector for +5 Volts :) |
|
|
| demogorgon |
and still as beautifully symetric :)
I'l whip up the scetch i promised now, and be back shortly.
Oh, and GCollier, an output board would not be as flexible as transistors connected only using wiers from the board.
wish i could have shown you a pic, but i have non available right now, but i'l be back though.
-Marius |
|
|
| metal |
Hello demogorgon
Glad to see that you like it demogorgon, I always try to do the best I can...
I will be waiting for your sketch :)
Gcollier:
Hey man, why hurry, as you can see, not all the members interested in this project are completely satisfied with this PCB design, so, late results, are better than nothing, Glad your up here now :) |
|
|
| demogorgon |
christ what a task i took on!
i'm gonna have to leave you hanging a bit more, sorry.
designing pcb's aint my strong side, or not something i'm cabapable to do in a flash anyway.
stay tuned :)
Marius |
|
|
| metal |
Ok demogorgon
I will wait for your work, no problem :) |
|
|
| Russ White |
| metal, what 10uf caps are you finding that will fit your PCB? Can you give me a manufacturer and model. I would love to find a good film cap at 10uf that is that small. |
|
|
| metal |
Hello russ,
Sorry man, I don't know a specifis manufacturer, I bought 25 of them from a shop where I used to live when electronics were young...Now I found some project to use them :) |
|
|
|
