Transistor equivalent

spec said:

Hi MrAl. I get the feeling that you are getting fed up with telling people about the LM358 corossover didtortion but, sure you can elliminate the gross crosover distortion with a pull-down resistor, but in the chips I had there was still some nasties around what would have been the crosover point. I think something was going on with the drivers. In a conventional AB amp you reduce crossover by biasing the op transistors at typically 20 to 100mA. But another aid to reducing it is by negative feedback but above around 10KHz, in a practical circuit, the LM358 doesn't have any gain so no negative feedback. It is all a matter of degree.

Click to expand...

Hi again,

You must be doing something different because for the way it should be done there is no crossover and no crossover point

Recall that the cross over is caused when one transistor is turning off and the other is turning on, so the second one has to pick up the load and there could be a slight delay before it is able to turn on to the proper level. Now if the upper is on and is turning off, then the lower is off and is turning on, but if we REMOVE the lower transistor, we can never have cross over distortion because there's simply NO crossover anymore.
The output biasing is set up such that the lower transistor is effectively never turned on, not even a little, and it never even comes close. That's how the cross over distortion is eliminated. If you see anything that resembles cross over distortion, then its' not being done right.
I'll see if i can find that app note. In short, the output is biased so that the current always comes from either the upper transistor or the resistor to ground. If the resistor to ground is too large, it may or may not work correctly depending on the load.
Way back when i designed a headphone amplifier and also a power amplifier based on the LM358 and the audio was very clear. It does depend highly on how the device is used however. You wouldn't think you could use it this way for a power amp (plus transistors of course) but it works.

Nikolai Petrenko said:

Transistors will drive subwoofer (35-40W) (to big?) => total 80W

Click to expand...

Not really a good question but I think i know what you mean. As long as you use the correct transistor for your design it will handle the bass frequencies no problem. It's at the high frequencies that power transistors have problems. It is the design of the amp that counts most as far as base is concerned. The base frequencies can be difficult for an amplifier because that is where the bass speaker will be at resonance. This implies either big current demand or possibly little current demand. The peak current you need is supply rail/8. it is generally accepted that this current capacity should be multiplyed by 1.2 upwards, depends who you speak to. My target is x2. The reason: sounds nice. Some really good amps have more than that. If you are saying you want to build a good 30W bass driver that would not be a problem and amounts to 5.5A and +-27V supply rails (you would go for +-30V probably)

spec said:

Ah! so you worked in the commercial world- dificult!

Click to expand...

Hi,

Yes, for several years in the past. Back then i dont think it was as competitive as it is now, but it always was a little. We did not like to keep a stock that was any larger than needed so we learned to design with parts that could be used in several designs, like the LM358

MrAl said:

Hi again,

You must be doing something different because for the way it should be done there is no crossover and no crossover point

Recall that the cross over is caused when one transistor is turning off and the other is turning on, so the second one has to pick up the load and there could be a slight delay before it is able to turn on to the proper level...

Click to expand...

Hello again to you,

You must wonder if I'm with it, and I can understand that- it's the way I talk. It's always getting me into trouble. You are absolutely right in what you say, but just to let you know I have spent years studying cross-over effects and ways to elliminate them from going to class A to the Bloomly approach and many other schemes. I'm currently thinking about a new approach. Perhaps I used the wrong term for you. My point is that you can get crossover distortion anywhere in an amp not just the output transistors.

But I'm not even saying that. What I am saying is this: there might be no cross-over in the output transistors but there appeared to be some abnormalty around the 0V op level. Perhaps I had better use that term for now on. Of course, I may have been measuring incorectly or even had bad batch of LM358s. They were some of the first out of the foundary. Oh yeah, they wern't even LM358s they were LM124s, the quad version, but as far as I know the individual amps are the same.

MrAl said:

Hi,

Yes, for several years in the past. Back then i dont think it was as competitive as it is now, but it always was a little. We did not like to keep a stock that was any larger than needed so we learned to design with parts that could be used in several designs, like the LM358

Click to expand...

I'm jealous of you and AG. I always fancied working at the sharp end; guitar amp and pickup design would have been the thing. With military projects often you never know what the equipment you are working on is used for. The other thing is that much of the white hot technology cant be used because it is not available in MIL SPEC. On the other hand, we did work on some really advanced stuff and I had a fantastic couple of years in researce. What sort of things were you doing?

Hello again,

Rather than reply with two replies again i'll reply for both in one.

When i worked in the industry i worked for a company that mostly built high powered synthesized sine wave converters. We made converters from 100 watts to about 30,000 watts three phase. I designed some of the newer ones because i got there late in the life of the company. We also did solar conversion/line tied applications for Sandia Labs. That was an interesting project. We had a solar panel as long as the building and some 20 feet across. We also did military stuff for a major aircraft company. In other companies i also worked with computer terminals and instrumentation. Sometimes had to travel to Canada to modify some kind of unit or another due to some original faulty design. That was interesting too i guess.

What you say about the LM358 is hard to decipher because i am not sure what level of 'distortion' you are talking about here. You know, like 10 percent, 1 percent, 0.1 percent, etc. All i can really tell you is that the crossover distortion in the output stage is eliminated with the right choice of output biasing, and it's not hard to imagine how this works. If there is something else going on then that must be a different issue and has to be handled separately. I can tell you though that once this is done, the output sounds clear enough for a headphone amplifier.
You can take a look at the schematic on the data sheet of the LM358 where they show the basic internal connections.
Any crossover distortion usually shows up as a blip or something, so if you have a spec on that you can state that as well if you like. I can tell you though that without the special biasing the cross over is quite pronounced and visible on the scope, and worse, audible

I would think the LM358 internals and the LM324 internals are similar, but i dont know if there is any different biasing going on due to the added two amplifiers. It doesnt take much though to set up a little experiment by wiring up an LM358 and pumping a low freq sine wave through it and testing for any noticeable crossover with a scope. You might also listen to the output too with headphones at various frequencies (but not too high) and various gains probably less than 10. There is a light chance i can do this myself later today but not sure if i can get ot it or not yet. I also have an LM432 which i believe has an two amps like the LM358 internally, as well as a voltage reference, and i just made a breakout board for it the other day but still need to put the male headers on it. One of the amps is dedicated to work with the voltage reference as an error amplifier, but the other amp is free to use like an LM358. Very interesting chip too BTW. I think it only comes in SMD package though so i had to make a breakout board for it to use it in tests with various micro controllers.

MrAl said:

Hello again,

Rather than reply with two replies again i'll reply for both in one.

When i worked in the industry i worked for a company that mostly built high powered synthesized sine wave converters. We made converters from 100 watts to about 30,000 watts three phase. I designed some of the newer ones because i got there late in the life of the company. We also did solar conversion/line tied applications for Sandia Labs. That was an interesting project. We had a solar panel as long as the building and some 20 feet across. We also did military stuff for a major aircraft company. In other companies i also worked with computer terminals and instrumentation. Sometimes had to travel to Canada to modify some kind of unit or another due to some original faulty design. That was interesting too i guess.

Click to expand...

Hi

Afraid I am replying in two posts. Thanks for the info about your career. You have done some stuff it seems. I too have had some great jobs- arrive early in the morning and before you know it's 6pm and you have to go home for dinner or you will be in trouble with domestic management! Sometimes we worked through the night. But as I went up the ladder it became more and more specification writing and meetings, which I don't mind but a lot of it was waffle. Did two years on proposals, which was more interesting than you may think- many visits to see the customer and the troops at the front line. Then department manager- that was fun!. Then formed a one- man company doing design consultancy.

Then back to company doing systems engineering, but they had caught managementitus by then. The way it was going you would have to have a meeting to decide if you could go for a pee or not. Every morning we had a meeting to describe what we had done the day before and what our objective were for that day. So it might go like this: "What did you do yesterday?." "Design work on the XYZ input module". What are your objectives for today?" "Continue designing the xyz input module". You get the picture. Pretty much every one apart from the Proative Managent Executive Team (I made that up) and all their many acolytes got cheezed, and slowly many of the good hardware design men drifted away. One day I was writing a huge report, which no one would read, and thought this is not for me, so I called it a day and retired at 58.

Since then, with a free hand. I have been up to all sorts of things, but perhaps too long for this post.

MrAl said:

... I also have an LM432 which i believe has an two amps like the LM358 internally, as well as a voltage reference, and i just made a breakout board for it the other day but still need to put the male headers on it. One of the amps is dedicated to work with the voltage reference as an error amplifier, but the other amp is free to use like an LM358. Very interesting chip too BTW. I think it only comes in SMD package though so i had to make a breakout board for it to use it in tests with various micro controllers.

Click to expand...

I have learned something today- didn't know about the LM432. As you say very intresting chip, presumably at a good price*. It is on similar lines to the LM10 (https://www.ti.com.cn/cn/lit/ds/symlink/lm10.pdf) which was unbelivable when it came out. It's one of my babies. But fairly soon National withdrew it so we never designed it into deliverable systems. It was also expensive. It was unavailable for a long time. Have to see if it is back now^.

I do the same as you, I use breakout boards for messing with surface mount components- works well. I slap small decoupling capacitors on the board too, and sometimes with opamps, 50 ohm isolating resistors in the output.

* £0.60 UK
^ not available

MrAl said:

What you say about the LM358 is hard to decipher because i am not sure what level of 'distortion' you are talking about here. You know, like 10 percent, 1 percent, 0.1 percent, etc. All i can really tell you is that the crossover distortion in the output stage is eliminated with the right choice of output biasing, and it's not hard to imagine how this works. If there is something else going on then that must be a different issue and has to be handled separately. I can tell you though that once this is done, the output sounds clear enough for a headphone amplifier.
You can take a look at the schematic on the data sheet of the LM358 where they show the basic internal connections.
Any crossover distortion usually shows up as a blip or something, so if you have a spec on that you can state that as well if you like. I can tell you though that without the special biasing the cross over is quite pronounced and visible on the scope, and worse, audible

I would think the LM358 internals and the LM324 internals are similar, but i dont know if there is any different biasing going on due to the added two amplifiers. It doesnt take much though to set up a little experiment by wiring up an LM358 and pumping a low freq sine wave through it and testing for any noticeable crossover with a scope. You might also listen to the output too with headphones at various frequencies (but not too high) and various gains probably less than 10. There is a light chance i can do this myself later today but not sure if i can get ot it or not yet.

Click to expand...

You sound like agood sport. I always get into discussions like this; it's just me. As I have said before, I agree with what you say. Just to clarify what I thought I saw was anomolies around about 0V with a split supply arrangement. Cant remember what the gain was , but probably 11 (2x 10K resistors with parallel feedback- is that right term?). In general xover effects are minimised at low frequencies , but as you go up they tend to get more pronounced. I didn't spend a lot on time doing the test because the LM358 was not designed for hifi. I do agree that the LM358 doesn't sound too bad with audio if the output transistors are biased to class A.

What I think I saw was a slight problem say about 2mV, but it was in the 1970s and I can't remember precisely.

Maybe the LM358 is one of your babies; I have many. I didn't mean to trash it. In fact it is a very usefull amp and I still use it now. I designed a car (auto) CD ignition using an an LM324 and it worked a treat. Before that I used an LM3900 Norton current differncing amp (https://uk.rs-online.com/web/p/oper...7633D424F5448267573743D4C4D333930304426&sra=p) which was the first cheap quad amp. Made my mate's Vaushall Victor pull like a train at low revs!

Hi,

Yeah i guess you could say i am a good sport, trying to stay reasonable even when people dont agree with me, and i realize most of us here work at the hobby level now mostly being away from industry for some time. I see that you are too and i know that both of us stay reasonable in the face of adversity
I also see that like me you are also quite tenacious when it comes to electronic stuff. That's good really. I hold on to my beliefs but if someone comes up with some good evidence to the contrary i will certainly willing adopt the new ideas or standards. Things change, and some viewpoints that were valid yesterday are not quite as good today.

Back to the LM358 and related...

I tested an LM358 at least theoretically, and it looks quite good. So far only one frequency and one gain setting, but i can do more easily. The results as well as the internal diagram are shown in the attachments. The second picture is a Zoom in of the crossover points.

Oh so the LM10 is no longer available either? Wow, that bites. What about the LM3900, that went out quite a while ago didnt it? I had some fun with that chip way back when. I am pretty sure the LM432 is still available. So many chips are going 'away' though, seems a little sad. It's almost like they became 'friends' of a sort.

The LM10 seems fine. Just one 14 pin SOIC version is gone.

MrAl said:

Hi,

Yeah i guess you could say i am a good sport, trying to stay reasonable even when people dont agree with me, and i realize most of us here work at the hobby level now mostly being away from industry for some time. I see that you are too and i know that both of us stay reasonable in the face of adversity
I also see that like me you are also quite tenacious when it comes to electronic stuff. That's good really. I hold on to my beliefs but if someone comes up with some good evidence to the contrary i will certainly willing adopt the new ideas or standards. Things change, and some viewpoints that were valid yesterday are not quite as good today.

Click to expand...

That's the way to be. Many of my buddies have views which I totally disagree with, not about electronics. At one time I used to argue about things like that. But, with some people, if you demonstrate quite clearly that they are wrong, they divert to ad homonym attacks. So now I divide people into two groups, ones that can be objective and just focus on the facts and ones that will not be wrong at any cost. We used to have a driving syndicate where four of us would take turns at driving the 20 miles to work. We had many wide ranging discussions, but once the driver had to stop the car because two of the passengers were getting so heated. It was about some trivial subject too. There is a rule in the UK which says in polite company you shouldn't bring up money, sex, politics, or religion- I find that rather sad. This may sound strange but I am genuinely pleased when someone points out an error, because it helps to get things right.

In electronics the first time this happened was when an older kid in our area got a job as a trainee TV service engineer. I was 12 and he was 17. Naturally I was quite interested in what he was doing and we discussed Ohms law and all that basic stuff, which I didn't really understand at the time. In passing he mentioned that current is the flow of electrons from negative to positive. I thought this was preposterous and told him he was wrong. When I found out I had to apologise, but from then on I always remembered how an electric current worked and that was the real start of my learning about electronics. More recently, I made some terrible mistakes, especially about Hi Fi and music. Some of them are: the wood that a solid guitar is made from doesn't affect the sound, loudspeaker cables make no difference, capacitors don't make any difference and as for resistors that's ridiculous, and so on. All proved to totally wrong. I will get my coat now.

KeepItSimpleStupid said:

The LM10 seems fine. Just one 14 pin SOIC version is gone.

Click to expand...

Thanks Keep, glad to hear that!

Can anyone post a 3055-2955 40w amplifer circuit? I found a good circuit but it's 60w on 4ohm load and 40w on 8ohm load,
I can't find small transistor MPS

mpsa06: https://www.utsource.net/search.aspx?keyWords=mpsa06&page=1 It's also available at digikey and mouser.

You found a copy of an extremely old circuit. Its 2N3055 transistors have a high frequency cutoff that is too low for low distortion at high frequencies. Maybe they can be used for a sub-woofer amplifier.

yes, 2n3055 very good at low frequency, I want to use it for subwoofer amplifier but mpsa06, mpsa56 and mps9632 can't buy in Vietnam,
MJ2955 or 2N2955, which can find high-quality easier?

Here, if an old part is still being manufactured then it is available at many places. I think the MPS transistors were made and used 50 years ago.
Sorry, I do not know what is available on the other side of the world.

spec said:

That's the way to be. Many of my buddies have views which I totally disagree with, not about electronics. At one time I used to argue about things like that. But, with some people, if you demonstrate quite clearly that they are wrong, they divert to ad homonym attacks. So now I divide people into two groups, ones that can be objective and just focus on the facts and ones that will not be wrong at any cost. We used to have a driving syndicate where four of us would take turns at driving the 20 miles to work. We had many wide ranging discussions, but once the driver had to stop the car because two of the passengers were getting so heated. It was about some trivial subject too. There is a rule in the UK which says in polite company you shouldn't bring up money, sex, politics, or religion- I find that rather sad. This may sound strange but I am genuinely pleased when someone points out an error, because it helps to get things right.

In electronics the first time this happened was when an older kid in our area got a job as a trainee TV service engineer. I was 12 and he was 17. Naturally I was quite interested in what he was doing and we discussed Ohms law and all that basic stuff, which I didn't really understand at the time. In passing he mentioned that current is the flow of electrons from negative to positive. I thought this was preposterous and told him he was wrong. When I found out I had to apologise, but from then on I always remembered how an electric current worked and that was the real start of my learning about electronics. More recently, I made some terrible mistakes, especially about Hi Fi and music. Some of them are: the wood that a solid guitar is made from doesn't affect the sound, loudspeaker cables make no difference, capacitors don't make any difference and as for resistors that's ridiculous, and so on. All proved to totally wrong. I will get my coat now.

Click to expand...

Hi again,

Sounds like you have many interesting stories to tell. It is people like you that make these forums interesting.

When i first learned about current it was just the opposite (literally the opposite ha ha): i knew about electron flow but didnt know there was such a thing as 'conventional' current flow until later which was used probably more often.

Funny you should mention a difference with respect to the type or make of resistors, because just the other day i found out that some of the 1 percent resistors i have actually have steel in their leads. Most of the ones i have only have copper, tin plated or something like that. I have to wonder how much that steel would come into effect at higher frequencies where the skin effect must be terrible for steel or iron wire. I have to test every one of my resistors now to see if they contain steel or iron in the leads by holding up to a magnet. If they stick, DC circuits only

As a side note:
WOW, the old old 2N3055 being brought up now in this thread...someone please tell me where do i get 1000 of them (har har). That's one part i dont miss.

Yep, Chemistry in college. You needed to know where the electrons were going. Then again in solid state physics we have holes and electrons.