Nelson,
Considering that DIYers can only buy so many fets, is there any real chance that the N and P mosfets can be matched in a DIY amp?
What is the maximum dissipation per IRFP240 that you are comfortable with, assuming the normal heatsink temperature of 50C? I am thinking that if I cannot get matched Ns and Ps, I might just use a larger heatsink for the mosfets in the amp now. I am hoping to dissipate 140-150W for six mosfets (3 pairs).
Thanks.
Considering that DIYers can only buy so many fets, is there any real chance that the N and P mosfets can be matched in a DIY amp?
What is the maximum dissipation per IRFP240 that you are comfortable with, assuming the normal heatsink temperature of 50C? I am thinking that if I cannot get matched Ns and Ps, I might just use a larger heatsink for the mosfets in the amp now. I am hoping to dissipate 140-150W for six mosfets (3 pairs).
Thanks.
I further reduced the supply voltage to 30v and increased the bias up to 2.5A. But beyond 2A I cannot tell any significant sonic differences. Is this your experience as well?
I realize that the front end of the 5400 is similar to that of the Aleph 0 ver.1.0-1.3. What is the lowest minimum voltage for the front end to work well? Right now mine is only at 35v.
I also see that the later versions (1.4-1.6) have a higher gain for the first stage but lower for the second. What are the sonic benefits of these changes? I might consider modifying mine to the later versions. Thanks.
I realize that the front end of the 5400 is similar to that of the Aleph 0 ver.1.0-1.3. What is the lowest minimum voltage for the front end to work well? Right now mine is only at 35v.
I also see that the later versions (1.4-1.6) have a higher gain for the first stage but lower for the second. What are the sonic benefits of these changes? I might consider modifying mine to the later versions. Thanks.
ctong said:
ctong said:
See the article on Class A biasing at www.passlabs.com (passdiy
section - "Leaving Class A"), where you can see the diminishing returns for
increased bias and also there is a discussion of parallel current source on the
output stage.
Ctong:
I am new to this forum, so that I can't send you PM. Could you please send the schematic to yangmicky@hotmail.com.
thanks a lot.
Jay
I am new to this forum, so that I can't send you PM. Could you please send the schematic to yangmicky@hotmail.com.
thanks a lot.
Jay
Crunchville dump strikes again!!
ctong, or anyone else- I would appreciate any schem or info on the 5400.
My wife brought a nice lookin' one home from the dump. It has 1 channel down with no toasted fets and no blown fuses... and no current flowing.
Also with that came a GTP-450 which seems to work fine
mpmarino at gmail dot com
Thanks....
ctong, or anyone else- I would appreciate any schem or info on the 5400.
My wife brought a nice lookin' one home from the dump. It has 1 channel down with no toasted fets and no blown fuses... and no current flowing.
Also with that came a GTP-450 which seems to work fine
mpmarino at gmail dot com
Thanks....
Nelson,
I have a question regarding the gate voltage of the cascode mosfet (IRF9610) in the front end. In the A75 it is set by a zener diode of 9.1v from the positive supply. But in the 5400 it is 8.2V. This reduces the working voltage of the transistor above it (IRFD9210) by nearly 1V (only 8.2-4=4v left). Is it an issue?
Another question is that the source resistor for the 9210 is 4.7ohms as compared to 47 ohms for the A75. What is the reason for the higher value in the A75? Is it to reduce the gain for this transistor? Thanks.
I have a question regarding the gate voltage of the cascode mosfet (IRF9610) in the front end. In the A75 it is set by a zener diode of 9.1v from the positive supply. But in the 5400 it is 8.2V. This reduces the working voltage of the transistor above it (IRFD9210) by nearly 1V (only 8.2-4=4v left). Is it an issue?
Another question is that the source resistor for the 9210 is 4.7ohms as compared to 47 ohms for the A75. What is the reason for the higher value in the A75? Is it to reduce the gain for this transistor? Thanks.
Thank you Nelson. I do not have the schematic for the power supply but the measured margin for the front end of the Adcom is about 13-15 V, comparable to that in the A75.
I replaced the zener diodes with 9.1V ones. The change is marginal, if real. I may also try changing the source resistors.
I replaced the zener diodes with 9.1V ones. The change is marginal, if real. I may also try changing the source resistors.
Strategy for optimizing amp performence
Nelson,
Recently I played around with the gain of the voltage gain stages and the amount of NFB in the amp and found that the sound is quite sensitive to these parameters. It seems that the sound tends to be dull and dry when the total gain is below 29db (more NFB). When the NFB is reduced the sound opens up, especially the treble. But too little NBF results in grainy treble, especially with bright recordings. I still do not have a good idea about the effects of the gain of the two voltage gain stages on the sound quality. How does one decide how much gain is needed for the the first and second stage? I have also tried reducing the first stage voltage gain (toward cascoding) as suggested in your paper on the A75.
With so many things to vary, I need your advise to come up with a good strategy to proceed. I would also appreciate the experience other DIYers have in fine tuning their amps.
Thanks.
CT
Nelson,
Recently I played around with the gain of the voltage gain stages and the amount of NFB in the amp and found that the sound is quite sensitive to these parameters. It seems that the sound tends to be dull and dry when the total gain is below 29db (more NFB). When the NFB is reduced the sound opens up, especially the treble. But too little NBF results in grainy treble, especially with bright recordings. I still do not have a good idea about the effects of the gain of the two voltage gain stages on the sound quality. How does one decide how much gain is needed for the the first and second stage? I have also tried reducing the first stage voltage gain (toward cascoding) as suggested in your paper on the A75.
With so many things to vary, I need your advise to come up with a good strategy to proceed. I would also appreciate the experience other DIYers have in fine tuning their amps.
Thanks.
CT
Hi ctong,
By playing with the amount of feedback, you are also affecting things like the stability of the amplifier. You really shouldn't be going in that direction without a good understanding of what is actually happening, and the test instruments is confirm what a calculator or computer is telling you.
This amplifier was designed to have a certain amount of feedback to be stable driving most, if not all speaker loads you are likely to run into. What may work fine with one set of speaker cables onto this set of speakers may very well self destruct if you change these basic parts of the design.
Amplifier design and modification appears to be straight forward to two types of people. The first are those who don't understand how things work. The second would be people like Nelson here, who have engineered several different amplifier types over the course of many years. I'm going to bet also that people who know as much as Nelson (for example) have forgotten how much knowledge they do have in relation to even a technician who has 25 years of experience. Much of everything he does, has been understood for so long that it's second nature. He does this in his head without too much effort. Also, people who really know what they are doing can take short cuts. The problem is that an outside observer does not see how much is happening between his ears.
These things are only easy to those who design for a living.
-Chris
By playing with the amount of feedback, you are also affecting things like the stability of the amplifier. You really shouldn't be going in that direction without a good understanding of what is actually happening, and the test instruments is confirm what a calculator or computer is telling you.
This amplifier was designed to have a certain amount of feedback to be stable driving most, if not all speaker loads you are likely to run into. What may work fine with one set of speaker cables onto this set of speakers may very well self destruct if you change these basic parts of the design.
Amplifier design and modification appears to be straight forward to two types of people. The first are those who don't understand how things work. The second would be people like Nelson here, who have engineered several different amplifier types over the course of many years. I'm going to bet also that people who know as much as Nelson (for example) have forgotten how much knowledge they do have in relation to even a technician who has 25 years of experience. Much of everything he does, has been understood for so long that it's second nature. He does this in his head without too much effort. Also, people who really know what they are doing can take short cuts. The problem is that an outside observer does not see how much is happening between his ears.
These things are only easy to those who design for a living.
-Chris
Thank you for the reply. That is not of much help, is it.
As in any experiments, (I am an experimentalist), hands on experience is needed to understand how things work. In this case how do different factors affect the perceived sound. Sure, Nelson has a lot of experience, but he had to start somewhere. I have very limited experience with this, but that does not mean I should not try to experiment. That is what DIY is about. However, advices from experienced people will be extremely valuable. That is why I am asking for advices.
CT
As in any experiments, (I am an experimentalist), hands on experience is needed to understand how things work. In this case how do different factors affect the perceived sound. Sure, Nelson has a lot of experience, but he had to start somewhere. I have very limited experience with this, but that does not mean I should not try to experiment. That is what DIY is about. However, advices from experienced people will be extremely valuable. That is why I am asking for advices.
CT
Hi ctong,
However, experimenting implies a structured experiment with clear objectives and some controlled parameters. What you are doing with this amplifier is not experimenting. You are simply messing around with a design done by someone else without an understanding of the basic concepts you need to understand so that you can learn further.
I'm not trying to be unkind, or unhelpful at all. I would much rather you learn and succeed with your project. Now, what would help you really learn would be to build some smaller gain stages. After doing that, then try varying different things in a planned way, much as you do plan on changing things with this amplifier. The goal here is to allow you to observe things without the really expensive smoke that accompanies a larger amplifier failing. So it's will be both less expensive and more instructive for you to be working on things that won't go bang in a big way.
I am well aware that you and I disagree on your approach. That's why I have stepped back. However, your last posts illustrate why you shouldn't just go ahead and mess with things. You ought to remember one saying that I've found very helpful over the years. If you look below the line in my posts, it says "Just because you can, doesn't mean you should"
What that really means is this. It is your amplifier. You can do with it what you will. However, some things you do will end badly. Therefore, you should really think about things before you attempt them. Now, I will ask you one question that I don't want to hear the reply from. This question is for you only. "If I do this, what is the probable outcome, and what exactly will I learn?" Asking this question is all about figuring out whether you should be doing this at all, or doing it another way. Being mostly self taught, as everyone outside the education system is, you have to find the best way for information to be gained. Learning on the bench after you have read something on what you want to learn is the best way to understand the knowledge. You will need certain equipment and supplies in order to build and test the idea under investigation. It doesn't have to be real fancy stuff for the basics, but you have to be able to observe changes in order to learn. Making a change and observing smoke really doesn't teach you a darn thing, except that you have no idea how to learn.
So, cuss at me or whatever makes you feel better. Just try to give some of this some thought. Maybe in some years you will see where I was coming from. That place was one where I am trying to help you.
-Chris
That depends. I'm trying to keep you from destroying an amplifier. A destructive failure that you would not learn anything from.
I completely agree whole heartedly with you.
However, experimenting implies a structured experiment with clear objectives and some controlled parameters. What you are doing with this amplifier is not experimenting. You are simply messing around with a design done by someone else without an understanding of the basic concepts you need to understand so that you can learn further.
I'm not trying to be unkind, or unhelpful at all. I would much rather you learn and succeed with your project. Now, what would help you really learn would be to build some smaller gain stages. After doing that, then try varying different things in a planned way, much as you do plan on changing things with this amplifier. The goal here is to allow you to observe things without the really expensive smoke that accompanies a larger amplifier failing. So it's will be both less expensive and more instructive for you to be working on things that won't go bang in a big way.
I am well aware that you and I disagree on your approach. That's why I have stepped back. However, your last posts illustrate why you shouldn't just go ahead and mess with things. You ought to remember one saying that I've found very helpful over the years. If you look below the line in my posts, it says "Just because you can, doesn't mean you should"
What that really means is this. It is your amplifier. You can do with it what you will. However, some things you do will end badly. Therefore, you should really think about things before you attempt them. Now, I will ask you one question that I don't want to hear the reply from. This question is for you only. "If I do this, what is the probable outcome, and what exactly will I learn?" Asking this question is all about figuring out whether you should be doing this at all, or doing it another way. Being mostly self taught, as everyone outside the education system is, you have to find the best way for information to be gained. Learning on the bench after you have read something on what you want to learn is the best way to understand the knowledge. You will need certain equipment and supplies in order to build and test the idea under investigation. It doesn't have to be real fancy stuff for the basics, but you have to be able to observe changes in order to learn. Making a change and observing smoke really doesn't teach you a darn thing, except that you have no idea how to learn.
So, cuss at me or whatever makes you feel better. Just try to give some of this some thought. Maybe in some years you will see where I was coming from. That place was one where I am trying to help you.
-Chris
anatech said:
Chris,
Although I am not an electrical engineer, I do understand the basics of electronics. I built numerous circuits when I was younger. So I know my way around circuits. But I thank you for the caution you provide.
What is relatively new to me is the high-end audio. Here the emphasis is listening tests-what sounds better, rather than what measures better (I used to believe that all amps that measured well sounded alike). I have learned much from reading this and other forums regarding factors that may affect the subjective performance of an amplifier, which provided guidance for my experiments. I have also learned much from experimenting with my Adcom amp, which now sounds nothing like the stock amp (It is stunning that Adcom can make a good design sound so bad). But I believe it can be further improved. At this point, I am running out of ideas to guide my further experiments. As I said I am an experimentalist and understand the importance of hypotheses in experiments. That is why I am asking for advices.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.