Hi guys,
I've been wanting to get into DIY audio for a long time now in the hope to achieve a sound similar to my 'ideal system' at a much lower cost.
(My ideal system is the best system I've listened to so far, which was a dealer demo of top end naim gear with PMC speakers, tried other equipment but that was the setup I enjoyed listening to the most)
I am currently studying Engineering, in my 4th year, and am a mature student (29)
With the anticipation that building speakers from scratch is a major challenge, I am planning on starting with building an amplifier.
I have begun by purchasing a broken AV amplifier, as I felt it offered the most bang for the buck (the torroid transformer alone should be worth what I paid for it.) and have purchased an old Denon AVC A1-D.
My current plan is as follows:
1. Strip out torroid and re-case it in a small shielded case with separate power cord (along with rectification)
2. a second shielded case would deal with smoothing, consisting of the existing capacitors (if they're still ok) wired in parallel with a large number of lower capacitance electrolytics, in the hope that increased capacitance will lead to improved smoothing, and parallel capacitors will improve ESR.
3. A third section with active cooling will provide voltage regulation
4. final section will be amplification and speaker control, with a last low value capacitor to smooth out any remaining ripple.
The current plan is to look into keeping the existing amplifier sections, whilst upgrading the components (possibly all of them, including mosfets). The final section will also be shielded as much as possible.
I am also looking at buying the DIY naim clone kits an T-class amplifier modules from ebay, in order to compare sound quality, though I anticipate I will design and develop my own amplifier eventually.
Any input as to what is/isn't worthwhile, questions or advice would really be appreciated.
For those who are interested, my current set up is as follows:
Digital output form onboard audio (motherboard, tried creative X-Fi extreme music card and it was awful!)
Digital lead into Denon AVC A10-SE amplifier, Van Damme 2.5mm blue speaker cable, celestion F35c centre channel speakers in all positions, celestion s10 subwoofer. all ebay bargains, hence the weird collection (I must say the tinny sound of the celestion titanium tweeter is doing my head in, but I don't see the point in upgrading them until I can afford PMCs or build something better)
first post, sorry if it's too long, or in the wrong place!
I've been wanting to get into DIY audio for a long time now in the hope to achieve a sound similar to my 'ideal system' at a much lower cost.
(My ideal system is the best system I've listened to so far, which was a dealer demo of top end naim gear with PMC speakers, tried other equipment but that was the setup I enjoyed listening to the most)
I am currently studying Engineering, in my 4th year, and am a mature student (29)
With the anticipation that building speakers from scratch is a major challenge, I am planning on starting with building an amplifier.
I have begun by purchasing a broken AV amplifier, as I felt it offered the most bang for the buck (the torroid transformer alone should be worth what I paid for it.) and have purchased an old Denon AVC A1-D.
My current plan is as follows:
1. Strip out torroid and re-case it in a small shielded case with separate power cord (along with rectification)
2. a second shielded case would deal with smoothing, consisting of the existing capacitors (if they're still ok) wired in parallel with a large number of lower capacitance electrolytics, in the hope that increased capacitance will lead to improved smoothing, and parallel capacitors will improve ESR.
3. A third section with active cooling will provide voltage regulation
4. final section will be amplification and speaker control, with a last low value capacitor to smooth out any remaining ripple.
The current plan is to look into keeping the existing amplifier sections, whilst upgrading the components (possibly all of them, including mosfets). The final section will also be shielded as much as possible.
I am also looking at buying the DIY naim clone kits an T-class amplifier modules from ebay, in order to compare sound quality, though I anticipate I will design and develop my own amplifier eventually.
Any input as to what is/isn't worthwhile, questions or advice would really be appreciated.
For those who are interested, my current set up is as follows:
Digital output form onboard audio (motherboard, tried creative X-Fi extreme music card and it was awful!)
Digital lead into Denon AVC A10-SE amplifier, Van Damme 2.5mm blue speaker cable, celestion F35c centre channel speakers in all positions, celestion s10 subwoofer. all ebay bargains, hence the weird collection (I must say the tinny sound of the celestion titanium tweeter is doing my head in, but I don't see the point in upgrading them until I can afford PMCs or build something better)
first post, sorry if it's too long, or in the wrong place!
Dont underestimate how cheaply shielding can be done. You can use cardboard with aluminum foil glued on with PVA, it takes a while to dry but does dry stronger than the cardboard was.
Knowing this I would keep the old box and torroid transformer mounting and shield it with the cardboard / foil ply solution to shielding.
I would suggest that active cooling will not be required if heat sinks of enough size and is not a great idea as the amp will make noise.
Personally I think fixing up a good amplifier that's broken is the most cost effective solution to a quality amp for a student. So I would encourage you to make your own amp but I would suggest that while DIY will save you money replicating hi end systems that's only if you consider the cost of buying the upgraded second hi end system. Building HiFi is sadly not as cheap as I would like both in time and money.
Personally I would not bother with a Niam board from ebay but explore some more DIY solutions such as the circlophone. I have not heard it sadly, but reading about the circlophone it does sound a very interesting project.
Knowing this I would keep the old box and torroid transformer mounting and shield it with the cardboard / foil ply solution to shielding.
I would suggest that active cooling will not be required if heat sinks of enough size and is not a great idea as the amp will make noise.
Personally I think fixing up a good amplifier that's broken is the most cost effective solution to a quality amp for a student. So I would encourage you to make your own amp but I would suggest that while DIY will save you money replicating hi end systems that's only if you consider the cost of buying the upgraded second hi end system. Building HiFi is sadly not as cheap as I would like both in time and money.
Personally I would not bother with a Niam board from ebay but explore some more DIY solutions such as the circlophone. I have not heard it sadly, but reading about the circlophone it does sound a very interesting project.
I started by reading Douglas Self's articles in Electronics & Wireless World.
The article is "Distortion In Power Amplifiers: Part 1 of 8" by Douglas Self Aug 1993.
I suggest you find the magazine in your university library.
I built his "blameless" design on a perforated board. It is very simple, sounds better than any reasonably priced commercial amplifier. It is tolerant of a wide variety of voltages, uses easily available parts or near substitutes. Parts on a perf board are easier to change than on a printed circuit.
Changing parts and listening to the sound changes has been interesting.
I learned that I by ear I could detect changes in the sound signature at levels below the noise level.
Imagine a small piccolo playing a wrong note in a large orchestra. Its sound is much below the noise level, but your ear will pick it up. This small sound will lost on an oscilloscope or a FFT trace. Your ear is a remarkable device.
After this experience, what people write on these pages makes more sense, and more fun.
I recommend building from scratch, rather than getting a kit.
If you can get hold a some commercial amplifiers that have a fault for very little money, do so. They are a great source of parts of not repairable. If you can easily repair it, compare it to what you build. Very revealing. Having a amplifier at home, where you can listen to at your leisure, is much better that listening in a shop. I have been there and done it.
The article is "Distortion In Power Amplifiers: Part 1 of 8" by Douglas Self Aug 1993.
I suggest you find the magazine in your university library.
I built his "blameless" design on a perforated board. It is very simple, sounds better than any reasonably priced commercial amplifier. It is tolerant of a wide variety of voltages, uses easily available parts or near substitutes. Parts on a perf board are easier to change than on a printed circuit.
Changing parts and listening to the sound changes has been interesting.
I learned that I by ear I could detect changes in the sound signature at levels below the noise level.
Imagine a small piccolo playing a wrong note in a large orchestra. Its sound is much below the noise level, but your ear will pick it up. This small sound will lost on an oscilloscope or a FFT trace. Your ear is a remarkable device.
After this experience, what people write on these pages makes more sense, and more fun.
I recommend building from scratch, rather than getting a kit.
If you can get hold a some commercial amplifiers that have a fault for very little money, do so. They are a great source of parts of not repairable. If you can easily repair it, compare it to what you build. Very revealing. Having a amplifier at home, where you can listen to at your leisure, is much better that listening in a shop. I have been there and done it.
Maybe the 1st step would be to get it working.
How old is it? Revamping the PS might have some benefit. It does give you some experience. It sounds like you are planning on TWO cases. I would use just one for the entire PS.
Some shielding for the torroid is ......ok, but no need to go overboard. A 'good' steel box is fine.
Maybe share with us your detailed schematic for the PS. Although if 50 members comment, you will likely get 49 different opinions.
"...e hope that increased capacitance will lead to improved smoothing" That is a BIG hope. BETTER to go with at least a floating capacitance multiplier, if not a regulated PS [with capacitance multiplier].
Parallel caps to reduce ESR? Some-one will likely smack me on the back of the head, but I would go for a cap with low ESR first.
Active cooling? Why? Did you work out the thermal dissipation? If there was anyway possible I would stay with passive cooling. Do you REALLY want an EMF generator [the fan] [after shielding the transformer!!!] creating EMF and audible noise?!
In '4.' I kinda understand "a last low value capacitor to smooth out any remaining ripple" but a well-designed PS would already have 'acceptable' ripple and noise.
If you really want to 'slap' in a BL**DY big cap on the amp side, then use the biggest low ESR cap you can afford/source/beg/borrow/pilfer AND re-think the cap values in the PS - no need for extreme cap values in the PS.
To twist a saying of mine, "We design in hope, and listen in despair"
How old is it? Revamping the PS might have some benefit. It does give you some experience. It sounds like you are planning on TWO cases. I would use just one for the entire PS.
Some shielding for the torroid is ......ok, but no need to go overboard. A 'good' steel box is fine.
Maybe share with us your detailed schematic for the PS. Although if 50 members comment, you will likely get 49 different opinions.
"...e hope that increased capacitance will lead to improved smoothing" That is a BIG hope. BETTER to go with at least a floating capacitance multiplier, if not a regulated PS [with capacitance multiplier].
Parallel caps to reduce ESR? Some-one will likely smack me on the back of the head, but I would go for a cap with low ESR first.
Active cooling? Why? Did you work out the thermal dissipation? If there was anyway possible I would stay with passive cooling. Do you REALLY want an EMF generator [the fan] [after shielding the transformer!!!] creating EMF and audible noise?!
In '4.' I kinda understand "a last low value capacitor to smooth out any remaining ripple" but a well-designed PS would already have 'acceptable' ripple and noise.
If you really want to 'slap' in a BL**DY big cap on the amp side, then use the biggest low ESR cap you can afford/source/beg/borrow/pilfer AND re-think the cap values in the PS - no need for extreme cap values in the PS.
To twist a saying of mine, "We design in hope, and listen in despair"
To everyone who has taken time to reply: thank you so much.
To expand on where I am up to now: The Denon AVC-A1-D is now in my possession, I payed a staggering £28 for it, not bad for an amp that retailed at £2000 and was very highly rated when it was released (albeit that was just over a decade ago now)
So I've opened the amp to see what was 'wrong' with it, and it basically works perfectly. The decoding is fine, power supply is fine, switching between inputs fine, case in great condition, etc.
The only problem is with the left audio channel, this cuts in/out intermittently, and seems to be some kind of loose connection. Because this is a problem ONLY on the left channel, whether using surround sound or headphones, I'm presuming the fault is with the pre-amp section of the left channel, however it is a bit odd that if i power the amp up with the case open, the left channel works fine (or produces a horribly low thump/pop) if you tap any part of the main power amplification section, even very gently. With that in mind I'll take the whole thing apart and look at it more latter, with a view to repairing this amplifier for personal use or just to sell on... at any rate, ripping out the PSU seems a bit much given that the main amp works well.
owenhamburg:
Thank you very much for your input, RF shielding isn't something that I've learnt the sceince to yet, and certainly is an area I need to learn more of. Though the tin-foil idea is very useful (if I ever have to screen two components that are close together, for example) I should also point out that I have 'free' access to a workshop and some pretty simple raw materials, which includes easy to chop/fold 1mm sheet steel - one of the excellent benefits of being at university!
Cornelius Spronk: Thank you so much for your reply, interestingly the best two books i could find in the library and took out the day after I wrote this post were both by douglas self, the first one, 'Audio Power Amplifier Design Handbook' deals almost exclusively with the blameless design. I really find douglas's ideas relating to subjective and objective testing interesting, even if I don't always agree with his logic/conclusions (i.e. he talks about how if you want a perfect amp your primary goal should be minimum THD and measurable distortion using electronic equipment - that physical measurements are better than your ears, but then goes on to say that certain percentages of distortion cannot be noticed with your ears so it doesn't matter if they're there... something I cannot agree with, though I do agree that much more work needs to be done to debunk design myths)
The second book is a compilation of audio engineering data, and is put together by douglas but not all authored by him, it's a much more interesting read, especially things like electrostatics compared to speaker designs etc.
In conclusion though, I am becoming convinced that I need to build something from scratch myself, probably in a modular enough manner that I can change components out/around and see (hear) for myself what the real difference is. So it looks like there's going to be yet another 'blameless' amplifier being made!
As i said above, i picked up the AVC A1-D for £28, and it pretty much works - what I learn from fixing it will be worth the many times what I paid.
My main worry with comparing amplifiers is I simply do not have a speaker system capable of revealing the differences!
KMossman:
Again thank you for your input, I didn't even realise that the fans would introduce more RF (can be a bit dumb sometimes) so the advice is really appreciated. I will have to put up some plans of what I hope to do at some point, but the problem is this changes and develops more as I see more people opinions/learn more, and unfortunately although douglas self claims to be entirely objective, he clearly has a very quirky agenda, and the time and money to really test the realities of what makes a difference is something I won't have enough of for years! (it is an ambition of mine to set up a true high fidelity audio company at some point in my life, i think real Hi-Fi is depressingly overdue for the masses!)
So... where I originally thought only a linear regulated power supply would do, douglas self would suggest a linear unregulated psu - another example of conflicting opinions in different text books, so i'll have to build both and figure it for myself.
Anyway, yes i will post some cad drawings and circuit diagrams of what I intend to do as I intend to do it.
In other news:
I'll be going to the 'sound and vision, the bristol show' in feb, any other member will be there that would like to talk geek over a drink?
It turns out the only (?!) engineer with an interest in HI-FI design at my university is the head-of-school, so i'll be meeting with him for a coffee soon and that will likely change my designs yet again
How on earth do i decide between the different available I.C's for the amplifier? Where's the best place to buy them?
Any body want to buy a 5-speaker set of celestion F35c's? the trebble is pi**ing me off no end!
Thanks everybody
To expand on where I am up to now: The Denon AVC-A1-D is now in my possession, I payed a staggering £28 for it, not bad for an amp that retailed at £2000 and was very highly rated when it was released (albeit that was just over a decade ago now)
So I've opened the amp to see what was 'wrong' with it, and it basically works perfectly. The decoding is fine, power supply is fine, switching between inputs fine, case in great condition, etc.
The only problem is with the left audio channel, this cuts in/out intermittently, and seems to be some kind of loose connection. Because this is a problem ONLY on the left channel, whether using surround sound or headphones, I'm presuming the fault is with the pre-amp section of the left channel, however it is a bit odd that if i power the amp up with the case open, the left channel works fine (or produces a horribly low thump/pop) if you tap any part of the main power amplification section, even very gently. With that in mind I'll take the whole thing apart and look at it more latter, with a view to repairing this amplifier for personal use or just to sell on... at any rate, ripping out the PSU seems a bit much given that the main amp works well.
owenhamburg:
Thank you very much for your input, RF shielding isn't something that I've learnt the sceince to yet, and certainly is an area I need to learn more of. Though the tin-foil idea is very useful (if I ever have to screen two components that are close together, for example) I should also point out that I have 'free' access to a workshop and some pretty simple raw materials, which includes easy to chop/fold 1mm sheet steel - one of the excellent benefits of being at university!
Cornelius Spronk: Thank you so much for your reply, interestingly the best two books i could find in the library and took out the day after I wrote this post were both by douglas self, the first one, 'Audio Power Amplifier Design Handbook' deals almost exclusively with the blameless design. I really find douglas's ideas relating to subjective and objective testing interesting, even if I don't always agree with his logic/conclusions (i.e. he talks about how if you want a perfect amp your primary goal should be minimum THD and measurable distortion using electronic equipment - that physical measurements are better than your ears, but then goes on to say that certain percentages of distortion cannot be noticed with your ears so it doesn't matter if they're there... something I cannot agree with, though I do agree that much more work needs to be done to debunk design myths)
The second book is a compilation of audio engineering data, and is put together by douglas but not all authored by him, it's a much more interesting read, especially things like electrostatics compared to speaker designs etc.
In conclusion though, I am becoming convinced that I need to build something from scratch myself, probably in a modular enough manner that I can change components out/around and see (hear) for myself what the real difference is. So it looks like there's going to be yet another 'blameless' amplifier being made!
As i said above, i picked up the AVC A1-D for £28, and it pretty much works - what I learn from fixing it will be worth the many times what I paid.
My main worry with comparing amplifiers is I simply do not have a speaker system capable of revealing the differences!
KMossman:
Again thank you for your input, I didn't even realise that the fans would introduce more RF (can be a bit dumb sometimes) so the advice is really appreciated. I will have to put up some plans of what I hope to do at some point, but the problem is this changes and develops more as I see more people opinions/learn more, and unfortunately although douglas self claims to be entirely objective, he clearly has a very quirky agenda, and the time and money to really test the realities of what makes a difference is something I won't have enough of for years! (it is an ambition of mine to set up a true high fidelity audio company at some point in my life, i think real Hi-Fi is depressingly overdue for the masses!)
So... where I originally thought only a linear regulated power supply would do, douglas self would suggest a linear unregulated psu - another example of conflicting opinions in different text books, so i'll have to build both and figure it for myself.
Anyway, yes i will post some cad drawings and circuit diagrams of what I intend to do as I intend to do it.
In other news:
I'll be going to the 'sound and vision, the bristol show' in feb, any other member will be there that would like to talk geek over a drink?
It turns out the only (?!) engineer with an interest in HI-FI design at my university is the head-of-school, so i'll be meeting with him for a coffee soon and that will likely change my designs yet again
How on earth do i decide between the different available I.C's for the amplifier? Where's the best place to buy them?
Any body want to buy a 5-speaker set of celestion F35c's? the trebble is pi**ing me off no end!
Thanks everybody
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.