| Peter Daniel |
It seems like more serious players are entering chip amps field now: http://www.6moons.com/industryfeatures/gill/gill.html
To keep up with the Jonses, I designed an integrated amp that pushes the envelope even further.
Starting with custom made toroids that have special windings and amorphous cores and dual mono design to the bridged amps output and TVC integrated volume control.
From the outside, the chassis is stylized by Levinson reference DAC and transport. |
|
|
| Peter Daniel |
From the inside, I was really influenced by Connoisseur preamps layout ;)
The front end consists of S&B102 TVC mounted on rubber dampers into Mardone Burl side panels. A stepping motor coupled to higher grade Elma switch allows remote volume control. The transformers act as phase splitters as well, allowing either balanced or single ended inuts and feeding the amps with balanced signal only.
The amps are basically Pateks in bridged configuration with a usual bypass scheme. The heatsink is made of bronze. Not a best material heatwise, but sonically better than other alloys. |
|
|
| Peter Daniel |
| The rear panel is pretty straightforward with dual signal sockets, which can be either configured for both RCA and XLR inputs, or one input and one output from the preamp, allowing preamp operation only. |
|
|
| Peter Daniel |
| What is interesting though, the rear panel is decoupled from the rest of the chassis with special, aircraft rubber grommets. This allows for the main heatsink assembly to be rigidly supported in one point only, with a cone, reducing unwanted chassis infuences. The other two points of support are of a semi-flexible type. |
|
|
| Peter Daniel |
| Here's the bottom view, showing 4 main cones supporting the chassis and the 5th cone, just for the heatsink. That cone is using special disc dumper support disk (you can also see it in a first picture) that further improves the sonics. ;) |
|
|
| Peter Daniel |
| Here you can see that disk under the amp. The front panel is not finalized yet, it will be made out of copper. |
|
|
| Peter Daniel |
A close up of the inside with top cover removed (as usually made of perforated aluminum for low mass).
I'm running the amp presently with 4 ohm speakers. In bridged mode it gets pretty hot and clips at louder levels. Eventually it will be used with 8ohm rated Fertins, but also as a headphone amp with AKG1000.
Regarding the TVC configuration, I opted for +6dB connection with a single primary, it sounds better this way.
What is interesting I have only 0.1uF coupling caps in my DAC and the bass is terrific, no loss at all.
Some of you might be curious how this compares to resistor based volume control I used previously. Well, not bad at all.
With that amp, the buffer after the DAC didn't work well. I much prefer direct input to TVC for more cleaner and pure sound.
The buffer worked really well with resistor based attenuator. Overall that setup had a bit more air extention, however it may also seem that it had some artifacts that were not completely natural. The current setup with TVC is very pure and natural. A bit highs shimmer is missing, but I enjoy both setups equally well (for now). |
|
|
| Netlist |
Very nice Peter. Remember, there are many types of bronze, I can't give you the details for now but my neighbour is an artists, working with bronze all day. He searched many years to find the good alloy for his work.
http://www.dirkdekeyzer.com/english.html
A bit off topic but a Gainclone of yours would fit nicely near one of those sculptures. :)
/Hugo |
|
|
| sinski |
Very nice amp !
Hi Peter,
I have some questions
1. What power and voltage have those toroids ?
2. I noticed some bigger caps on the rectifier PCB ???
This weekend i finished Your premium kit , it works great but i want more !
http://photobucket.com/albums/y252/praganj/?action=view¤t=gc_prototype1.jpg
It is a prototype and both pcb are mounted in such a way, that i can play with some components. Later i will put it in one nice case.
Now i am inspired and thinking about two monoblocks.
Could You post some more technical details ?
Cheers, Tomek |
|
|
| Peter Daniel |
This looks like Shigaraki amp.
Toroids are 300VA, 2 x 22V AC, single primary wounded over secondaries.
There are only 100uF caps beside the chips, the ones at the rectifiers are 1000uF.
Not to create further consternation between more technically oriented guys, I will not mention what brand it is. |
|
|
| karma |
There are only 100uF caps beside the chips, the ones at the rectifiers are 1000uF
i tryed that with my lm4780 chipamp i was happy with the sound
lots of power. ;) before i changed it again:D |
|
|
| pinkmouse |
:cop:
Bronze heatsink debate is now here
:) |
|
|
| Dave Cigna |
| Peter, I'm curious about the wire you've used. The power and ground wires look like fairly light gauge (#22?) solid hookup wire. Is that what you normally use? |
|
|
| Peter Daniel |
| I alwyas use Cardas chassis wire, 19.5ga, for power connections. To me, that wire sounds the best in that application. The connection to binding posts is made with DH Labs hook up wire and signal wire is DH Labs solid core, 26ga silver. I'm not sure about the choices here yet. |
|
|
| KT |
Peter,
I foresee your integrated amp being a big hit and an amazing performer! Looks very, very promising!
The thick wood is very classy.
How would you say the bronze heatsink compares with copper or aluminum sonically? I understand there's some sort of debate over the use of bronze - I haven't read the thread yet, but I know there's some connection to the C37 theory as posited by Deiter Ennemoser. I haven't tried the bronze in anything, but the C37 lacquer was miraculous on my Lowther drivers. Big thumbs up.
Is there any problem with the switching motor interfering with the TVCs?
Best,
KT |
|
|
| Peter Daniel |
I'm not getting any interferences from the motor, the control circuit is isolated by using separate PS. While turning, the mechanism make nice clicking and definitely isn't of a quiet type.
Comparison between all those heatsinks isn't certainly an easy one. Maybe you remember that my first monoblock amp was using large copper piece acting as heatsink and structural bar. Initially, I liked the sound, but after CES, we came to conlusion that the air was somewhat missing in that amp. I was also using a clamping bar to attach the chip to the heatsink. After removing that calmp and replacing it with a regular screw things improved, but eventually we decided for aluminum heatsink as it provided more air extention, with copper it was a bit less. Those are not only my observations, but also from the people who worked with me while developing the amp.
I later built the first Patek using bronze and I found that that material provided nicely balanced sound: it was less zingy and harsh than aluminum, but didn't damp higher frequencies as much as copper. It seemed that the sound gained more refinement as well.
Brass also worked well, but somehow I didn't feel convinced about using it, I'm not sure why, maybe because of the qualities of the metal iself.
The amount of damping and air extention will depend on the mass of the heatsink and cannot be easily estimated untill tried in an actual design. It will also depend how the hetasink is attached and supported.
Also, with such small amps, I prefer some weight, so usually aluminum is out of question as it's not heavy enough. Bronze is too expensive and I can only have it in the amps for my personal use. Brass is rather hard to work with, so the only reasonable choice is copper. When used in smaller sizes, it's almost as good as bronze. |
|
|
| eddog |
| Thank you for finding a use for my old cracked drumset cymbals, all bronze and now about to melted down into heatsinks(as soon as I educate myself about the probably hazardous process!) I have built the 3875 kit and am floored by the quality of sound. |
|
|
| DocLorren |
Peter,
Again a creation of art! I do wonder if you have an explanation for the difference in sound when using different heatsink materials. After all it is only meant to dissipate some (not even that much) heat? Can you hypothesize why this apparent sonic difference might occur? |
|
|
| Peter M. |
Microphonics and vibrational properties maybe?
Have you tried annealing the copper to make it softer?
Bronze is much harder than copper, and church-bell bronze is alot harder than most other bronze alloys, a friend of mine got some of it from a bell maker. Allmost impossible to work with ordinary tools. It rings like a :D,ahem,, a bell when you tap it with a hammer :smash:, much lower damping than other alloys.
BTW, cool design :up:
Regards,
Peter |
|
|
| Peter M. |
Another idea for heatsinks.
Tin and zinc are have high damping properties, something to try?
I see you have some kind of plastic (PE?) to mount the caps in your amp, what about using wood instead? Would it make any difference?
Regards,
Peter |
|
|
| JesseG |
Hi all...
I can see this thread is heading toward a discussion on tghe sonic properties of heatsink materials. That's fine, but before it wanders too far of Peter's original post, I would like to ask:
Peter: The amp you have shown looks great - up to your usual level of design and execution. I am very interested in the stepping-motor volume control. Could you provide some detail (maybe a schematic), either here or on another thread?
Eddog: To melt spun bronze (your cymbals) you will need a furnace that can sustain 2,200C, a crucible that can hold liquid amorphous amalgams (Bronze is brass and tin, remember) and a cooling chanber for the molten metal which will allow cooling to occur over about 18 hrs. Otherwise, oxydation will cause your ingots to break (explode) when they drop thru 1,250 - 1,200C, unless you can cool it in an oxygen-free, inert-gass environment. This project is extremely delicate and hazardous and should proovide you and your friends with a fun-filled weekend :xeye:
Jess |
|
|
| tade |
That is a sweet volume control. Too bad it is 450 bucks though...
I love the amp.
I am considering making gainclone. Does it feel strange to spend way more work on the case than the electronics, the gainclone being reasonably simple?
Wonderful job. |
|
|
| jleaman |
| quote: | Originally posted by tade
That is a sweet volume control. Too bad it is 450 bucks though...
I love the amp.
I am considering making gainclone. Does it feel strange to spend way more work on the case than the electronics, the gainclone being reasonably simple?
Wonderful job. |
It's 450$ but well worth it. :smash: |
|
|
| tade |
sweet : )
actually yes. That will save you a lot of getting up! |
|
|
| JesseG |
| quote: | From Peter
For that amp, I used the board from Bentaudio. It is exactly the same ... |
Thanks, Peter. I had found that offering from DIYCable. It is rather pricey and it set me off on a hunt for a circuit I could DIY. I am playing with some floppy disk drive stepper motors and a cute little IC from Texas Instruments. When I have something, be sure I will post it.
Also, I just finished my 4766 stereo chipamp. Works great at 40wpc. I posted it here:
http://www.diyaudio.com/forums/show...5358&highlight=
Not up to your exacting enginerering standards, Peter, but have a look - the case and heatsink are interesting.
Jess |
|
|
| Russ White |
| quote: | Originally posted by JesseG
Thanks, Peter. I had found that offering from DIYCable. It is rather pricey and it set me off on a hunt for a circuit I could DIY.
Jess |
You may actually be interested in this... :)
http://www.diyaudio.com/forums/show...8119#post848119 |
|
|
| Peter Daniel |
| quote: | Originally posted by JesseG
Thanks, Peter. I had found that offering from DIYCable. It is rather pricey and it set me off on a hunt for a circuit I could DIY. I am playing with some floppy disk drive stepper motors and a cute little IC from Texas Instruments. When I have something, be sure I will post it.
|
The $450 is for the stepper motor assembly/control board AND the DACT attenuator. Without DACT attenuator, the price is about 50% of that. The Elma switch is needed to control setting on the transformers and it's about $60 from Percy.
It definitely can be done much cheaper, but at that time I didn't want to be bothered with developing another circuit. |
|
|
| jh6you |
| quote: | Originally posted by Peter Daniel
Comparison between all those heatsinks isn't certainly an easy one. Maybe you remember that my first monoblock amp was using large copper piece acting as heatsink and structural bar. Initially, I liked the sound, but after CES, we came to conlusion that the air was somewhat missing in that amp. I was also using a clamping bar to attach the chip to the heatsink. After removing that calmp and replacing it with a regular screw things improved, but eventually we decided for aluminum heatsink as it provided more air extention, with copper it was a bit less. Those are not only my observations, but also from the people who worked with me while developing the amp.
|
Wow . . .
The op amp chip must be a great magician . . .
On what color of the chip, the sound could be the best . . . ? :clown: |
|
|
| eddog |
"Eddog: To melt spun bronze (your cymbals) you will need a furnace that can sustain 2,200C, a crucible that can hold liquid amorphous amalgams (Bronze is brass and tin, remember) and a cooling chanber for the molten metal which will allow cooling to occur over about 18 hrs. Otherwise, oxydation will cause your ingots to break (explode) when they drop thru 1,250 - 1,200C, unless you can cool it in an oxygen-free, inert-gass environment. This project is extremely delicate and hazardous and should proovide you and your friends with a fun-filled weekend "
Hahahaha!! More confirmation that I should just hit my cymbals less hard, then I wouldn't feel compelled to do something with the damn things when they break. At upwards of$200 each you just hate to have them hanging in the garage collecting dust! |
|
|
| jleaman |
| quote: | Originally posted by Peter Daniel
The $450 is for the stepper motor assembly/control board AND the DACT attenuator. Without DACT attenuator, the price is about 50% of that. The Elma switch is needed to control setting on the transformers and it's about $60 from Percy.
It definitely can be done much cheaper, but at that time I didn't want to be bothered with developing another circuit. |
Id do it but i don't want to step on any one's toe's. Is any one interested in such a device ? 450$ is a little steep. |
|
|
| JesseG |
| quote: | From jleaman
Id do it but i don't want to step on any one's toe's. Is any one interested in such a device ? 450$ is a little steep. |
I for one would be VERY interested in a remote volume setup that I could DIY for less that the known kits - hopefully LOTS less.
Jess |
|
|
| agent.5 |
| quote: | Originally posted by jleaman
Id do it but i don't want to step on any one's toe's. Is any one interested in such a device ? 450$ is a little steep. |
yes |
|
|
| Peter Daniel |
It's not only steep, but DACT attenuator is nothing really special either.
I already mentioned that the better volume control is under development: http://www.diyaudio.com/forums/show...9425#post839425
I'm working on it with Veteran, who does programing and development of control board, pictured below.
This setup should be able to compete with any currently available attenuator kit on a market, the price will reflect it, but as usually, it will be still affordable. |
|
|
| JesseG |
| quote: | From Peter Daniel
This setup should be able to compete with any currently available attenuator kit on a market, the price will reflect it, but as usually, it will be still affordable. |
I am very interested in this type of control, as I said before. I have this vision of being able to touch a button from the comfort of my chair to adjust the volume on my chipamp. But, I like the sound of my chipamp and I don't want to loose it to the innards of a volume control chip.
I have seen some very simple motorized pot setups that didn't cost much, but the weak point was, of course, the pot.
The other 'but' is that I got in to chip amps because of they are phenomenally inexpensive. If I wanted to spend hundreds of dollars, I would buy something with a warranty so I had someone to yell at if it didn't work properly. However, for me, the fun of building a working amp with descent preformance from junk-bin parts and a couple of hours is right up my alley. $200 for a volume control would be more than double what all 3 of my chipamps cost - $450 would probably exceed my entire investment in hifi.
So, guys, I am looking for a good, CHEAP remote stepped attenuator solution - probably an oxymoron, but then so am I. :cannotbe:
Jess |
|
|
| karma |
| peter will it have a LED numeric display? will it be a option? |
|
|
| Peter Daniel |
| Yes, there will be LED numeric display, as well a volume knob and remote control. No options here, it will be standard. |
|
|
| karma |
| quote: | Originally posted by Peter Daniel
Yes, there will be LED numeric display, as well a volume knob and remote control. No options here, it will be standard. |
nice;) |
|
|
| sinski |
| quote: | Originally posted by Peter Daniel
This setup should be able to compete with any currently available attenuator kit on a market, the price will reflect it, but as usually, it will be still affordable. |
How much ?
;) |
|
|
| dariusku |
Nice work:)
We already know almost everything about new amplifier could You tell what resistors are used in amplifier?
& am thinking to order several different type resistors for experiments very interesting to try Your favourite combination. |
|
|
| filholder |
Hi if you want a relay/resistor controlled volume on the cheap with source selector then you could do a lot worse than the lite audio stuff they sell at diyclub.biz/
this setup :-
http://eshop.diyclub.biz/product_in...products_id=112
is what i am currently using. I would abvise on getting the transformer locally to save on shipping since the weitght on them really pushes up the shipping cost. For the price its a great setup I have used it here in my AD815 based preamp |
|
|
| karma |
| not bad a little far away, i wana see what peter comes out with.;) |
|
|
| Peter Daniel |
| quote: | Originally posted by sinski
How much ?
;) |
We need to build it first to determine the pricing.
| quote: | Originally posted by dariusku
Nice work:)
We already know almost everything about new amplifier could You tell what resistors are used in amplifier?
& am thinking to order several different type resistors for experiments very interesting to try Your favourite combination. |
I use Caddock MK132 in feedback position and Riken for gain setting. No series resistor, but if I would need to use one, it would be a nude Vishay. |
|
|
| jleaman |
| quote: | Originally posted by karma
not bad a little far away, i wana see what peter comes out with.;) |
Id like one :) details ? email me ? |
|
|
| Peter Daniel |
| The most advanced GainClone amp just got better: |
|
|
| Peter Daniel |
| Since I'm expanding my system I needed to add source selector. This one is based on silver contact Toco switches: 2 XLR and 3 RCA inputs. |
|
|
| Peter Daniel |
| The switching module is an add on and can be removed if not used. |
|
|
| Peter Daniel |
| The two toggles beside volume knob are for polarity switching. |
|
|
| Rob Thomas |
| quote: | Originally posted by Peter Daniel
The most advanced GainClone amp just got better: |
Good gracious my friend, that is a work of art..., or, more precisely..., Peter.
-Rob
(Viva La Lurkers!) |
|
|
| Peter Daniel |
Here's one more pic of the rear.
The connectors beside binding psts will be used for headphones, with toggle switch sorting ouputs.
That amp was actually originally intended as K1000 amplifier. |
|
|
| CarlosT |
Terrific work, Peter.
I had an artistic question related to all the GC PCB kits out there. The LM3875/LM3886 chip is set up with offset pins at 90 degrees to the PCB plane. How could one use one of the ready made PCBs like your mono kit and turn the chip another further 90 degrees so that it would be parallel with the plane of the PCB?
The answer could be as simple as bending the whole affair another 90 degrees but that hits me as brutish.
A more complicated answer would some kinda 90 degree PCB solderable connector with the odd 1.7mm/5mm pitch...maybe this doesn't exist.
Any ideas? |
|
|
| Peter Daniel |
Just bend the pins, here's one example:
 |
|
|
| CarlosT |
Nice...except I need the bend the pins 90 degrees the "other" way :D
In my design the LM3875 would lay parallel to the PCB hanging out the side...not bent in tucked under the PCB.
So...same approach? Bend-N-Pray? :D |
|
|
| CarlosT |
Peter:
A sorta related question to this thread...
Have you ever considered designing and making an inverted GC kit? Your PCBs super compact and simply stunning elegant.
BTW I keep buying more of your kits because I need the rectifying PCBs...not too cost effective eh? :D ;) |
|
|
| CarlosT |
| double post...sorry |
|
|
| Peter Daniel |
| It will work the other way as well |
|
|
| Peter Daniel |
| And this how the pins need to be formed: |
|
|
| Peter Daniel |
| quote: | Originally posted by CarlosT
Have you ever considered designing and making an inverted GC kit? Your PCBs super compact and simply stunning elegant.
BTW I keep buying more of your kits because I need the rectifying PCBs...not too cost effective eh? :D ;) |
I started with invering Clones but later switched to NI designes as I prefer them.
You should've told me you need just the PS section, I could've arranged for that. The kit shipped yesterday already. |
|
|
| CarlosT |
| quote: | Originally posted by Peter Daniel
And this how the pins need to be formed: |
Pretty malleable...this is all silly putty under your trained hands, eh? :D |
|
|
| CarlosT |
| quote: | Originally posted by Peter Daniel
...You should've told me you need just the PS section, I could've arranged for that. The kit shipped yesterday already. |
That's OK...now I'll have 4 AudioSector LM3875 kits that will need rectifying boards in the future :D
Thanks again. |
|
|
| sharpi31 |
Thanks Peter - I too have been loving your LM3875 pcb. I built my first one using the standard component pack and was blown away.
The second amp i built is a bit over the top (in spec, not in build!): huge bank of panasonic fm (12000uF per rail) with 100uf black gate N on the amp pcb. I've also got a 0.1uf black gate NX right on each power supply pin. Cheap ebay stepped attenuator, alps source selector. Each of the 3 analogue inputs has different coupling caps (horses for courses, and i never need more than one line in so i have the choice). I've also got a monica 2 dac board inside the chassis with separate transformer, rectifier bridge and psu cap bank.
Anyway, your work really impresses me - thanks again :-) |
|
|
| KT |
| quote: | | Originally posted by Peter Daniel
Just bend the pins, here's one example:
|
Hi Peter,
Great work, as always.
A couple of questions:
1) When bending the chip pins back a la the Patek, how do you get the bolt through the chip to secure it to the heatsink? Seems that the board is in the way.
2) I have enough boards to make the bridged dual-mono version the the Gainclone. Any pointers on how to connect the two sides of the chip together to parallel them? I'm inspired by the dual-mono Patek that was favorably reviewed on 6moons.com.
I may also email you about ordering a couple of the AudioSector PS boards - I have the v.1 Chipamp boards from the initial PD/Brian GT group buy and would like to take advantage of the wider pin spacing of the AudioSector PS board in order place the 1000uF caps at that location.
Thanks,
KT |
|
|
| twitchie |
| quote: | Originally posted by Peter Daniel
The switching module is an add on and can be removed if not used. |
Hi Peter, those look like Audio Research toggle switches, can you share where you got those from? I've been trying to get a replacement switch for months now and the local dealer want's alot of money for one. Are they the same ones?
Thanks,
Stephen |
|
|
| Peter Daniel |
| Those are TOCOS with silver contacts available from percyaudio.com They perform very well here. |
|
|
| Peter Daniel |
| quote: | Originally posted by KT
1) When bending the chip pins back a la the Patek, how do you get the bolt through the chip to secure it to the heatsink? Seems that the board is in the way.
2) I have enough boards to make the bridged dual-mono version the the Gainclone. Any pointers on how to connect the two sides of the chip together to parallel them? I'm inspired by the dual-mono Patek that was favorably reviewed on 6moons.com.
|
I install the chip first, then the board. To remove the chip, you need to remove the board. But only 5 pins need to be soldered.
The wiring diagram for bridged/parallel version of LM4780 chip was posted here: http://audiosector.com/lm4780%20amp.pdf
Doing it with LM3875 is the same.
Srajan still uses those amps, here the link to a more recent review when they've been featured: http://6moons.com/audioreviews/esoteric4/trio_2.html
 |
|
|
|
