With 2 oz. copper it is rated at about 12 ampsI can wiggle it for about 20A, maybe widen the board too.
Yes I know Sprint gives you an indication, but then why are the rails that fat?
Anyway, I like the ground path following the way back the signal path takes, others prefer connecting it to star GND (think this is what Andrew is referring to). But IF you provide a GND header for the Spk connection it should have nearly the same width (=resistance) like the signal path... just my opinion BR,
Holgi
I like fat low resistance rails, and the layout with the caps as big as they where made it that way. Personal choice, without any special theory or magic. IF you have alternative thought, please let me know, I am always open to learning something.
I think what you are saying is reflected in the layout now.
I think what you are saying is reflected in the layout now.
Attachments
Thanks Hugh, it is just a amateur imitation of a Alex MM layout, he is the BOSS.
What is the diameter of the wire that will be used to make the Output Inductor. The one on the layout is way too small. Any suggestions?
I see a few issues.
1) PCB dimensions: Length of 268mm will force using 11” or 12” heat sink, and all the power devices are sitting on one side of the heat sink. This doesn’t seem to be the best way of utilizing an expensive heat sink. 268mm X 92mm dimension yields 8 boards off a standard 18”x24” material sheet. You can easily yield 10 boards by reducing height from 92mm to 81mm. The height of 92mm is a bit excessive anyway even for a 6” length heat sink extrusion. The power devices would all located closer towards the edges of the heat sink than towards the middle which is less optimal for heat spreading.
2) MOSFET placement: PCB outline indicates the PCB blocks the way of the mounting hardware both in and out. Once the PCB is soldered in place it cannot not be detached from the heat sink as an assembly by simply undoing the mounting screws. It could have been more prototyping/service friendly. The placement also indicates bending the MOSFET leads at about 10mm off the package, that leaves about 4mm clearance for the mounting hardware, that may not be sufficient, especially when using thick large washers for an even pressure spreading. Either cutout or clearance holes should be used for the mounting hardware access.
3) Placement of the e-caps risking short circuit to the MOSFET leads when there is excessive solder pump or lead stub. They could be easily avoided by offsetting the placement slightly.
4) All TO-126 transistors seem to have wrong pin-out
5) All 2SC2240 and 2SA970 transistors seem to have wrong pin-out
6) TO-126 heat sinks should not sit over traces unless they are elevated. These heat sinks should be mounted to the PCB instead of being supported by the transistor leads and their solder joints. They seem to have a solder-able pin in the middle. A plated hole could be used perhaps.
7) The trim pots probably better be side adjust type for easy access, as the PCB most likely sits vertically in a chassis. In that case TO-126 heat sinks and other tall parts should be out of the way.
8) There is not a clearly defined grounding system, not much to comment about but three points: A) when local OPS rail filter cap banks are used, usually the main GND should be located on the PCB to ensure the most tightly enclosed possible load current loop. An off-the-board main GND, which seems to be the case here, negates such purpose. B) the front end GND should have its own fast-on blade. This amp has low impedance feedback network which carries significantly larger sample current than conventional amplifier. You do not want to re-use the input GND lead lobbing the large sample current back to the RCA jack and from there to the main audio GND. The common GND impedance may cause you trouble. C) Direct capacity loading the output node to GND (upstream the output inductor) should be avoided especially for an amp with high voltage transition rate design, as doing so may decrease stability margin, although in our particular case it may not be critical.
9) PCB mounting points don’t have to be at the corners. They should be placed close to the connectors to help control the stress applied on the board at engaging and disengaging these connectors.
10) Are you sure this amp is going to terminate anything at all
?
Nattawa
Thanks for your input, I have not indicated that it is done. I still have to go over it to find the errors such as you have brought forth. I really appreciate your taking the time to do this.
This board is was not designed for anything other than testing and development. I am almost certain that there will be many minor and maybe major changes to the topology, schematic and overall topology. It is not ready for prime time.
That being said a few of these will be built and tweaked and maybe fried, and once Dadod is satisfied that the design is ready, a new board will be done and maybe we all can enjoy the fruits. I worked at National Semi many years ago in the linear IC R&D section, I think you would be surprised at how many iterations of every device that was made, some IC's were more than two years in R&D. Vector bread boards and flea clips with bus wire, that's what I am talking about.
I am sure the excellent SMD board you laid out is indicative of your considerable expertise, I have never done one. I have never soldered a SMD part to any board. So stay close and teach me some of your insights into all of this. I am still trying to learn some new tricks.
To show you how long in tooth I am, my first Amplifier Kit was a 35 watt Knight-Kit from Allied Radio, 100 North Western ave. Chicago. The first from scratch was the Leach Low Tim.
So give me a minute or two and I will make corrections so that some first prototypes can be hacked and burnt. If I am able to get this board done, then I will put 1 pair of outputs on it and bring that up with a Variac, slowly and once the get to cooking I will add more. I would not think of putting 4 pair on to start with. SLOW and Steady as she goes.
It seems like to me we are a good way from a chassis ready PCB, but that is where we are headed.
Thanks for your input, I have not indicated that it is done. I still have to go over it to find the errors such as you have brought forth. I really appreciate your taking the time to do this.
This board is was not designed for anything other than testing and development. I am almost certain that there will be many minor and maybe major changes to the topology, schematic and overall topology. It is not ready for prime time.
That being said a few of these will be built and tweaked and maybe fried, and once Dadod is satisfied that the design is ready, a new board will be done and maybe we all can enjoy the fruits. I worked at National Semi many years ago in the linear IC R&D section, I think you would be surprised at how many iterations of every device that was made, some IC's were more than two years in R&D. Vector bread boards and flea clips with bus wire, that's what I am talking about.
I am sure the excellent SMD board you laid out is indicative of your considerable expertise, I have never done one. I have never soldered a SMD part to any board. So stay close and teach me some of your insights into all of this. I am still trying to learn some new tricks.
To show you how long in tooth I am, my first Amplifier Kit was a 35 watt Knight-Kit from Allied Radio, 100 North Western ave. Chicago. The first from scratch was the Leach Low Tim.
So give me a minute or two and I will make corrections so that some first prototypes can be hacked and burnt. If I am able to get this board done, then I will put 1 pair of outputs on it and bring that up with a Variac, slowly and once the get to cooking I will add more. I would not think of putting 4 pair on to start with. SLOW and Steady as she goes.
It seems like to me we are a good way from a chassis ready PCB, but that is where we are headed.
Last edited:
By the way we still have some power supply boards to do, front end and back end, and schematic with BOM and such. The Honey Badger was started in July of 2011 and has just had it latest revision done. Lots of work to do.
This is supposed to be fun
Follow excellent Nattawa suggestions and we will have very good amp.
Kris, sorry if I sounded overwhelming. That's how we get our job reviewed where I work. We design and build boards and systems for customers in an industry full of competition and perhaps got much too used to the motto "afford nothing but do it right the 1st time around".
How is going with your layout?
If we use my regulated power supply there are two leds, one indicating power on, other protection kicket on. Post #201 and #205.
By the way, for a LED to shines 2 to 3 mA is quite enough and for that you need 0.5 W resistor max.
I was looking for the diameter of the WIRE that is used to MAKE the coil(inductor). i. e. 18 gauge or 14 gauge...
Kris, sorry if I sounded overwhelming. That's how we get our job reviewed where I work. We design and build boards and systems for customers in an industry full of competition and perhaps got much too used to the motto "afford nothing but do it right the 1st time around".
Doing it right the first time is what I am trying to do. I am reviewing the board post haste. Your astute input will give me a starting point to start going over the board with a fine tooth comb. With everyone's input and eagle eyes it will get done RIGHT the first time we send it to the PCB maker. You went thru several iterations of your board already and I hope to see yours in real life one day. I appreciate your enhancements of the topology and contributions to the project and hope to listen to this and maybe other boards SOON. Stay the course, and enjoy the fruits...