 |
|
--> Heading: Everything your mother never told you about "High Quality" PC Board construction...
|
|
|
| |
PC Board construction or Point To Point? Mary-Ann or Ginger? from Gilligan's Island, Godzilla or Mothra? Many a debate has taken place over these subjects but the "Information Cocktail" presented below by Andy Marshall, President and CEO of THD Electronics, Ltd. www.thdelectronics.com is quite simply, stellar.
Andy Marshall and Reinhold Bogner have been friends for many, many years and we thought the information below was so well written, articulate and concise that we asked Andy if we could post it here. Since THD and Bogner share many of the same construction techniques and design philosophys we also think Andy and Reinhold are part of the Psyhic-Friends-Network! So go get a tuna sandwich and settle into your bean-bag chair for a good read, by the way I was always partial to Mary-Ann :)
From Andy Marshall:
Not all manufacturers choose to use PC boards just to save money. We use them for consistency more than for price, but making a somewhat affordable amplifier is a nice benefit. I don't think that someone should have to be a lawyer or Microsoft Millionaire to be able to afford a new amplifier that is hand-built, reliable and sounds and feels good to play.
If a PC board is designed correctly and the correct components are used, the amplifier production should be absolutely consistent from one unit to the next. No re-routing of traces should ever be necessary to make an amp function or sound right. If you find it necessary to change and re-rout wires in your amps, then you are not in production, but are just making a series of unstable prototypes. Treble reduction to the point where it reduces the clarity of the amplifier is not an acceptable stabilizing technique for either a PTP or PCB amplifier.
Recently, we got a call from a tech complimenting us on our old Plexi model amplifier (that we built between 1990 and 1995), but he said that it was just a little bit "stiff in the high-end" compared to a real Marshall Plexi. To back up his point, he told us that he had a real Marshall Plexi on the bench next to ours and was comparing the two side by side. What he did not seem to realize was that no two Marshall Plexis sound the same. They were terribly inconsistent with their component sources and values, not to mention the inconsistencies in wire routing.
Taking a point to point or a turret-board amplifier, if one moves the wires around, the entire sound and character of the amplifier can change, often dramatically. This is a well-recognized phenomenon.
If you understand these interactions well, you can design a PC board to sound and feel any way you want it to. Furthermore, every one will sound the same. How many times have you plugged into an old Marshall-50 watt head, only to be terribly disappointed by the sound and feel of the amplifier? While this may be caused by poor tubes, at least in part, inconsistencies in the internal layout of the amplifier often play a significant role.
If you understand how one component affects the component next to it and how one trace affects the trace next to it, then you should be able lay out a circuit board correctly the first time, not by building 10 and picking the best one. Mind you, it takes many years of experience to develop the sort of understanding of the capacitive and inductive interrelations involved. In the old days, I did this for a living for other companies, designing circuit boards for the audio sections of amplifiers, mixing consoles, signal processing equipment, etc... While I am under confidentiality agreements with almost all of my former clients, I can tell you that there is hardly a professional recording studio in the US or Europe that does not have some audio circuit board with my layout in some piece of equipment. After a few hundred such projects, one develops an intricate understanding of how traces and components interact.
A number of years ago, Guitar Player magazine did a review of one of our amplifiers. They stated that they, as a general rule, do not care for circuit board amplifiers, but also said that I had addressed every one of their concerns, and that they had nothing bad to say regarding our use of circuit boards. It felt good to see someone start to understand what it is that we do and why.
Certain components throw a rather large field. Others do not. Some components are very susceptible to the fields from other components, while some are not. Components can affect the signal passing through traces, and traces can affect the signal passing through components. It ends up being an enormous network of positive and negative feedback between components within each other's sway. This is why the distance between specific components on the board and the physical orientation of the components relative to one another (rotational orientation, as well as lateral placement) cannot be ignored. Furthermore, which traces are parallel to one another and at what distance, which traces are perpendicular to one another and that what distance, and the amount of ground plane in-between them can seriously affect the overall sound and feel of the finished amplifier.
Most people design circuit boards either haphazardly or for the greatest parts density/easiest and least expensive manufacture. Neither of these methods belongs in a high-end amplifier, and such approaches give PC Board designs a bad name.
If you know what you are doing, a thicker board is better than a thinner board (ours are .093" or 3/32", most are .062” or 1/16”) and that thick copper is a good idea (ours is 4 oz, most use 1/2 oz or 1 oz). One of the greatest problems facing most circuit board amplifiers is board flex. Board flex creates metal fatigue in the copper. As the copper cannot really "break", it just crystallizes and makes tons of noise. This is much worse in combo amps, of course. We go to the trouble to support our boards ever few inches. Our design standard is that 100 pounds of force on a 1/4" diameter probe should not be able to flex the board more than 20 thousandths of an inch at any point on the board. All of our amps designs must pass this test. For comparison, most Marshall and Fender circuit boards would break under such force, and would flex more than 3/8 of an inch just before breaking.
Through-plated holes are an absolute must, with solder pads on both sides. This makes it much harder for a repairman to inadvertently lift a pad or a trace by overheating or from poor technique. The way that we have addressed this is to start with boards that are clad with 2 oz copper, and in the through-hole plating process we add another 2 ounces. This leaves us with traces and ground planes of 4 ounces, and through plated holes with 2 oz copper in the holes themselves. I have seen some other people start with 3 oz copper, plating on an additional 1 oz, and I have not like the results I have seen. The through-holes pull out too easily.
Contrary to popular belief, “Orange Drop” film capacitors are far from great. They are OK for certain position in certain circuits, but their consistency from one to the next is atrocious. Maybe this is part of why so many people who use them in PTP amps find the need to make wire adjustments. This is a big part of what I mean by using the right components.
As for PCB solder joints becoming problematic with time, this is no more a problem than on PTP. A good solder joint with absolute minimum stress on it (using the right component with the right lead length and the right mounting technique) will yield the longest and most consistent life. Assuming that the flow-solder machine is correctly set up, the right solder, right flux, right solder temperature, right flux temperature, right pre-heat, right cooling, etc… are done, a flow-soldered board will last longer and have higher quality solder joints than a hand-soldered board. If you doubt this, ask yourself the following questions: How do you decide what solder to use? Do you choose SN60, SN63, SN96, Savebit or some other? How do you decide what flux to use in your solder and how much? How do you decide what temperature to set your iron at? It all makes a HUGE difference in the quality and consistency of your solder joints. If you cannot answer all of these questions, then you cannot even have a clue about the lon
g-term consistency and life expectancy of your products. This, along with countless other points, is part of what separates the hobbyist from the professional.
In a PTP amp, the entire surface of the solder joint is exposed to air, and thus, to corrosion. In a through-plated PCB amp, only the top and bottom surfaces of the solder joint are exposed to corrosion, not the majority of the joint, which is within the through-hole, which is where most of the contact is made.
We use only FAA-approved aircraft assemblers in every stage of our manufacturing. They have to understand all of these points completely. The FAA is even more stringent than the military. Also, the aircraft industry is just about the only industry left that uses PCBs for the electronic components wired to chassis-mounted electro-mechanical components like the controls and connectors. They do this because countless FAA tests have shown that devices built this way last longer and are more reliable and consistent than any other method, even taking cost out of the picture entirely. This is, of course, why we use the exact same methods.
In closing, I absolutely believe that circuit boards, when they are well-designed and laid out, are in all ways superior to other manufacturing techniques when one is building amplifiers in quantities. If I did not believe this firmly, I would not be doing it. This said, I think it is a terribly expensive and cumbersome method for hobbyists to attempt. If you don't have a great deal of experience under your belt designing circuit boards, you won't like the results. Point to point and turret-board techniques offer the hobbyist and the small-scale amp shop the opportunity to easily tweak their designs, as is so often necessary. So, unless you're going to be building 50 amps a month or more, it is probably best to stay away from circuit boards.
Andy Marshall President, CEO THD Electronics, Ltd. www.thdelectronics.com
|
| |
|
|
|
Tips, Availability, Hotline: (33) 676 877 860, 7/7 from 8:30 AM to 9 PM !
|
|
Service | Contact | Basket | Home
|
| |
| © Audio Tube Tech 2003 |
|
|
Service
