I can tell you the same story: a guy that was specializing on RF made for himself an audio amp, on PCB with ground plane, and surrounded pars of the amp by grid of shields between them, but could not get rid of hum and parasitic oscillations. I took his amp and resoldered it on a prototyping board using thin wires. Oscillations stopped, hum ceased, the guy was surprised, and I had to lead him step by step through his schematic, ground wires, power connections.
For everyone else, the reason that we gold plated the boards was because we did not want them to tarnish with time. Normally circuit boards are protected with something. Either a solder-coating, or some hi temp plastic insulator that is based on a less than perfect, (from a dielectric isolation point of view, material. Apparently Teflon was out of the equation, as it might have been acceptable.
We started out with a solder-coating, just like the standard JC-80 used at the time. I resisted using plastic based protective coatings, because I was concerned with the potential of effecting the sound. As the solder-coated boards aged, they turned spotty and even discolored. We had to clean the boards with Cramolin before we could even use them and they did not stay pristine. Finally, we settled on gold plating, and at first it was a great success. With later batches, as the price of gold started to increase, the gold thickness got too thin to cover properly, and we had mixed results. Today, with the price of gold 5 times more than we first used it, I doubt if it is worth the cost to have an appropriately thick layer of gold deposited on the boards.
When it came to the CTC Blowtorch, Carl Thompson finally convinced me to use the latest technology films, developed for high speed computers (they had problems with the old stuff too, apparently) to bypass the gold, and to reduce soldering shorts, etc. As the CTC Blowtorch was a success, we have used similar technology for all subsequent products. We look back to gold as 'pretty', stable, and a good choice at the time. Tektronix used it as well, in the same way, in the old days. It might be noted that our gold boards are a dull sheen rather than bright. This shows that there is no nickel coating to 'shine back' and that our gold layer must be significantly thicker to work at all as a protective barrier.
We started out with a solder-coating, just like the standard JC-80 used at the time. I resisted using plastic based protective coatings, because I was concerned with the potential of effecting the sound. As the solder-coated boards aged, they turned spotty and even discolored. We had to clean the boards with Cramolin before we could even use them and they did not stay pristine. Finally, we settled on gold plating, and at first it was a great success. With later batches, as the price of gold started to increase, the gold thickness got too thin to cover properly, and we had mixed results. Today, with the price of gold 5 times more than we first used it, I doubt if it is worth the cost to have an appropriately thick layer of gold deposited on the boards.
When it came to the CTC Blowtorch, Carl Thompson finally convinced me to use the latest technology films, developed for high speed computers (they had problems with the old stuff too, apparently) to bypass the gold, and to reduce soldering shorts, etc. As the CTC Blowtorch was a success, we have used similar technology for all subsequent products. We look back to gold as 'pretty', stable, and a good choice at the time. Tektronix used it as well, in the same way, in the old days. It might be noted that our gold boards are a dull sheen rather than bright. This shows that there is no nickel coating to 'shine back' and that our gold layer must be significantly thicker to work at all as a protective barrier.
Last edited:
Actually, no. That is incorrect.
It is more likely that you have a matte finish as opposed to a bright finish.
A matte finish is what you will get if you remove the organic levellers from the solution.
I've purchased matte finish as well, as there were solderability requirements that had to be met. A bright finish had those organics within, they made solderability using R type flux very difficult to pass.
Nickel did not "shine back". Rather, the surface finish of the gold layer determined the characteristics of the reflection, specular or diffuse.
Cheers, jn
Gold!
If you put a really thin plating of gold over copper (5-10 microinch, called a flash) you will have green spots in no time as the copper migrates through the gold. The nickle plating is to prevent this. If you put enough gold (100 microinches+) on it won't happen but few shops will do the work and it will cost a fortune.
The first gold plated PCB's I saw were in all of the HP stuff from the 70's. They were all made in HP's internal shop. We tried for years to get in with no success. They had to abandon it sometime in the late 80's I think for cost reasons. We did get into Tektronix's shop and got some pretty exotic boards made using a special unreinforced polycarbonate substrate (I think) with very good electrical properties. However it would dissolve in most cleaners instantly. Tek closed that shop a few years after we started using them.
We used gold boards in the Spectral DMC10 for reliability as much as anything. The heavy tin/solder plating used as an etch resist at the time would reflow and mess up the solder mask. Also those board would have solderability issues after a few months. The gold has a very long shelf life unlike the tin/solder plating. You can also find old mil-spec parts with gold flash on the leads to ensure good solderability over time.
If you put a really thin plating of gold over copper (5-10 microinch, called a flash) you will have green spots in no time as the copper migrates through the gold. The nickle plating is to prevent this. If you put enough gold (100 microinches+) on it won't happen but few shops will do the work and it will cost a fortune.
The first gold plated PCB's I saw were in all of the HP stuff from the 70's. They were all made in HP's internal shop. We tried for years to get in with no success. They had to abandon it sometime in the late 80's I think for cost reasons. We did get into Tektronix's shop and got some pretty exotic boards made using a special unreinforced polycarbonate substrate (I think) with very good electrical properties. However it would dissolve in most cleaners instantly. Tek closed that shop a few years after we started using them.
We used gold boards in the Spectral DMC10 for reliability as much as anything. The heavy tin/solder plating used as an etch resist at the time would reflow and mess up the solder mask. Also those board would have solderability issues after a few months. The gold has a very long shelf life unlike the tin/solder plating. You can also find old mil-spec parts with gold flash on the leads to ensure good solderability over time.
If any of you have ever worked with quality audio connectors. You might remember seeing both bright gold and dull gold versions.
The 'bright gold' versions had the nickel barrier, and the dull gold versions used either silver as an interface, or directly applied with a thicker gold film on copper. They were the most expensive versions from Tiffany connectors, for example.
The 'bright gold' versions had the nickel barrier, and the dull gold versions used either silver as an interface, or directly applied with a thicker gold film on copper. They were the most expensive versions from Tiffany connectors, for example.
The dull was a result of a specifying a matte finish gold.
The bright (or shiny) was a result of specifying a bright finish gold.
Both pertain to the gold plating process. Not as a result of the barrier metal or lack thereof.
jn
ps. Wow, I've been googling gold platers. The companies I used are gone gone gone. googling matte finish gold produces not too much information...sigh..
Last edited:
I would have people actually look at two connectors made by Vampire, the company who used to make the connectors for Tiffany. If you just look at a female chassis mounted RCA connector, you can note that they make two versions that appear almost identical, except that one is direct plated to copper, without the nickel interface, and the other has a nickel interface. It is difficult to see the difference with just digitized images, but if you compare the two, you will see that the one that is 3 times more expensive has a duller sheen. This is due to the lack of nickel 'shining' (not my term, I learned it elsewhere) through.
The term is both misleading and inaccurate.
It would be better to know what the plating specifications were. When I specified and purchased plated parts, it did indeed cost more to have a matte finish, but if I recall, it was a result of quantity. Matte finish was not the typical plating specification, so it was a special process. If I also recall accurately, matte required a thicker coat for solderability shelf life due to the lack of a barrier for oxygen.
There are other options, like having another metal within the plating of course.
Then again, only the vendor can really tell us why the cost difference.
Given the fact that auplater and SY are here, maybe they can weigh in? My experience is based on specifying, purchasing, testing, and using in production the different plating schemes..they could speak on the chemistry aspect..
jn
Last edited:
I am not in a position to really argue. I did NOT make the boards, I just paid for them. In my experience, the first boards had few problems. Later gold boards had breakthrough from the copper underneath, and had to be cleaned before we would stuff them. The gold film was obviously thinner. However, the original boards look amazingly almost as fresh as new after 25 years. People interested might look at the boards photographed at the VendettaResearch website. They are about 24 years old.
On a connector, a gold layer thin enough that the base metal shines
through would be gone after the first mating attempt.
I have a plating solution here for experimental boards that works
without current, and the deposited layer can be only a few tin
atoms thick, but the copper does not shine through. Not a bit.
(SENO Glanz-Zinn, quite good. The test boards corrode more slowly
and look better. Gibt's in .de beim blauen Klaus. )
through would be gone after the first mating attempt.
I have a plating solution here for experimental boards that works
without current, and the deposited layer can be only a few tin
atoms thick, but the copper does not shine through. Not a bit.
(SENO Glanz-Zinn, quite good. The test boards corrode more slowly
and look better. Gibt's in .de beim blauen Klaus. )
Last edited:
How do the solder fillets look? The biggest problem with the gold plated boards is that the gold dissolves into the solder, and that makes the solder joint brittle. After a long time, there can be lots of cracks, especially with elevated temperatures.
jn
So far as I know, my solder fillets look great. No complaints as of yet, but it has only been 24 years. Maybe, in future, somebody will complain. I doubt it, however, because although you make a good point, it takes a certain percentage of gold to get there and there is just not that much gold on the surface when the solder joint is made. I do believe, that the 2% silver in the solder also limits the scavenging of other materials into the solder.
Recently, a member here gave me a power amp design of mine from 30 years ago to repair.(I will NEVER give a 'lifetime' warranty again.) What was the biggest problem? The plastic surrounding the Supertex mosfets that I used in the driver of the design. Replacing all of the Supertex mosfets essentially fixed the amp up to new-like performance.
John,
I already raised that issue. The pictures I looked at appeared to be hand soldered with a non-temperature controlled iron. They also had some joints that appeared to be redone. There also were a few spots with the green tinge. Slightly more joints did show signs of a tight solder mask.
Now for 30 years ago those cards were quite advanced. Now that time has shown us what techniques don't really last.
Different design techniques in the circuit design will affect the design rules for the PC layout. If you are running very high impedance (which has certain advantages and liabilities) then the PC techniques shown are probably not for you. Same with high current designs.
One old grizzled manager used to say, "If you are doing it the way you did twenty years ago you are wrong! Things change and you should have learned better ways."
I always wondered about compatability of solder on gold boards and long term reliability. I just gave away a Tek 453a scope with gold boards and it still worked great but some of the joints looked a little corroded. Fortunatetly i work in audi0 not medical or military electronics. leave that to the pros.
Some of the green tinge IS on purpose. It is a tiny ring of solder-mask to help prevent solder shorts in the second stage board, made about 1988, whereas the first stage board was designed in 1984 or earlier. Please note the extra spacing around the solder points as well. That is all the solder-mask that I would allow at the time.
- Status
- Not open for further replies.