i'm amazed you are unable to see the baiting going the other way…
also, if we are to believe John, while well known and getting free Bentley rides, he doesnt make a lot of money; Sy on the other hand in his senior engineering position in materials sciences; is likely to do alright...
Actually that would make for a poor connection. There is a lot of intermetallic bonding (sort of local alloying) at the solder joint. Preventing solder from getting to the joint will degrade the connection.
In normal production environments parts are rarely old enough to have much surface oxidation. Usually anything that is more than 6 months from manufacture (in commercial production) is rejected. In high rel (avionics, implantable medical . . . .) I suspect the time is shorter and the storage is more controlled. The hi rel military spares you occasionally find on the surplus markets may have gold plated leads. These are to ensure solderablity under adverse conditions.
If one is using NOS (new old stock) parts cleaning the leads may be important. Or use a stronger flux and wash the board afterwards. Also the board will deteriorate with time. In the old days we would bake them to get the crud out of the holes. Today those vendors would be without customers. Board quality is much better.
should you step back and look around,
you may notice what is the result of this sort of thing -
I do. I use a quantum-based material that comes out of classified research. It's called flux.
se
Okay, if the surfaces were tinned very lightly just prior to soldering, after scraping; the pressure would then force excess solder from between the contact surfaces - still a problem ...?
Probably. You need enough space for the solder to wick up between the connection. If you want to use something like ultrasonic bonding or welding the story is different. Welding will change the metal at the weld. I don't know if that's good or bad. Or a good gas tight crimp is good but not possible directly to a PCB.
The intermetallic bond is the important part. I'm sure some physicist here can explain why that will be better than a mechanical contact can ever be.
I'm sure some here are up on the latest RoHS solders etc. that give good results. Beware of tales from old coots who reject the new stuff from those young whippersnappers. . . Sometimes that new stuff really does work.
The intermetallic bond is the important part. I'm sure some physicist here can explain why that will be better than a mechanical contact can ever be.
I'm sure some here are up on the latest RoHS solders etc. that give good results. Beware of tales from old coots who reject the new stuff from those young whippersnappers. . . Sometimes that new stuff really does work.
I keep a ball of steel wool in my soldering kit for the extreme cases, just to pinch it around the leads and give the component a couple of spins with the other hand.
Don't do this over the work area, of course. Do it over the waste can.
Having clean copper on the board is just as important as the component leads, IMO.
I've also always heard a good mechanical connection is the first step to a good soldered connection.
Don't do this over the work area, of course. Do it over the waste can.
Having clean copper on the board is just as important as the component leads, IMO.
I've also always heard a good mechanical connection is the first step to a good soldered connection.
John, I was going through some of your previous posts here on amplifier design.
What's the deal with output coils ?
Are they
a) a bad idea, so don't use them or
b) they're a great precaution so use them if you like.
why ?
If you have addressed this some time ago I apologise - it's difficult parsing this thread for the nuggets.
Last edited:
Well, we got some interesting questions.
Answer 1. I have not ridden in Jack's Bentley but a few times, and not for a year or more. I did SEE Jack's Bentley in his garage, recently. What is the big deal?
Answer 2. I think that careful soldering requires a minimum separation between metals and you have to use GOOD SOLDER, that we clean with Cramolyn (Caig) before inserting in solder dispensers. I personally always use SN62 solder. I am sure there are some excellent lead free solders as well.
Answer 3. Jack recommends laser welding for critical junctions. I have never used it, but it seems like a good idea.
Answer 4. Circuit board construction and quality is very important. I use either polyamide (Vendetta) or Teflon (Blowtorch) for my best circuits, FR-4 for power supplies. I ALWAYS clean each board with Cramolyn and Isopropyl alcohol before adding any parts.
Answer 5. I always try to clean the flux from the Boards after soldering. Usually a combination of manual removal with a dental pick, then a wash with Freon to remove the final stuff. It doesn't take much Freon, if you have removed the bulk of the flux, mechanically.
Answer 6. I stopped using output coils about 25 years ago. Before, I used the usual 2uH paralleled by a 1-10 ohm resistor. A large debate on this website occurred between Bob Cordell and me over this. It is on another thread, I'm pretty sure. I think that they have the potential to reduce sound quality.
Answer 1. I have not ridden in Jack's Bentley but a few times, and not for a year or more. I did SEE Jack's Bentley in his garage, recently. What is the big deal?
Answer 2. I think that careful soldering requires a minimum separation between metals and you have to use GOOD SOLDER, that we clean with Cramolyn (Caig) before inserting in solder dispensers. I personally always use SN62 solder. I am sure there are some excellent lead free solders as well.
Answer 3. Jack recommends laser welding for critical junctions. I have never used it, but it seems like a good idea.
Answer 4. Circuit board construction and quality is very important. I use either polyamide (Vendetta) or Teflon (Blowtorch) for my best circuits, FR-4 for power supplies. I ALWAYS clean each board with Cramolyn and Isopropyl alcohol before adding any parts.
Answer 5. I always try to clean the flux from the Boards after soldering. Usually a combination of manual removal with a dental pick, then a wash with Freon to remove the final stuff. It doesn't take much Freon, if you have removed the bulk of the flux, mechanically.
Answer 6. I stopped using output coils about 25 years ago. Before, I used the usual 2uH paralleled by a 1-10 ohm resistor. A large debate on this website occurred between Bob Cordell and me over this. It is on another thread, I'm pretty sure. I think that they have the potential to reduce sound quality.
I hope that Demian is right, and that relatively fresh components will not have any significant flux buildup. Demian and I also work for Constellation, and they cannot clean the leads the way that I would prefer, unfortunately. Of course, Parasound has the same compromise as Constellation, but for myself, when I am in control, I chose mechanical scraping. Drives the techs nuts!!
Thanks for that - I looked around and found your discussions on this on the Bob Cordell Negative Feedback Thread -- approximately in this vicinity http://www.diyaudio.com/forums/soli...terview-negative-feedback-29.html#post1287213 posts-wise...
Demian
From the slides I’ve seen and the papers I have read, I have not seen data for "local alloying" at a solder joint. There is a lot going on within the eutectic solder but not btn solder and components surfaces, at least for Pb-Sn solder. I would appreciate any relevant data.
It is counter-intuitive but tight mechanical coupling, forces resin and flux to wet the touching surfaces through capillary action. They can flow against gravity pull and always toward the point of higher temperature.
At the same time, tight mechanical coupling makes for proper transfer of mechanical forces.
From my limited experience, solder joints in commercial audio equipment are frequent victims of operational temperatures themselves and of mechanical stresses generated by delta temp.
George
I saw this advice from you here several years ago and have been scraping my leads ever since
On "peeking" I disagree, John has repeatably said for him the differences disappear under DBT. This is meerly a restatement of what John has already said.
If you believe a reviewer that always knows what they are testing that's your problem.
Demian
From the slides I’ve seen and the papers I have read, I have not seen data for "local alloying" at a solder joint. There is a lot going on within the eutectic solder but not btn solder and components surfaces, at least for Pb-Sn solder. I would appreciate any relevant data.
It is counter-intuitive but tight mechanical coupling, forces resin and flux to wet the touching surfaces through capillary action. They can flow against gravity pull and always toward the point of higher temperature.
At the same time, tight mechanical coupling makes for proper transfer of mechanical forces.
From my limited experience, solder joints in commercial audio equipment are frequent victims of operational temperatures themselves and of mechanical stresses generated by delta temp.
George
Do some searches with keywords such as gold tin solder intermetallics embrittlement.
There are a lot of problems that can result from using gold in conjunction with tin, for example, both in solder joints and in mechanical connectors.
Somewhere in the articles or posts that such searches will find, there are some that discuss the problem in more-general terms. I remember seeing a list of metal types, sorted such that physically connecting metals that were distant from each other, on the list, could cause problems.
I may have found the relevant papers while researching the famous (at the time) case where a computer manufacturer was forced to replace memory modules because they had mated gold-plated connectors with tin-plated connectors. That may have led me to also find the solder-joint alloy-formation info.
Last edited:
John,
My own experience of output coils leads me to believe that if you want avoid using o/p coils in amplifiers with global feedback without incurring sonic penalties, then it is necessary to design an amplifier which does not suffer from ANY internal instability when driving a capacitive or a capacitor like load.
Just removing an output coil from a regular 3 stage, follower output design will probably make things worse rather than better.
Not understanding this distinction caused me some misunderstanding for a while.
I followed that other thread that discusses this for a while but I never got to understand the nature of problem that coils can bring.
I avoided using coils for a long time but when I actually got around to trying them I found that with my current design adding a coil ( 1.5uH + 6.7R ) brought very clear advangtages and no disadvantages that I can distinguish.
I think I will not be able to get to the bottom of this issue to my own satisfaction until I build an amp that is so stable that it genuinely does need any help from a coil - only then will I be truly be able to assess any disadvantages the coil may be bringing - so for now, keeping an open mind.
mike
Last edited:
Or you could like the rest of the electronics industry use the IPC specs and training material, and not have to second guess how to do a reliable solder joint:
http://www.ipc.org/4.0_Knowledge/4.1_Standards/SpecTree.pdf
I'm up for it..I neglected to take the advice freshman year, when they insisted I take English as a second language...
I share that opinion. Good advice.
Too tight, and there will be problems. 1. The flux may not penetrate the space, which eliminates any possibility of a solder joint, and 2. If the flux gets in, the liquid metal advancing in to form the intermetallics may be unable to push the residual flux and byproducts out. Leaving some bad stuff buried. The typical activator for RMA flux is zinc chloride, and it's breakdown produces hydrochloric acid...
Mil spec solderability was typically 6 months shelf. After 6 months, the parts must be able to be soldered, and the test was the use of R type non activated flux, and a solder pot of eutectic lead/tin at 250 degrees C, with a specific immersion, dwell, and removal requirement. Plated lead/tin finishes cannot pass this, only well made reflowed surfaces. Also, gold plated bright is a problem due to organic levelers, and matte plated is porous.
Old stock is best pre-tinned in a solder pot using the alloy of the final board.
At first dip, use R type flux, and withdraw very slowly from the pot. The slower the withdrawal rate, the thinner the coat will be. This will show if the surface is solderable or not. Withdraw fast, and the coat will be thicker, and it will cover any solderability problems.
If the slow withdrawal shows lots of pockmarks indicative of dewetting, you can either re-dip multiple times with R flux, or go to an RMA flux and multiple redips. Do this only for solid leads, stranded will wick the flux, you will never clean it out fully.
Dewetting is caused by two things. Inability of the flux to activate the surface, or outgassing of embedded chemistry. Wires are typically made by drawing the copper through progressively smaller dies, and in many cases mineral oil is used to lubricate the process. I have found that it sometimes requires up to 5 dips to recover copper which has mineral oil embedded within.
Be careful, as sometimes the surface will coat, then outgassing will passivate either the first copper/tin intermetallic, or the second.
Also, if you use a cloth or paper towel to wipe off excess solder when liquid, you will remove the free lead/tin and expose the top intermetallic layer...that layer passivates in air, and will not respond to R type flux nor many RMA fluxes. DO NOT WIPE MOLTEN SOLDER off any surface. If you do, you have to remove it to copper.
The flux is a key part of the process.
All solder joints require the intermetallics. copper/tin has two intermetallic layers. Cu3Sn, and Cu6Sn5. (cool, subscripts worked...)
Oops, leaves me out..
Ask Toyota owners about ROHS and unintended acceleration..
Yup.
There are. But the MOST IMPORTANT thing is the flux, and it's ability to activate the surface metal. Nickel and copper require different chemistries.
Google copper tin intermetallics.Demian
From the slides I’ve seen and the papers I have read, I have not seen data for "local alloying" at a solder joint. There is a lot going on within the eutectic solder but not btn solder and components surfaces, at least for Pb-Sn solder. I would appreciate any relevant data.
You are correct in capillary action. But the geometry of the joint does play a part, as does the ability of the flux residue to be forced out by wicking. Not a small deal, just ask the LHC guys. They designed a joint which was about 3/4 inches wide, 6 inches long, and did not provide any method for either introducing solder in, or allowing flux residues to leave the joint. (two of six flaws in their interconnect joints). It was a mess..
Yes. And it is by far the worst for large surface mount devices. The removal of lead made this problem even bigger, as the extra 38 degrees C causes even more differential expansion between solidus and operational temperature.
Also, lead/tin shrinks 15% during solidification(edit: I read the article in 1984, so will have to find it again, it may be 15% bismuth and 25% shrink, but the idea's the same..), so puts the solder in tension. That, low cycle fatigue, and lead tin creep tendencies cause lots of "top of the pole" (telephone pole) failures. Back in the mid 20th century, Bell labs setup a wave machine with tin/lead/bismuth to eliminate that shrinkage upon solidification problem. Nowadays, some people are revisiting bismuth to reduce the solidification temperature raised by the lack of lead, I believe they are backing into the good effect of reduction of density change from liquid to solid.
Totally. I've seen how this causes cracked solder joints. Also, free gold on tin can slowly form gold-tin intermetallics...look up purple plague..
jn
ps...sorry for the length of my post.
Last edited:
- Status
- Not open for further replies.