I have a question for those of you who have worked with SMD devices in the past.
I am reworking some of the Sonic Impact 5066 boards, and replacing some of the SMD 0603 resistors on it. Desoldering it is no problem. I hit them with my rat-shack desoldering iron, then used a copper wick to soak up any left over solder.
To solder the devices, I wet one of the pads with some .022" diameter rat-shack silver-bearing rosin-core solder. I then tacked down the device on one side, waited for it to cool, then soldered the other. Finally, after waiting 20-30 seconds, I reflowed the original tack.
Anyhow, after soldering down 4x 20k resistors, they all seem to measure 16.5k or so with my DMM. I know these resistors are still good, because if I desolder one and measure it, it measures within 1% of 20k.
I'm trying to make sure everything is working properly and to spec before putting this amp back together, but these values not showing up correctly is driving me nuts. For comparison, I measured some of the SMD 20k resistors in an unmodified sonic impact, and they measured correctly.
Any advice the more experienced folks on here can provide would sure be appreciated!
Thanks in advance!
Mike
I am reworking some of the Sonic Impact 5066 boards, and replacing some of the SMD 0603 resistors on it. Desoldering it is no problem. I hit them with my rat-shack desoldering iron, then used a copper wick to soak up any left over solder.
To solder the devices, I wet one of the pads with some .022" diameter rat-shack silver-bearing rosin-core solder. I then tacked down the device on one side, waited for it to cool, then soldered the other. Finally, after waiting 20-30 seconds, I reflowed the original tack.
Anyhow, after soldering down 4x 20k resistors, they all seem to measure 16.5k or so with my DMM. I know these resistors are still good, because if I desolder one and measure it, it measures within 1% of 20k.
I'm trying to make sure everything is working properly and to spec before putting this amp back together, but these values not showing up correctly is driving me nuts. For comparison, I measured some of the SMD 20k resistors in an unmodified sonic impact, and they measured correctly.
Any advice the more experienced folks on here can provide would sure be appreciated!
Thanks in advance!
Mike
robbie b said:Consider the fact that you will measure the resistor within a circuit. This can and probebly will cause the difference.
Rob
I considered this initially, so I popped open an unmodded amp for comparison. The resistors in the unmodified amp measure spot on to their value.
The resistors I'm replacing are the input and feedback resistors. The input resistor goes to to the inverting input of the amplifier, while the feedback resistor goes between the op-amp output and the inverting input.
I'm just wondering if I'm doing a crappy soldering job and either not getting a good electrical connection or if it's something else that I'm missing entirely.
You may be having a differential expansion issue. The board is not hot, so as the component cools, it undergoes strain.
Try heating the board to 100 to 150 C, then reflow the resistors.
It may be possible to reflow both pads by touching one side.
After the whole shebang cools, re-measure.
Cheers, John
Try heating the board to 100 to 150 C, then reflow the resistors.
It may be possible to reflow both pads by touching one side.
After the whole shebang cools, re-measure.
Cheers, John
jneutron said:
Thanks John. Short of putting the board in the oven, I can't think of an easy way to heat it up. Unfortunately, I have a rather small 1 bedroom apartment, so most of this work is taking place on the breakfast bar attached to my kitchen
I can indeed reflow both pads by touching one side, but the little buggers like to stick to the iron when i do that. It's been an wonderful exercise in patience using needle-tipped tweezers to hold them in place while soldering.
marin.weigel said:
I'm replacing the parts to reset the gain from 1.8 to 1.0 on the amp. I don't think it's the opamp -- the opamp is actually part of the tripath chip, and like I said, these measure fine on the other amps.
The only time I run into this problem is when replacing the part. I'm almost certain i'm screwing something up here in the soldering process.
SY said:
Hmm.. I thought about this as well. I've tried scrubbing the parts down with 100% alcohol and a toothbrush. Is there a better way to was rosin out?
Thank you all so much for your help. I hope I can figure this out. SMD seems to be the way everything is going, and I don't want to be left out
Build the $20 hot air pencil (of if you have $$$, buy one). I don't recommend doing SMD with a contact iron. These parts are too fragile and uneven heating has too much risk of damaging them.
Use no-solids flux. When you are done, pour some hot alcohol on the PCB (microwave the alcohol, don't put it on the stove, especially if it's gas) and brush any flux left with a small brush (stiff plastic bristles). To avoid the quickly drying alcohol from depositing some leftover flux, rinse with hot water before the alcohol dries out (this way you also waste less alcohol). Dry the board quickly, though, as plastic parts absorb some of the water.
Use no-solids flux. When you are done, pour some hot alcohol on the PCB (microwave the alcohol, don't put it on the stove, especially if it's gas) and brush any flux left with a small brush (stiff plastic bristles). To avoid the quickly drying alcohol from depositing some leftover flux, rinse with hot water before the alcohol dries out (this way you also waste less alcohol). Dry the board quickly, though, as plastic parts absorb some of the water.
motherone said:
I can indeed reflow both pads by touching one side, but the little buggers like to stick to the iron when i do that. It's been an wonderful exercise in patience using needle-tipped tweezers to hold them in place while soldering.
Hi Motherone, here is a site that shows you how to make a quick and dirt cheap smd hold down for soldering. I've used it and it works great.
http://www.psnw.com/~kd7s/smdhd.html
Stan
One can easily solder SMT passives with a fine tip soldering
iron. It should be temperature controlled. I use an adjustable
Weller iron with a 1/32" tip. Temp is set at about 730 deg F.
This can easily handle down to 0805 packages. Much better
is a fancy iron like the Metcal, but usually a DIY'er doesn't
want to spend $500+ for a soldering iron.
Use a very fine rosin core solder. Make sure the solder whets
to the part and the board.
Things you might have done wrong: The SMT resistors have
the resistive element on top, so if you poke the top too hard
with the tweezers while holding down the part for soldering,
you might poke through the top coating and damage the
resistive element. You could possibly have cracked the resistor,
though that usually (but not always) creates an open circuit.
Also if you are not careful, you might have separated the metal
end cap (the thing you solder to) from the ceramic substrate
and the resistive element.
Also, be especially careful about not overheating SMT ceramic
caps. It's not overly hard to induce stress cracks due to the
uneven heating and the cap shorts, opens, or changes value
dramatically w/o any obvious external signs.
iron. It should be temperature controlled. I use an adjustable
Weller iron with a 1/32" tip. Temp is set at about 730 deg F.
This can easily handle down to 0805 packages. Much better
is a fancy iron like the Metcal, but usually a DIY'er doesn't
want to spend $500+ for a soldering iron.
Use a very fine rosin core solder. Make sure the solder whets
to the part and the board.
Things you might have done wrong: The SMT resistors have
the resistive element on top, so if you poke the top too hard
with the tweezers while holding down the part for soldering,
you might poke through the top coating and damage the
resistive element. You could possibly have cracked the resistor,
though that usually (but not always) creates an open circuit.
Also if you are not careful, you might have separated the metal
end cap (the thing you solder to) from the ceramic substrate
and the resistive element.
Also, be especially careful about not overheating SMT ceramic
caps. It's not overly hard to induce stress cracks due to the
uneven heating and the cap shorts, opens, or changes value
dramatically w/o any obvious external signs.
BrianL said:
I picked up some smaller diameter solder. I have a 1/32" tip for my weller, so I'm going to try that with the new solder in the next couple of days. Like I said in the posts above, if I desolder the resistor, it still measures properly. I'm starting to lean towards maybe rosin shorting it out or something. Anyhow, I have probably 40 or so of these 0603 resistors left, so I'll keep trying 'til I get it right (or order more if I screw up that many times).
I stress my previous suggestion. Take a look here:
http://www.diyaudio.com/wiki/index.php?page=SurfaceMountDeviceRework
And the linked from there page:
http://www.usbmicro.com/odn/documents/46.html
Or buy a new commercial one for as little as $60.
If you had bothered to search even only within diyaudio, you probably could have avoided this thread completely.
http://www.diyaudio.com/wiki/index.php?page=SurfaceMountDeviceRework
And the linked from there page:
http://www.usbmicro.com/odn/documents/46.html
Or buy a new commercial one for as little as $60.
If you had bothered to search even only within diyaudio, you probably could have avoided this thread completely.
Prune said:
Thanks Prune. I'm obviously a drooling moron who's never googled for instructions on this thing called the "interweb." In fact, I'm so dense that I didn't bother searching the very forum that I read regularly for assistance before attempting this project or posting this topic.
While I appreciate you trying to help out, condescending comments don't help anyone -- it only encourages others to ignore the merit of what you may have to contribute. Perhaps I'm misunderstanding the tone your last sentence, but it rubbed me the wrong way. Out of all the replies, this is the only one with a negative tone to it.
However, if I'm still struggling with the contact iron (this is, afterall, only 8 solder joints), I'll look at acquiring a pump for my desoldering iron and see about building the iron mentioned in your second link.
SMD resistors changing value
Hi Guys
I have had the same problem with SMD resistors adopting a permanent value change after soldering (obviously compared out of circuit and all other things being equal). The problem appears to be most marked on 'special' resistors eg high voltage withstand types or precision types.
Panasonic ECHU series SMD caps seem prone to a similar phenomenon - after soldering the cap value drops to about 20% of it's nominal!
I've also had trouble with leftover solder paste which gets squashed under the component and doesn't snap-melt onto to the pcb land. This causes spurious resistive paths in the circuit and on one occasion this resulted in an ADC buffer amp having THD+N of -60dB instead of its nominal -90dB! Apart from my own SMD bumblings at home and at work, I've encountered these problems with batches of pcbs supposedly 'professionally assembled' by subcontractors in both UK and China.
In all cases I've found a solution is to use a soldering iron with a very small tip, run it at a low temp like 250C and be as quick as you can in the soldering process. And avoid re-using SMD parts you've removed, with the obvious exception of pricey stuff. But I think the best method of all is to find some really skilled tekkie geezer who's a wizard with manual SMD assembly and bribe him with a case of beer to do your SMD assembly as a PJ.
Cheers all
John Hope
Hi Guys
I have had the same problem with SMD resistors adopting a permanent value change after soldering (obviously compared out of circuit and all other things being equal). The problem appears to be most marked on 'special' resistors eg high voltage withstand types or precision types.
Panasonic ECHU series SMD caps seem prone to a similar phenomenon - after soldering the cap value drops to about 20% of it's nominal!
I've also had trouble with leftover solder paste which gets squashed under the component and doesn't snap-melt onto to the pcb land. This causes spurious resistive paths in the circuit and on one occasion this resulted in an ADC buffer amp having THD+N of -60dB instead of its nominal -90dB! Apart from my own SMD bumblings at home and at work, I've encountered these problems with batches of pcbs supposedly 'professionally assembled' by subcontractors in both UK and China.
In all cases I've found a solution is to use a soldering iron with a very small tip, run it at a low temp like 250C and be as quick as you can in the soldering process. And avoid re-using SMD parts you've removed, with the obvious exception of pricey stuff. But I think the best method of all is to find some really skilled tekkie geezer who's a wizard with manual SMD assembly and bribe him with a case of beer to do your SMD assembly as a PJ.
Cheers all
John Hope
A year or so ago I googled DIY ans SMD and found references to using a cheap toster oven to do the work rather than and iron on individual components. See: Wagner Lipnharski - UST Research Inc.
http://www.ustr.net/smt/
Note the comment a bout rapid heating and cooling damaging SMD components. Perhaps this relates to some of the problems mentioned.
http://www.ustr.net/smt/
Note the comment a bout rapid heating and cooling damaging SMD components. Perhaps this relates to some of the problems mentioned.
Another SMD oven link: Have you seen my new soldering Iron?
http://www.seattlerobotics.org/encoder/200006/oven_art.htm
http://www.seattlerobotics.org/encoder/200006/oven_art.htm
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.