how to repair this circuit board?

[Abe D]

I have a broken circuit board and accompanying topside components here:

https://imgur.com/a/Pbt1HSU

The red circled areas show the cracked sections. In a pinch, I have
glued the broken parts back together with JB Quick Weld. After several
hours, it seems stable enough to handle.

Now, the question is, what's the best way to rejoin the broken traces?
Initial thought was to just jumper over the solid sections and join the
jumpers between soldered components. Not so sure what to do about the
large screened sections (probably serving as a ground screen for the
flyback topside?). I welcome any tips or thoughts. If I start trying to
scrape the conformal coating, it may come apart again.

Unfortunately, the board cannot be replaced.

Thanks in advance.

Abe

[micky]

In sci.electronics.repair, on Sun, 2 Feb 2020 02:47:18 -0500, Abe D

<abedn...@verizon.net> wrote:

>I have a broken circuit board and accompanying topside components here:
>
>https://imgur.com/a/Pbt1HSU
>
>The red circled areas show the cracked sections. In a pinch, I have
>glued the broken parts back together with JB Quick Weld. After several
>hours, it seems stable enough to handle.
>
>Now, the question is, what's the best way to rejoin the broken traces?
>Initial thought was to just jumper over the solid sections and join the
>jumpers between soldered components.

Sounds good.

> Not so sure what to do about the
>large screened sections (probably serving as a ground screen for the
>flyback topside?).

I'm no pro, but why not the same thing. Does it matter if the edges of
the screen are touching if the screen itself is all connected to itself.
Then you can see if the thing it comes from works or not.

40 years ago I had a roommmate whose father threw a clock-tv at him. I
neve saw another one. So cute. 9" TV, I would have liked to have had
one. Board was broken in about 20 places but it worked after I jumpered
each of them.

[John-Del]

It's too late now but repairing board with a topical coating of epoxy is not the way to stabilize it. The epoxy acts like a top hinge and the bottom of the board will flex away from the crack. To make it more stable, you have to add some epoxy on the other side and hope it adheres.

The proper way to repair a fractured board is to remove any component(s) that crosses the break, and fit the board pieces back together where the seam doesn't even show. You may actually have to break the board a bit more to get it to fit perfectly.

At this point, apply FRESH cyanoacrylate so it wicks inside the fracture and bonds the board completely together from the inside, not just one side.

If done properly, the cracks will be barely visible. Scrape the green mask off the trace just at the break and half the distance the trace is wide. A drop of solder across each land will finish it.

In your case, because the pieces don't fit that well, scrape back the mask and hand wire from point to point and across the fracture with some fine copper stranded wire (one or more strands depending on width of the trace), and coat the entire trace with solder using the copper strand as a sort of "rebar".

Deflux the board and apply either a conformal coating or get some green nail polish to protect the lands.

[Abe D]

On 2/2/20 8:32 AM, John-Del wrote:
> On Sunday, February 2, 2020 at 2:47:22 AM UTC-5, Abe D wrote:
>> I have a broken circuit board and accompanying topside components here:
>>
>> https://imgur.com/a/Pbt1HSU
>>
>> The red circled areas show the cracked sections. In a pinch, I have
>> glued the broken parts back together with JB Quick Weld. After several
>> hours, it seems stable enough to handle.
>>
>> Now, the question is, what's the best way to rejoin the broken traces?
>> Initial thought was to just jumper over the solid sections and join the
>> jumpers between soldered components. Not so sure what to do about the
>> large screened sections (probably serving as a ground screen for the
>> flyback topside?). I welcome any tips or thoughts. If I start trying to
>> scrape the conformal coating, it may come apart again.
>>
>> Unfortunately, the board cannot be replaced.
>>
>> Thanks in advance.
>>
>> Abe
>
> It's too late now but repairing board with a topical coating of epoxy is not the way to stabilize it. The epoxy acts like a top hinge and the bottom of the board will flex away from the crack. To make it more stable, you have to add some epoxy on the other side and hope it adheres.
>
> The proper way to repair a fractured board is to remove any component(s) that crosses the break, and fit the board pieces back together where the seam doesn't even show. You may actually have to break the board a bit more to get it to fit perfectly.
>
> At this point, apply FRESH cyanoacrylate so it wicks inside the fracture and bonds the board completely together from the inside, not just one side.

Not totally clear to me here. Am I going to try applying more epoxy on
the board side of the crack, cyanoacrylate, or both?

>
> If done properly, the cracks will be barely visible. Scrape the green mask off the trace just at the break and half the distance the trace is wide. A drop of solder across each land will finish it.

What about the large, outer section of the board appearing more like a
screen rather than solid trace? I doubt I could find stranded wire wide
enough and pretty fragile there to try scraping. Also, what's the best
tool for the scraping?

>
> In your case, because the pieces don't fit that well, scrape back the mask and hand wire from point to point and across the fracture with some fine copper stranded wire (one or more strands depending on width of the trace), and coat the entire trace with solder using the copper strand as a sort of "rebar".
>
> Deflux the board and apply either a conformal coating or get some green nail polish to protect the lands.

Nail polish would be a go, but what to use for defluxing?

Thanks.

[Ralph Mowery]

In article <r16ua5$7ce$1...@dont-email.me>, abedn...@verizon.net says...

>
> > In your case, because the pieces don't fit that well, scrape back the mask and hand wire from point to point and across the fracture with some fine copper stranded wire (one or more strands depending on width of the trace), and coat the entire trace with solder using the copper strand as a sort of "rebar".
> >
> > Deflux the board and apply either a conformal coating or get some green nail polish to protect the lands.
>
> Nail polish would be a go, but what to use for defluxing?
>
>

Depends on the flux used.

Most of the time use IPA . That is alcohol that is 90% or better with
no addatives that is often in common rubbing alcohol.

The common rubbing alcohol that is 70 % would be ok except most of it
contains other chemicals that leave a residue on the boards.

[et...@whidbey.com]

On Sun, 2 Feb 2020 02:47:18 -0500, Abe D <abedn...@verizon.net>
wrote:

I think some JB weld formulations have very fine metal powder in them
for strength. Maybe you should see if the stuff you used is conductive
before you apply power to the board again.
Eric

[John-Del]

On Sunday, February 2, 2020 at 11:44:57 AM UTC-5, Abe D wrote:
> On 2/2/20 8:32 AM, John-Del wrote:
> > On Sunday, February 2, 2020 at 2:47:22 AM UTC-5, Abe D wrote:
> >> I have a broken circuit board and accompanying topside components here:
> >>
> >> https://imgur.com/a/Pbt1HSU
> >>
> >> The red circled areas show the cracked sections. In a pinch, I have
> >> glued the broken parts back together with JB Quick Weld. After several
> >> hours, it seems stable enough to handle.
> >>
> >> Now, the question is, what's the best way to rejoin the broken traces?
> >> Initial thought was to just jumper over the solid sections and join the
> >> jumpers between soldered components. Not so sure what to do about the
> >> large screened sections (probably serving as a ground screen for the
> >> flyback topside?). I welcome any tips or thoughts. If I start trying to
> >> scrape the conformal coating, it may come apart again.
> >>
> >> Unfortunately, the board cannot be replaced.
> >>
> >> Thanks in advance.
> >>
> >> Abe
> >
> > It's too late now but repairing board with a topical coating of epoxy is not the way to stabilize it. The epoxy acts like a top hinge and the bottom of the board will flex away from the crack. To make it more stable, you have to add some epoxy on the other side and hope it adheres.
> >
> > The proper way to repair a fractured board is to remove any component(s) that crosses the break, and fit the board pieces back together where the seam doesn't even show. You may actually have to break the board a bit more to get it to fit perfectly.
> >
> > At this point, apply FRESH cyanoacrylate so it wicks inside the fracture and bonds the board completely together from the inside, not just one side.
>
> Not totally clear to me here. Am I going to try applying more epoxy on
> the board side of the crack, cyanoacrylate, or both?
>

I had two avenues of advice: the first how to properly repair a cracked or totally broken PC so it's virtually invisible, and advice directed to you having already attempted a repair with epoxy.

Because you have already epoxied the board together, it's really too late to use the cyano adhesive, particularly if you put epoxy into the crack. We never want to use cyano as a filler. The beauty of cyanoacrylate is that it doesn't require a gap in order to put the epoxy in. With cyano, you can fit the board together perfecly and the cyano will wet *into* the PC (either glass or phenolic) and bond it with no gap whatsoever. Properly done, there is virtually no gap across the broken foil and the foil ends butt together almost perfectly.

If you can remove the epoxy you already put down, and there is no epoxy in the break line, then you can refit the pieces so the crack is perfectly joined and virtually invisible. At this point, you can use cyanoacrylate to permanently bond the board internally, as opposed to using external adhesive.

If done properly, no jumpers are even required as the mask can be scrapped off right at the crack and a dot of solder put across the gap. Because the board is bonded internally, there is is virtually no flex and the solder ball will make a perfectly clean and virtually indestructible connection across the break.

If you can't remove the epoxy, you're better off adding a layer to the bottom to reduce the "hinge" effect you will get with epoxy on one side or the other allowing flexing.

As for the large areas of damaged foil, you can add strips of solder wick braid to add some physical strength as well as complete the circuit.

[Tempestinatesttube]

Too well bonded now, even though not perfectly aligned. And I did use
some cyano on the board side just a little while ago to fill in. Not
the best idea for sure and it's only about 49 F in the room where the
board is, so I better wait a full day before doing anything else.

>
> If done properly, no jumpers are even required as the mask can be scrapped off right at the crack and a dot of solder put across the gap. Because the board is bonded internally, there is is virtually no flex and the solder ball will make a perfectly clean and virtually indestructible connection across the break.

The board rejoining is too imperfect for solder dots. Looks like I will
be using jumpers and solder braid, removing flux afterwards, and
covering with nail polish.

> If you can't remove the epoxy, you're better off adding a layer to the bottom to reduce the "hinge" effect you will get with epoxy on one side or the other allowing flexing.

I'm going to glue a couple of round wooden sticks along each length of
the board on the underside. There will be one horizontal stick along
each edge. This won't interfere when I reinstall the board. Hopefully
it will help augment any strengthening needed.

> As for the large areas of damaged foil, you can add strips of solder wick braid to add some physical strength as well as complete the circuit.

Solder braid, good idea. Don't have any though, will have to order
some. What's the best way to remove the conformal coating?

[Tempestinatesttube]

No conductivity, so thumbs up there. Thanks for pointing that out.
Would have hated to go through all this only to realize I was
shortchanged by the epoxy.

[Tempestinatesttube]

Have both 70 and 90, so thumbs up, thanks!

[Tempestinatesttube]

Maybe not so thumbs up when I read more clearly. Standard ISO is all I
have. What about Everclear?

[et...@whidbey.com]

Yeah, that would suck.

[John-Del]

On Sunday, February 2, 2020 at 1:58:34 PM UTC-5, Tempestinatesttube wrote:

> Solder braid, good idea. Don't have any though, will have to order
> some. What's the best way to remove the conformal coating?

I just use a jeweler's screwdriver that's not knackered up. If it has a crisp edge, you can easily scrape off the conformal coating (if the board has it) as well as the green solder mask. Expose the copper and use a bit of flux to improve the wetting and bonding of the solder to the copper.

[Tempestinatesttube]

Very good, thanks again. Will tackle this tomorrow once I have the
braid in hand.

[Tom Biasi]

JB weld is conductive.

[Fox's Mercantile]

On 2/2/20 7:51 PM, Tom Biasi wrote:
> JB weld is conductive.

I found that out the hard way.
I replaced a thermal fuse on a crystal heating oven.
Turned the power on and Blam!
It took a bit to clean that mess up.

--
"I am a river to my people."
Jeff-1.0
WA6FWi
http:foxsmercantile.com

[three_jeeps]

This is exactly what I would recommend. For small traces, 22awg or 24awg copper wire would work fine. For larger traces that carry more power, add two or three strands of wire instead of 1 strand. Not sure of the board mounting arrangement, or how much stress the board may be subject to, but, I'd suggest you consider applying a board stiffener along each side of the board. the stiffener could be wood/plastic and could be screwed (#4 machine screw?) or epoxied in place. If you mount with screws, make sure the holes do not short copper traces. Nylon screws should work well.
This is a general recommendation. An exact spec would be based on seeing the board and mounting arrangement.

[John Robertson]

I agree with the above repairs - we have done similar with other HV
Xformers PCBs successfully that have lasted for many years.

This PC Board was broken either by stress from the weight of the flyback
or bad handling if the board was not mounted. Perhaps rough handling
during a move or the monitor was dropped a foot or so to a floor...

To reduce the risk of future fractures make sure the weight of the HV
transformer/flyback is properly supported otherwise it will break again.

Take care that HV route between the switching transistor and Xformer are
kept clear of any potentially (yeah, I know) conductive material when
you are supporting it - as it sees back EMF of up to 1500VDC. Other HV
traces also need to be kept isolated from the supports. Note that any
added supports can accumulate dust over the coming years and that this
dust can provide a conductive path...

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
(604)872-5757 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."

[Tempestinatesttube]

I picked up some desoldering braid and used that across the broken
traces. I was careful to first scrape away any coatings with a flat
blade screw driver. After maybe 20 min of work, all of the gaps were
bridged and time to test. Once plugged in and switched on, whola! All
working again.

Thanks for the group's help. Lots of helpful suggestions. I really
didn't know if it would work even after doing all this as it had arrived
that way through the mail, and the person I got it from sent it parcel
post with no insurance and in a single flimsy box. About the worst way
anyone could send fragile electronics. However, it is working again and
I am happy. I do want to add supports across the ends of the board
horizontally for more strength, but otherwise I think it's all good now.

By the way, some of you might want to check JB Weld's website. Only
certain of their epoxies are conductive and others are not. The one I
used (JB Quick Weld) was not. Just FYI.

Thanks again!

[Jeff Urban]

I could do that in a half hour. I should show you the regulator board in my bench receiver, it is twice as busted up as that. Send it over and I'll do it for, umm, forty bucks.

First, get a scraper. I got a special tool to remove the soldermask. (the green shit)Then bridge most of the bigger foils with solder - only. Then take cyanoacryilate (super glue) and glue it together from the top. Then go to each connection you bridged, of course to made it to the right position, I have had to use a hammer to pound them back together at times, and that is why you do not want the extra thick glue, you want low viscosity. And at each connection reflow the solder blob and embed used solder wick in it. It will lend some strength and therefore longevity.

After that get to the smaller connections, any which way. The short distance makes it easy. Some you can do with the cutoffs from the leads of components you have installed in other stuff. (that is REALLY what the tray on your soldering station is for :-)

Shit, the one you got is so easy, if you can get to Cleveland bring it over with a case of Bud and I will just do it right in front of you.

[J-J]

My prior post inquired as to a cheap ink and paper that could be used to
test out my recently acquired barograph from an estate sale.

After some trial and error and with the group's help, I have managed to
get some accurate traces:

https://imgur.com/a/kKWKhQh

However, at least one issue still remains: the large ink spot you see
occurs within an hour of me refilling the tiny ink well on the device
arm. Once all that ink "hemorrhages", the trace then proceeds as you see
with no more spots. Any idea what could be causing this? Am I adding
too much ink each time?

One other thing I'd like to ask about is paper. For the chart, I had
printed them out on a piece of Georgia Pacific standard 20 lb weight
8.5x11 multipurpose paper and backed it with a couple of strips of
packing tape in the event a bleed-through occurred (and good thing I
did!). Suggestions for maybe a thicker paper I could try would be
welcome, with the idea to increase absorbency and still keep cost low.
In a pinch, I've been thinking of just doubling up the sheets I have
except the new thickness would not be uniform and glue would decrease
absorbency.

Thanks in advance.

JJ

[Jeff Liebermann]

On Wed, 5 Feb 2020 11:32:23 -0500, J-J <no...@none.non> wrote:

> https://imgur.com/a/kKWKhQh
>
>However, at least one issue still remains: the large ink spot you see
>occurs within an hour of me refilling the tiny ink well on the device
>arm. Once all that ink "hemorrhages", the trace then proceeds as you see
>with no more spots. Any idea what could be causing this? Am I adding
>too much ink each time?

Was the drum moving when the blob appeared, or was it stopped? If the
clock mechanism had stopped, I could see why there was a blob.
However, if it was moving, as in the visible trace in the photo, I
would ask what is different between when the blob formed and when it
did not? Other than a stalled or erratic clock mechanism, it could be
a "bubble" in the paper, where it was warped or curled, some defect in
the paper itself, or one of the problems mentioned in the article at:
<https://www.metcheck.co.uk/blogs/barographs/156686919-how-to-get-a-good-barogram>
"The two halves of the nib bucket should be of even length..."
If only one point is touching the paper, I think you might get a blob.

>One other thing I'd like to ask about is paper. For the chart, I had
>printed them out on a piece of Georgia Pacific standard 20 lb weight
>8.5x11 multipurpose paper and backed it with a couple of strips of
>packing tape in the event a bleed-through occurred (and good thing I
>did!). Suggestions for maybe a thicker paper I could try would be
>welcome, with the idea to increase absorbency and still keep cost low.
>In a pinch, I've been thinking of just doubling up the sheets I have
>except the new thickness would not be uniform and glue would decrease
>absorbency.

Dunno. I was wrong about the paper needing to be absorbent (see above
Metcheck URL). I suggest you try a variety of paper types and see
which works best. Coated, not coated, 22 lb, vellum, photo paper
(glossy or matt), butcher paper, wax paper, premium clay coated
plotter paper, etc.
<https://plotterpaper.plotter-paper-rolls.com/2014/07/17/premium-coated-plotter-paper/>
Visit an art store and see what they have to offer.
Also look for "pen plotter paper":
<https://www.graytex.com/pen-plotter-paper.htm>
The pen is critical because most "plotter paper" is for inkjet
plotters.


--
Jeff Liebermann je...@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558