Topics

uBITX No Power Output all of a sudden #ubitx

Jim Sheldon
 

While prepping a V3 uBITX I had already built up to give to a friend of mine, all of a sudden the RF Power output went to ZERO.  Thought for sure I'd somehow blown 1 or both of the IRF510's, but I learned (and taught) electronic equipment troubleshooting during the later years of my 20 year US Army career and one point I learned was that nothing is always as it seems.

First thing I checked was voltage to the drains of the IRF510's -- ZERO (yes it was hooked up and turned on).  Tracing the voltage back, I had nothing on the supply ends of L8 OR L9.  I checked continuity between those same points of L8 and L9 and had good connections there. 

I found that I had +12V on the PA Power pin of P1 but nothing on the supply end of L8 which gets connected by a short trace to a very small VIA through the board and a short trace to L8 on the underside.  Looking at the underside, the trace was not burned but it turns out that tiny little VIA was open -- poor plating through the hole.  I put a piece of 30 gauge wire wrap wire through the via and soldered it on both sides.  This brought the +12 back to the supply side of L8, L9 and subsequently the drains of the IRF-510's .  Put it all back together and I now had full power again.

I've noticed in the past a few people having similar problems and not sure if they ever did find them.  This might be the cause and is definitely worth checking as there will be no visible indication of anything bad between the PA Power pin on P1 and the rest of the PA circuit.

Hope this is helpful to some of you.

Jim Sheldon, W0EB

 

Thanks for the tip Jim.

Joel
N6ALT

n4eev
 

Nice fix Jim! Thanks for the write up.

all my best,

Mike N4EEV

Clark Martin
 

Poor plating or did normal plating act as a fuse??


Clark Martin
KK6ISP

On Jul 20, 2018, at 1:10 PM, Jim Sheldon <w0eb@...> wrote:

I found that I had +12V on the PA Power pin of P1 but nothing on the supply end of L8 which gets connected by a short trace to a very small VIA through the board and a short trace to L8 on the underside.  Looking at the underside, the trace was not burned but it turns out that tiny little VIA was open -- poor plating through the hole.  I put a piece of 30 gauge wire wrap wire through the via and soldered it on both sides.  This brought the +12 back to the supply side of L8, L9 and subsequently the drains of the IRF-510's .  Put it all back together and I now had full power again.

 

Jim,

Probably not enough safety margin of current carrying capacity in the PTH was considered.

Raj

At 21/07/2018, you wrote:
Looking at the underside, the trace was not burned but it turns out that tiny little VIA was open -- poor plating through the hole. I put a piece of 30 gauge wire wrap wire through the via and soldered it on both sides. This brought the +12 back to the supply side of L8, L9 and subsequently the drains of the IRF-510's . Put it all back together and I now had full power again.

Jim Sheldon
 

I suspect poor plating.  Rig had been working just fine and I had power out set to 5 watts into a Microwave Associates 20 watt dummy load on my YouKits DP-1 digital wattmeter on 40 meters.  I checked the resistance on two of the other V3 uBITX boards I have and it varies slightly as you press on the board with an insulated tool near that VIA. 

Don't know about the V4 boards, but during any V3 build, I'd suggest running a piece of 30 gauge wire wrap wire through that hole and soldering it on both sides as a precaution before mounting the board.  You will probably have to carefully scrape the solder mask from both top and bottom of the VIA pad.  I'll try and take and post a good hi resolution photo of it later in the day so you can positively identify the spot.

Jim - W0EB

Jim Sheldon
 

Whatever, Raj, I recommend soldering a bit of 30 gauge wire wrap wire through the VIA on both sides at least as a precaution.

Jim

Jim Sheldon
 

As promised, pictures of the offending VIA.

W0EB

David Wilcox
 

So you had to figure out that it was a VIA and not an unused plated through hole solder point?  I have never thought about that. It seems that many boards have unused holes for parts that later were not needed.  Are there any tip offs that a point on the board is a VIA?

Dave K8WPE

On Jul 21, 2018, at 10:26 AM, Jim Sheldon <w0eb@...> wrote:

As promised, pictures of the offending VIA.

W0EB<Img_5100.jpg><Img_5101.jpg>

Jim Sheldon
 

Parts on factory PC boards are 99 percent of the time outlined by silkscreened legends.  VIA's are unmarked.  Hole far too small for a component lead (30 gauge wire barely fit the hole), what would you think?  Get real.

On Jul 21, 2018, at 5:12 PM, David Wilcox via Groups.Io <Djwilcox01@...> wrote:

So you had to figure out that it was a VIA and not an unused plated through hole solder point?  I have never thought about that. It seems that many boards have unused holes for parts that later were not needed.  Are there any tip offs that a point on the board is a VIA?

Dave K8WPE

On Jul 21, 2018, at 10:26 AM, Jim Sheldon <w0eb@...> wrote:

As promised, pictures of the offending VIA.

W0EB<Img_5100.jpg><Img_5101.jpg>

David Wilcox
 

Get real?  I am 73 years old and have been building kits since the Heathkit days and every day I learn something new makes it a good day to be alive.  I am of the old school persuasion that there are no dumb questions. Most of the kits I have built lately have had unused holes but the kits worked and I never even thought about it.  Once I was sent a prototype board by mistake by a very reputable kit seller and it took a month of Sundays and the help of the board designer to figure that out and fix the board for me.  Another good day. St. Peter don't you call me as I have 10 kits that aren't finished yet.

Dave K8WPE

On Jul 21, 2018, at 6:25 PM, Jim Sheldon <w0eb@...> wrote:

Parts on factory PC boards are 99 percent of the time outlined by silkscreened legends.  VIA's are unmarked.  Hole far too small for a component lead (30 gauge wire barely fit the hole), what would you think?  Get real.

On Jul 21, 2018, at 5:12 PM, David Wilcox via Groups.Io <Djwilcox01@...> wrote:

So you had to figure out that it was a VIA and not an unused plated through hole solder point?  I have never thought about that. It seems that many boards have unused holes for parts that later were not needed.  Are there any tip offs that a point on the board is a VIA?

Dave K8WPE

On Jul 21, 2018, at 10:26 AM, Jim Sheldon <w0eb@...> wrote:

As promised, pictures of the offending VIA.

W0EB<Img_5100.jpg><Img_5101.jpg>

Kevin Timm
 

Jim Sheldon : Get Real?  That was way out of line.

ajparent1/KB1GMX
 

Vias are used to get from one side to another or as a top ground to bottom.
They are just plated through and often solder masked.

Like leaded component holes they are just plated through and can be defective or 
marginal.  When a soldered lead is there it tends to mask that but a via being 
unfilled by solder can fail.  

Its east to test for such a failure, ohm from point A to point B and if the resistance
is not very small (miilliohms typically) it is suspicious. 

Generally the via is sized to have at least the current carrying ability as the trace
leading to it or more.  When the work I've done on the my board I've had traces
vaporize but the vias were intact.

I chalk this one up to one of those odd failures that can happen in large board runs.
Unless its a more than .5% event its likely anomaly.

Allison

Clark Martin
 

Vias are frequently smaller than most component holes.  I suspect the small size tends to prevent the plating solution from getting in thoroughly due to surface tension. 

Clark Martin
KK6ISP

On Jul 23, 2018, at 8:17 AM, ajparent1/KB1GMX <kb1gmx@...> wrote:

Vias are used to get from one side to another or as a top ground to bottom.
They are just plated through and often solder masked.

Like leaded component holes they are just plated through and can be defective or 
marginal.  When a soldered lead is there it tends to mask that but a via being 
unfilled by solder can fail.  

Its east to test for such a failure, ohm from point A to point B and if the resistance 
is not very small (miilliohms typically) it is suspicious. 

Generally the via is sized to have at least the current carrying ability as the trace 
leading to it or more.  When the work I've done on the my board I've had traces 
vaporize but the vias were intact.

I chalk this one up to one of those odd failures that can happen in large board runs.
Unless its a more than .5% event its likely anomaly.

Jerry Gaffke
 

Board fab shops give their design rules, perhaps something like:
   minimum 10 mil drill size
   minimum 5 mil annular ring of copper pad around each drill hole
   minimum 6 mil trace width, minimum 6 mil spacing 
That's the rules for OSH-Park, where a mil is 1/1000'th of an inch

The minimum bid shops might be a bit more sloppy than OSH-Park.
I've seen boards where the drill job wasn't centered on the pads, to where
the edge of the hole was well beyond the edge of the pad.
In that case, you won't get as good a connection between layers through that via,
though even so it's probably good enough for our purposes in most cases..
Or could be that the plating process was rushed, and they didn't get enough material in there.
Generally speaking, you're best off backing off 10 or 20% from what a fab shop claims
they can do on a good day.

The via that failed was being asked to take 2 or 3 amps into the IRF510 drains.
That's asking a lot for one via.
Might be better to make it a 40 mil hole, has greater circumference, greater current capacity.
Though as someone has already suggested, the wimpy via does serve as a sort-of poorly spec-d fuse.

Jerry, KE7ER





On Mon, Jul 23, 2018 at 02:00 PM, Clark Martin wrote:
Vias are frequently smaller than most component holes.  I suspect the small size tends to prevent the plating solution from getting in thoroughly due to surface tension.