|
sBITX V2 #12 arrived here around two weeks ago and has been waiting for me to make time to remove the front panel and fit the RPi4. That finally happened yesterday, prior to my taking both the DE and the V2 to my club's two-weekly technical night meeting. Fitting the RPi went easily, so I moved on to making sure the phones, key and microphone plugs? were firmly connected, then gave the entire board an admiring closer examination.
I quickly noticed that the clamp holding the P.A. trannies to the heat sink did not look to be sitting flat. I recalled somebody reported finding the mounting screws on their V2 required tightening, so reached for an appropriate screw driver to check mine. I was surprised to find that the entire clamp pivoted sideways as the screwdriver touched it. What's more, one the screw would not tighten. It quickly became apparent that the problem lay in the fixed nut, "nutsert" or whatever the trade term may be. It had obviously been pressed into a blind hole in the heat sink but for some reason, had? now climbed out of that hole, presumably at some time during transit from Hyderabad. Whatever, it is no longer a press fit. Please see photos.
Given that the clamp serves no electrical purpose, merely applying physical pressure to hold the transistors against the heat sink, I am thinking that I should be able to glue the captive nut back into the hole using something like JBWeld or other steel treated epoxy. Ordinary epoxy might also do the job, although I have used JBWeld in other heat prone areas with success. I did think of using one of the gap filling thread-lock products from Loctite and other suppliers but not sure that heat might be an issue.
Happy to hear other thoughts.
Mine might be a one-off failure but I suggest that consideration be given to tapping the blind hole and screwing the clamp down directly to the heatsink, adding a spacer if necessary, Ashhar. The press fit approach is clearly not fully reliable.
Bill? ?VK7MX
|
Is it possible to just drill through and put a nut on the far side? Perhaps a locknut with the plastic inserts so ?it won’t work loose?
toggle quoted message
Show quoted text
On May 17, 2023, at 19:58, Bill Maxwell via groups.io <wrmaxwell@...> wrote:
?sBITX V2 #12 arrived here around two weeks ago and has been waiting for me to make time to remove the front panel and fit the RPi4. That finally happened yesterday, prior to my taking both the DE and the V2 to my club's two-weekly technical night meeting. Fitting the RPi went easily, so I moved on to making sure the phones, key and microphone plugs? were firmly connected, then gave the entire board an admiring closer examination. I quickly noticed that the clamp holding the P.A. trannies to the heat sink did not look to be sitting flat. I recalled somebody reported finding the mounting screws on their V2 required tightening, so reached for an appropriate screw driver to check mine. I was surprised to find that the entire clamp pivoted sideways as the screwdriver touched it. What's more, one the screw would not tighten. It quickly became apparent that the problem lay in the fixed nut, "nutsert" or whatever the trade term may be. It had obviously been pressed into a blind hole in the heat sink but for some reason, had? now climbed out of that hole, presumably at some time during transit from Hyderabad. Whatever, it is no longer a press fit. Please see photos. Given that the clamp serves no electrical purpose, merely applying physical pressure to hold the transistors against the heat sink, I am thinking that I should be able to glue the captive nut back into the hole using something like JBWeld or other steel treated epoxy. Ordinary epoxy might also do the job, although I have used JBWeld in other heat prone areas with success. I did think of using one of the gap filling thread-lock products from Loctite and other suppliers but not sure that heat might be an issue. Happy to hear other thoughts. Mine might be a one-off failure but I suggest that consideration be given to tapping the blind hole and screwing the clamp down directly to the heatsink, adding a spacer if necessary, Ashhar. The press fit approach is clearly not fully reliable. Bill? ?VK7MX
  
|
Guess that is possible Gordon but I would first have to mill away some of the fin area on the far side to enable seating of the nut. The fins are closely spaced, as you see on your DE, which has the same heat sink. I assume that is why HF Signals went for the blind hole/nut set approach as it enables single-sided mounting.
|
I have the same issue. Inserts are not fitting tightly. Epoxy maybe? Drill and tap??
I still don’t have a working radio anyway. Hopefully it’s resolved soon.?
--
Joe McConnaughy
AG7XH
|
my solution can be read here:
-- Gyula HA3HZ
|
Rats. Then I would drill and tap. I’ve used those inserts before in my life and I don’t really like them.?
It should be aluminum. The problem of course is the threads are not going to be very strong. That’s why they use the insert I’m sure?
toggle quoted message
Show quoted text
On May 18, 2023, at 01:51, HA3HZ <gyula@...> wrote:
?my solution can be read here:
-- Gyula HA3HZ
|
Gyula did a very nice job! ??
toggle quoted message
Show quoted text
On May 18, 2023, at 05:43, Gordon Gibby via groups.io <docvacuumtubes@...> wrote:
? Rats. Then I would drill and tap. I’ve used those inserts before in my life and I don’t really like them.?
It should be aluminum. The problem of course is the threads are not going to be very strong. That’s why they use the insert I’m sure?
On May 18, 2023, at 01:51, HA3HZ <gyula@...> wrote:
?my solution can be read here:
-- Gyula HA3HZ
|
|
|
Jb weld will work just make sure not to get it in threads. Clean with alcohol before applying. Put screw in threaded insert before pressing in then take screw out after inserting and let dry. That should keep most of it out of the threads.?
These inserts are never a good option in my opinion.?
|
Probably the mechanical designer thought it would be sturdy. 2mm is just not enough for pressing.
A longer pressed part is worth more, with a longer pressing. By the way, the aluminum heatsink here is a very soft material.
Electrically, I can imagine that the FETs can't transfer the heat, which is why there will be a distorted radio frequency output over time.
--
Gyula HA3HZ
|
Unfortunately I just experienced what you all have been talking about. One of my inserts came loose (V2 #89). My quick solution was to put back the stud but secure it from behind with a M3 30 mm screw and a nut at the outside back of the heat sink. It's not elegant and it took some difficult fin grinding with a Dremel, but it works. If the other insert comes out I'll try drilling and tapping for a 3 mm screw. My tapping set is only for SAE--time to get another metric set.
So far the heat sink seems to work. Some of my PA MOSFET failures may have been from overheating but every time I've replaced them the clamp was on tight--it only came loose when tightening the last replacements. I now also add heat conducting grease on both sides of the mica insulator--that may have been the weak spot.
By the way a recent failure occurred just during the PA_BIAS potentiometer adjustment and I know the bias voltage was below 4 volts--and one of the finals shorted after only about 1 second so it couldn't have been from thermal stress. These IRFZ24Ns from Mouser. Perhaps it is a bad batch.
Jack, N6LN
|
My solution was use a .3124 inch end mill bit in a drill press and do a
plunge cut to the base aluminum though the fins.? That clears the fins
and provides a flat base for the nut.? With that a screw and lock washer
and nut.
The mechanical designer should have used a T-nut from the finned side
as they can't pull? out.? A larger diameter screw with less fine thread directly
into threaded aluminum is fairly common way to do it.
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
|
I agree Allison, end-milling a suitable flat base for a nut would be the better mechanical solution but at the time, I didn't want to take the heatsink off the case to work on it, being anxious to power up the V2 to get familiar with it. So far my less elegant JB-Weld fix is holding up but I will be going down the mechanical route if it fails in future. Of course, not everybody has access to an endmill though.
Bill, VK7MX
|
I don't know the exact cost, but the end mill toolbit shouldn't
be exorbitantly expensive, & it's being used in an ordinary
drill press, not a milling machine.? I used to work in a machine
shop, so I have an idea of what's going on.
Gerry Sherman
Sent by the Windows 11 Thunderbird
toggle quoted message
Show quoted text
On 2023-06-19 19:32, Bill Maxwell via groups.io wrote:
I agree Allison, end-milling a suitable flat base for a nut would
be the better mechanical solution but at the time, I didn't want
to take the heatsink off the case to work on it, being anxious to
power up the V2 to get familiar with it. So far my less elegant
JB-Weld fix is holding up but I will be going down the mechanical
route if it fails in future. Of course, not everybody has access
to an endmill though.
Bill, VK7MX
|
A good carbide end mill in that size is about 15-20$, tool steel a bit cheaper.
I am talking about just clearing the fins so we can get a nut or a t-nut/PEM-nut
can be pressed in from the back?if the through hole is correct diameter.?
In mine the surface of the component side was extremely rough, I had to resurface
the hard way on a surface table using 320, then 400, finish with 600 emery paper.
A knee mill would be easy but I haven't the room for one.
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
|
Bill, the endmill is a tool bit.? I have a very cheap $14 HSS straight/square end multiple bit set from Amazon?ASIN ? : ??B00ZZYJU2U
That includes the 5/16".? If you have a drill press, consider purchasing bits like this for doing plunge cuts on Aluminum, plastic.
I'm not a SBITX owner, but from seeing the heatsink problems: Another possible unvetted solution is to use a self tapping threaded insert.
If it is a M3, something like this should work?ASIN ? : ??B07X9JQ79Q
I don't know how much Al heat sink stock is remaining after the pressed-in insert is pulled out.
|
I thought so, as did Gary, so I replaced the short set screws with longer ones. See hereIn a message, Ashar wrote that in newer devices, the threaded pin is pressed in from the side of the rib, so when screwed in, it is on the opposite side of the rib. -- Gyula HA3HZ
|
Hi Gary
Yes, I'm familiar with endmills, having fairly much a complete set in both metric and imperial sizes, plus the same in slotdrills. Certainly worth having on hand, as you suggest.
Bill, VK7MX
|
Bill, I’ve never used an end mill but it looks as though a full sized drill press would be necessary. Is there a way to safely use an end mill without a drill press?
Jack,N6LN
|
YES, with a Milling machine.? NO, by other means.
IT would likely bounce around in a hand drill and cut the user.
You don't need an industrial drillpress just one that can chuck the tool and
that's usually 3/4ths or 1/2"? Use the slowest speed.
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
|
HA3HZ
John Terrell, N6LN