Vacuum-dehydrating hygroscopic components...
- Thread starter Stickman393
- Start date
Stickman393
Well-known member
Stick man. Bang rocks.
Ahem...
So, I've spent a good amount of time dehydrating these things. Like, 12 hours. I'm not noticing any improvement in leakage...not even incrementally speaking.
BUT!! That got me thinking. Tin whiskers. Fuckin...high humidity is going to affect that. I wonder...
Well, let's take a look. Behold!
Poo.
Well I'll be damned.
I'm gonna be trying something else out here. SCIENCE!!!!!!!!!!!!!!*
*this is not science. Stickman disavows all talk that what he engages in is science. At best, it's hackery conducted by a mind that is far too easily distracted by the nearest shiny object.
darwin999
Well-known member
@Stickman393 -
So did you re-measure them after your vacuum treatment?
Did any change?
Experimental data always wins!
If nothing changed much, I suspect they are still hermetically sealed.
When something is chemi-sorbed you generally need high vacuum** + heat to get any out (for that exp(-ΔE/kT)*** factor), and typically you will have important temperature constraints that limit how much can be removed. E.g., too much heat for too long will definitely damage those Ge transistors. There may be solder on the leads inside the can, dopant atoms in the germanium can diffuse (move) and damage the device, the case itself may have temp limits, etc, etc. I doubt it's worth trying. But if you do I'd be really careful, look at some spec sheets for those or similar transistors and stay well below any listed max heat values, especially since you might warm it for an hour or so under vacuum. If you want to experiment, just try it w/ one transistor, starting with the worst leaker. Measure before and after.
Sorry, I'm just a nerd...
But I admire your experimental curiousity!
** @Stickman393 asked about the requisite vacuum, which is a very good question - and unfortunately hard to quantify. You try to think about what types of background gas atoms you're exposing it to, and what damage they might cause. The usual lab approach is to make the vacuum so damn good that even if every background gas atom stuck, the resulting impurity level would be too low to matter. That usually leads to ultrahigh vaccum. In my prior science days, we made thin films samples at ultra-high vacuum for physical study. Our system had a base pressure of <1x10^-10 torr, or roughly 10^-7 micron.
*** k = Boltzmann's constant, T= temperature (°Kelvin)
So did you re-measure them after your vacuum treatment?
Did any change?
Experimental data always wins!
If nothing changed much, I suspect they are still hermetically sealed.
When something is chemi-sorbed you generally need high vacuum** + heat to get any out (for that exp(-ΔE/kT)*** factor), and typically you will have important temperature constraints that limit how much can be removed. E.g., too much heat for too long will definitely damage those Ge transistors. There may be solder on the leads inside the can, dopant atoms in the germanium can diffuse (move) and damage the device, the case itself may have temp limits, etc, etc. I doubt it's worth trying. But if you do I'd be really careful, look at some spec sheets for those or similar transistors and stay well below any listed max heat values, especially since you might warm it for an hour or so under vacuum. If you want to experiment, just try it w/ one transistor, starting with the worst leaker. Measure before and after.
Sorry, I'm just a nerd...
But I admire your experimental curiousity!
** @Stickman393 asked about the requisite vacuum, which is a very good question - and unfortunately hard to quantify. You try to think about what types of background gas atoms you're exposing it to, and what damage they might cause. The usual lab approach is to make the vacuum so damn good that even if every background gas atom stuck, the resulting impurity level would be too low to matter. That usually leads to ultrahigh vaccum. In my prior science days, we made thin films samples at ultra-high vacuum for physical study. Our system had a base pressure of <1x10^-10 torr, or roughly 10^-7 micron.
*** k = Boltzmann's constant, T= temperature (°Kelvin)
iamjackslackof
Well-known member
Is the TC-1 reliable for leakage measurements? I feel like I remember hearing it was not, but I don't know for sure.
Stickman393
Well-known member
Similar thought. Discharge a cap between the case & EBC bundled together.
right now I'm maxed out at like 36 volts unless I do some shady bullshit with a bridge rectifier I have laying around and mains. Nah. Don't have the work holding setup for that. No thank you. Not gonna die today.
36v and a 160v 100uF cap got it there. Kinda.
Huh. You know...I actually have a megaohmmeter in my garage somewhere. Up to 1000vdc. That'll arc real good. Let's make some fireworks.
owlexifry
Well-known member
mine won’t even recognize germanium transistors as BJTs. just comes up with diodes.
fucken waste of $20
Stickman393
Well-known member
Absolutely. But only up to like, 100uA. Its remarkably close to measured values using the RG keen method below 100uA. Anything more and it all goes out the window. These have a min hfe of 60, and they all measure under 20 or over 250, depending when I measure them.
Its not reliable if you're looking for definitive numbers. But its a quick and easy way to identify if a device measures way out of spec.
Shoot, I have two. I love em, cause I can just pop a transistor in and know the pinout. No need to look up the datasheet. 5 seconds. Bingo. Good for quick capacitor tests, and for any diodes which don't make their polarity immediately clear in their casing.
They're not ideal for testing germaniums. That's for sure.
tegendemuur
Well-known member
If anything, Germaniums make for cool little retro futuristic stickmen.
owlexifry
Well-known member
that’s pretty much all i’ll be able to with it.
but at least it does silicon BJTs alright.
i wouldn’t be recommending them to people though.
Now we're cooking!
They're a tool. Like any tool, if you use it without the bounds of its limitations, it's fine.
Now, back to abusing that megaohmmeter...
owlexifry
Well-known member
well, to be fair, it does have a leg up over the multimeter for measuring pF values, and a lot of scenarios where using leads isn’t convenient.
Stickman393
Well-known member
tegendemuur
Well-known member
Stickman393
Well-known member
Save a low E string.
Ride a stickman.
Ride a stickman.
Stickman393
Well-known member
A'ite, did a little more testing. Keep in mind I'm working with potentially deadly voltages here kiddos. Do not try this at home!!
Anywho: let's melt some whiskers!!!
Upshot? I melted the whiskers. But our leakage values are still crazy high. I'm curious if, perhaps, maybe I build a burn-in circuit. Just to get some current flowing through this thing. See what happens.
My curiosity here is likely exceeding my grasp. I want to understand to a larger extent if there are ways to bring a shitty Ge transistor back from the dead. If there are processes that occur over time that are reversible.
For instance: right now I'm sitting with my positive lead on my collector and my negative lead on my emitter, and I'm measuring resistance. I started around 90 ohms...but as I've sit, that value has increased. Now at about 320. The smallest bit of heat from my fingers causes the resistance to drop again. Reverse biasing shows a similar result
Hmm. I'm tempted to see if maybe a little current flowing through this guy might do more to bring down my leakage values. Who knows.
Anywho: let's melt some whiskers!!!
Upshot? I melted the whiskers. But our leakage values are still crazy high. I'm curious if, perhaps, maybe I build a burn-in circuit. Just to get some current flowing through this thing. See what happens.
My curiosity here is likely exceeding my grasp. I want to understand to a larger extent if there are ways to bring a shitty Ge transistor back from the dead. If there are processes that occur over time that are reversible.
For instance: right now I'm sitting with my positive lead on my collector and my negative lead on my emitter, and I'm measuring resistance. I started around 90 ohms...but as I've sit, that value has increased. Now at about 320. The smallest bit of heat from my fingers causes the resistance to drop again. Reverse biasing shows a similar result
Hmm. I'm tempted to see if maybe a little current flowing through this guy might do more to bring down my leakage values. Who knows.
Stickman393
Well-known member
Whelp, I've been having fun.
Had a couple of sparks, just started grabbing at random values from the datasheet in terms of max Vce, max Ic, yada blah...
Figures that my first real attempt to sit down with a breadboard and do some work would be torture testing transistors that were likely to be otherwise useless.
So...yeah, wattage? What the hell is that? I can drive 9V across Vce easily if I'm not connecting the base! Or so I thought. Holy hell that thing got hot. But, it survived.
So I got smart, did the math, and used appropriately sized resistors to choke down the current, and monitored it while driving the base bit by bit. I left it to cool down.
This morning I took a look:
Huh. Well I'll be damned.
Granted, ambient was probably a good 10-15 degrees colder this morning. When I check again today I'll see it it's shifted. Maybe torture test it a little more.
Preliminary evidence supports my hypothesis that part of the leakage observed in this old ge transistor could be related to whiskers. Sorta. Making a lot of inferences there without really having a firm grasp on the physics involved. Stick man. Bang rocks. Spark spark ooohhh. Hell yeah.
Had a couple of sparks, just started grabbing at random values from the datasheet in terms of max Vce, max Ic, yada blah...
Figures that my first real attempt to sit down with a breadboard and do some work would be torture testing transistors that were likely to be otherwise useless.
So...yeah, wattage? What the hell is that? I can drive 9V across Vce easily if I'm not connecting the base! Or so I thought. Holy hell that thing got hot. But, it survived.
So I got smart, did the math, and used appropriately sized resistors to choke down the current, and monitored it while driving the base bit by bit. I left it to cool down.
This morning I took a look:
Huh. Well I'll be damned.
Granted, ambient was probably a good 10-15 degrees colder this morning. When I check again today I'll see it it's shifted. Maybe torture test it a little more.
Preliminary evidence supports my hypothesis that part of the leakage observed in this old ge transistor could be related to whiskers. Sorta. Making a lot of inferences there without really having a firm grasp on the physics involved. Stick man. Bang rocks. Spark spark ooohhh. Hell yeah.
Last edited:
Stickman393
Well-known member
Nah dude. No dice. This was the component earlier:
This thing would make a fantastic Q1 in a fuzz face if I could keep it below 50 degrees 100% of the time. At that point it's like, HFE 75, leakage 100uA.
But the temperature coefficient on this thing is super touchy. Like, huge swings just by touching the things.
Hmm. I must read. For my next test: asking it nicely to stop leaking!
This thing would make a fantastic Q1 in a fuzz face if I could keep it below 50 degrees 100% of the time. At that point it's like, HFE 75, leakage 100uA.
But the temperature coefficient on this thing is super touchy. Like, huge swings just by touching the things.
Hmm. I must read. For my next test: asking it nicely to stop leaking!