Dogsled, if permander boards you can check the group buy thread from a few years ago. I have the pdf posted there if needed. I, too, am way behind on my CSX1 build - all parts still in box!
BK
BK
Hi All. Any chance getting T-shaped brackets now? Pre-drilling is not obligatory, just a chunk of black anodized aluminium.
Wow! Moscow might be a bit of a challenge but it's easily found on this side of the pond.
6063-T52 Aluminum T-Bar - Online Metal Store
Regards,
Dan
Here you go.
Tee Aluminum Extrusions from SAF Metals in GA & CA SAF – Southern Aluminum Finishing Co, Inc.
Regards,
Dan
Aluminium Tee
I don't see why you need the Tee bar to be black when it is inside the case. Black radiates heat better than clear aluminum, but I am pretty sure clear aluminum will conduct about the same as black.
We are trying to dissipate heat outside of the case by transferring through conduction from the Tee bar to the heatsinks (outside case.)
All the heat that is dissipated on the inside will raise the inside temperature even more.
I made my own Tee bars 2x2x1/4" and tried to blacken them with a chemical wash. It made them gray in color and not all that even.
Then as I started thinking about it, I came to the above conclusion.
So I'll bet someone on this forum knows a lot about temperature radiation (electromagnetic) and conduction as it pertains to the color of aluminum and if my assumption is correct to get the heat outside the case..
So chime in.
Rush
I don't see why you need the Tee bar to be black when it is inside the case. Black radiates heat better than clear aluminum, but I am pretty sure clear aluminum will conduct about the same as black.
We are trying to dissipate heat outside of the case by transferring through conduction from the Tee bar to the heatsinks (outside case.)
All the heat that is dissipated on the inside will raise the inside temperature even more.
I made my own Tee bars 2x2x1/4" and tried to blacken them with a chemical wash. It made them gray in color and not all that even.
Then as I started thinking about it, I came to the above conclusion.
So I'll bet someone on this forum knows a lot about temperature radiation (electromagnetic) and conduction as it pertains to the color of aluminum and if my assumption is correct to get the heat outside the case..
So chime in.
Rush
well, yes, you right. That's the simplest way and most effective. I just hoped there's a stock of ready bars from previous order.
you want all the heat out of the case ?
make it water cooled , that's exactly what i will do when the new VFET kit will be available at the store 😉
.
how high you are going to bias vfets to implement water cooling? Bet there's no reason for it. I just put there thermal breaker at 75C which breaks the mains power if vfets got overheated.
the reason to make it water cooled , is the pleasure i will take in doing it , and i am pretty sure it won't harm 😉
regarding the bias , i have no idea , i will probably ask here for some good advice 😀
.
Okay guys, I have a little problem. At least, I hope it is just a little problem.
I completed my VFET kit this past weekend, and I been checking biases and output offset the past couple nights after a couple hours of warm up each time. Last night, my test lead for T1 accidentally touched the chassis ground. The chassis is grounded to mains safety ground. It smoked R32, but the inlet fuses did not break.
Do I need to check any other parts or do I only replace R32?
Cheers
Curtis
I completed my VFET kit this past weekend, and I been checking biases and output offset the past couple nights after a couple hours of warm up each time. Last night, my test lead for T1 accidentally touched the chassis ground. The chassis is grounded to mains safety ground. It smoked R32, but the inlet fuses did not break.
Do I need to check any other parts or do I only replace R32?
Cheers
Curtis
Pass DIY Addict
Joined 2000
Paid Member
Yikes- I've done something like this myself in the past. In this case, my first reaction would be that you've completely shorted R32 and it's not a surprise that it burned. But, you might just be lucky here. I see two paths forward, both starting with replacing R32.
1) replace R32, then compare other resistances across your working and non working boards after replacing R32. Look for big differences in all things that come close to touching R32. If you pass this test, then I'd use a variac to ramp up the PSU and see what happens with your test voltages as you ramp up. This could lead to other problems if more than R32 was damaged...
2) replace R32 and remove the vFets. Go back to testing the board as Nelson describes in the article. When you are satisfied, replace the vFets and then ramp up the PSU slowly and measure your test points as you do so. This is the more conservative route to follow and probably safer given the rarity of the vFets.
1) replace R32, then compare other resistances across your working and non working boards after replacing R32. Look for big differences in all things that come close to touching R32. If you pass this test, then I'd use a variac to ramp up the PSU and see what happens with your test voltages as you ramp up. This could lead to other problems if more than R32 was damaged...
2) replace R32 and remove the vFets. Go back to testing the board as Nelson describes in the article. When you are satisfied, replace the vFets and then ramp up the PSU slowly and measure your test points as you do so. This is the more conservative route to follow and probably safer given the rarity of the vFets.
Connecting T1 to ground pulls every node connected thereto to ground.
My guess would be that R32 drew a whole lot of current (~28/0.1 A) and died.
The other subcircuits connected to R32, such as negative bias supplies for Q8 (Q14) and Q9 (Q12) lacked their supply during that time, which may have caused trouble. Q9 may simply have turned off, no biggie. But without the negative bias voltage at the gate, VFET Q9 may have been turning on more and more, since it's a normally open part (you need negative bias to shut it off). With a little luck C5 (and C4) held enough charge to keep the bias voltage alive.
The way I see it, it's either Q9 or nothing that's damaged.
Edit: Careful with the variac, don't turn it up too slow or the bias voltage may lag behind. There is a remark in this regard in the article: "I suggest that you raise the AC line to voltage over a period of less than 5 seconds."
My guess would be that R32 drew a whole lot of current (~28/0.1 A) and died.
The other subcircuits connected to R32, such as negative bias supplies for Q8 (Q14) and Q9 (Q12) lacked their supply during that time, which may have caused trouble. Q9 may simply have turned off, no biggie. But without the negative bias voltage at the gate, VFET Q9 may have been turning on more and more, since it's a normally open part (you need negative bias to shut it off). With a little luck C5 (and C4) held enough charge to keep the bias voltage alive.
The way I see it, it's either Q9 or nothing that's damaged.
Edit: Careful with the variac, don't turn it up too slow or the bias voltage may lag behind. There is a remark in this regard in the article: "I suggest that you raise the AC line to voltage over a period of less than 5 seconds."
Last edited:
To learn a bit more what VFET / SITs are I found this:
AMPLIMOS V-Fet & SIT Amplifiers, One Stage Amplifiers
There are links to this forum which still works…...but long time since the threads were active. At least I am a bit "wiser" on this topic.
AMPLIMOS V-Fet & SIT Amplifiers, One Stage Amplifiers
There are links to this forum which still works…...but long time since the threads were active. At least I am a bit "wiser" on this topic.
Thanks for the quick replies everyone.
I removed R32, and it tested in the Mohm range on a DMM. It's totally dead.
I added a wire to replace R32. I am using a variac (< 5 sec ramp from 0 to 120V) and 1.25A fast blow fuses. The regulator (Q14) nearest to R32 is still giving me 24V at T4. The voltages at R5 & R6 look good (1.5V). T18 to GND is slightly high (400mV), but I would expect it to get to 100mV after warming up.
The voltages at T16 and T17 are the same as the good channel. I think this is expected since the VFETs that I received had the same Vgs values for each channel (ie. 93 for both 2SK82s and 103 for both 2SJ28s). DC offset at the output is 20mV. I just can't measure the exact bias current until I get an R32 replacement.
It looks like I am okay.
I removed R32, and it tested in the Mohm range on a DMM. It's totally dead.
I added a wire to replace R32. I am using a variac (< 5 sec ramp from 0 to 120V) and 1.25A fast blow fuses. The regulator (Q14) nearest to R32 is still giving me 24V at T4. The voltages at R5 & R6 look good (1.5V). T18 to GND is slightly high (400mV), but I would expect it to get to 100mV after warming up.
The voltages at T16 and T17 are the same as the good channel. I think this is expected since the VFETs that I received had the same Vgs values for each channel (ie. 93 for both 2SK82s and 103 for both 2SJ28s). DC offset at the output is 20mV. I just can't measure the exact bias current until I get an R32 replacement.
It looks like I am okay.