The AM Forum
March 28, 2024, 12:56:11 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: ANODIZED Aluminum  (Read 6740 times)
0 Members and 1 Guest are viewing this topic.
KK4RF
Member

Offline Offline

Posts: 158


« on: July 23, 2021, 07:58:14 AM »

Gentlemen:
     I am rebuilding an old Palomar DC-30 30 amp 13.8 volt power supply. I took it all apart. I plan to add a crowbar circuit. I pulled the pass transistors off and tested them all. 2N6258. All good.  I was surprised that there was no mica insulator between the TO-3 transistor cases and the black aluminum heat sink. On the side of the heat sink inside the cabinet, there were little insulator washers that insulate the mounting screws. Hmmm...  I suspect these heatsinks are anodized. I googled the topic and note that these anodized surfaces are NOT SUPPOSED to conduct electricity, but I wondered what the ham community thinks about this. I suspect some of you guys have thought about this and wanted to get some opinions. I plan to order some mica TO-3 insulators, but wondered if they are necessary. Also, I note there are newer TO-3 insulators made of a silicone rubber type stuff. Any opinions on these also would be appreciated. Gentlemen, thanks for your thoughts and have a great weekend.---Marty, KK4RF
Logged
KD1SH
Member

Offline Offline

Posts: 737



« Reply #1 on: July 23, 2021, 09:15:14 AM »

A thick anodize can be a surprisingly good insulator, even the stuff that looks like bare aluminum, but I'm surprised that a manufacturer would rely on it alone for insulation. Unless they intended for those pass transistor collectors to be tied together. If they intended them to be isolated, they'd also have nylon bushings and washers on the mounting screws. What does the schematic show?
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
KK4RF
Member

Offline Offline

Posts: 158


« Reply #2 on: July 23, 2021, 09:27:38 AM »

I don't have a manual on this old power supply, but again no insulator between the case of the pass transistors and the heat sink. There are nylon bushings on the flip side of the heat sink around the mounting screws. Thanks for your ideas---Marty, KK4RF
Logged
KD1SH
Member

Offline Offline

Posts: 737



« Reply #3 on: July 23, 2021, 10:26:58 AM »

It does sound like the manufacturer intended to isolate the transistor cases. If it was my design, I wouldn't have relied on the anodize alone to provide insulation; that concept would never fly in a shop that made medical or aerospace stuff, since most of that stuff is subjected to insulation resistance and dielectric breakdown (hipot) testing.

I don't have a manual on this old power supply, but again no insulator between the case of the pass transistors and the heat sink. There are nylon bushings on the flip side of the heat sink around the mounting screws. Thanks for your ideas---Marty, KK4RF
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
kc9pcp
Member

Offline Offline

Posts: 11


« Reply #4 on: July 23, 2021, 12:33:16 PM »

Makes sense that the screws would have nylon bushings since the may "bite" trough the insulating anodized layer.  I'm assuming the transistors are all connected by wires to the mounting screws, so the heat sink is not used for connecting them.  Also wonder if the heat sink is grounded  if not than that would be safer when using only the anodation as an insulator.  Either way I would add the mica insulators for peace of mind and be done with it.
Logged
KK4RF
Member

Offline Offline

Posts: 158


« Reply #5 on: July 23, 2021, 01:18:49 PM »

Yes, all 4 pass transistors are connected in parallel with wires which are fastened to the mounting screws of the transistors. I'm ordering some mica insulators today. I found it interesting, however, that Palomar put these transistors in without insulators and wondered if this is standard in the industry. My Astron 35M uses mica insulators for the TO-3 pass transistors. Thanks, guys, for the help. ---Marty, KK4RF
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #6 on: July 23, 2021, 07:20:32 PM »

It's a Tripp Lite design I believe.

In those the negative rail is regulated.

I rebuilt mine using positive rail regulation and newer higher powered xisters.

If it's a Tripp Liter design you may find that the balance resisters, one is used as the current shunt for shutdown.

I may be cometely off base, but that's how my Tripp Lite 60A supply is wired.

--Shane
KD6VXI
Logged
KK4RF
Member

Offline Offline

Posts: 158


« Reply #7 on: July 23, 2021, 07:52:31 PM »

Shane,
     Thanx for the info. I'm not actually rebuilding this supply using the same schematic as the original, Palomar DC-30.  I have an  Astron 35M regulator board which I intend to use. I'm going to wire this all up like as if it is an Astron supply. I just need to figure out how to power the regulator board (in the Astron, the regulator board is powered by a separate set of taps on the big power transformer.) I hope I can get it all to work. This should be a fun project. Will check out the Tripp Lite circuit. Thanks again---Marty,K4RF
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #8 on: July 24, 2021, 10:27:12 AM »

Sounds like the transistors are electrically connected to the heatsinks - IF the heatsinks are isolated from the
chassis with shoulder washers.

I have never had any luck with the silicone insulators or the kapton.
So I stick with mica.

Nominally anodize, Aluminum Oxide is an insulator.
But only hard anodize is used as a reliable electrical insulator.
That's because it makes for a stronger thicker layer than standard anodize.
So, do not expect standard anodize to be a reliable insulator.

So, IF the heatsinks are electrically isolated, do not add insulators to the transistors, as
that reduces the thermal conductivity, which might be necessary to stay within SOA wrt
thermal control.

An ohmeter ought to tell the story.

                 _-_-bear
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
KK4RF
Member

Offline Offline

Posts: 158


« Reply #9 on: July 24, 2021, 10:38:43 AM »

Thanks,  Bear,
     I've ordered some mica insulators already. Thanks for the other ideas and insights. I'll  update everything once I get this thing going. 73s---Marty, KK4RF
Logged
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #10 on: July 24, 2021, 12:02:24 PM »

Marty,

  The 2N6258 has a real good SOA (Safe Operating Area). The high SOA means lots of volts and Amps at the same time when used as a linear regulator. They are better than the more common 2N3055. Adding a Crowbar circuit might be a good idea to protect a precious load, but make sure the SCR circuit does not blow anything up. The 2n6258's, if one shorts, should have an emitter resistor sized to be a fuse when the Crowbar SCR turns on.

   Back in 1981, I designed a magnet power supply for an ion implanter machine. I needed 0-200 ampere linear regulator into two large magnet coils in series. Took a total of 40 volts at 200 amps, or 8KW! At the half way point, the heatsink had to dump several kilowatts, so it needed to be water cooled. I forget how many pass transistors I used, but 28 comes to mind. I used 2N6258's! At the time, those transistors were among the top candidates for a job as a high power linear regulator. I think I still have a box full of them...In my application, there was excess capacity, and a high need to keep on going if a transistor failed. Each transistor emitter was individually fused where the fuse was in series with the emitter resistor.

Jim
Wd5JKO
Logged
KK4RF
Member

Offline Offline

Posts: 158


« Reply #11 on: July 24, 2021, 03:56:13 PM »

Jim,
     Thanks for the interesting story. As for emitter resistors as a fuse, when I was working on an Astron 35M this past month, I did something, not sure what, but I ended up blowing one of the 4 pass transistors (2N3771) and to my amazement ALL 4 of the emitter resistors, each 0.05 ohms, 5 watt rating,   were blown. I've never seen that happen before. I had to replace all 4 of those and went ahead and put all new pass transistors ( 4 of them as well.)  Anyway, I got the power supply working again but it dawned on me that these resistors had acted like fuses.
     I checked  the data sheets on 2N6258 and 2N3771 and note that the former has 250 watt dissipation and the 2N3771 has 150 watt. I would presume the 2N6258 would be more rugged. For fun, I might go ahead and use them  in my rebuild  of the Palomar supply. Again, I plan to try to use the Astron 35M regulator board which has a crowbar circuit. I'm going o hook  up the pass transistors just as they are in the Astron 35M. Again, Astron uses emitter resistors of 0.05 ohms, 5 wat rating.  Fingers crossed here...-Marty, KK4RF
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #12 on: July 24, 2021, 09:45:12 PM »

Marty,

  The 2N6258 has a real good SOA (Safe Operating Area). The high SOA means lots of volts and Amps at the same time when used as a linear regulator. They are better than the more common 2N3055. Adding a Crowbar circuit might be a good idea to protect a precious load, but make sure the SCR circuit does not blow anything up. The 2n6258's, if one shorts, should have an emitter resistor sized to be a fuse when the Crowbar SCR turns on.

   Back in 1981, I designed a magnet power supply for an ion implanter machine. I needed 0-200 ampere linear regulator into two large magnet coils in series. Took a total of 40 volts at 200 amps, or 8KW! At the half way point, the heatsink had to dump several kilowatts, so it needed to be water cooled. I forget how many pass transistors I used, but 28 comes to mind. I used 2N6258's! At the time, those transistors were among the top candidates for a job as a high power linear regulator. I think I still have a box full of them...In my application, there was excess capacity, and a high need to keep on going if a transistor failed. Each transistor emitter was individually fused where the fuse was in series with the emitter resistor.

Jim
Wd5JKO

Reminds me of a project I've gone back and forth on out in the shop.

28 V @ 300A or 14V at 600A.

It's a BIG xformer.  Came out of a Dyno.  Has two 18V @ 300A secondaries and a 240 input.

--Shane
KD6VXI
Logged
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #13 on: July 25, 2021, 12:39:30 PM »


Reminds me of a project I've gone back and forth on out in the shop.

28 V @ 300A or 14V at 600A.

It's a BIG xformer.  Came out of a Dyno.  Has two 18V @ 300A secondaries and a 240 input.

--Shane
KD6VXI

Don't mean to hijack the thread, but...

The transformer for the big magnet supply had a special requirement not typically found.

The insulation between the primary and secondary needed to withstand 150KV continuous!!

This made the core a large 'C' core in two pieces, about 2 1/2 feet long and 2' high.

The prototyping was iterative since that much insulation causes a huge amount of leakage inductance primary to secondary.

This leakage inductance caused significant secondary voltage droop under load.

So there we were in this WW2 Quonset Hut building covered with asbestos insulation on the ceiling.

We were in a room that did not have 208vac outlet, but we did have two 120v outlets where each represented a separate phase of 208 vac 3 phase power.

There we were, core off, add or subtract the primary turns, core on, test and then repeat.

After a large number of tries, we were close. One more change, and test...We forgot to clamp the two core halves!! Boom!!

There was a huge 60hz groan, and the lights dimmed. It turned out that the 120v circuits were not breaker'd, and went straight to the big 480v transformer feeding the building.

The whole building went dark, and maybe more. Had to fess up to the boss.

Ever do something like that to three phase 400hz circuit? Makes a really cool noise.

Back to thread.

Jim
Wd5JKO



 

Logged
KK4RF
Member

Offline Offline

Posts: 158


« Reply #14 on: July 25, 2021, 05:31:02 PM »

Jim, Shane,
     Thanks for the interesting and amusing stories. Sounds like these were some huge transformers. Also, sorry to hear about the asbestos exposure. I suspect many of us had asbestos exposure in the past.
     Waiting for parts from Mouser to finish up this power supply. Will keep you guys posted.  I'm going to use the 2N6258 transistors. 73s all---Marty, KK4RF
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #15 on: July 26, 2021, 10:11:36 AM »

Thanks,  Bear,
     I've ordered some mica insulators already. Thanks for the other ideas and insights. I'll  update everything once I get this thing going. 73s---Marty, KK4RF

Still somewhat unclear - are the heatsinks electrically isolated from the chassis or not?

IF they are isolated, no point in using mica insulators at all.
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #16 on: July 26, 2021, 10:50:43 AM »

Fantastic interesting accounts!

Heatsink can be isolated, but still, some compound or filler is needed, right?

According to these articles, anodization can have better thermal conductivity than a separate insulator, but not always.

"Measured thermal conductivities fell in the range of 0.53–1.62 W m−1 K−1."
https://www.sciencedirect.com/science/article/abs/pii/S0254058413004409  (login as academic or purchase)

Advanced Thermal Solutions public article:
https://www.qats.com/cms/2010/11/09/how-heat-sink-anondization-improves-thermal-performance-part-1-of-2/
https://www.qats.com/cms/2010/11/10/how-heat-sink-anodization-improves-thermal-performance-part-2-of-2/



Silicone-fiberglass composite is my favorite. No grease and inexpensive.

NTE Termo-Pads
thermal conductivity 1.2 W/m-deg k
breakdown >4KV
0.009" thick
NTE - nr. TP0001 - $2.33 each
https://www.nteinc.com/specs/thermopad/thermopads.pdf

AAVID Thermasil III
thermal conductivity 0.92 W/m-deg c
breakdown >4KV
0.006" thick
$1.55 each
https://www.mouser.com/datasheet/2/2/Aavid-Board-Level-Heatsinks-Catalog-2018-1507171.pdf page 103






Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
KK4RF
Member

Offline Offline

Posts: 158


« Reply #17 on: July 26, 2021, 11:39:28 AM »

Patrick,
     Thanks for the links and info on thermal conductivity. This is an interesting topic.

Bear,
     The whole power supply is apart. The supply has actually TWO heatsinks that form the 2  sides of the the case itself. Each heatsink holds 2 of the 4  pass transistors. There is a bottom plate that that forms the "chassis" to which the power transformer, the bridge rectifier... is bolted to. The bottom plate and the top plate  actually just slide into grooves in the heat sinks and the whole thing is held together with a number of sheet metal type screws that attach the front  and rear panels to the heatsinks. There are no insulators per se. Again the pass transistors had no TO-3 insulators between the case of the transistor and the heatsink.
     At some point, I will send pictures of the power supply. ---Marty, KK4RF
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #18 on: August 02, 2021, 02:00:03 PM »

Marty, my question was simple, and I can't tell from your description exactly. Measured with an ohmeter the heatsinks are on ground?

And, unless the collectors of those transistors are intended to be grounded, it will NOT WORK without an insulator.
The screws would have to have shoulder washers or go to transistor sockets that have shoulders to provide a spacer
to the heatsink...

Patrick, if the surfaces are flat, no "goo" is actually needed for a direct connection. I'd put some in anyhow because it will extrude
out between two dead flat surfaces anyhow.

                    _-_-bear
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
KK4RF
Member

Offline Offline

Posts: 158


« Reply #19 on: August 02, 2021, 07:48:08 PM »

Gentlemen,
     First, an embarrassing admission. Over the weekend, I began cleaning off all the old heat sink grease on the 2N6258 transistors, and found the mica insulator (placed over 40 years ago) stuck to the bottom surface of the transistor. I had originally only removed one of the pass transistors. I just did not see it in all the old gunk. So, admittedly, my original question was based on erroneous information. I apologize to all who likely developed a migraine trying to figure this all out.
     Again, the heat sinks are not  presently mounted to the chassis, but this is a moot point at this time, as the Palomar engineers clearly intended for the pass transistors to be insulated from the heat sink.
     Again, my apologies for the wrong initial information. I wanted to set the record straight on this one. 73s---Marty, KK4RF
Logged
KK4RF
Member

Offline Offline

Posts: 158


« Reply #20 on: September 15, 2021, 04:45:18 PM »

Gentlemen,
     Wanted to give a follow up on my Palomar DC-30 power supply re-build. It is basically all done and it works well. Testing it with a bank of power resistors, it puts out 17 amps easily with good regulation. I don't have more resistors to try to see if it will put out 30 amps (as its name would suggest), but I think it is working well at this point.
     A few points that I would like to make:
     1. From reading I have done, when aluminum is ANODIZED, it does produce a surface that does NOT conduct electricity. I could not find anywhere where a person recommended NOT insulating pass transistors from the heat sink, however. Interestingly, when aluminum is anodized, the process greatly increases the surface area of the heatsink due to the increase in the porosity of the aluminum, hence the ability  to release heat is much increased.
     2. I rebuilt the Palomar using a regulator board from an old Astron 35M. It works great for this (so far.) The Astron regulator board has a crowbar circuit for over-voltage protection and the old Palomar did not.
     3. The heat sinks are not electrically continuous with the chassis  as measured with my VOM/DVM.
     4. Lastly, the final word on the original pass transistors, I actually found that 2 of them had the mica insulators and 2 did not. I suspect somebody else had worked on this power supply years ago as I  found different hardware on these transistors. I suspect  that person failed to put the mica insulators back in.  The power supply was given to me by a ham, now a SK, and it was non-functioning when I got it. I had no information available as to what had been done to it over the years.
     5. I installed a 15 volt DC voltmeter on the front panel and have it wired  to measure voltage and also current in amps, using a switch and resistors. It was an old Radio Shack meter, a hamfest pickup for 3 dollars.
     6. I have not yet tried it with a transceiver hooked to a real antenna, so I am not sure if I will have any problems with RF getting into the supply and aggravating the LM723 regulator chip. I just got my 80 meter dipole back up so I intend to do this and will hope for the best. Keeping my fingers crossed.
     Thanks again to all for the comments, suggestions and questions. I am a hobbyist only, and not an engineer, but this was a fun project, and I learned quite a bit. Sorry for the misinformation at the beginning of this thread.  73s to all---Marty, KK4RF
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.065 seconds with 19 queries.