i'm on a roll... the kgsshv

[kevin gilmore]

Rule #1 Never mess with high voltage at the end of a day when you are tired.

Rule #2 Always remove the lead from the AMP terminal on the DVM as soon as you are done

measuring current.

Yep, i fried a meter once. It was more than 20 years ago.

Half a dozen DVM's around the house, including a couple of $19.95 cheapies makes

things a lot easier.

Edited January 8, 2011 by kevin gilmore

[Craig Sawyers]

Rule #1

Yup. Solid rules those

Half a dozen DVM's around the house

Yeah - I had a couple. Which weren't really enough for setting up a T2, so I went on an eBay raid. So now got 4 8060A's, an 8024B and an 87V. Put 1mm jacks on the test points on the tube side of the board and made up some 4mm to 1mm test leads. So I can set the batteries up by just taking the top plate off and plugging straight in. Put a ground wire on the board so another pair of meters go from ground to the stax jack. Switching those between DC and AC allows the noise voltage to be measured durning final battery tweaking.

[kevin gilmore]

i asked this to marc and craig in PM's, but i might as well ask here.

Anyone know the voltage ratings of a TO220F package?

Specifically if i put a metal screw in the hole, how many volts does it

take for the package to break down between the screw and the collector

tab inside.

Has to be well into the 1.5 to 2kv range right?

I've found some new polycarbonate (lexan actually) screws that seem to

be extremely strong, but still metal screws are better.

[digger945]

I remember seeing that the minimum air gap for 1500vdc is .008", somewhere.

This part has a 2500vrms insulation withstand voltage between all 3 pins and heatsink.

[Kerry]

The calculations I use is .1" for 500V + 0.00012" for each volt above that. So 1.5KV would be about .22". This is for uncoated conductors below 10,000 Ft. There is an FAQ on this website FAQ near the bottom. I can't vouch for the accuracy.

[Craig Sawyers]

I've found some new polycarbonate (lexan actually) screws that seem to

be extremely strong, but still metal screws are better.

I've never found a plastic screw (and I tried a few different types on my T2 build) that will take anything like the recommended torque for a TO220 (around 1.1Nm, 8-10 inlb). The best reference for semiconductor mounting I've come across is Onsemi's AN1040D http://www.onsemi.com/pub/Collateral/AN1040-D.PDF .

I ended up with a long Aavid bush that went most of the way through the 2mm alumina insulator, and then used steel fastenings.

[kevin gilmore]

I want to make the kgsshv absolutely trivial to build, and want to make sure that people know how to mount the semiconductors.

For all the TO220F transistors, a hex head #4-40, a #4 plastic washer, the transistor, a thermal pad, the heatsink, then preferably a belleville washer (which i can't find stock on) and a nut.

So for the 2sa1968 version, that leaves only the ixys parts and the low voltage regulators.

The small heatsinks are #6-32 tapped.

So you have to go with a plastic screw.

For the ixys parts, thats the plastic screw, the transistor, the aluminum oxide insulator, and the heatsink

For the low voltage regulators, all you need is a screw and a regular thermal pad.

The #6-32 lexan screws are pretty strong it seems.

unless someone knows where to get the screw tabs that are #6-32 on one end and #4-40 on the other end.

[livewire]

I'm probably not the best one to chime in on this due to the fact

that I have a few scorched parts on my psu board at the moment.

What I did with the small heatsinks was to use #4-40 stainless screws installed backwards.

They fit through the tapped #6-32 hole without threading and I nutted them from the frontside,

over the metal tab of the sand. To do this I had to grind the screw head flat on two opposing sides

in order for it to fit between the fins of the backside of the heatsink.

Since I didnt have the alumina oxide insulators handy for the high voltage devices,

I stood the small heatsinks up off of the pcb about 1/4 of an inch, by re-drilling the tab mounting hole

and then pulled the mounting pins off of the bottom of the heatsink

so that the heatsink would not make physical contact with the ground plane of the pcb.

A dab of silicone glue under the outside corner of the heatsink helps hold it in position

so it couldnt be knocked over and potentially make contact with the pcb.

The small heatsinks are light enough for the leads of the sand device to support it

freestanding in the air. If desired, I can take a pic tomorrow and post it.

Here again, this is not a "goof-proof" way of doing it, just a work-around kludge

that may not be acceptable in the proper view of building a circuit.

[kevin gilmore]

yeah... please don't do it that way. The alumina insulators really are required.

The plastic #6-32's definitely work plenty fine.

[livewire]

Since I will be placing another order with Mouser and others soon,

I will be ordering the alumina insulators and new small heatsinks and will be

re-doing the install correctly. I understand that it is not prudent to potentially

having 450VDC floating around on top of the heatsinks of the bare metal tab devices.

Big shock hazard there if one were to brush up against it while say testing other parts on a live circuit.

I'm still not a fan of plastic screws to affix / clamp these devices.

Craig Sawyers mentioned earlier about a longer plastic shoulder washer that protrudes

beyond the metal tab of the sand device, into the hole of the alumina insulator.

These may also be prone to shorting out due to the thin side wall which may or may not have

sufficient dielectric withstanding strength to resist an arcy-sparky jumping from the tab to the metal screw threads.

So plastic screws it is.

Is there a general consensus on what is the best kind to use, strength wise?

At a decent price point and availability? Vendor name and part numbers please?

[livewire]

So I guess I've proven (the hard way) that the current limited kgsshv power supply

is not immune to a dead short between it's ground and one of the 450VDC rails.

I'm curious to see what fried. It is still outputting -200VDC on that rail.

Time to remove it from the chassis and have a look.

[Craig Sawyers]

Craig Sawyers mentioned earlier about a longer plastic shoulder washer that protrudes

beyond the metal tab of the sand device, into the hole of the alumina insulator.

These may also be prone to shorting out due to the thin side wall which may or may not have

sufficient dielectric withstanding strength to resist an arcy-sparky jumping from the tab to the metal screw threads.

It is the Aavid 7721-3PPSG. Bush length is 3.18mm (1/8"). Wall thickness is 0.36mm (14 thou) and the material is 40% glass filled polyphenylene suphide, which has a dielectric strength of 385V/thou - so 14 thou = 5.4kV.

I've used these exclusively on the T2, and we're talking 50 devices mounted to grounded heatsinks this way, without any difficulty.

[livewire]

Thank you sir for clarifying that!

When I read your post before, I remembered that there was still some doubt

as to the application of this shoulder washer.

I would still need to use my custom ground backwards #4-40 stainless screws.

But that shouldnt be a problem.

[Craig Sawyers]

Thank you sir for clarifying that!

When I read your post before, I remembered that there was still some doubt

as to the application of this shoulder washer.

Well first off - don't believe for one minute that my build of the T2 was without problems! But the shoulder washers wasn't one of them. The only thing to watch is that these are one thou larger in bush diameter than a regular TO220 bush. So they go into some transistors real easy, and in others they are pretty tight (technically an interference fit) - but they do go in with some persuasion - depends on where in the tolerance band the bushes and transistor holes lie.

[kevin gilmore]

The kgsshv power supply is NOT current limited. To do that the power supply board would

have to be about 1.5 inches bigger in the one direction... Adding 2 x ixys parts and 2 more

heatsinks. Making the board this much bigger would guarantee that the thing would never

fit in one box.

Maybe for a 2 box second version.

Still the amount of energy stored in the output caps, subjected to a dead short results

in fireworks no matter what you do. But at least the semiconductors would not have

popped.

I may have come up with an even better solution for the mini heatsinks. have to try it first.

[justin]

no optional fuse holders? shame on you

[jcx]

I'm curious what people do about the lead spacing creepage/clearance - TO-220 only has ~0.028" min spacing at the lead entrance to the package - not enough to meet any consumer safety standard without coating

obviously lots of stuff gets built without extra coating, sleeving, lead forming but not so much at 450 Vdc

the TO-247 package has more reasonable lead spacing,

~2 mm but still dosen't work with the 0.6 mm+0.005 mm * Vpk IPC-9592 formula

I once found but have since lost a reference to kV blocking sub 100mA fuses - not real common

and why is my font so small?

[kevin gilmore]

no optional fuse holders? shame on you

and the ones you use have fuses that are rated at 250 volts.

I'll bet they arc real nice, and might even explode at 500 volts when they pop.

The 600 volt fuses are huge.

[dreamwhisper]

I want to make the kgsshv absolutely trivial to build,

So I don't need to understand every post in this thread? Awesome...

[Craig Sawyers]

I'm curious what people do about the lead spacing creepage/clearance - TO-220 only has ~0.028" min spacing at the lead entrance to the package - not enough to meet any consumer safety standard without coating

and why is my font so small?

Try Permissible RMS Voltage Between the Legs of a TO220 package? for the answer. And your font is small because you used Times New Roman. New Romans must have had good eyesight.

[Craig Sawyers]

I'm curious what people do about the lead spacing creepage/clearance - TO-220 only has ~0.028" min spacing at the lead entrance to the package - not enough to meet any consumer safety standard without coating

Creepage and clearance is all to do with product safety. Where the regulations apply is when a hazardous condition can arise through breakdown or shorts - for example around a power transformer, either conventional or switching, or in common-mode inductors in power line filters for example. This is particularly important because power lines are subject to high voltage transients through industrial users, power line fault handling, and environmental disturbance (lightning, solar weather etc). Creepage and clearance has to take these conditions into account.

Internal parts in which electrical breakdown does not give rise to an external hazard condition (such as via the chassis, exposed connectors or cabling) are not covered. This thinking is well considered in the European Low Voltage Directive Electrical Safety: Low Voltage Directive (LVD) - Electrical engineering - Enterprise and Industry

In any event, the lead spacing between the wide part of the pins on a TO220 varies considerably depending on the device. Specification and caliper measurement on three devices rated at 800-900V have minimum interlead spacing between 1.02mm (0.040") and 1.35mm (0.053"). The highest voltage device, a 2SA4686A with 1200V Vceo, has 1.65mm (0.065"). In fact there is a lot of evidence that semiconductor manufacturers know about all this, and chose a lead frame that takes device voltage handling into account.

Air breakdown is 3kV/mm, and so these devices are all comfortably derated, with the absolute maximum device voltage being between a third and a quarter of the air breakdown voltage.

If you are still concerned, you can go over the leads, package entry and PCB pads with a conformal coating pen - a bit like solder resist, apart from the fact that you can solder through this stuff if you need to do a repair.

[jcx]

safety analysis may have assume the Q are shorts to the ps - which may work out depending on lots of other assumptions - including ES headphones/cable construction meeting double insulation standards - which I don't see stated in their literature

even if not endangering the user - the large energy storage of the ps C with a internal short is likely to damage the amp itself - so knowing the "real" creepage/clearance limits is valuable

contamination/condensing conditions are not just industrial issues - here in costal NE condensing conditions happen indoors when cold soaked parts in my lab stroage bins get hit with warm humid air from hurricane remnants

and that sea breeze smell includes salts and microbe guts to glue it to surfaces

so what may work on the bench in a climate controlled lab may not continue to work after a decade in the wider world where "indoors" and "shirtsleeve" may not be so controlled as central US homes typcially are

when practice doesn't meet the standards by factors of 5-10x I think its worth looking into the issue a little more

[Craig Sawyers]

so what may work on the bench in a climate controlled lab may not continue to work after a decade in the wider world where "indoors" and "shirtsleeve" may not be so controlled as central US homes typcially are

when practice doesn't meet the standards by factors of 5-10x I think its worth looking into the issue a little more

Heavens - lets just knock the whole idea of DIY on the head, why not? After all, what was good for the death of Heathkit ought to be good enough for the death of all the DIY projects on Head Case.

Let's get a little perspective here, please! And rely on a seasoned electronics professional like KG to know his stuff.

[jcx]

if the spacing is a real issue then some solutions can be quite DIY - nail polish or RTV Silicone and and bending the source/collector lead out of line for greater pcb pad spacing isn't going to kill the project

wasn't a complaint about some ES amps (origninal T2?) that some have died due to dust/contamination shorts? - I don't think the problem is complete fantasy

and occasional safety questions/reminders when the project involves near/over 1 KV delta ps and >100 J energy storage seems prudent in a DIY thread

[luvdunhill]

But what about all the ES headphone cables you would still have to replace!