question about power rating on a potentiometer

[mort]

The pot is providing a "programming" voltage to your FET, and the FET is passing current to your tubes. Back of the envelope calc gives me about 1/10 watt dissipation in the pot. For safety I would avoid mounting that pot on chassis metal, put it on a piece of non conductive plastic instead. Plexiglas, Lexan, something like that. Keep in mind there's not going to be much space inside the pot between its shell and the track & electrodes. Breakdown voltage of air is 8000V per inch. The 350V in your circuit could hop across a millimeter gap. I'd place the pot inside the chassis, with a screwdriver slot on end of shaft, on the other side of a hole with a plastic insulator grommet on it. I can see a knob on this sort of thing would be an invitation to a bad shock.

I'd use a separate switch for the amp power. What sorta genius thought to put the power switch on the back of a pot doing this function...

Also with the switch separate, you may be able to find a pot that has a higher power rating. In the hi-vacuum surface studies lab we'd sometimes have to rig up something like this. Those tough ol' Allen-Bradley or Ohmite pots were rated 2 watts, some came with pre cut screwdriver-adjust shafts. Snoop the surplus sources like ApexJr.

Hmmm.. I'll take all of this into consideration. Mounting a power switch somewhere else isn't an issue, but the control will need to be readily accessible to the user. I suppose I could cut out a small square area of the chassis where the pot is going and put in a small piece of plexi and mount the pot to that. That would at least get it off of the chassis.

[Leo_Gnardo]

Hmmm.. I'll take all of this into consideration. Mounting a power switch somewhere else isn't an issue, but the control will need to be readily accessible to the user. I suppose I could cut out a small square area of the chassis where the pot is going and put in a small piece of plexi and mount the pot to that. That would at least get it off of the chassis.

Still have to be concerned about a possible arc/short within the pot - connected to 350VDC - to the user. Somebody wants to fiddle faddle with that control, doesn't want to use a screwdriver with well insulated handle, I expect. Any knob with a set screw still invites the hazard. A push on knob made of definitely non conductive material, lexan or plexi again. Or a shaft extension made of lexan/plexi, we did those in circuits up to 25 KV. Then you could use an ordinary looking knob. You'll also want to heatsink your FET, and make sure that heat sink is well insulated/isolated. What do the people that offer the VVR kit use? Wonder if they've thought out the hazards & how to avoid 'em?

[mort]

I never found a kit. Just sourced the parts and built them onto the main eyelet board. Was planning to mount the FET to the chassis wall and put some heat grease in between, as per the advice of the guy that did the schematic. But if you're saying that the heat sink needs to be isolated then maybe I need to rethink more than I thought.

[Leo_Gnardo]

I never found a kit. Just sourced the parts and built them onto the main eyelet board. Was planning to mount the FET to the chassis wall and put some heat grease in between, as per the advice of the guy that did the schematic. But if you're saying that the heat sink needs to be isolated then maybe I need to rethink more than I thought.

I dont' know how your IRF is packaged, but even if only 3 metal leads stick out of a plastic rectangle, there's 350V on them & what's the insulation properties of the plastic pak? I'd hafta say you're whistling in the graveyard to just bolt the FET to the chassis.

FWIW I remember some experimenters using LM317 regulators for hi voltage reg supplies in hi fi home builds 1980's, they were having problems similar if they didn't dope out the hazards first. (Audio Amateur magazine.)

[mort]

This is new ground for me.. never worked with transistors of any type.

I've got some room in the chassis to work with. How do I size/select a heat sink? Here's one of the transistors, came in a bag of 5.

[Tom Phillips]

I agree with ditch the switch on the pot.
Also note that a linear taper is specified on the schematic.
If it was my build I would just use a high quality pot such as http://media.digikey.com/pdf/Data%20...tyle%20RV4.pdf
See the spec section that lists 900V Dielectric strength and 500V operating voltage. That's the pertinent spec for your application. The 2W power rating is just along for the ride.
Proper parts selection is just plane old good engineering. Why try to make a silk purse out of a sow's ear?

They are available from Digikey at RV4NAYSD105A Precision Electronics Corporation | Potentiometers, Variable Resistors | DigiKey

Also remember that the pot body and shaft will be grounded to the earthed chassis.

[mort]

Thank you for taking the time to help. I appreciate it.

[J M Fahey]

P(ower)=V^2 (V squared)/R so admissible V is:

V=(Square root of) (P*R) = Sqrt (0.25W*1000000)=Sqrt 250000=500V

So with 500V applied, end to end, full track will dissipate rated 250mW .

Since Fet gate does not take current, things stay as calculated above.

As of:

Always fun times when Juan uses Mega-Font and his nuke-bomb picture...

Justin

you are absolutely right, how about this picture?
Still applies to amps with 250VDC at the output:


don't like to post gory pictures though, so won't show the CSI forensic pictures of the unlucky singer who was standing just in front of those speakers, when voice coils were shot forward at 1250 fps (.44 Magnum speeds)

Just google "rail gun" or "coil gun"

[Leo_Gnardo]

They are available from Digikey at RV4NAYSD105A Precision Electronics Corporation | Potentiometers, Variable Resistors | DigiKey

PEC pots, same co. as the 'spensive guitar & amp controls, made in Canada if I'm not mistook. Similar enough the A/B & Ohmites I mentioned. With specs like that should be safe enough to chassis mount.

The transistor, metal tab on there, I wouldn't chassis mount. Isolated heat sink & I'm not the worlds best at figuring out which would be a good heat sink to buy. Another back of the envelope calculation says you'll have up to 40 watts of heat to dissipate. You can always score heavy duty ones cheap at surplus, or if you rip up a dead solid state amp for parts. Look for some phenolic standoff legs to mount it, not plastic/nylon as they might bend & melt with the heat. Digikey might have these, might as well look if you're going to buy the control pot there. Same with the hardware holding that transistor to heatsink, I'd just bolt it on there with a schmear of heatsink grease, no plastic parts as they may melt & who needs that.

[Leo_Gnardo]

don't like to post gory pictures though, so won't show the CSI forensic pictures of the unlucky singer who was standing just in front of those speakers, when voice coils were shot forward at 1250 fps (.44 Magnum speeds)

IOW about 850 miles per hour. So you get a sonic boom besides. Note guitarist appears to be squirting lighter fluid into the victim amp. Some amps deserve this treatment, was this one a Gorilla? Or a Stage?

[Justin Thomas]

No, it's a Crat-

(Ducks, runs out of room, as Juan throws charged lytics and a blown OT at me)

Justin

[J M Fahey]

No, it's a Crat-

(Ducks, runs out of room, as Juan throws charged lytics and a blown OT at me)

Justin

Hey !!! , I LOVE and RESPECT Crates !!!!!!!!!!!!

Unjustly despised by armchair Guitar Gods by the way ..... or Harmony Central/My Les Paul/etc. guys, which are the same.

The best and most complex SS distortion I ever saw (although the original one was invented by Carvin's Murphy something).

And the Bass amps are straight Ampegs (same thing under the hood, different label and Tolex).

Maybe you were thinking about something guru human?

[Justin Thomas]

My first amp (I don't count the Gorilla!) was a GX-30M. Got it in 1993. It's still working just fine, after I sold it to my aunt about 10 years ago. And it still sounds good! Not as good as my Concert, but seriously, is that a fair comparison?

And yes, I was poking the bear...

Re: Ampeg bass amps, I liked the oldies, before the proliferation of 5-strings and the pervasiveness of what I call "Rap Bass." And they still get mopped up by my 50W Bassman with periods-correct 2x15." You know, from when bass was an INSTRUMENT and not just a set of frequencies that are faked, processed, and overdone. And I will STILL put a real SVT up against anything. We need to be going in the opposite direction, instead of kilowatt power and inefficiency.

Justin

[Tom Phillips]

...Was planning to mount the FET to the chassis wall and put some heat grease in between, as per the advice of the guy that did the schematic. But if you're saying that the heat sink needs to be isolated then maybe I need to rethink more than I thought.

You need to realize that the mounting flange is connected to the drain pin so you can't just bolt the MOSFET package to the chassis. If you did then the HV would be shorted to to the chassis. The solution is to use the proper insulating mounting materials. I suspect that you misunderstood the guy that did the schematic or he used a completely different package or he never really built it. See the attached figures and also note that the metal extends across the back of the package to provide a larger surface area for heat transfer. If you use a heat sink then you need to either insulate the device from the heat sink or you need to electrically isolate the heat sink from the chassis and place it where no one will come in contact with it. There are certainly lots of potential pitfalls in DIY electronics. Lots to learn to get all the details sorted out. Thankfully the information is available on the internet and in the excellent books that are available.
Cheers,
Tom

[66 Kicks]

I use an Aavid 43-77-20G pad between power MOSFET's and whatever they are mounted to. These pads will fit devices much larger than an IRF830, but they will work fine with the IRF830 and bigger is better, right?

I wouldn't isolate a heatsink and bolt directly to it because somewhere, sometime somebody will touch the thing not suspecting that it has the full B+ voltage on it. I would suggest drilling a hole through the center of a Wakefield 558-75AB heatsink and mount that on the outside of the chassis with the MOSFET on the inside of the chassis bolted through that hole in the heatsink and using the Aavid pad.

From the looks of the datasheet, I would say that the IRF830 is going to get real hot. If it burns up, you could use a STW12NK90Z which will definitely hold it and should only need the chassis as a heatsink. It also has zeners built into it so you don't need that external one. A 7721-7PPSG shoulder washer makes the mounting pretty slick.