Blew my speaker dummy load

[diydidi]

I just blew my dummy load's one side, the 8ohm side. It consists of some ARCOL 2ohm 50 Watt resistors in series.
I mounted the on a HUGE heatsink and a fan.
This time around one of my techies forgot to turn on the fan while he was testing. The culprit was a Behringer iNuke 6000 driven to almost clipping. Hell knows why this happened. Still pissed.
Anyways, I took it apart. The heatsink is fairly huge with some thin machine cut grooves cut into it, running across the length of the heatsink.
Maybe these 4 or so grooves running under each resistors is weakening the thermal contact of the res/heatsink. I did use thermal paste under them.
Would this play a big role, or was I just unlucky??
It there a way to fill these little grooves, or what can I use to cover them before installing resistors again.
I have plenty of Brass shim stock of varying thicknesses lying around. Would it work.
Any ideas welcome.

[Jazz P Bass]

Picture?

[teemuk]

The little grooves in the sink probably are much more insignificant cause for failure than driving a 100W dummy load with more than 1600 watts!

[diydidi]

some pics,

[Silvertone Jockey]

What cools better than a fan and heat sink?..............oil. In the RF amp world dummy loads are cooled with mineral or better yet transformer oil. The Heathkit 'Cantenna' built with a relatively low wattage resistor mounted inside the lid of a 1 gallon paint can filled with oil is a good example. Don't remember exactly but I believe the resistor is around 50-100 watts and immersed in oil it will withstand 1000 watts for about 10 minutes with a 50% duty cycle. (10 on 10 off)

[g1]

Electric kettle full of water. Make sure to get the right impedance one, or connect multiples in series as required .

[km6xz]

The resistor was carbon film on a 1.5" ceramic hollow tube 6 or 7 inches long and was about 50 watts on air. If there was a circulation pump it could have been used continuously but with just the gallon of oil there was not enough time at high temperature to create convection current circulating up along the resistor, flowing over the top of the oil and sinking along the can metal that was radiating the heat away, causing the cooling oil to sink. The resistor had a lot of surface area, compared to those metal clad resistors so they would be less efficient than a flat or edge wound power resistor with a ceramic hollow core. I build a series of amplifiers using clear transformer oil to cool the whole output tubes(exterior anode transmitting tubes), driver tubes and their driver pc board all sitting in a glass column filled with transformer oil.Worked great.

[Silvertone Jockey]

The resistor was carbon film on a 1.5" ceramic hollow tube 6 or 7 inches long and was about 50 watts on air. If there was a circulation pump it could have been used continuously but with just the gallon of oil there was not enough time at high temperature to create convection current circulating up along the resistor, flowing over the top of the oil and sinking along the can metal that was radiating the heat away, causing the cooling oil to sink. The resistor had a lot of surface area, compared to those metal clad resistors so they would be less efficient than a flat or edge wound power resistor with a ceramic hollow core. I build a series of amplifiers using clear transformer oil to cool the whole output tubes(exterior anode transmitting tubes), driver tubes and their driver pc board all sitting in a glass column filled with transformer oil.[ATTACH=CONFIG]30192[/ATTACH]Worked great.

Nice work on that amp. So two 4CX250B's (judging by appearance) are good for 500w, what were you able to get out of them oil cooled? Myself I just used to go with a bigger tube

[km6xz]

I went with 3cx300ae made by Svetlana intended for audio. They stopped making them after I built this and a hi-end manufacturer started getting orders for the mono block. They dropped because no manufacturers designed gear around it. But when this one, the first, was shown at a Japanese hi-end international hi-fi show a bunch of small makers jumped on it but it was too late, Sveltana had no clue marketing or how manufacturers of end units worked.
There were some direct conduction tubes, without internal fins, but a flat side to be bolted on a heatsink that would have been even more effective but not as cool looking. I had to generate a good grid swing being triodes but wanted that sound, although tetrode would have been easier to drive.
I made a wilder 833 class A amp for the same show but two years before that was a monster. Each monoblock weighed 110lbs and each had an external power supply that also weighed 105lbs so a stereo pair was 430 lbs but looked really cool Those directly heated cathodes were so bright, no room lighting was needed, all in a cage made of 1/4 stainless steel rods molded into a "U". I tried to fine some old 750TH tubes to really freak out the Japanese hi-fi press, a pair of those glass monsters would look like two bulbs from coastal lighthouses., only bigger. Each tube was about 2 feet tall and a bit over 1 foot diameter in the center where they were widest. They required a pretty high Z plate transformer which made them less practical due to the high plate voltage needed. The external anode transmitting tubes in the 100-500 watt range are much more practical, with manageable voltages but still needed elaborate interlocks for safety since the whole exposed anode is the plate, at plate voltage.

The amp above was only 300 watts in that version but I made a version running cooler bias that generated 600 but the transformers were expensive to have built, and weighed a ton. These run hot mide between Class A/B and A. The 600 watt'er was A/B and used as a bass amp and is still in use years later, with the original tubes. The cooling by convection currents really worked well, better than calculated. I found some plastic beads with the right mass to be neutral buoyancy so mixing some in to the oil allowed me to measure the circulation speed to make dissipation calculations. My GF at the time wanted little plastic fish added to the oil for finished units.

[Silvertone Jockey]

Man you're bring back memories. Continuing the thread drift, I know well how a room lit only by tube filaments looks, and really although ceramic tubes are more robust, nothing screams coolness like large glass tubes with a foot or two of filament glowing. I have built RF amps with 3CX3000, 4-1000, 3CX5000A etc. on down to 813's, 811's, and 572's etc. There are many plans for AM modulators using 813's etc. in the Radio Handbook by Bill Orr and the ARRL handbook. All a modulator is for those who don't know is nothing more than a large audio amp. Replace the modulation transformer with a output tranny and there you go. Can't find a big one? use a large power transformer of the right ratio and be amazed how good the frequency response is, seriously!

Sometimes thread drift is a good thing

[Mick Bailey]

The little grooves in the sink probably are much more insignificant cause for failure than driving a 100W dummy load with more than 1600 watts!

That's the problem exactly. Not enough resistors to handle the load. The other consideration when running a dummy load close to its maximum or beyond, is the thermal resistance of the resistive element to it's case, and the thermal resistance of the heatsink. You may be able to run quite a high sustained load well beyond the rating of the resistor, but a sudden increase can defeat the ability of the resistive element to pass that heat to its surroundings and it will burn out.

It doesn't matter how well you can cool the heatsink, or (beyond a certain minimum limit) how much cooling area or mass of aluminium you have, it's the internal temperature of the resistor that matters.

[diydidi]

So IOW, i need to blow air DIRECTLY onto the resistors?

[J M Fahey]

Dunk them in a water bucket.
No kidding.

[Tom Phillips]

So IOW, i need to blow air DIRECTLY onto the resistors?

Actually that would not be a significant improvement in cooling the resistors. The case of that model resistor is not designed to dissipate a significant quantity of the heat into the surrounding air. The main heat dissipation path is through the mounting base and into the heatsink. Post #3 diagnosed the problem. If the fan has been running it just would have delayed the time it took for the damage to happen.

[Jazz P Bass]

How about using an electric water heater element.