ECP Torpedo III [indexed in first post]

The adhesive is to provide long term and/or shipping stability.

You can get 748 from McMaster-Carr.

 

Perfect bazelio..! The more directly you can measure on the resistor element itself, the better... Really hope they are not crazy hot after some hours with signal.

As we know the Amtrans were ''toasty'' to the touch, and when we feel this it's typically around 55-60 degrees. These are 1/5 of the power rating and have a much smaller surface area to radiate the heat, so they will be hot for sure, but hopefully not too bad... If they are in the ''safe zone'' I might buy a set as well, they are very special resistors. And with all the money thrown into the T3, who cares about another $100

Where did you buy them at a discount?

 

Morten, did you see my post above where I mentioned the TX2575-2 part number? With two chips in line they would spec out max voltage at 350V, and also just less than double the peak power of the single chip (which is what I have). Texas Components will tell you they spec these things very conservatively.

EDIT: Delivery is 2 weeks ARO on the stacked parts, and cost ~$100 (4pcs) direct from TC.

 

Yes, I know they make the stacked versions, but they don't seem to be in stock, and they are quite expensive. I might spend $100 on resistors (because I know they matter), but I'm not spending $200 or more...

Where did you find them at a discount?

 

Please don't touch high voltage parts.

 

That's right... you're supposed to start with the low voltage parts, to slowly build up your immune system.

 

Wow, what a great improvement with all six CCS boards installed. Loving the sound. Thanks so much, Doug and Tom!

 

Ok, Torpedo heat and plate loading resistors. Ahem....

Measurements taken using a Flir E8 directly on the naked foil surface. Thermal imagery to follow. My Comcast is down right now.

We said 0.484W peak, if *ALL* 2.2mA flowed through one side. Very much the absolute worst case scenario. Not even sure if it is possible under normal operation? Doug might be able to comment.

My Z Foils are running at about ~80C after a couple hours of active signal, a couple seconds after removing the chassis cover.

Looking at the Texas Components spec temp derating curve, that puts them at roughly 175% the rated peak power (0.3W) or 0.525W peak.

So 0.484W peak is absolute worst case and the resistors have a (conservative*) rating of about 0.525W peak at ~80C.

Interestingly, the CCS boards are the hottest component in the chassis at a blazing 135C. Hotter than my 12AZ7 tubes which are running at about 105C.

*TC tells me they are already spec'ing these things conservatively (via email).

 

That is pushing the temperature. Maybe more than we really want Those transistors should not be more than 150C. Possible solutions are adhering a small sink of some sort to the CCS board to help with dissipation, changing out the resistor to lower the current slightly (even one mA would probably be enough), or even increasing the resistance of the plate load resistors (or increasing the tube's CCS current) which would lower the transistor's voltage and thus decrease the voltage drop across the CCSes. Lowering the CCS current is probably the easiest solution, though I know some of you are not going to enjoy the process.

The most likely way that a resistor will see the entire 2.2mA is if one side fails for some reason (blown resistor, damaged tube, etc). Then the CCS would pull all the current through the other side.

 

Yeah, Doug, the CCS board is up there. After another hour it has crept up to 140C while the plate resistors are holding at 80C. Let me run overnight and take another reading in the AM.

Edit: It also appears as though the middle two are running 5-7C cooler than the outer two boards. The middle two are below vent slots, while the outer two have solid metal directly above. After removing the cover for 5 minutes, they are measuring roughly 127C. I'm thinking again that extra ventilation is warranted. Maybe a couple holes directly above the outer boards and a few along the side to draw room temp air in.

 

Were it mine, I wouldn't leave it on overnight. We know it will be hot

I am going to look for some small clip on heatsinks that will fit as a first option.

 

Thanks for doing the temp measurements..! The 80 deg for the plate loads is what I would consider the safe zone, they should last. Great news! The CCS is more of a concern...

Please don't increase the plate loads Dough. Most of us use 12AT7 tubes and we have a relatively low voltage and current for these already. The small heat sink sounds like a great idea. If there are no clip on's that fit, then maybe something that can be soldered to the boards back? Small piece of copper or similar that's easy to solder?

 

Thermal images.

My hottest plate load resistor:




My second hottest CCS board (unfortunately, I found one that is sitting at 150C) :




My tube:



PS @MortenB you are welcome.

 

The easiest/quickest solution is, ironically enough, increasing the current through the tube from 2.2mA to perhaps 2.6-2.8mA. This would mean changing the ~30R1 current setting resistor to something more like 25R. This does not decrease the dissipated heat (in fact it increases it slightly) but it moves some of the CCS heat to the output transistors. It also has the (possible) downside of changing the tube's operating point. Whether that's an issue would need to be determined experimentally.

Alternatively I think reducing the output current by 15-20% is probably the best solution. The downside is that that involves changing 4 hard to reach resistors.

I'll poke around for some clip on heatsinks, and maybe think about designing one. A small piece of aluminum with some little fins would not be too hard. There area also lots of little glue on heatsinks on Amazon. Attaching them would be the trick. Thermal epoxy perhaps.

 

Just ordered some GE 12AZ7A tubes, but wanted to try a pair of Amperex but couldn't find any on eBay. Suggestions? Other variants that anyone recommends? I'm currently using Mullard 12AT7 but it's a bit bass lean with the new CCS boards.

Planning to buy and install the Mundorfs next. Reliable sources in the US?

@TomB: As I am modding an existing build, is there way to make/buy spacers like the ones you ship out now?

 

This TO-220 heatsink with clips might work. Fitment could be awkward, as it's a bit wider than the CCS boards and it might end up making contact with a screw head on the existing heatsinks on the boad.

 

All you need to do is ask. I'll get some in the mail to you.

 

Contact with the screw is the concern. I ordered several different small sinks as well as both thermal tape and thermal epoxy. We'll let you know if this works in a few days.

 

I don't have any fancy thermal imaging equipment (really cool btw!), but my amp doesn't really feel too much hotter than before. I do have a huge hole cut out of the top, though haha

 

If you don't mind soldering leads on to the Mundorfs, you can have the ones I removed from my amp. Send me shipping costs and they're yours.

Tubes:
1. See this post.
2. I did a quick Amperex run-through last night, and concluded the Amperex 12AT7 is very similar to the 12AZ7A. I still prefer the 12AZ7A, primarily because it's a bit faster on the attack, but the 12AT7 isn't too far off. It'll give you a good taste of the 12AZ7A sound. I also concluded that the Amperex 12AU7 is never going to get a second listen.