I recommend removing the amplifier before trying that. I'd also point out that tossing the amp chassis on the campfire is an irreversible process ... and a smelly one with the plywood glue. Just saying.
Wood is fine for a DIY chassis. It's certainly easy to work with and the tools are cheap. One can RF shield the wood by adding a grounded copper foil or copper mesh (see your arts & crafts store) on the inside.
I'm not personally a fan of wood for amplifier enclosures. It rips and splits when I work with it and I can't cut threads in it. This probably has more to do with my woodworking skills (and metalworking skills relative to woodworking skills) than the material itself, though.
Tom
Because electronics used to be full of hot resistors that would gather dust, catch on fire, and burn your house down. I also recall radios full of air core inductors stuffed with little parafin soaked sponges to prevent the inductors from changing shape. Thanks to SDR we don't have to deal with those anymore.
I'd never use wood in a production flow. UL94 being only one of the reasons. I recall a Danish HiFi company in the mid 1990ies that made everthing out of rosewood, including a turntable platter. The platter I saw had a huge crack in it because the wood had split. I think they've gone belly-up since then. Maybe the warranty claims did them in.
Tom
"Some".
I'd run the amp at half the rated output power with a sine wave (worst case for power dissipation) for an hour and measure the heat sink temperature to find out. Keep a fire extinguisher nearby.Tom
"Some".
Tom
I agree "wonky" is a good word to describe the holey box. And it's also true that I feel more comfortable working with wood, and don't really have any skill or tools for working with aluminum. The premade aluminum chassis that I see referenced here is certainly a good deal too,.... I just really like the nostalgic look of the wooden amp.
I think that maybe if I place the holey box under my 4 channel Mod86 (attached) it will add to it's aesthetics. Their wooden face-plates are from the same sheet of wood, and same width too.
It's been a while, I went through all of the math for heatsinks when I built the four-channel.
I'm not super good at the math, but after reading on Neurochrome -Taming LM3886, I had a good understanding of the basic principals, some formulas to work with, and also some references and explanation of the formulas on the TI LM3886 Data Sheet. I fed all of this data into an Excel sheet that I made to do the math for me. It takes into account duty cycle based on crest factor, as well as corrections for length and ambient temperature. With this information, I was well aware that I was making some conservative choices erring on the side of buying more heat sink than I need.
At any rate, the huge 7" heatsinks on my four channel were originally purchased with the intention of building a six channel amp.
I later changed my plan and built a four channel, and there are only 2 mod86 on each 7"sink (instead of 3 mod86). I can play loud for hours and they will barely luke-warm to the touch. Luke-warm might be too strong of a word for it, they don't seem warm at all really. After hours of loud music, it's difficult to tell by touch if they are any warmer than they would be if the amplifier were turned off completely! This is driving 4 ohm speakers.
Now on the holey box, they are 4" heat sinks, will have only 1 Mod86 on each one, and will drive 8 ohm speakers. I have not fed the numbers into my excel sheet for the holey box. My instinct tells me that based on the above information, I should have more than sufficient heat sink. I don't really expect much if any higher ambient temp around the heat sink, especially with the holes. I guess that remains to be seen.
I think it will be very smart to drive a sine wave at 50% power for an hour, just to be sure. Primitives like me still have a healthy fear of fire.
Would it make any difference if I use an old test speaker, or would it be better to use an 8 ohm resistor across the output during the sine-wave test?
Attachments
Awesome. You're ahead of many already then! I'm glad you found my Taming the LM3886 series helpful as well.
That's in line with my expectations. I'd expect a heat sink of that profile and length to reach maybe 30 ºC if you really crank it. Nothing wrong with keeping the amp cool.
You'll be fine with the 4" length. The heat sink won't reach lukewarm during normal operation and I doubt it'll get beyond 40 ºC even at max dissipation – unless the holes significantly reduce airflow.
I agree. Do note that such an experiment only tests for what happens during normal operation, not what happens during a fault condition. I'm not saying I'm overly concerned in your case, or I would have said so. I'm just pointing out the facts.
Dumping 20 W into a speaker will be LOUD. I'd use a load resistor. If you're looking at buying a load resistor, I'd look for a set that can be configured for 4 Ω (and maybe 2 Ω). For example I use ten non-inductive 25 W 0.8 Ω resistors (Vishay NH-25) in series to form 8 Ω (250 W). A tap in the middle gives me 4 Ω (125 W). Connect the two ends together and drive the middle tap to get 4 Ω || 4 Ω = 2 Ω (250 W). I have two such strings on a fan cooled heat sink. I can put the two 8 Ω strings in parallel for 4 Ω (500 W).
If you do get fancy, put banana jacks on the load resistor. Put the jacks in this order: Middle – End-1 – End-2 and put them on 3/4" (19.05 mm) centres. That way you can use banana plug adapters to access both 8 Ω and 4 Ω.
Tom
Last edited:
I know what you mean Bill about being busy. It's too bad that with computers I can do work for my employer from home, as I worked both Saturday and Sunday. At least Monday will be an easier day for it when I go to the office, but another weekend has passed with no progress made on the amp.
By the way,.. I didn't get a chance to congratulate you on the new born, so "Congrats!"
By the way,.. I didn't get a chance to congratulate you on the new born, so "Congrats!"
Thank you. It's sad that I can produce offspring faster than I can make a case for an amp, but there you go
Producing offspring is far more fun. Raising them is a different story. I can only imagine the lifestyles of the kings & queens of yore! My wife and I are currently entertained by a US production series named "Reign" - a story of the Queen Mary of Scots...
Sorry for the off topic...back to season 4!
Best,
Anand.
I guess you found one that isn't regulated.
Looking at the market a bit closer it appears that some SMPSes for audio use are essentially transformer emulators. It looks like they're running at a constant duty cycle, so the output voltage is a fixed ratio of the input voltage. If the input mains voltage varies, so does the output voltage. Hypex's supplies are like that (at least the ones I've looked at). I'm not really sure why. Implementing a feedback loop would take a few more components and result in a fixed output voltage, lower output impedance, etc. Not much additional cost for a lot of advantages.
The Connex Electronic supplies I recommend for the MOD86 and MOD286 are the SMPS300RE and the SMPS500R. I've tested both the '86 and '286 with the SMPS300RE. I have not tested with the SMPS500R as I don't currently own one.
Tom
Looking at the market a bit closer it appears that some SMPSes for audio use are essentially transformer emulators. It looks like they're running at a constant duty cycle, so the output voltage is a fixed ratio of the input voltage. If the input mains voltage varies, so does the output voltage. Hypex's supplies are like that (at least the ones I've looked at). I'm not really sure why. Implementing a feedback loop would take a few more components and result in a fixed output voltage, lower output impedance, etc. Not much additional cost for a lot of advantages.
The Connex Electronic supplies I recommend for the MOD86 and MOD286 are the SMPS300RE and the SMPS500R. I've tested both the '86 and '286 with the SMPS300RE. I have not tested with the SMPS500R as I don't currently own one.
Tom