Hi Guys,
I rearranged the output devices on my heatsink so that they run across the bottom of the heatsinks and have the fans blowing upward through them. I measured the temps of the heatsink and after about an hour of running, the heatsink measures about 58C but the devices themselves are at about 75C. Is this temp OK for them? I used TO3's on my first one and they seem to run a little closer to the temp of the heatsink. Is it normal for the flat-packs to run that much higher than the heatsink?
Thanks, Terry
I rearranged the output devices on my heatsink so that they run across the bottom of the heatsinks and have the fans blowing upward through them. I measured the temps of the heatsink and after about an hour of running, the heatsink measures about 58C but the devices themselves are at about 75C. Is this temp OK for them? I used TO3's on my first one and they seem to run a little closer to the temp of the heatsink. Is it normal for the flat-packs to run that much higher than the heatsink?
Thanks, Terry
Parts GB'ers-
No Digi-key or Welbourne yet. I will be calling them today to see what's the issue or get a tracking #.
And thanks Phil for your comment earlier.
Edit:
Emailed Welbourne Labs.... Their reply. I sure hope they did their job and got our parts out last week.
"We will be closed this week while we attend the Rocky Mountain Audio Fest.
Our website will remain open 24/7.
Thanks for your interest in our products.
Welborne Labs"
Edit 2: Called Digi-Key. They tracked the package through UPS. Package scheduled for delivery today.
No Digi-key or Welbourne yet. I will be calling them today to see what's the issue or get a tracking #.
And thanks Phil for your comment earlier.
Edit:
Emailed Welbourne Labs.... Their reply. I sure hope they did their job and got our parts out last week.
"We will be closed this week while we attend the Rocky Mountain Audio Fest.
Our website will remain open 24/7.
Thanks for your interest in our products.
Welborne Labs"
Edit 2: Called Digi-Key. They tracked the package through UPS. Package scheduled for delivery today.
still4given said:
Yep,
thermal resistance between case and heatsink is much higher for a TO247 than a TO3, and thermal resistance of the insulator disc is also higher for the TO247.
Temperature difference is lower depending on the type of insulator material used; Al2O3, Kapton, mica or silicon(silpads).
For class A operation nothing beats TO3's, best plastic package is the 2-bolt MT200 Sanken i am using for the Krells.
Thermal resistance of case to heatsink for a TO247 is 0.5 C/W, at 1.8 amps bias and 37.5 Vdc rails that alone is good for +11 C temperature difference.
The remainder is because of the type of insulator used.
Minimum for a TO3 is 0.3, mounted straight to the heatsink with thermal paste.
80 C is OK.
aamof, a lot of the familiar full class A amplifiers seem to have used 80 C for dimensioning the output stage.
When i started this hobby design information was not always easy to obtain, so i tried to gather emperical info.
I used to estimate heatsinks sizes from brochure pictures, and size heatsink dimensions at audio shows.
Combined with output device datasheets, bias and voltage ratings, that resulted in design transistor die temperatures.
80 C was the regular outcome, be it BJT or Mosfet output stage.
If you make sure you have adequate venting holes to keep the temperature of the electrolytics low, 58 C for the heatsink shouldn't be much of a problem.
aamof, a lot of the familiar full class A amplifiers seem to have used 80 C for dimensioning the output stage.
When i started this hobby design information was not always easy to obtain, so i tried to gather emperical info.
I used to estimate heatsinks sizes from brochure pictures, and size heatsink dimensions at audio shows.
Combined with output device datasheets, bias and voltage ratings, that resulted in design transistor die temperatures.
80 C was the regular outcome, be it BJT or Mosfet output stage.
If you make sure you have adequate venting holes to keep the temperature of the electrolytics low, 58 C for the heatsink shouldn't be much of a problem.
If you throw out the sil-pads and use the Thermalloy ceramic/alu. oxide insulators as I have been doing you can cut your transfer loss loss in half vs. using even micas. This gets the TO-247 package close to what the TO-3 does as far as dissipation. Sil-pads are really meant for production and the ceramic stuff while more expensive is alot better design. Just be sure the heat sink is perfectly flat and there are no metal shavings under the ceramic pad or it'll crack.... but they are tougher than you might think.
Mark
Boy Howdy!
All of the Group Buy boards have now been shipped out. Although I know that many have already received them lets give all these enough time to reach their destinations and then in a few weeks let me know if you ordered boards but did not receive them. The longest transit time seems to be to Italy and Canada and the shortest transit time other than the U.S. is to the United Kingdom. It'll be interesting to see how long it takes for them to get to Dubai, South Africa, Austrailia, and New Zealand!! Thats where some of our members are from!! This is an amazing and diverse bunch of audioaholics.
Best of luck to all of you building this longest running DIY project ever... I think it'll be a long, long time before any other thread gets towards 500 pages in length. This has been very well worth the effort and everyones input and assistance has been much appreciated by everyone involved. My first KSA-50 has been running now for almost the entire year totally trouble free! Most importantly please post pictures of yours while under construction and then post your amplifiers results in the builders WIKI!!
Special thanks to Jan for creating the first board design(we haven't forgotten you!) and a special thanks to Al (Pinkmouse) for taking the time to improve on that design with the present board, to Brian GT for a bullet proof Power Supply board that seems to have at least a pound of copper to it, to LGreen for starting the builders WIKI... I think he did, and to all that ordered during the GB for your patience during the repeated proto tests and all my absences during long work trips. This should have happened alot faster but had to be fit in around my now hectic work schedule.
I may be up for another run of these boards in the near future if enough new interest develops!! I will also do an audit to check what funds are left over once I know for sure that everyone received their boards. Any left over money will be donated to DIY Audio.
Mark
All of the Group Buy boards have now been shipped out. Although I know that many have already received them lets give all these enough time to reach their destinations and then in a few weeks let me know if you ordered boards but did not receive them. The longest transit time seems to be to Italy and Canada and the shortest transit time other than the U.S. is to the United Kingdom. It'll be interesting to see how long it takes for them to get to Dubai, South Africa, Austrailia, and New Zealand!! Thats where some of our members are from!! This is an amazing and diverse bunch of audioaholics.
Best of luck to all of you building this longest running DIY project ever... I think it'll be a long, long time before any other thread gets towards 500 pages in length. This has been very well worth the effort and everyones input and assistance has been much appreciated by everyone involved. My first KSA-50 has been running now for almost the entire year totally trouble free! Most importantly please post pictures of yours while under construction and then post your amplifiers results in the builders WIKI!!
Special thanks to Jan for creating the first board design(we haven't forgotten you!) and a special thanks to Al (Pinkmouse) for taking the time to improve on that design with the present board, to Brian GT for a bullet proof Power Supply board that seems to have at least a pound of copper to it, to LGreen for starting the builders WIKI... I think he did, and to all that ordered during the GB for your patience during the repeated proto tests and all my absences during long work trips. This should have happened alot faster but had to be fit in around my now hectic work schedule.
I may be up for another run of these boards in the near future if enough new interest develops!! I will also do an audit to check what funds are left over once I know for sure that everyone received their boards. Any left over money will be donated to DIY Audio.
Mark
Heat
Spent some time listening to the Krell today, sounds great as usual. Hey, in the interest of science took some notes on temp from the front panel temp controller, about every 3 minutes over 2 and a half hours. Not playing too loudly, 60WPC class A 8 ohms (according to K amps spreadsheet), blah blah blah you all know the amp by now.
** Temps are from sensors mounted on the Heat sinks for the indicated devices (drivers and LM have thier own heat sinks as you prob. recall, if not check out the web site below), lost this when editing the chart for the last time. Temps are not from the devices themselves, although the output devices are TO-3 mounted directly onto heatsink with thermal goop and no insulators.
Anyhow, after about 2.5 hours it still had not quite reached thermal equilibrium...I think. I'm tired, need sleep! Check out the heat profile. Wish I had a DMM with data logging so I could plot bias V as well. Anyhow, looks like my output devices track relatively closely in temp, might indicate a close bias match, no? Not sure what to make of the data, so have at it.
edit- the R channel output heat sink is the positive rail just like the left, omitted from the legend.
An externally hosted image should be here but it was not working when we last tested it.
Spent some time listening to the Krell today, sounds great as usual. Hey, in the interest of science took some notes on temp from the front panel temp controller, about every 3 minutes over 2 and a half hours. Not playing too loudly, 60WPC class A 8 ohms (according to K amps spreadsheet), blah blah blah you all know the amp by now.
** Temps are from sensors mounted on the Heat sinks for the indicated devices (drivers and LM have thier own heat sinks as you prob. recall, if not check out the web site below), lost this when editing the chart for the last time. Temps are not from the devices themselves, although the output devices are TO-3 mounted directly onto heatsink with thermal goop and no insulators.
Anyhow, after about 2.5 hours it still had not quite reached thermal equilibrium...I think. I'm tired, need sleep! Check out the heat profile. Wish I had a DMM with data logging so I could plot bias V as well. Anyhow, looks like my output devices track relatively closely in temp, might indicate a close bias match, no? Not sure what to make of the data, so have at it.
edit- the R channel output heat sink is the positive rail just like the left, omitted from the legend.
What you don't show it whether or not there are changes in the ambient temps in the room. That may have something to do with it as well.
Also, playing music through it would have to have some bearing on the changes. It's not like you were playing a sine wave. I would think some types music would drive the amp harder than others.
Blessings, Terry
Also, playing music through it would have to have some bearing on the changes. It's not like you were playing a sine wave. I would think some types music would drive the amp harder than others.
Blessings, Terry
playing music...
Terry,
IIRC all the time the amp stays in class A, the power going into the speaker is 'diverted' from the heatsink, so if you really, really play it hard and loud, you could see the temp drop...
Theoretically, up to a point, the harder you play the music, the cooler it will run.
With the exception of really hard loads dropping the amp into class B, silence is the worst case...
Stuart
Terry,
IIRC all the time the amp stays in class A, the power going into the speaker is 'diverted' from the heatsink, so if you really, really play it hard and loud, you could see the temp drop...
Theoretically, up to a point, the harder you play the music, the cooler it will run.
With the exception of really hard loads dropping the amp into class B, silence is the worst case...
Stuart
right
Right, the room did heat up a few degrees which I didn't track. Perhaps this accounts for the gradual rise of some of the temps. If we really want to be picky the AC was occassionally cycled on so when the room was cooled wonder if it affected the results.
Also I listened to music as I can't sit there for a few hours and listen to nothing or a sine wave. Not so loud it would have come out of class A.
Yes Jacco, the fans were fixed RPM over the entire duration. They don't change speed with heat (you have to dial up speeds on the front panel). Luckily it remembers the prior settings.
So my thinking is that this is real results in a real room with real music. Might be useful for something or if not it was fun having a few drinks and taking the data.
Right, the room did heat up a few degrees which I didn't track. Perhaps this accounts for the gradual rise of some of the temps. If we really want to be picky the AC was occassionally cycled on so when the room was cooled wonder if it affected the results.
Also I listened to music as I can't sit there for a few hours and listen to nothing or a sine wave. Not so loud it would have come out of class A.
Yes Jacco, the fans were fixed RPM over the entire duration. They don't change speed with heat (you have to dial up speeds on the front panel). Luckily it remembers the prior settings.
So my thinking is that this is real results in a real room with real music. Might be useful for something or if not it was fun having a few drinks and taking the data.