Distortion will be a whisker higher. Power output and efficiency are unchanged.
You could also argue that unbalanced operation--since it won't cancel common mode hum picked up in the cables--will be a little noiser, but that will be dependent on the installation. If you live next door to a radio station, it may be a high priority. Otherwise, it'll be a listen-really-carefully sort of thing.
If you live in a (nearly) pre-technological part of the world like I do, it's a non-issue. Around here, people are just now getting used to light bulbs. Under most circumstances, this is a real nuisance, but in this one narrow instance, it's a benefit. Never say that clouds don't have silver linings...sometimes they're just really, really small silver linings.
Grey
You could also argue that unbalanced operation--since it won't cancel common mode hum picked up in the cables--will be a little noiser, but that will be dependent on the installation. If you live next door to a radio station, it may be a high priority. Otherwise, it'll be a listen-really-carefully sort of thing.
If you live in a (nearly) pre-technological part of the world like I do, it's a non-issue. Around here, people are just now getting used to light bulbs. Under most circumstances, this is a real nuisance, but in this one narrow instance, it's a benefit. Never say that clouds don't have silver linings...sometimes they're just really, really small silver linings.
Grey
I hear that people in Aiken travel on horseback...
Grey
P.S.: For those who are unaware--particularly those from overseas--this nation is big enough that there are wide variations in economics, education, and all the other factors that go in the pot to make up a local population. By any and all metrics, South Carolina is near the bottom of the list of states here in the US. It's shocking to see the differences as you travel to other states. The US is far from the homogenous culture that many assume. Sadly, many of the stereotypes about people in this corner of the country are true. I have spent time volunteering in the schools and such trying to make some small changes, but there's a huge amount of momentum to be overcome. Ignorant parents raise ignorant children, who in turn grow up to take jobs that don't pay very well because they weren't well educated. Then they have children of their own and the cycle begins anew. It's tragic to see this go on and on.
And, mind you, I live in one of the more enlightened areas of the state.
Grey
P.S.: For those who are unaware--particularly those from overseas--this nation is big enough that there are wide variations in economics, education, and all the other factors that go in the pot to make up a local population. By any and all metrics, South Carolina is near the bottom of the list of states here in the US. It's shocking to see the differences as you travel to other states. The US is far from the homogenous culture that many assume. Sadly, many of the stereotypes about people in this corner of the country are true. I have spent time volunteering in the schools and such trying to make some small changes, but there's a huge amount of momentum to be overcome. Ignorant parents raise ignorant children, who in turn grow up to take jobs that don't pay very well because they weren't well educated. Then they have children of their own and the cycle begins anew. It's tragic to see this go on and on.
And, mind you, I live in one of the more enlightened areas of the state.
Grey,
You could say the same thing about most all of the rural areas of the South.
m.
P.S. ....Go Gators!
(sorry, had to write that)
((members in Europe scratch their heads... Gators? Go??))
P.S.S Built a Balanced Zen Line Stage and you'll be able to run
your Aleph-X balanced, the way Pass intended it! LOL
You could say the same thing about most all of the rural areas of the South.
m.
P.S. ....Go Gators!
(sorry, had to write that)
((members in Europe scratch their heads... Gators? Go??))
P.S.S Built a Balanced Zen Line Stage and you'll be able to run
your Aleph-X balanced, the way Pass intended it! LOL
Re: DC offset
On mine, V2 adjusts the absolute offset. That is the measurement between +input and ground. It also brings up the bias a bit while doing so. V1 & V3 adjust the bias for each side. V1 & V2 give some control of the relative offset which is what the speaker sees.
Hope that helps.
To all,
What type of cap are you using for the input caps? Electo, polly?
Thanks, Terry
yoke said:
On mine, V2 adjusts the absolute offset. That is the measurement between +input and ground. It also brings up the bias a bit while doing so. V1 & V3 adjust the bias for each side. V1 & V2 give some control of the relative offset which is what the speaker sees.
Hope that helps.
To all,
What type of cap are you using for the input caps? Electo, polly?
Thanks, Terry
Hi guys,
Well, I basically finished my Aleph X last night. I didn't have the input caps so I have left them off for now. I didn't see or hear any ill affects but I do plan to install them soon.
I wanted to post some of them things I found to make sure I'm allright.
First, I changed to just one bridge and I'm now getting +/-14.75 rails.
The bridges are measuring 68C.
I doubled up the 0R1 resistors in my CRC so now have 0R05 each. They are measuring 78C. Pretty hot.
My main heatsinks are setting right at 58C but I measured the outputs themselves and they are 107C.
I'm not sure why the devices are so much hotter than the heatsink. I used Kapton tape and grease. Should be very good heat transfer.
I was able to get absolute offset down to less than .5V and relative offset is right at 24mV. This is with 0R15 source resistors set at 0.5V and 4 outputs per channel.
I got to listen to it for about a 1/2 hour so I can't give a real accurate evaluation of the sound. It is less powerful than any of the other amps I've built so comparing is tough.
I has a pleasant sound but didn't wow me. It definately drive 4ohm better than 8, which is what I wanted.
Any input on the readings I got above will be greatly appreciated.
Blessings, Terry
Well, I basically finished my Aleph X last night. I didn't have the input caps so I have left them off for now. I didn't see or hear any ill affects but I do plan to install them soon.
I wanted to post some of them things I found to make sure I'm allright.
First, I changed to just one bridge and I'm now getting +/-14.75 rails.
The bridges are measuring 68C.
I doubled up the 0R1 resistors in my CRC so now have 0R05 each. They are measuring 78C. Pretty hot.
My main heatsinks are setting right at 58C but I measured the outputs themselves and they are 107C.
I'm not sure why the devices are so much hotter than the heatsink. I used Kapton tape and grease. Should be very good heat transfer.
I was able to get absolute offset down to less than .5V and relative offset is right at 24mV. This is with 0R15 source resistors set at 0.5V and 4 outputs per channel.
I got to listen to it for about a 1/2 hour so I can't give a real accurate evaluation of the sound. It is less powerful than any of the other amps I've built so comparing is tough.
I has a pleasant sound but didn't wow me. It definately drive 4ohm better than 8, which is what I wanted.
Any input on the readings I got above will be greatly appreciated.
Blessings, Terry
still4given said:
First, I changed to just one bridge and I'm now getting +/-14.75 rails.
The bridges are measuring 68C.
I doubled up the 0R1 resistors in my CRC so now have 0R05 each. They are measuring 78C. Pretty hot.
My main heatsinks are setting right at 58C but I measured the outputs themselves and they are 107C.
I'm not sure why the devices are so much hotter than the heatsink. I used Kapton tape and grease. Should be very good heat transfer.
The bridges at 68C will probably be fine. You may need to derate their current capability by 40% or so.
Resistors typically can handle much higher temperatures than semiconductors, so 78C is fine there as long as you don't burn yourself.
Your mosfet output devices, on the other hand, are near death. If I have interpreted your post correctly, you are running about 50W per FET. This means that the junction temperature is about 150C (add 0.8C/W to the case temperature for the junction to case thermal resistance). That means their power handling capability is nearly zero. In fact, I am surprised they haven't failed already. The temperature difference from the heat sinks to the case is much higher than it could be with better thermal contact. You could probably get it down to heatsink T+20C. In spite of that, your FETs will still have an untimely appointment with the grim reaper unless you cut them back to 30W.
Jeremy
Jeremy,
Terry is not using garden variety irfp240 or similar - he is using a hefty Onsemi device that has a large die and better heat transfer.
That said, it seems like the temperature rise from sink to device is high.
Terry,
It might be helpful to remind us what your output devices are. Did you use a lot of grease or just enough that you can tell grease has been wiped on the tape? Too much conventional grease actually hurts heat transfer. Make sure that the devices are flat on the sinks - did something get under them? Use a fender washer or similar to spread the mounting force.
Check your measurement technique - I remember that some devices would give false high readings looking down on a hot surface - they are calibrated to not receive convection heating.
And to think that you thoguht you might be able to get away with half that much heat sink. Was it NP who said figure out how much you need, double it and you won't quite have enough.
Terry is not using garden variety irfp240 or similar - he is using a hefty Onsemi device that has a large die and better heat transfer.
That said, it seems like the temperature rise from sink to device is high.
Terry,
It might be helpful to remind us what your output devices are. Did you use a lot of grease or just enough that you can tell grease has been wiped on the tape? Too much conventional grease actually hurts heat transfer. Make sure that the devices are flat on the sinks - did something get under them? Use a fender washer or similar to spread the mounting force.
Check your measurement technique - I remember that some devices would give false high readings looking down on a hot surface - they are calibrated to not receive convection heating.
And to think that you thoguht you might be able to get away with half that much heat sink.
Was it NP who said figure out how much you need, double it and you won't quite have enough.BobEllis said:
Thanks, Bob. I had forgotten that detail. It looks like the devices he is using are MTW32N20E. They do seem to be a bit tougher than the IRFP240's. They still have a maximum junction temperature of 150C, but the junction to case thermal impedence is a slightly better 0.7C/W (claimed to be the maximum value). They also have a maximum power of 180W - again somewhat better.
If the case temperature really is 107C, the junction temperature may be as high as 142C, which gives a limit for the device power of about 11W. So, if the case temperature can be lowered a bit or is actually lower than measured, you are right, these devices shouldn't have to be turned back to 30W.
Jeremy
kropf said:
Hi Jeremy,
Thanks for taking the time to answer some of my concerns. According to Nelson Pass, The devices I'm using ,MTW32N20E , are rated for higher temps, and that I should use fewer of them due to their high capacitance.
According to the AXE-1 speadsheet, I am running about 6.5A, and 48W per fet. The junction temp should be around 108C which is almost exactly what I read on my gauge. However, the heatsink temp from the spread sheet says 49C and I'm a bit higher than that. I have more FET's that are a close enough match to add more if necessary.
Blessings, Terry
EDIT:
Sorry, you guys answered before I finished posting.
The way I measured is with a probe meter. I touched the tip against the device right next to the mounting screw. I checked the heatsink the same way. I did put a pretty good coating of grease on them. I will try adding washers as well. I was pretty happy that I could actually leave my hands on the fins of the heatsink. Way better than before.
Blessings
Wipe off most of that "pretty good coating" - all you really want is enough to fill in the imperfections in the mating surfaces. Anything more acts as a thermal insulator. (Arctic Silver a possible exception)
You're not measuring junction temperature - you're measuring case temperature, since you cannot get inside the devices to get at the junctions. Junction temps are higher than the case. Maybe measure the mounting screw temperature - it might give you a closer feel for the mounting surface temp.
You're not measuring junction temperature - you're measuring case temperature, since you cannot get inside the devices to get at the junctions. Junction temps are higher than the case. Maybe measure the mounting screw temperature - it might give you a closer feel for the mounting surface temp.
still4given said:
Terry,
There is a subtle, but important, point here regarding transistors that I believe is often overlooked. The temperature that matters for transistor operation is the junction temperature. However, you cannot easily measure the junction temperature, seeing as it is inside the case. You can, on the other hand, measure the case temperature.
In order to estimate the junction temperature from the case temperature, the data sheet usually gives the thermal impedence from the case to the junction. For your transistor, it is 0.7C/W or less. So, add 0.7C/W * 48W, or 34C to your measured case temperature to get the junction temperature. That is 141C and very close to the maximum junction temperature of 150C. If you derate the transistor power according to the datasheet, you get 180W - 1.44W/C * (141C-25C) = 13W maximum power.
Probably reducing the grease to the minimum necessary to cover the surfaces and using a larger washer will decrease your case temperature by enough that everything will be fine.
Cheers,
Jeremy