I remember in one of Nelson's articles he did a distortion comparison between the IRFP040, 140, and 240. He noted that the IRFP040 was the preferred part (at the time). So, I looked around until I found a few of them. By the time I read Nelson's article, I had already built my first two ACAs with the IRFP240s. I have seen other postings since recommending the IRFP044. It seems that more than a couple of folks here like the 44N10 part better than the IRFP parts. I'm more confused than ever about which parts to use for this set. I will eventually upgrade my original two ACA amps (again).
Hi Rocket
all my powerFETs are tested (matched) as you can see in my compare list pages before. this helps me to have in my tiny housing for nearly the same temp. management for Q1 +Q2.
and i think that from a batch of e.g. IRFP240 you should use the parts with the most gain....i compared my irfp240 with that i got 2 years ago from the shop for my M2!
Test 1:
I compared the 240 and 140 and i have to say that the 140 is more "darker" and give me more high end resolution.
Test 2: 140 vs 044 its not my final result but the 044 seam to have more punch and get a bit better sound stage in the deepness but loose a bit of high end resolution as the 140.
as other members remind me is that the ideal power fet should have less Ciss and a high gfs = transconductance ("gain") so the best in that is the 44N10 from fairchild. still not available. as tungsten found there is a little brother -->FQA28N15.
i will wait for the 44N10.
if you read there are some Cree parts e.g. C3M0120090D (not cheap!) with nearly the same gfs as the IRFP240 but with very low ciss,cout...
its on my bench and i will try this as Test 3 next weeks and report back.
as a lot members wrote to me.
the changes from gfs and Ciss on paper is not as huge in listening as expected....nevertheless i like to play around
i never tried to use different FETS at Q1+Q2
chris
all my powerFETs are tested (matched) as you can see in my compare list pages before. this helps me to have in my tiny housing for nearly the same temp. management for Q1 +Q2.
and i think that from a batch of e.g. IRFP240 you should use the parts with the most gain....i compared my irfp240 with that i got 2 years ago from the shop for my M2!
Test 1:
I compared the 240 and 140 and i have to say that the 140 is more "darker" and give me more high end resolution.
Test 2: 140 vs 044 its not my final result but the 044 seam to have more punch and get a bit better sound stage in the deepness but loose a bit of high end resolution as the 140.
as other members remind me is that the ideal power fet should have less Ciss and a high gfs = transconductance ("gain") so the best in that is the 44N10 from fairchild. still not available. as tungsten found there is a little brother -->FQA28N15.
i will wait for the 44N10.
if you read there are some Cree parts e.g. C3M0120090D (not cheap!) with nearly the same gfs as the IRFP240 but with very low ciss,cout...
its on my bench and i will try this as Test 3 next weeks and report back.
as a lot members wrote to me.
the changes from gfs and Ciss on paper is not as huge in listening as expected.
...nevertheless i like to play aroundi never tried to use different FETS at Q1+Q2
chris
Last edited:
Hi Rocket,
Each of these chips have their own sound. I know that I went through a few of the Cree chips and many others, and they each have a different sound character. The Cree chips did have a remarkable mid-range and upper clarity but less impact in the lower registers. I tried so many different chips and combinations, but after an intense period of trying chip after chip, looking for the "Best" one, I came finally to the conclusion there is not a best one, there is only the one that you like the best, and that is the one for you. There isn't a perfect chip, just the one that does the best in the most categories that matter to you.
DIY for me is figuring out how I want the music to sound, and then trying to find the parts that get the closest to that sound. You can go crazy chasing other people's preferences. Of course there is satisfaction in trying new parts, it is exciting, but in the end, you have to decide what you like best, that is the major battle
DIY for me is figuring out how I want the music to sound, and then trying to find the parts that get the closest to that sound. You can go crazy chasing other people's preferences. Of course there is satisfaction in trying new parts, it is exciting, but in the end, you have to decide what you like best, that is the major battle
Thanks everyone for the thoughtful replies. I guess it is time to start experimenting. The IRFP240 in Q2 and IRFP044N in Q1 did not quite do it for me. That was using the standard ACA part values. For the premium part build, I will probably start with the IRFP140 in Q2 and the FHQ44N10 in Q1 using 28VDC linear power. Then upgrade the original set of ACAs with something different. Then it will be on to the Aleph J builds. I hope everyone has a safe and enjoyable weekend!
David
David
Desoldering the big transistors from the board is a lot of work, and could damage the board. Best to unscrew the transistors from the heatsink.
Note: trying to desolder the transistors while they are still attached to the heatsink is extremely difficult or even impossible with a low wattage iron.
Note: trying to desolder the transistors while they are still attached to the heatsink is extremely difficult or even impossible with a low wattage iron.
Last edited:
Desoldering the big transistors from boards is easiest if you clip the leads flush with the board on the side of the transistor, so now you have 3 short pieces of copper that you can heat individually and pull out. Much easier than heating all 3 longer leads at once and yanking all out at once when some will be cooler than others. Obviously if you want to save a transistor this won’t work, BUT it’s a whole bunch easier and much less likely to damage the board.
Yep. cut the leads and deal with removing them from the PCB one at a time - it's what I do as sacrificng a device is a good trade for keeping my stress levels under control. I have adopted this as a general rule when I need to remove anything with more than one connection to a PCB.
Don't forget that the PCB might only tolerate a snall number of reheatings before the copper tracks start lifting off the board so if you want to do some sand-rolling some sort of plug/socket arrangement might be good, at least until you settle on your preferred device.
Don't forget that the PCB might only tolerate a snall number of reheatings before the copper tracks start lifting off the board so if you want to do some sand-rolling some sort of plug/socket arrangement might be good, at least until you settle on your preferred device.
Hey there, I'm asking for a little help sourcing the correct parts:
is this the correct cap for C1?
Access to this page has been denied.
They are running low so I want to grab them, there looked like 7 different options of the cap.
For R11 and R12, these are out of stock any other suggestions?
https://www.mouser.com/ProductDetai...ibsBBZgv0ne99aHA9wG4rrr%2BEJ/LznDpgJ30IEYGA==
is this the correct cap for C1?
Access to this page has been denied.
They are running low so I want to grab them, there looked like 7 different options of the cap.
For R11 and R12, these are out of stock any other suggestions?
https://www.mouser.com/ProductDetai...ibsBBZgv0ne99aHA9wG4rrr%2BEJ/LznDpgJ30IEYGA==
I think this is the better one for C1
Nichicon Capacitor 4700uF 35Vdc KG Gold Tune
As this is the coupling cap, could we go even bigger with this part? I have a very early SAE amp and the coupling caps are the size of soda cans.
Nichicon Capacitor 4700uF 35Vdc KG Gold Tune
As this is the coupling cap, could we go even bigger with this part? I have a very early SAE amp and the coupling caps are the size of soda cans.
Ok, Awesome, thanks, and yes I did go through all the pages and found the parts list and mouser numbers and have found everything I need. That was so helpful.
I'm looking for a little clarification on the power transistor, in this modification do you switch Q1 from 240 to 140, and also Q2 from 240 to 140? or just one of them. I couldn't quite follow.
I'm looking for a little clarification on the power transistor, in this modification do you switch Q1 from 240 to 140, and also Q2 from 240 to 140? or just one of them. I couldn't quite follow.
I am new to the forum so there is a delay when I post as the post is waiting for approval, during this time after I asked the question about parts I went through the rest of the thread and found the spreadsheet with all of the Mouser part numbers, so thanks to whoever put that list together!
My remaining two questions are:
1) the parts list shows replacing Q1 or both Q1 and Q2 with 140's, which configuration is recommended? and does it still make sense to change these transistors if you are still just using the Meanwell PS?
2) For C1 as a coupling cap would increasing it further to 6800uf extend the frequency response in the bass or would it make it tend to sound boomy or bloated?
My remaining two questions are:
1) the parts list shows replacing Q1 or both Q1 and Q2 with 140's, which configuration is recommended? and does it still make sense to change these transistors if you are still just using the Meanwell PS?
2) For C1 as a coupling cap would increasing it further to 6800uf extend the frequency response in the bass or would it make it tend to sound boomy or bloated?
I finished up the latest pair of ACAs last night. I was not able to get the 28v supplies connected, but the improvement in performance even with the 24v SMPS is not small. I ended up using IRFP044N for Q2 and FHQ44N10 for Q1. Now, I have to "upgrade" the other two amps. Then its on to the pair of Aleph Js. Many thanks to the members who have provided some very sound advice.