I finally got the twin case to the one I built one channel for AN, I got a good jump on installing today. Baring any big issues, it should be functional Monday-Tues. My machinist kinda hosed me on one part, He took it upon himself to blend a part that I didn't want done. I looks slick and all, but it won't work because of the decal size. Now I will have to start over with a raw piece of aluminum and polish and shape it before I can apply the decal... that will likely cause me an extra day.
I had some rather massive Clarity Cap CSA laying on the bench and am putting one in for testing. If I like it, I'll replace the other stock too. Kinda funny, looking at the board with a cap on it which is 1/3 the size of the board. For us old school guys, think Flintstones with that massive slab of ribs tipping the car. lol
I had some rather massive Clarity Cap CSA laying on the bench and am putting one in for testing. If I like it, I'll replace the other stock too. Kinda funny, looking at the board with a cap on it which is 1/3 the size of the board. For us old school guys, think Flintstones with that massive slab of ribs tipping the car. lol
Sorry, I watch this closely but I have no idea what this means. Could you please explain in “standard English”?
X, can you show me a pic of the reverse side of a set of these completed snubber boards? (See attached pic.)
I'm just wondering what implications there are in terms of the silicone wires connecting to the right-angled MOSFET pins - given I plan to clamp the MOSFETS (and the snubber boards) between a CPU cooler pad and a 1/4" aluminium support bracket.
Thanks,
Andy
Attachments
Last edited:
Sorry, I never took photos of the other side. It’s mounted in the chassis and all buttoned up now. Not much to say other that the wires are soldered onto the pads directly for the lowest profile connection to allow a clamp if needed. What you can do is use a clamp bar that is only 15mm wide and it won’t interfere with the pins at all. As an example, here is my recent clamp of a MOSFET and a BJT for the Glass Harmony. Using a 15mm square aluminum extrusion that is very stiff and has channels for mounting to other framework. This is using the MakerBeam brand extrusions for DIY robotics. Very handy stuff.
Although I did not use snubber boards here, you can imagine that they would fit perfectly well and not interfere. They would in fact, provide a much more even pressure distribution from the clamp bar to the MOSFET.
Here they are clamped to that really compact Dell CPU cooler (which, btw, Vunce has verified has no problem cooling 150w thermal load):
Here is the CPU cooler viewed from above:
And here is the CPU cooler (with 90mm Noctua fan) mounted into a framework formed by the MakerBeam 15mm extrusions mounted inside a PC chassis. The chassis is a ThermalTake Core VI which is turning out to be an ideal monoblock chassis for CPU cooler based Class A amps, including the ABBB or any Alpha for that matter:
Here is what the monoblock chassis looks like all buttoned up, plain but very practical and compact. More importantly, because of its consumerism based PC-gaming heritage, cost is unbelievably low compared to any audio grade chassis you can imagine that will handle a 150w+ cooling load:
More info on MakerBeam XL:
Specifications MakerBeam and OpenBeam - MakerBeam
You can buy it on Amazon etc...
Although I did not use snubber boards here, you can imagine that they would fit perfectly well and not interfere. They would in fact, provide a much more even pressure distribution from the clamp bar to the MOSFET.
Here they are clamped to that really compact Dell CPU cooler (which, btw, Vunce has verified has no problem cooling 150w thermal load):
Here is the CPU cooler viewed from above:
And here is the CPU cooler (with 90mm Noctua fan) mounted into a framework formed by the MakerBeam 15mm extrusions mounted inside a PC chassis. The chassis is a ThermalTake Core VI which is turning out to be an ideal monoblock chassis for CPU cooler based Class A amps, including the ABBB or any Alpha for that matter:
Here is what the monoblock chassis looks like all buttoned up, plain but very practical and compact. More importantly, because of its consumerism based PC-gaming heritage, cost is unbelievably low compared to any audio grade chassis you can imagine that will handle a 150w+ cooling load:
More info on MakerBeam XL:
Specifications MakerBeam and OpenBeam - MakerBeam
You can buy it on Amazon etc...
Attachments
Last edited:
You know X, this technique of cooling a large mosfet with 150W+ of dangerous heat really is the bees knees. It really is inexpensive, and with temp sensors and Noctua fans which are highly sophisticated and reliable this has revolutionalised Class A construction.
You are to be commended for adding to the expertise in this tricky area. Computers have come of age! If only HDD could last forever....... I had a failure recently of a 3TB HDD with 3k CDs on it, and I'm going through the pain of rebuilding.
Hugh
You are to be commended for adding to the expertise in this tricky area. Computers have come of age! If only HDD could last forever....... I had a failure recently of a 3TB HDD with 3k CDs on it, and I'm going through the pain of rebuilding.
Hugh
Sorry to hear about your HDD Hugh. Years ago, I had a series of HDD failures within months of one another. I was losing a main drive every 2-3 months and they were all Western Digital. That was 20yrs ago but to this day, I only use Seagate (server quality), Hitachi or Fujitsu drives and have not had a problem except on the laptop drive 2 years ago. However, I set up an automatic backup script to do incremental backups every week to a second HDD. Simple batch script I have used for years.
There is a trick used by data recovery people to get drive working to get critical files off. Put it in freezer for a few hours and try using it. Sometimes that will let it work for 15min or so so that you can backup the few critical files you worked on last.
(I think Seagate bought WD, but I still won’t use WD branded drives).
There is a trick used by data recovery people to get drive working to get critical files off. Put it in freezer for a few hours and try using it. Sometimes that will let it work for 15min or so so that you can backup the few critical files you worked on last.
(I think Seagate bought WD, but I still won’t use WD branded drives).
Hugh, Danny and XRK,
Great work with the recent measurements. Given Danny's recent build, I think a very reasonable set of projects is to get a toroid spec'd for both designs, a 5U chassis, dual SLB supplies and you are off to the races. A Toroidy 800VA with 4 dual secondaries will do just fine! My plan, like Danny is to build the Alpha 20 first and then the Nirvana.
Best,
Anand.
Great work with the recent measurements. Given Danny's recent build, I think a very reasonable set of projects is to get a toroid spec'd for both designs, a 5U chassis, dual SLB supplies and you are off to the races. A Toroidy 800VA with 4 dual secondaries will do just fine! My plan, like Danny is to build the Alpha 20 first and then the Nirvana.
Best,
Anand.
@Francios It's a bit of custom machining and hard to explain. Basically, I put a plate on the front of the amp to fill a milled out section of the face to allow mounting of a switch and space for a decal.
@ Zman It shall get done, soon!
@X I have something similar in the works as well. I duuno if I will use snubber or not, but If I do, I will use a Molex with the 90 degree turn to extend it straight out and keep it sway from where you mount the base of the heatsink... if that makes any sense to you.
@ Zman It shall get done, soon!
@X I have something similar in the works as well. I duuno if I will use snubber or not, but If I do, I will use a Molex with the 90 degree turn to extend it straight out and keep it sway from where you mount the base of the heatsink... if that makes any sense to you.
Last edited:
Hi Anand,
I would go straight for the Nirvana with FQA mosfets.
From what Hugh And X have reported the Nirvana is an upgrade to the Alpha20.
When simulating in LTspice both the Alpha and the Nirvana, I see that the Nirvana is better,
it has a higher damping factor, reacts better to 4R loads (lower distortion) and clips at higher voltage.
I've used two Toroidy transformers of each 400VA and 2x18v, my rails after the LT4320 bridge are stable at 25.2VDC, with SLB you need higher voltage specified transformers.
Grts,
Danny
I would go straight for the Nirvana with FQA mosfets.
From what Hugh And X have reported the Nirvana is an upgrade to the Alpha20.
When simulating in LTspice both the Alpha and the Nirvana, I see that the Nirvana is better,
it has a higher damping factor, reacts better to 4R loads (lower distortion) and clips at higher voltage.
I've used two Toroidy transformers of each 400VA and 2x18v, my rails after the LT4320 bridge are stable at 25.2VDC, with SLB you need higher voltage specified transformers.
Grts,
Danny
@danny without making the correction for a 4R load it does not react well to them. It's fine at low volumes, but if you try and give it any gas it doesn't react well at all. That's not a bad thing, as the design was not made for 4R. There is in fact, a schematic and parts for a 4R version here, I think Andy or one of the guys built/building it.
More details about the 8R in the next couple of days when I button up my AN monos.
More details about the 8R in the next couple of days when I button up my AN monos.
Yes, the 4R version is when you need 4R at large voltage swing, that is 4R speakers with low sensitivity.
But for my 93db speakers I usually listen at 1-2vpp, so the 8R will do for me,
X did the calculation.
Looking forward for your build, don't forget: always triple check before powering on
But for my 93db speakers I usually listen at 1-2vpp, so the 8R will do for me,
X did the calculation.
Looking forward for your build, don't forget: always triple check before powering on
I don't know Hugh, perhaps there is something wrong with the one I built because the J2 drives them fine, but with the AN the speakers where breaking up at lower levels... Let me finish the other and see if they are the same and go from there. Again, perhaps I got something wrong.
Can you describe the symptoms more clearly? At lower volume, you get breakup in sound output? The AN should not sound like it’s breaking up at all. What is your bias current?
Hello Hugh, 1.72A bias.
What I mean is at relatively low db I started to hear distortion almost like the sound of a blown speaker. I didn't hear anything wrong under about 80db, when it started to get rock'n is when I heard the breakup. I only did a real quick session with it, so I don't have a lot to add.
The J2 will drive the Tektons to almost a 90db before it starts to give up. Speakers are are easy to drive at 98db efficient.