PWatts said:Hi guys,
Anyone going to use cans (i.e. MJ21193/4) instead of flatpacks? I managed to lay my hands on original Krell heatsink cubes, so natrually cans are preferred.
What would be the best: Keeping the output boards on the main PCB and take the transistor wires straight from there, or leave the boards altogether and hardwire the resistors, diodes etc at the cans themselves?
Thanks for all the pioneering work
Pierre
Pierre,
I will be using metals cans in a wind-tunnel. Stuart is helping me with some circuit mods, I will be doing a high biased class-AB with higher power. Rails about 82vdc. plan to use 4 pairs of mj21195/96. All OP wiring will be point to point on the heatsink. the input VAS will be on the group boards. Thats the plan so far...
thanks! K-
"Mark,
nice photo of your setup. Well done.
The fans appear to be blowing (definately better than sucking).
Can you invert the sink so that natural convection augments the fan pressure?
The heatsink will have slightly higher efficiency if you can mount the output transistors about 40% of length up from the bottom. Some sources say middle of length but I disagree. Any comments?"
Absolutely! Tha fans will end up on the bottom of the sinks. Today, time permitting I will fasten the two sinks together at the sides and mount the fans to the bottom..... a bot more fabrication to do there.....
Didn't give any thought to the actual position of the devices on the sinks. A bit taller standoffs under the board would allow me to move them up towards the center of the sink. Efficiency is everything with this amp.... Thanks for the tip!
Mark
nice photo of your setup. Well done.
The fans appear to be blowing (definately better than sucking).
Can you invert the sink so that natural convection augments the fan pressure?
The heatsink will have slightly higher efficiency if you can mount the output transistors about 40% of length up from the bottom. Some sources say middle of length but I disagree. Any comments?"
Absolutely! Tha fans will end up on the bottom of the sinks. Today, time permitting I will fasten the two sinks together at the sides and mount the fans to the bottom..... a bot more fabrication to do there.....
Didn't give any thought to the actual position of the devices on the sinks. A bit taller standoffs under the board would allow me to move them up towards the center of the sink. Efficiency is everything with this amp.... Thanks for the tip!
Mark
NUTTTR said:I'm using MJ21194 can's.. easy, and they look nice
82v rails?! Maybe you mean 41vdc rails?
82vdc rails would kill the amp i would have thought...
Aaron
Aaron, I actually do mean +/-82vdc
. I have some Toroids lying around for this. Yes as it is it would kill the amp but we would need to reduce the bias, change component values and increase OP devices. Stuart is da man doing the calcs for me. I have 6 boards, so 2 will be clones, the rest of the 4 will be some kind of KSA-100 or KSA-350 (class-A or AB have not decided yet) .... depends......
Am listening to mine tonight for the first time in stereo... Just on my NS-10's for now but the NS-10's sound better than they ever have. Output device temp is about 57.5 deg c. with the fans running on about 5 volts and the draft sucking up through the fans..... a temporary arrangement. I am pretty comfortable with that temperature though. INitial impressions are that it soulds alot mike my KSA-80 did in the mids and top end.
Mark
Mark
Attachments
"Are these the 4312 or the 4312L, Mark ?
The regular 4312 does 48dB at 2800 rpm."
Jacco Vermeulen,
These are the M version. At 4.5 volts you have to have your ear right next to them to hear any noise. This might equate to a noise level in the mid 20's....? Keep in mind that 2800 rpm is at 12 volts. I've found im my work that three bladed fans are ALWAYS alot quieter than fans with more blades, such as 5 or 6. Those type fans wuill actually whine somewhat at normal operating speeds. These 4312 M's have been laying around since my power follower days about 5 years ago.....
Mark
The regular 4312 does 48dB at 2800 rpm."
Jacco Vermeulen,
These are the M version. At 4.5 volts you have to have your ear right next to them to hear any noise. This might equate to a noise level in the mid 20's....? Keep in mind that 2800 rpm is at 12 volts. I've found im my work that three bladed fans are ALWAYS alot quieter than fans with more blades, such as 5 or 6. Those type fans wuill actually whine somewhat at normal operating speeds. These 4312 M's have been laying around since my power follower days about 5 years ago.....
Mark
K-amps said:
Aaron, I actually do mean +/-82vdc
From my experience, reducing the bias on KMA200 output stage from about 100W pure class A (couldn't go any higher with my heatsinks) to about 32W class A (my standard "summer" setting) is absolutely minimal. Frankly I cannot say that I can hear the difference. MAYBE the higher bias sounds a tad smoother ... maybe not. I can't do the blind test so I cannot be competely impartial - I know higher bias SHOULD work better, so ...
All said, with reasonably efficient speakers you will rarely go above 30W anyway, and even that will be for very short periods. I kind of see why Krell went to active bias schemas later ...
So go ahead, make that KSA 350, I'm sure it will be fine. As far as I know, there was KMA-400, but I've never seen schematics for one. My guess is it would again double on KMA-200 (that is 16 pairs of MJE15003/4). That is a LOT of drilling (I know, I have that many TO3's in my amp - for 2 channels though), and will be capable of LOT of current. Make sure rectifiers, capacitors and wiring do it justice ! I used 8 gauge OFC copper to supply rails to output cans (you will need a GOOD soldering iron for that!), and have some good ultra low ESR high voltage polypropilenes (2.2 uF) right on output stage (+V to -V is much more effective than +V to ground and -V to ground).
You will need some really high voltage devices for the input though, and 6431/33 are not available anymore in original (300V) spec ...
Bratislav
Mark,
the only square 3-blade Papst models i know of are the:
4550 - 4580 - 4850
The 4850 is the least noisy of the three, 30dBA at 1750 rpm.
A 5-blade Papst 4890 does 28 dBA at 1450.
Seems to me that RPM has a much greater influence on sound level than the number of blades.
My thought would be that a 5-blade vent at a lower speed produces the same airvolume as a 3-blade model at higher rpm, but with lower noise level.
The material used seems a factor too, the 4312 uses a GRP hausing and rotor, other Papst models are fully made in alloy.
I heard a KSA50 once, afraid i listened to the KMA400 too.
(big boy, measures 23" wide by 28" deep)
I found the KSA50 very noisy, anyone know what kind of vent was used in the little Krill ?
On the picture you posted is a second duct in the background.
It looks as that you placed inductors inside the tunnel.
Intentionally ?
That looked like a great idea, integrating powersupply inductors inside a vented heatsink.
Jacco.
the only square 3-blade Papst models i know of are the:
4550 - 4580 - 4850
The 4850 is the least noisy of the three, 30dBA at 1750 rpm.
A 5-blade Papst 4890 does 28 dBA at 1450.
Seems to me that RPM has a much greater influence on sound level than the number of blades.
My thought would be that a 5-blade vent at a lower speed produces the same airvolume as a 3-blade model at higher rpm, but with lower noise level.
The material used seems a factor too, the 4312 uses a GRP hausing and rotor, other Papst models are fully made in alloy.
I heard a KSA50 once, afraid i listened to the KMA400 too.
(big boy, measures 23" wide by 28" deep)
I found the KSA50 very noisy, anyone know what kind of vent was used in the little Krill ?
On the picture you posted is a second duct in the background.
It looks as that you placed inductors inside the tunnel.
Intentionally ?
That looked like a great idea, integrating powersupply inductors inside a vented heatsink.
Jacco.
Bratislav said:
Thanks for your thoughts Bratislav, I am not too worried about the caps.. I have enough, however one thing I did not think about was the rectification. I do have 4 80 amp ultrafast / softrecovery MUR8060's Do you think those would do? I know the Chipamp guys swear by the MUR860's (8 amps), but don't know of anyone using the MUR8060's. (Heck I can't find the datasheets anymore either)
K-amps said:
That's plenty enough. Peak currents are drawn from the capacitor bank, but you want rectifiers not to be stressed with "topping up" and inrush. 80A is generous. As far as ultrafast, I've noticed that Nelson Pass "converted" to ultrafast camp after years of following the mainstream. I'm not quite yet sold to that idea (yes, you need fast diodes at switching frequencies, but 50/60Hz ... ?), but I'm keeping my mind open.
I wouldn't use 8A rectifiers in high power class A amps. You will draw almost half of its capacity at all times, and inrush will stress it to the point of having to pray and hope every time you switch it on.
Bratislav
I will get the auction up tonight. I have one person who STILL has not received their boards.
I have another set available from a person who signed up and then changed their mind when they realized the pwr supply and chassis cost.
So the total of available pairs will change if the pair in transit is delivered.
I have had a measured response to the board requests since I have few left and people still not rx'ing theirs.
Sorry and please be patient.
I need to know if ANYONE has NOT gotten thier boards by now.
I have one person as of now. After this request the spares will be fair game!!
Please reply ASAP if you have NOT received your boards!! They will ALL be gone next week!