I have a question. Is class A really worthy? I had a chance to buy a KSA250 for about $2000 before, I decided against it after I calculated how much heat it radiates out in the living room and drawing 12A from 110V outlet. $2000 is nothing compare with the running cost of the amp. Not only the power, the cost of electricity for AC. I am sure I have to put in a separate AC in the living room instead of using the central air as it will cost so much more. I won't be surprised to burn over $1000 of electricity a year if I use it 4 hours a day.
You guys talking about water cool!!!
It better be really worth while to run this!!!
You guys talking about water cool!!!
It better be really worth while to run this!!!
I'm not sure if you are trolling, but I will try to address your concerns.
I cant recall the specifics, but if its idle power is 250W X 2 then its 500 Watts. To run this 4 hours a day in NZ is NZ$10 a month. I think our power charges are comparable with the rest of the world. That's not a lot to pay if you are passionate about your music. On the other hand, you may prefer a class AB amp or a class D. Most people building or buying this amp don't think its a big deal to spend this amount on power. How big is your car engine, and do you use public transport😉
I am not trolling. I was very serious in buying that. If you calculate a 250W class A, idle power is more than double.
Let's do some calculation, for 250W per channel
W=(1/2)I^2 X R => 250W=(1/2) I^2 X 8ohm => I=7.9A peak
Voltage swing for 8ohm 7.9A= 63.25V. So rail has to be +/-65V or 130V total.
To be class A, Idle current is half of peak current or 7.9/2=3.95A
So idle power is 130V X 3.95A=513.5W per channel
For 2 channel, it is 1027W. You still have power dissipation across the rectifier. You are close to 1100W idle. 1100W heater in the living room, you need AC except in cold winter.
It is scary.
Last edited:
OK, so you did the math, but I still dont think its huge. I have built a few amps and disassembled them as I prefer my aleph 2s. They consume 300W per chanel for 100W out.
My KSA50 was a nice amp too, why not build that if you don't want to consume a KW?
Then again, there are lots of AB amps as well. P101 was a very good amp too.
My KSA50 was a nice amp too, why not build that if you don't want to consume a KW?
Then again, there are lots of AB amps as well. P101 was a very good amp too.
You read my mind, I attracted to this thread because you guys are talking about KSA50 or 100. The 50 is manageable.
What is Aleph 2S and P101?
I am thinking of doing a happy balance of using many output transistors in parallel and run about 100mA to get a few class A watts before going into class B. And even going into class B, lots of output transistor reduce crossover distortion. I try to keep the heat to about 60W per side.
BTW, after I posted last night and went to bed, I was thinking, the KSA250 might be worst than calculation as the 12A input current was from the spec I read somewhere. I was calculating for 8ohm. That means if the speaker drop below 8ohm, the amp will go into class B if idle current is 3.95A only. I calculated the class A power when driving 4ohm speaker, it's only about 140W before going into class B if you only have 3.95A idle current. If KSA250 spec higher power when driving 4ohm, they got to have higher idle current. that means even more heat.
Thanks
Last edited:
Aleph is a nelson pass design, quite well known in this forum. P101 is a class ab mosfet amp. It sounds identical to my perreaux amp. There are many good designs on this forum.
Last edited:
Is the KSA250 a fixed bias output stage, or some form of sliding bias output stage.
If it's fixed bias for 50W of ClassA into 8ohms, then the bias should be set to around 1.8A.
That would result in ~300W of dissipation per channel.
Is the KSA250 with passive cooling sized to get rid of that heat?
Or, as more likely, it uses sliding bias.
If it's fixed bias for 50W of ClassA into 8ohms, then the bias should be set to around 1.8A.
That would result in ~300W of dissipation per channel.
Is the KSA250 with passive cooling sized to get rid of that heat?
Or, as more likely, it uses sliding bias.
The bias circuitry seems complicated as it auto adjusts for variations in line voltage and other operating conditions but is not sliding. The S series which came after uses a sliding bias configuration with fixed steps.
Cooling appears to me as quite adequate and perhaps even better than the S series which was designed to run with lesser bias most of the time.
Cooling appears to me as quite adequate and perhaps even better than the S series which was designed to run with lesser bias most of the time.
Anyone has the schematic for KSA250 particularly the bias circuit. The KSA100 schematic I have just have a round thing that say "bias". I have no idea how the bias is done. I just assume fixed bias.
I read a few article like this saying 250W class A and idle current 12A. I don't know how accurate is that. That sway me away from a very good deal.
Krell KSA-250 power amplifier Specifications | Stereophile.com
I read a few article like this saying 250W class A and idle current 12A. I don't know how accurate is that. That sway me away from a very good deal.
Krell KSA-250 power amplifier Specifications | Stereophile.com
The KSA100 is fixed output bias.
It is set to ~2.6A to allow 5Apk of ClassA output.
5Apk into an 8r0 resistive load is equivalent to 100W.
The KSA50 is also fixed output bias.
It is set to ~2.6A to allow 5Apk of ClassA output.
5Apk into an 8r0 resistive load is equivalent to 100W.
The KSA50 is also fixed output bias.
that's why it spec 12A AC idle current. If 250 is fixed bias and it spec 500W wattage at 4ohm, this means bias current is 8A!!! I question how true this is as the dissipation is over 2000W. can you make any sense of this?
How do people deal with amp like KSA250 that constantly generates 1300W or heat? I have a Nakamichi PA-7 and I did at one point adjust it to 70mA per stage. So for 7 stages, it's about 0.5A. Supply is 150V total, that's only 75W of heat per side and 150W total. The room got really warm in the summer. Air conditioning does not even help because the thermostat is in another room. If I turn down the temperature, the other room will be too cold.
You are talking at least 8 times the heat with the KSA250. How do you deal with it? I had to back off the bias to factory spec of 40mA per stage of the Nakamichi to keep the room cooler. I read here people use water cool, isn't that over board?
Can the heatsink on the KSA250 able to dissipate 1300W in air only?
When I build my power amp, I think I would bias to have a few watts in class A and class AB at higher power. I can't take the heat in the room.
You are talking at least 8 times the heat with the KSA250. How do you deal with it? I had to back off the bias to factory spec of 40mA per stage of the Nakamichi to keep the room cooler. I read here people use water cool, isn't that over board?
Can the heatsink on the KSA250 able to dissipate 1300W in air only?
When I build my power amp, I think I would bias to have a few watts in class A and class AB at higher power. I can't take the heat in the room.
Last edited:
I have two KSA250s in a biamped system working into a large room of 41' X 13'. The heat is nice during the cooler months. For the summer I use a different pair of amps since I do not run air conditioning.
Hot climates, ClassA and air-con does not suit Builders with thin wallets.
They need lots of money to supply the electricity for the ClassA and even more money for the energy used to cool the room air back to comfortable.
I would use ClassA for normal British weather/temperatures all year round and turn to ClassAB when we get an exceptionally warm summer. But that's pretty rare here in the hills of Scotland. I design for Tamax=28°C
They need lots of money to supply the electricity for the ClassA and even more money for the energy used to cool the room air back to comfortable.
I would use ClassA for normal British weather/temperatures all year round and turn to ClassAB when we get an exceptionally warm summer. But that's pretty rare here in the hills of Scotland. I design for Tamax=28°C
You lucky!!!
For us, we have to have separate air conditioning because central air usually have one thermostat for the whole house, you have a big heat generator like this, you disturb the whole house. In winter you make it comfortable in this room in winter, other people will be freezing in the other rooms as you literally have to turn off the central heat.
Summer is easier already, you just put in a small air conditioning in the room.
I was so tempted to buy that KSA250, it pained me to give that up.
Is a big class A really worth it? 1300w is like a hair dryer on all the time. I know that it can heat a room. Water cooling may make sense as you can pipe it to a radiator outside the room/house.
Thanks for the circuit. It's a good one, no need to thermal coupling between spreader and power transistor to control bias.
One thing, in a sense, it is still a fixed bias as you see the bias current and using feedback to keep it constant regardless of temperature. I thought what they talked about auto bias is like adjusting the bias according to the power output needed. that means the bias stay low for low volume music, then goes up as signal goes higher. Am I miss reading this?
Thanks