O.K. AndrewT, because I need to know how to adjust the amp (3 output pairs per channel) for my speakers (4 Ohm) I took my calculator and figured out the following:
Iq = 2,55A for 50W into 4Ohms.
Bias current per pair = 2,55/3 = 0,85A (or 850mA)
Bias voltage = 0,85 x 0.68 = 0,578V (or 578mV)
Is this correct? I have 0,47 and 0.22 Ohm 5W (MPC71) resistors ad hand, so these two in series gives me 0,69 Ohm. The result will differ slightly from the above calculation but I took the 0,68 resistor as an example. Can I use only a 0,47 Ohm resistor by the way?
Cheers
Iq = 2,55A for 50W into 4Ohms.
Bias current per pair = 2,55/3 = 0,85A (or 850mA)
Bias voltage = 0,85 x 0.68 = 0,578V (or 578mV)
Is this correct? I have 0,47 and 0.22 Ohm 5W (MPC71) resistors ad hand, so these two in series gives me 0,69 Ohm. The result will differ slightly from the above calculation but I took the 0,68 resistor as an example. Can I use only a 0,47 Ohm resistor by the way?
Cheers
AndrewT said:
Hi Andrew,
Thanks for all of that. I have 3 pairs per channel.
I set the voltage across the Re resistors yesterday at 0.4. Man doe it run hot. I'm not sure the fan I have is sufficient or maybe the heatsinks are too small. I can barely hold my hand on the heatsink and that is with no load. I can't hold my finger on the devices. Too hot for that. I’m not sure how hot this thing is suppose to run. Can you usually keep your hand on a Class A heatsink?
Also, I am running a 30-0-30 trafo. It is putting out about 42-44V rails. Is this contributing to the heat? Would it run cooler with only 4 pair per channel?
Should I be biasing for higher output since I have higher rails?
Thanks, Terry
jacco vermeulen said:
OK, is there an easy way to reduce the voltage? I don't really want more fan noise. What's weird is that the air coming out of the heat sink doesn't really feel all that warn. You lay your hand on the sink though and you get a whole different picture.
Are these suppose to run that hot?
Will lowering the bias cool it any?
Thanks, Terry
Re: Setting the Bias
You guys forgot about the trusty Class-A calculator SS I posted several moons ago: Mr. Green Here it is for 8 ohms and 50 watts:
Inputs
Number of output devices: 4 Pairs
Voltage rails (per rail): 38 volts
Emitter resistance (per device): 0.499 ohms
Bias voltage per Emmiter resistor: 221 mv
Speaker ohms 8 ohms
Results
Idle bias per device: 0.443 Amps
Total Amplifier bias (per rail)** 1.77 Amps
Total Amplifier bias (both rails) 3.54 Amps
Total Dissipation (per channel) 134.6 Watts
Total Dissipation (per Device) 16.8 Watts
Class-A output: Peak 100.4 Watts peak
Class-A output: RMS 50.2 Watts RMS
Efficiency 37.30 %
Max Class-AB RMS 74.00 Watts RMS
and for 70 watts class-A
the voltage across the 0.49 emitter resistors is 261mV up from the 221mV for 50 watts class-A.
Inputs
Number of output devices: 4 Pairs
Voltage rails (per rail): 38 volts
Emitter resistance (per device): 0.499 ohms
Bias voltage per Emmiter resistor: 261 mv
Speaker ohms 8 ohms
Results
Idle bias per device: 0.523 Amps
Total Amplifier bias (per rail)** 2.09 Amps
Total Amplifier bias (both rails) 4.18 Amps
Total Dissipation (per channel) 159.0 Watts
Total Dissipation (per Device) 19.9 Watts
Class-A output: Peak 140.1 Watts peak
Class-A output: RMS 70.0 Watts RMS
Efficiency 44.05 %
Max Class-AB RMS 74.00 Watts RMS
for 4 ohms, you need double the bias current....
For 50 watts the ER voltage drop would be: 315mV and for 70 watts class-A into 4 ohms the ER voltage drop would be about 369mV across each 0.49 ohm emitter resistor.
K-Amps
lgreen said:
You guys forgot about the trusty Class-A calculator SS I posted several moons ago: Mr. Green Here it is for 8 ohms and 50 watts:
Inputs
Number of output devices: 4 Pairs
Voltage rails (per rail): 38 volts
Emitter resistance (per device): 0.499 ohms
Bias voltage per Emmiter resistor: 221 mv
Speaker ohms 8 ohms
Results
Idle bias per device: 0.443 Amps
Total Amplifier bias (per rail)** 1.77 Amps
Total Amplifier bias (both rails) 3.54 Amps
Total Dissipation (per channel) 134.6 Watts
Total Dissipation (per Device) 16.8 Watts
Class-A output: Peak 100.4 Watts peak
Class-A output: RMS 50.2 Watts RMS
Efficiency 37.30 %
Max Class-AB RMS 74.00 Watts RMS
and for 70 watts class-A
the voltage across the 0.49 emitter resistors is 261mV up from the 221mV for 50 watts class-A.
Inputs
Number of output devices: 4 Pairs
Voltage rails (per rail): 38 volts
Emitter resistance (per device): 0.499 ohms
Bias voltage per Emmiter resistor: 261 mv
Speaker ohms 8 ohms
Results
Idle bias per device: 0.523 Amps
Total Amplifier bias (per rail)** 2.09 Amps
Total Amplifier bias (both rails) 4.18 Amps
Total Dissipation (per channel) 159.0 Watts
Total Dissipation (per Device) 19.9 Watts
Class-A output: Peak 140.1 Watts peak
Class-A output: RMS 70.0 Watts RMS
Efficiency 44.05 %
Max Class-AB RMS 74.00 Watts RMS
for 4 ohms, you need double the bias current....
For 50 watts the ER voltage drop would be: 315mV and for 70 watts class-A into 4 ohms the ER voltage drop would be about 369mV across each 0.49 ohm emitter resistor.
K-Amps
Terry lowering the bias will definitely cool it down. A good place to start is bias it 10 watts class-A as most of the signal is below the 10 watt range (unless you hear ungodly music 
)
So set the bias to drop about 0.18v or 180mV across the 0.68's and you will dissipate only 70 watts per channel in heat and get 10 watts pure class-A. This will represent about 50% reduction of heat.
Also as long as the devices are below 65C (heatsink about 60C) and steady, you should be ok.
K-
)So set the bias to drop about 0.18v or 180mV across the 0.68's and you will dissipate only 70 watts per channel in heat and get 10 watts pure class-A. This will represent about 50% reduction of heat.
Also as long as the devices are below 65C (heatsink about 60C) and steady, you should be ok.
K-
I must have missed that calculator, I am trying to calculate the VD across RBE when R=.68 and Vin is +/- 36VDC across 3 pairs of output devices into a 6 Ohm Load. Can I use the High low Bias setting to represent the same output into either a 6 or 8 Ohm load?
BTW does not the input ground need to be connected to ground to get the correct voltage measurements to ground as Mark posted? This is the only ground point the first boards offer, so if you have no input connected there is no possible chance of forming a ground circuit.
Regards
Anthony
BTW does not the input ground need to be connected to ground to get the correct voltage measurements to ground as Mark posted? This is the only ground point the first boards offer, so if you have no input connected there is no possible chance of forming a ground circuit.
Regards
Anthony
still4given said:
Th higher the difference between temperature of the air passing the heatsink and temperature of the sink, the easier the air can take heat away from the sink.
As the air moves past the heatsink fast it doesnt really have much time to get hotter.
It takes a lot of heat away because of the large volume of air flowing past it.
So you feel a lot of air, but it isn't that hot, the heatsink is !
doh...
...you are definitely on a different page to me, I was of course talking about the output stage...
Since what we really want to talk about is the drivers...
Readers digest version: max of 20w per driver transistor.
Here's my analysis:
If ridiculously high current is drawn from the output stage, by a really reactive load, the gain of the transistors will diminish...If the Hfe drops to, say 50, and the output stage is delivering 50amps, the driver will need to pump an amp into the output transistors, while the output voltage is still close to 0...not likely, but we are talking worst case here...
If the drivers are pulling an amp from a rail that hasn't collapsed under the load, the worst peak power scenario will see about 40V at 1A, or 40w, across 2 transistors, gives 20w peak each...so perhaps 14w RMS..If you have higher rails scale accordingly...
HTH
Stuart
...you are definitely on a different page to me, I was of course talking about the output stage...
Since what we really want to talk about is the drivers...
Readers digest version: max of 20w per driver transistor.
Here's my analysis:
If ridiculously high current is drawn from the output stage, by a really reactive load, the gain of the transistors will diminish...If the Hfe drops to, say 50, and the output stage is delivering 50amps, the driver will need to pump an amp into the output transistors, while the output voltage is still close to 0...not likely, but we are talking worst case here...
If the drivers are pulling an amp from a rail that hasn't collapsed under the load, the worst peak power scenario will see about 40V at 1A, or 40w, across 2 transistors, gives 20w peak each...so perhaps 14w RMS..If you have higher rails scale accordingly...
HTH
Stuart
Stuart,
You are absolutely correect about the drivers. The way Krell got the later amps to drive those rediclously low impedances at the power levels they did was by paralleling driver transistors and splitting the output stage between drivers. Beginning with the KMA-100 the driver stage also ran at a higher voltage (regulated to boot) Even late versions of the KSA-100 had two sets of drivers per channel anad that helped the situation with the big Apogees. If the driver stage can't supply enough drive current then the situation is hopeless at lower impedances. Split driver stages and 2 kva power supplies could also be incorporated into the KSA-50.......
Mark
You are absolutely correect about the drivers. The way Krell got the later amps to drive those rediclously low impedances at the power levels they did was by paralleling driver transistors and splitting the output stage between drivers. Beginning with the KMA-100 the driver stage also ran at a higher voltage (regulated to boot) Even late versions of the KSA-100 had two sets of drivers per channel anad that helped the situation with the big Apogees. If the driver stage can't supply enough drive current then the situation is hopeless at lower impedances. Split driver stages and 2 kva power supplies could also be incorporated into the KSA-50.......
Mark
jacco vermeulen said:
Well it is huge, 14" sqaure and 24" tall without the legs.
Think of it as a giant wind tunnel with all the electronic bits on the inside and that's how I built it. There are up to 4 2" fans pushing air in and a 5" pulling it out. I will be starting with just 2 2" fans pushing at first. The Fan circuits are on a 55 DEG thermostat mounted on the heatsink. A 70 Deg thermostat controls a softswitch that can shut down the DC to the Amp but keeps the fans running.
Regards
Anthony
