emailists said
What was the answer regarding the exact Wattage of class A? Is it in fact 40 watts?
It is about 1W into an 8 Ohm load. Every increase of idling current of a factor of 1.414 will increase the class A power by a factor of 2. By doing this, it will quickly increase the heat sink temperature to values that will shorten the life of the components inside the amp. As I have said in a previous post, the sound of this circuit is surprisingly independent of the actual idling current and class A power within reason and the idling current it is set at makes for exceptional sound and reliability.
BHK
Is it 1W indeed? Or is there something missing after the ā1ā? It canāt be 1W if the amp is idling at 150W.
emailists said
What was the answer regarding the exact Wattage of class A? Is it in fact 40 watts?
It is about 1W into an 8 Ohm load. Every increase of idling current of a factor of 1.414 will increase the class A power by a factor of 2. By doing this, it will quickly increase the heat sink temperature to values that will shorten the life of the components inside the amp. As I have said in a previous post, the sound of this circuit is surprisingly independent of the actual idling current and class A power within reason and the idling current it is set at makes for exceptional sound and reliability.
BHK
Is it 1W indeed? Or is there something missing after the ā1ā? It canāt be 1W if the amp is idling at 150W.
Well, actually it is! The idling current is nominally 250 mA per phase and with a supply voltage of about +/-42V, the output stage idling dissipation is about 42W per channel or 84W. Add the solid state front end power, the tube heater and B+ power, and power supply inefficiency, and it does add up to 150W off the AC line. Now, if I havenāt posted this yet, to a first approximation, one allows the idling current to double for a signal peak. When this happens, the doubled current goes into the load for one half cycle of signal current and the other output device just cuts off, a basic definition of class A. Now, this happens for both phases and we have a current of 0.5A travelling through 8 Ohm from the plus phase into the minus phase for this half cycle of consideration. 0.5A x 8 Ohm is a 4V peak voltage across the load. Power is peak voltage squared divided by twice the load resistance for 16/16 or 1W. One can demonstrate this with a two device output stage by monitoring the device source current waveforms and measuring the output powers that result when the waveforms just begin to cut off
Interestingā¦ My Pass Labs x250 āeatsā 300W. According to Nelson Pass, it leaves Class A at 15W. 2 times difference in idling wattage equals 15 times difference in Class A wattage. Of course, you canāt compare them directly, but stillā¦
Bascom, out of curiosity. Could you please tell us the parameters of the 300W monoblocks? In terms of power consumption, Class A wattage, etc?
Alekz said
Interesting... My Pass Labs x250 "eats" 300W. According to Nelson Pass, it leaves Class A at 15W. 2 times difference in idling wattage equals 15 times difference in Class A wattage. Of course, you can't compare them directly, but still...
Bascom, out of curiosity. Could you please tell us the parameters of the 300W monoblocks? In terms of power consumption, Class A wattage, etc?
Nelson Pass invented one of the methods of āsliding biasā a long while ago. In such an arrangement, the bias is dynamically changed over the signal cycle where when one device is turning on and the complement is turning off, the bias lets the turning on device conduct more than twice the idle bias current amount while keeping the turning off device from turning off all the way. this fulfills one definition of class A where the turning off device does not turn off. The waveforms of the device currents still look like unidirectional half sine waves at higher output but close looking of the waveforms show the turning does not turn off. This has the good benefit that devices that donāt turn off all the way donāt store charge to the same extent and are more easily turned back on. So it would be possible to have this condition where the turning off device actual turns off might well be 15W or more with a overall power dissipation or power draw that is way less than what full ātrue class Aā might require.
My understanding of the 300W monoblock setup is to have the same idling current per phase as the stereo amp has in one channel. The AC power draw would be similar as would be the Class A power. The difference is that the whole power supply is devoted to this one channel with twice the number of output devices and hence the ability to drive difficult loads with more ease. It is like one had paralleled the channels of the stereo amp.
If the BHK Signature 250 (stereo) amp will produce 250 wpc into 8 ohms and 500 wpc into 4 ohms, and if the BHK Signature 300 (mono) amp will produce 300 wpc into 8 ohms and 600 wpc into 4 ohm, approximately how many watts per channel will each amp produce into a 2 ohm load? The reason for the question is, my speakers present an amp with an 8 ohm nominal load but dip down to 3 ohms within their 30Hz to 30kHz frequency range.
Letās make it more extreme. I have Martin Logan Summits whose impedance gradually drops from 4 to 0.5Ohm Is the stereo version stable/powerful enough or the monoblocks will be recommended in this case?
If the BHK Signature 250 (stereo) amp will produce 250 wpc into 8 ohms and 500 wpc into 4 ohms, and if the BHK Signature 300 (mono) amp will produce 300 wpc into 8 ohms and 600 wpc into 4 ohm, approximately how many watts per channel will each amp produce into a 2 ohm load? The reason for the question is, my speakers present an amp with an 8 ohm nominal load but dip down to 3 ohms within their 30Hz to 30kHz frequency range.
Best Regards, Kevin
Hi Kevin. This response is going to apply to your question and the following ones that ask what the amps will do into difficult loads like 2 Ohm, less than one Ohm & Electrostatic loads. First, the steady state power output into loads of 4 Ohm and below is going to be limited first by the protection circuitry and then by the output stage supply rail fuses. The best way to test this is with tone bursts which relate more to playing music. When we get more of these amps built and testing conditions permit, we will test for some of these things.
Thank you for following up! So the Signature 250 and Signature 300 will only be specced at 250 / 300 wpc into 8 ohms and 500 / 600 wpc into 4 ohms respectively, correct?
Thanks again, Kevin
Hi Kevin. We may do some further testing to add to the specs. For instance, we might spec the maximum tone burst current into a 1 Ohm load and possibly more.
Hi Bascom, understood. Thank you so much for taking the time to provide your insight! It is truly amazing to have an opportunity to converse with the Designer/Developer of a product. Take care, Kevin. PS. I wish I were in Santa Barbara.
Yeah, looks like the same old issue with the digits and measurements. You can improve the sound by looking at the measurements (think DS and John Atkinsonās measurements), but you canāt guess the sound based on the specs or measurements
In the specs the 300W amp is not very different from the 250W amp. Doubling the amount of transistors in the output stage did not double the power, and does not change the Class A āregionā. Doubling the the size of the power supply also did not double the power. Doubling the price usually does not double the SQ. The sound quality should definitely improve, but you canāt quantify it.
So, let me guess. We can expect better stability, more control, more dynamics, better sound sources (or instruments) separation, right? Provided the power line is thick enough.
Also, monoblocks can be placed closer to the speakers to keep the speaker cables short or even connect the internal speaker wires directly to the amp terminals (one of the typical Magnepan tweaks ).
Nelson Pass invented one of the methods of āsliding biasā a long while ago.
...
My understanding of the 300W monoblock setup is to have the same idling current per phase as the stereo amp has in one channel. The AC power draw would be similar as would be the Class A power.
Interesting, this is what Nelson says about his amps:
Nelson Pass said
As a practical matter, this means that our X (Class AB) amplifiers are biased to dissipate roughly half of their rated output power. The XA (Class A) amplifiers are biased to dissipate roughly three times their rated output power.
So, for the 30W pure Class A amp, the idling power will be 180W (two channels)
Comparing to Class A Accuphase:
A-20V 20Wpc idle at 160W (x4)
A-45 45W idle at 235W (x2.6)
A-60 60W idle at 300W (x2.5)
A-200 100W (mono) idle at 300W (x3)
And something completely different:
Ongaku 20Wpc max 200W, they do not specify the idle power consumption, but still x5
So this is why Iām surprised to know that an amp idling at 150W leaves Class A at 1W (x75!!!).
" Tube and MOSFET stages are completely isolated, with tubes turned off in standby, and everything else allowed to idle at 80ā100W. 15 minutes warm-up is ideal for the class-A/B output stage, which runs in class-A for the first 20W. "
Bascom is telling the truth. Our definitions of class A differ and because the amp runs about that much idle current I have always referred to that as āclass Aā bias because the stage is on and conducting by that much. Bascom subscribes to the more traditional, and technically accurate, definition of class A that he previously described to this group and corrected me. Heās a stickler for accuracy. I bow to his words.
So far Iām a bit worried about the low impedance speakers. From what I have read in the forum, BHK amp should not like them. Even @4 Ohm the wattage is 400W, not 500W. I raised this concern before, and as far as I remember the amp was not tested on capacitive loads (e.g. Martin Logan speakers).
Is the stereo version stable/powerful enough or the monoblocks will be recommended in this case?
