Obviously the acoustic power will increase quite linearly according to electric power. If we douple the electric power, acoustic power will be doupled too.
Does somebody disagree ?
My example has nothing to do with my or your hifi system.
It also has nothing to do with current drive amplifier/speaker system in general. It is an example how the power delivery from current generator type amplifier takes place.
It shows how the bass resonance can be emhasized when the output impedance is increased. It is pure mathematics.
You said you have listened "one of those systems".
Any listening experience with one high output impedance amplifier-speaker system has nothing to do with this subject.
I simply tried to point out how the frequency response of the whole system varies when the output impedance of the amplifier is changed from low to high.
It doesn't happen even with another amplifier. In fact they also make a zero feedback solid state amplifier with a Zout of 2.5 ohm which mirrors the typical zero feedback tube amp. This one was published with the kit I was talking about earlier. I understand that this destroyes your beliefs but that is!
To respond also to your previous message you are a little bit optimistic about your modelling of the real world......
There is now two alternatives: You do not understan what I try to say or you try to invent new physics law.
I just summarize my points:
- When constant current generator is used to supply power to variable resistor, the delivered power increases when load impedance increases.
Do you agree ?
- When a power supplied to the speaker is increased the acoustic power will also increase, not remain same.
Do you agree ?
I just summarize my points:
- When constant current generator is used to supply power to variable resistor, the delivered power increases when load impedance increases.
Do you agree ?
- When a power supplied to the speaker is increased the acoustic power will also increase, not remain same.
Do you agree ?
Last edited:
I think you are a little bit arrogant for two reasons:
1) You don't know those systems at all
2) Maybe I know physics better than you....
For the last time: the point is using the right drivers with right reflex tuning. It's not inventing laws just solving the problem in another way which is not the conventional one. It is VERY clear this is the only one you know and you are not able to think in another way. This is the reason why you will never understand. You can go on forever with your limited thinking. Bye....
matejsirk, nice build (strange side conversation...), but I have one question about the amp picture. What are two the silver things in near the back? Potentiometers or a switch?
Hi,
my resultsv (I have only multimter and osciloscope):
- Closed loop gain = 25,2 dB-----> NFB = 9 dB Yes, around 24dB
- Pmax = 13 W @ THD = 9,6 % I Got around 20V PP on 8Ohm before clipping (I belive that is around 12W)
- THD = 3,1 % @ 2 W (I can not measure that)
- Zout = 4,7 ohms. (I can not measure that)
- Quiescent current of GU-50 = 70 mA, Ig2 = 5 mA. On cathode I got 66mA
Can you make simulation with B+ 550V and 60mA on cathode? What power can I expected and what THD?
When the anode voltage of GU50 is 550 V, then 470 ohms cathode resistor gives some 60 mA cathode current.
Power is increased a lot but distortion increases too.
The anode impedance should be higher than 3k5 if 550 V supply voltage is used.
Some results:
Uin Pout THD
0.35Vrms 4.7 W 7.5 %
0.64Vrms 14,3 W 13,3 %
0.78Vrms 20.2 W 15.5 %
20 Vpp across 8 ohms is only some 6.2 W.
THD looks "a little high". What about if I take (I have them) C3G at B+ 390V, G2=150V, plate voltage 220V and red LED bias? I do not have skills to use LTspice. Artosalo - could you send me your LTspice simulation on mail?
Distortion comes from the output stage, not from voltage amplifier.
So using C3G instead of 6J9P or 6J49P changes nothing.
I think the original supply voltage (390 V) is guite optimal with your 3k5 transformer.
If you want to get better distortion readings, why don't you use global NFB ?
What do you want to be simulated ?
So using C3G instead of 6J9P or 6J49P changes nothing.
I think the original supply voltage (390 V) is guite optimal with your 3k5 transformer.
If you want to get better distortion readings, why don't you use global NFB ?
What do you want to be simulated ?
No, not in terms of optimal output power
The cathode current of his amplifier is 66mA @ plate voltage of 390V (Cathode Voltage assumed as ground, OPT-DC-loss and Screen Current neglected)
The nearest ideal OPT value (for a Pentode) in that specific operation point would be 390V/66mA = 5.9k
With 3.5k, the triangle below the load line is a lot smaller and so is the output power. Furthermore the distortion will rise
With your OPT (3.5K) you will get most power out with the following U and I:
P=U^2/R ; P=I^2*R ; with Pa-max=40W this yields:
374V and 106mA
and a theoretical output power of nearly 20W
---The discussion about current driving of speakers is hopeless. I tried that once with my "normal" speakers and it was awful. So, I have no more experience in this field. Nonetheless I can't imagine that it works that well - simply because I don't know anybody who does it. (Yeahyeah, I know, this is a hardcore Argumentum ad populum)
BTW: Anybody read that book ?
http://www.amazon.de/Current-Driving-Loudspeakers-Eliminating-Distortion-Interference/dp/1450544002
Higher load impedance will reduce distortion essentially, but since the thread started with already built amplifier I assumed the transformer will not be first component to be changed.
Concerning the current drive of the speakers, I think one reason why non-feedback amplifiers are often told to sound so good is the relatively high output impedance
and the loudness-effect this creates with typical speakers. Especially the boost at the bass resonance frequency can be many dB:s.
Concerning the current drive of the speakers, I think one reason why non-feedback amplifiers are often told to sound so good is the relatively high output impedance
and the loudness-effect this creates with typical speakers. Especially the boost at the bass resonance frequency can be many dB:s.
Last edited:
I have a pair of 10K:8 OPT. I have also one power trany 350V AC (around 450V DC I belive). This amp is still prototype. I will also try 6J52P and D3A. All this tubes I have. Thx to everyone.
1. I did on output tubes - should I do that on 6J49P too?
2. I did to 110K. Should I try something arounf 70K
3. I will try 10K
4. I will try.
The point is, to decrease distortions of an output stage and decrease it's output resistance you first use higher load resistance to trade off voltage swing for less distortions, shunt it's cathode resistor decreasing feedback by current, then increase feedback by voltage decreasing value of resistor. If resulting amplification factor is too low shunt driver's cathode resistor, or/and use a tube with higher transconductance.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.