In our amps gate to source zeners are 15V IW....and gate driver voltage cannot exceed 15V [not because of Zeners but due to special driver arrangement]under overload/overdriven circumstances...The test was done with no current limiting circuitry in amps involved...I think I already mentioned that....
Bob Cordell said:
In the mean time, I found a preliminary and incomplete description of my distortion analyzer. If you like, I can mail it to you.
Cheers, Edmond.
andy_c said:
Andy, your observation about the current limits being the same into 4 ohms and 2 ohms is exactly correct. However, these are continuous average power measurements, so if, for example, he was using a VI limiter of some sort with a decent time constant to allow for good burst power capability, that would not show up in this data.
I certainly don't want to sound like an apologist for Bruce Candy, but I do want to try and be sure we sort out as best we can what we know and what we don't know. As you know, I have previously expressed concern that there is no evidence out there that shows that he is really meeting his claimed very low THD specifications. The Stereophile data strongly suggests that he is not, even given some of the measurement caveats that JA has cited.
Bob
Re: DM88 distortion
Yes, you are exactly right. I think the twin tone measurements also suggest that he is not meeting his numbers. If he was meeting his numbers, you would not see anything in a conventional test unless there was a setup artifact or distortion.
Bob
PHEONIX said:
Yes, you are exactly right. I think the twin tone measurements also suggest that he is not meeting his numbers. If he was meeting his numbers, you would not see anything in a conventional test unless there was a setup artifact or distortion.
Bob
Workhorse said:
This is very impressive. Try it into a 1-ohm load on a two-cycle tone burst at various powers up to full power.
Also, as mentioned, run it at full power continuous at 1 kHz into an 8-ohm load and then short the output.
Bob
Ok I will try 100Hz for dual cycles into 1 ohm....
But i have already done the short circuit test...is there any difference involved...
Kanwar
But i have already done the short circuit test...is there any difference involved...
Kanwar
Well done Kanwar, I doubt that a SUSTAINED short circuit, without protection would do anything but melt the silicon, sooner or later, or blow the building fuse.
That reminds me of a story about short circuit testing.
Back in 1978, I was commissioned by Matti Otala to make a regulated power supply for his future Harmon Kardon power amp. Well, I made the supply, but I also had to make it short circuit proof. So, I ran some tests on it, much like you are doing. Finally, I blew the circuit breaker at the fusebox. I thought this was quite an accomplishment at the time. How about that as a limiting condition?
That reminds me of a story about short circuit testing.
Back in 1978, I was commissioned by Matti Otala to make a regulated power supply for his future Harmon Kardon power amp. Well, I made the supply, but I also had to make it short circuit proof. So, I ran some tests on it, much like you are doing. Finally, I blew the circuit breaker at the fusebox. I thought this was quite an accomplishment at the time. How about that as a limiting condition?
The best measurement for the Halcro DM38 is in the June 2004 'Hi Fi News'. At 300W/4 ohms at 1KHz 3rd was -116dB, 5th at -142dB. With an 8 ohm load (150W) 3rd was -128dB and 5th was -137dB Residual was about -143dB in this test.
Pretty darn good results. I again give them a 'gold star'.
Pretty darn good results. I again give them a 'gold star'.
john curl said:
Hi John,
Thanks for this info. I don't receive Hi Fi news, so I was unaware of this test. What kind of a setup did they use to achieve a -143 dB residual? Was it just an AP feeding the distortion output to a spectrum analyzer, or did they do something more or different?
Did they have any THD-20 tests there?
Thanks,
Bob
john curl said:
Hi John,
Got another question for you. Do you know how many output pairs Halcro uses in the DM88? I don't, but I wonder.
Bob
Speaking of Halcro, by the way. Does any of you know what is at the other end of it, ie. what the input stage looks like? I have seen a schematic in a Halcro patent application that uses a CFP diff amp at the input. Since we have been discussing such input stages in another thread recently and had a hard time coming up with amplifiers actually using it, I wonder if Halcro does, or if that was only used in the patent app.? That is, I am not really asking about that specifc amp being discussed, but rather about Halcro amps in general.
I think that they use feedback pairs around their differential input devices. This lowers distortion even further. I'm pretty sure that the input stages look a lot alike Self's designs, except for more feedback pairs, etc.
john curl said:
Hello John
From the Halcro website the DM38 specifications are :THD @ 1kHz <-130dB (<300 parts per billion), they still do not meet their specs.
Regards
Arthur R
Read John Atkinson´s follow-up in Stereophile. It is not easy to measure such ultra low distortion (test load banana plug oxidation). Under -110dB at high power, one can easily make a small mistake.
DM38 distortion
Hello PMA
You are right , they ultimately measured lower distortion further on in the article.
Regards
Arthur R
Hello PMA
You are right , they ultimately measured lower distortion further on in the article.
Regards
Arthur R
Quote:
Ok I will try 100Hz for dual cycles into 1 ohm....
quote:
"Also, as mentioned, run it at full power continuous at 1 kHz into an 8-ohm load and then short the output."
But i have already done the short circuit test...is there any difference involved...
Kanwar
Hi Kanwar,
Any difference would be in the speed of the event. I tried to find a reference to the rise time for a sudden short, such as would be produced by a screwdriver shorting the output when the unit was operating, but couldn't find anything. If this is similar to an ESD strike though there could be a problem. I do know that ESD circuits are typically designed to cope with a current rise time of about 200ps. This is a good deal faster than most protection devices can respond and tends to leave the circuit at the mercy of the parasitic elements unless steps are taken to slow the event down to a point where the protection circuits can operate.
I'm interested in this because I want to understand why John Curl's experience is that FET devices tend to be fragile with shorts. Maybe there's something here, maybe not.
Tony
Ok I will try 100Hz for dual cycles into 1 ohm....
quote:
"Also, as mentioned, run it at full power continuous at 1 kHz into an 8-ohm load and then short the output."
But i have already done the short circuit test...is there any difference involved...
Kanwar
Hi Kanwar,
Any difference would be in the speed of the event. I tried to find a reference to the rise time for a sudden short, such as would be produced by a screwdriver shorting the output when the unit was operating, but couldn't find anything. If this is similar to an ESD strike though there could be a problem. I do know that ESD circuits are typically designed to cope with a current rise time of about 200ps. This is a good deal faster than most protection devices can respond and tends to leave the circuit at the mercy of the parasitic elements unless steps are taken to slow the event down to a point where the protection circuits can operate.
I'm interested in this because I want to understand why John Curl's experience is that FET devices tend to be fragile with shorts. Maybe there's something here, maybe not.
Tony
Cellardoor said:
Great Strike.....!
The special arrangement of output coil and special PCB layout help us in achieving neccessary delay to slow down the fast rising current slope.
I have made this point here several times before.....but no one seems to understand its importance..
Kanwar