Mark A. Gulbrandsen said:
Mark, I guess I could install a socket and install an LED every 5 to 7 years? Not a big deal to me if it fades. If it sounds better than the Zener or LM current regulators on noise, then I think the trade off is worth it.
How would the fade off affect the current and or resistance of the LED? Lower it or increase it?
Bill
LED lifespan?
Mark, I just checked on line and most LED's have a lifespan of 50,000 to 100,000 plus hours. That would be a max of 11 years continuously if you left the amp on 24 hours 7 days a week.
So, are you sure there would be a problem. I do not think I would ever use my amp 1/10th that time. The power bill would kill me before the LED would die!!!
Bill
Mark, I just checked on line and most LED's have a lifespan of 50,000 to 100,000 plus hours. That would be a max of 11 years continuously if you left the amp on 24 hours 7 days a week.
So, are you sure there would be a problem. I do not think I would ever use my amp 1/10th that time. The power bill would kill me before the LED would die!!!
Bill
LED as a voltage reference
After reading some more on diyaudio, I have found recommendations to use purposely old dim worn out LEDs, because apparently their resistance goes down, so actually with age the LED works better for this voltage reference application.
Here is the quote of what I found searching:
"Post #16
OK, if you have trouble getting parts, then use an MJE340 as a crude CCS in place of R2. There's no need to fuss matching LEDs - and you don't want good ones! You want the cheapest nastiest oldest dimmest red LEDs you can find. The nice bright new ones have a much higher internal resistance...
There's no need to put a capacitor in parallel with the LED. Assume 5R internal resistance and that you want the capacitor's reactance at 100Hz to be equal to the LED's internal resistance. That's 330uF. It's hardly worth the trouble."
http://www.diyaudio.com/forums/show...&highlight=LED+voltage+reference&pagenumber=2
It seems to me LEDs have been used before and are fine.
Bill
After reading some more on diyaudio, I have found recommendations to use purposely old dim worn out LEDs, because apparently their resistance goes down, so actually with age the LED works better for this voltage reference application.
Here is the quote of what I found searching:
"Post #16
OK, if you have trouble getting parts, then use an MJE340 as a crude CCS in place of R2. There's no need to fuss matching LEDs - and you don't want good ones! You want the cheapest nastiest oldest dimmest red LEDs you can find. The nice bright new ones have a much higher internal resistance...
There's no need to put a capacitor in parallel with the LED. Assume 5R internal resistance and that you want the capacitor's reactance at 100Hz to be equal to the LED's internal resistance. That's 330uF. It's hardly worth the trouble."
http://www.diyaudio.com/forums/show...&highlight=LED+voltage+reference&pagenumber=2
It seems to me LEDs have been used before and are fine.
Bill
We have our freedom to go extreme.
But, I wonder a little bit whether the noise from zener is a real concern for the power amp operation . . . so that I am very interested in your result.
My concern? is more about the traffic and wind noise smearing from out of the window.
. . . . . . joke!
Regards
But, I wonder a little bit whether the noise from zener is a real concern for the power amp operation . . . so that I am very interested in your result.
My concern? is more about the traffic and wind noise smearing from out of the window.
. . . . . . joke!Regards
Mark A. Gulbrandsen said:
Mark-I'll not bother with that "led longevity "issue......world is full with CCSes with led as reference,even in medical instruments....
if led based CCS is good enough for Bill ,I don't see any further reason for concern.......except that he will need soon (if not already) new set of pcbs for his Aleph,tnx to extensive tweaking........
After a wonderful baked cajun catfish dinner tonight courtesy or Emeril's wonderful spices we all thought we ought to post some photos to re-inforce whatr Bill WW has been up to.... and I would say he has created one of the very best sounding Alephs that I've ever heard. Incredible presence, incredible soundstage... in both width and depth The JFET input is definately one of the best additions I've heard happen to an Aleph!
Here are some photos taken at 1 watt output, 1 khz into 4 ohms.....
Here are some photos taken at 1 watt output, 1 khz into 4 ohms.....
Attachments
Square wave ringing?
Mark and I worked on the Aleph 3 jfet conversion late in the evening, before I needed to go home.
We found the amp tends to oscillate with any compensation cap. We tried the .001 UF and also 10 pF and 15 pF.
We decided the .001 UF was way too much. The 10 pF seemed the best we had tried, but the channels both seemed somewhat unstable with oscillations.
Any ideas? With no compensation caps you can see the very first photo Mark posted with a ringing peak on the leading edge. When we tried anything else, it seemed to be very unstable.
Any ideas on what to do or check?
This amp seemed very incredible at Mark's tonight, and has excellent potential. I was concerned afterwards about the square wave response at 10Khz and 1Khz with compensation caps seeming to make the amp unstable with too much capacitance.
I do not have any 43 ohm resisters on the LTP. Could that be the cause?
Ground loops?
It seemed strange that oscillations only occured immediately after placing the .001 or 10 or 15 pF caps in the amp.
Regards, Bill
Mark and I worked on the Aleph 3 jfet conversion late in the evening, before I needed to go home.
We found the amp tends to oscillate with any compensation cap. We tried the .001 UF and also 10 pF and 15 pF.
We decided the .001 UF was way too much. The 10 pF seemed the best we had tried, but the channels both seemed somewhat unstable with oscillations.
Any ideas? With no compensation caps you can see the very first photo Mark posted with a ringing peak on the leading edge. When we tried anything else, it seemed to be very unstable.
Any ideas on what to do or check?
This amp seemed very incredible at Mark's tonight, and has excellent potential. I was concerned afterwards about the square wave response at 10Khz and 1Khz with compensation caps seeming to make the amp unstable with too much capacitance.
I do not have any 43 ohm resisters on the LTP. Could that be the cause?
Ground loops?
It seemed strange that oscillations only occured immediately after placing the .001 or 10 or 15 pF caps in the amp.
Regards, Bill
jfet j176 as current source
Would any Engineers have more information on if this is okay.
While adding the LEDs in place of the zener 9.1 volt regulator, I used two 2.6 Volt 10 Mcd 28 mA red LEDs in series. This rougly regulated at about 3.6 volts. How do I tell what voltage these jfets should be running at for the current source regulation? I am using the j176 interfet brand, I believe from www.allelectronics.com.
I initially tried one 3.0 volt 15 mA LED, but found it regulated at 1.9 volts, but then the DC offset was too high to compensate. So, then I raised it with the two 2.6 volt LEDs in series and this made the DC offset within range to adjust. It seems like the left channel is more touchy then the right on regulation. I have found by adjusting the 4.75K ohm resister on the bottom of the former Zener voltage reference (Aleph 3 schematic), I can vary the amount of voltage reference.
Perhaps I should adjust it to 3.5 volts instead of 3.6 to make it a little bit more stable on regulation? While at Mark's adjusting my DC offset, on the left channel, I noticed the LED voltage reference seemed to dim and become unstable so perhaps it would be better to lower the 3.6 volt reference down a bit to make sure it keeps regulating?
Bill
Would any Engineers have more information on if this is okay.
While adding the LEDs in place of the zener 9.1 volt regulator, I used two 2.6 Volt 10 Mcd 28 mA red LEDs in series. This rougly regulated at about 3.6 volts. How do I tell what voltage these jfets should be running at for the current source regulation? I am using the j176 interfet brand, I believe from www.allelectronics.com.
I initially tried one 3.0 volt 15 mA LED, but found it regulated at 1.9 volts, but then the DC offset was too high to compensate. So, then I raised it with the two 2.6 volt LEDs in series and this made the DC offset within range to adjust. It seems like the left channel is more touchy then the right on regulation. I have found by adjusting the 4.75K ohm resister on the bottom of the former Zener voltage reference (Aleph 3 schematic), I can vary the amount of voltage reference.
Perhaps I should adjust it to 3.5 volts instead of 3.6 to make it a little bit more stable on regulation? While at Mark's adjusting my DC offset, on the left channel, I noticed the LED voltage reference seemed to dim and become unstable so perhaps it would be better to lower the 3.6 volt reference down a bit to make sure it keeps regulating?
Bill
Mark A. Gulbrandsen said:
Mark forgot to add the bass of my Aleph 3 sounded much lower than the Aleph 30 perhaps due to my heavy biasing? Still not as great as the Krell KST-100 or KSA-50 in my opinion. It still blended well with the Aleph though and you would not notice the bass unless directly comparing as we had done.
Normally, Mark will rotate amps every 10 minutes, but this time we just kept listening to the Aleph jfet conversion for 90 percent of the day. We hardly got around to comparing any other amps for some reason. Lyndon and Mark just kept wanting to keep listening to my Aleph 3 jfet conversion, while mumbling incredible, unbelievable, simply amazing etc..., I could quite frequently hear being repeated over and over again.
Bill
choky said:
Mark-I'll not bother with that "led longevity "issue......world is full with CCSes with led as reference,even in medical instruments....
if led based CCS is good enough for Bill ,I don't see any further reason for concern.......except that he will need soon (if not already) new set of pcbs for his Aleph,tnx to extensive tweaking........
Actually Mark and I commented while testing the Aleph that my circuit board no longer looked like a printed circuit board, but more like a point to point wiring board!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.