So this is a finished now and the bias has been set at 100ma.
I took it out of my work room where it has been tested days over and attached to my KEF cresta's.Hard to tell in such a small space how good the sound is, so i took them down stairs and plugged them in via my marantz CD 6000ose and fyne audio 303's, now thats a different kettle of fish.
I tried it against a standard 3020A, and all i will say is that it appears 'clearer', especialy the trebel side of it.
The bass doesnt seem as wooly as the original, so id say a success.
Those FETS get real hot dont they!
The heatsink is about 56c at almost half volume (10 mins)
so we went from this
to this
I took it out of my work room where it has been tested days over and attached to my KEF cresta's.Hard to tell in such a small space how good the sound is, so i took them down stairs and plugged them in via my marantz CD 6000ose and fyne audio 303's, now thats a different kettle of fish.
I tried it against a standard 3020A, and all i will say is that it appears 'clearer', especialy the trebel side of it.
The bass doesnt seem as wooly as the original, so id say a success.
Those FETS get real hot dont they!
The heatsink is about 56c at almost half volume (10 mins)
so we went from this
to this
56C isn't bad and ultimately there isn't really a lot of metal in the heatsink to absorb and radiate the heat. They shouldn't really be any hotter or cooler than a bjt for the same voltage and currents. W = I *V and all that 🙂
0.5A flowing in a transistor that has a 20 volts across it generates 10 watts. The heat produced is the same whether it is a lateral FET or a 2N3055.
100ma bias and lets say a -/+28 volt supply produces 2.8 watt per device. So that's over 11 watts in total. Hotter than you might think perhaps.
The deal is that you could never run those original 2N3055’s with that high a bias at idle with no emitter resistors. They’d probably have to go up to an ohm or more, since EF thermal stability is somewhat dependent on Rth.
Not ofter it would be run at that volume, and here it is in its final place. It's been a good project and it will now replace my current 3240pe amp
Attachments
Yamaha A-700 increases the standard bias from 20mA to 100mA in "auto A-class" mode.
Half volume (pot knob at 12:00) would constitute to 10% of the signal level?
My guess is that the amp is operating in that signal level in A-class mode hence the heat dissipation.
If it gets too hot then you have to reduce the bias or add a fan to increase ventilation.
A few more ideas, practiced on the 302:
https://www.diyaudio.com/community/...ing-the-220-volt-version.368883/#post-7734043
https://www.diyaudio.com/community/...ing-the-220-volt-version.368883/#post-7734043
so back to this temporarily.I was talking to @Mooly about some loud 'hiss' that is coming through the loudspeakers in my lounge when a music load is applied.
They were attached to a set of fyne audio 303's
I took the amp upstairs to have a look with the scope but when i attached them to the speakers in my repair room, a pair of KEF CRESTA-NO NOISE
2 way, 2 driver loudspeaker system
• Frequency Response: 50Hz to 20kHz
• Recommended Amplifier: 10 to 70W
• Crossover Frequency: 3000Hz
• Impedance: 8Ω
• Sensitivity: 88Db
• Bass: 1 x 100mm long throw
• Tweeter: 1 x 25mm silk dome
Just wondering if there is anyone out there who can expain why this would be?? because i am at a loss.
They were attached to a set of fyne audio 303's
| System Type | 2 way, rear ported |
| Recommended amplifier power (Watt RMS) | 35- 150 |
| Continuous power handling (Watt RMS) | 75 |
| Sensitivity (2.83 Volt @ 1m) | 91dB |
| Nominal impedance | 8 Ohm |
| Frequency response (-6dB typical in room) | 32Hz- 28kHz |
| Drive unit complement | 2 x 150mm multi-fibre bass/ mid 1 x 25mm polyester dome tweeter |
| Crossover frequency | 3.2kHz |
I took the amp upstairs to have a look with the scope but when i attached them to the speakers in my repair room, a pair of KEF CRESTA-NO NOISE
2 way, 2 driver loudspeaker system
• Frequency Response: 50Hz to 20kHz
• Recommended Amplifier: 10 to 70W
• Crossover Frequency: 3000Hz
• Impedance: 8Ω
• Sensitivity: 88Db
• Bass: 1 x 100mm long throw
• Tweeter: 1 x 25mm silk dome
Just wondering if there is anyone out there who can expain why this would be?? because i am at a loss.
There have been so many of these (NAD's) I can't just recall exactly where it was all up to but if the noise is load dependent (one speaker type does it and one doesn't) then you wonder over a stability issue. If the hiss is 'loud' then something is definitely going on somewhere.
Try adding a series resistor of a few ohms to the speaker feeds and see if that kills the noise. If it does some stability issue is possible,
Try adding a series resistor of a few ohms to the speaker feeds and see if that kills the noise. If it does some stability issue is possible,
July when the last posts were to now is like a million posts in the past 🙂 It means picking it all up from the beginning.
So I've solved the issue with the hiss,but I'm going to ask why it would be caused by a bias issue. When I removed the driver sections and tested it,I left it around 100ma. When I checked it,the current was around 160ma,which explains why it was hot,so I put it back down to 100ma,but the hiss was still there,so I lowered it to around 70ma,and the hiss has gone. Why would this be?
Hiss can also be associated with an instability (sometimes more than one simultaneously). Sometimes when troubleshooting RF oscillations, the presence of multiple spurs simultaneously results in the entire noise floor coming up. One of them may be out of range of the spec-an (or scope if looking at audio). Quench it and the hiss goes away (or noise floor falls). You could have some well above band local oscillation at the higher bias.
Indeed it can. I know poundy has a decent (digital) scope, 100MHz I think so you think it would winkle something out if there was anything going on.
I did check it out with the scope, but there was no hiss without a musical passage, and a more experienced person could possibly have seen it amongst all of the other noise, so i couldnt see the hiss on the scope, although im sure it was there obviously.
That's where you need a clean test signal to try and see what is happening. The fact signal seems to be a trigger is another tick box checked in the possible oscillation scenario.
Your phantoms may be way above the BW of the scope. If it is oscillating up that high the only clue you’ll often get is the increased noise floor. The likes of Anatech swear by a 300(+) MHz analog scope, and one of the old HP green screen spectrum analyzers that have the band shift at 2.9 GHz. These tools let you get a really good look at the VHF/UHF and L bands. But not everybody has them of course. I’ve had to troubleshoot oscillations that wouldn’t show up on my TDS1012 or the boat anchor 7904. Those LatFets are RF-capable, and it may be as simple as adding RF decoupling to the scheme that’s already in place. Like small ceramic disc capacitors to the back of the board. It may take a combination of class 1 and class 2 caps. On RF EVBs we have to do that, to make the transistors stable enough to test. My issues were with tubes - using IF amplifiers for audio stages. Have to be more careful than with 12AX7’s, but at $3 a pop one can live with the lower mu, higher gm and fT.
May be totally unrelated and not oscillating, but should be explored anyway to rule it out.
May be totally unrelated and not oscillating, but should be explored anyway to rule it out.
I guess the question is then is it worth chasing? The amp appears to works well enough at around 70ma current, without any hiss at all.