Rainy Sunday afternoon was devoted to a careful evaluation of AN capabilities using several DACs, all my 8ohm speakers sets and variety of recordings, old and new, including hi-res FLACs and even a few junk 128kbps MP3s. Well, dynamics is there and addictive, you want to live it on and listen more. I do not own any hi-end amp but as far as I can tell based my other systems, AN lower spectrum is definitely present but does not obscure vocals or high percussions. It makes my other amps sound a tiny bit flatter.
Conclusion:
me: goosebumps.
wifey: 'sounds pretty good for a stove'
.
Jacques
Conclusion:
me: goosebumps.
wifey: 'sounds pretty good for a stove'
.Jacques
Nice builds of Jacques and Minek !
Minek's is really a great "out of the box" concept
I'm still listening to my Alpha20, now playing: Julia Holter - Have You In My Wilderness (24b 96k), sounds fabulous.
Only problem is that I need to take the Alpha20 apart to install the Nirvana PCBs, but it sounds so good ...
Minek's is really a great "out of the box" concept
I'm still listening to my Alpha20, now playing: Julia Holter - Have You In My Wilderness (24b 96k), sounds fabulous.
Only problem is that I need to take the Alpha20 apart to install the Nirvana PCBs, but it sounds so good ...
Nice builds of Jacques and Minek !
Minek's is really a great "out of the box" concept
I'm still listening to my Alpha20, now playing: Julia Holter - Have You In My Wilderness (24b 96k), sounds fabulous.
Only problem is that I need to take the Alpha20 apart to install the Nirvana PCBs, but it sounds so good ...
I agree Danny, the box isn't what is used to be.
Well, today I finished all the wiring up and fired up the first AN 4R. No magic smoke! 😊 Yay!
One heatsink is running slightly hotter than the other – is that a problem? Also the 2x 8D-20s (one on the primary of each traffo) run very hot – too hot to pinch between finger and thumb, for long!
Given the comment someone made about shiny aluminium giving a too-high reading and a black surface is best, I have shone my IR heat gun down between the CPU cooler’s heatsink and the aluminium bracket it’s bolted to – so the ‘red spot’ hits the side of the JFET nearest the top of the bracket. This is the P-channel MOSFET – temp reads 94 deg on one side and 89 deg on the other – this is after it’s been on for about an hour.
Is this a problem … do I need to turn the fan up? (At the moment, I am using the ‘noise-reducing cable’ extension which Noctua told me incorporates a resistor to reduce the fan speed to 1700rpm (instead of 2200rpm).)
Note: I've just shone the 'red spot' on the top edge of each of the SLB transistors. The readings came up similar - in fact a few degs hotter than the MOSFETs - yet I can touch these transistors and the top surface is only just warm! So maybe my IR temp device is not actually reading the temperatures correctly?
DC rails on both channels are the same: +/- 21.7v.
DC offset (with the present setting of the Bourns pots) is:
* -4mV on one channel, and
* -2mV on the other.
So that’s great! 😊
Now I need to:
1. switch off and remove the Bourns pot in the DC offset circuit.
2. measure them.
3. replace them with a fixed resistor, and
4. lacquer the underside of the two PCBs.
Can someone tell me how I measure the ripple on the DC supplies? I'm hoping that, with a CRO, I can foresee whether or not I'm going to have any hum, before I put it into my system.
As you might be able to see from the attached pic, the mains leads to the traffo primaries come down the middle of the case from the back to the front - between the 2x SLB PCBs. I'm worried these leads might induce hum in the DC rails - if they do then I would have to run each mains lead inside an earthed shield.
Andy
One heatsink is running slightly hotter than the other – is that a problem? Also the 2x 8D-20s (one on the primary of each traffo) run very hot – too hot to pinch between finger and thumb, for long!
Given the comment someone made about shiny aluminium giving a too-high reading and a black surface is best, I have shone my IR heat gun down between the CPU cooler’s heatsink and the aluminium bracket it’s bolted to – so the ‘red spot’ hits the side of the JFET nearest the top of the bracket. This is the P-channel MOSFET – temp reads 94 deg on one side and 89 deg on the other – this is after it’s been on for about an hour.
Is this a problem … do I need to turn the fan up? (At the moment, I am using the ‘noise-reducing cable’ extension which Noctua told me incorporates a resistor to reduce the fan speed to 1700rpm (instead of 2200rpm).)
Note: I've just shone the 'red spot' on the top edge of each of the SLB transistors. The readings came up similar - in fact a few degs hotter than the MOSFETs - yet I can touch these transistors and the top surface is only just warm! So maybe my IR temp device is not actually reading the temperatures correctly?
DC rails on both channels are the same: +/- 21.7v.
DC offset (with the present setting of the Bourns pots) is:
* -4mV on one channel, and
* -2mV on the other.
So that’s great! 😊
Now I need to:
1. switch off and remove the Bourns pot in the DC offset circuit.
2. measure them.
3. replace them with a fixed resistor, and
4. lacquer the underside of the two PCBs.
Can someone tell me how I measure the ripple on the DC supplies? I'm hoping that, with a CRO, I can foresee whether or not I'm going to have any hum, before I put it into my system.
As you might be able to see from the attached pic, the mains leads to the traffo primaries come down the middle of the case from the back to the front - between the 2x SLB PCBs. I'm worried these leads might induce hum in the DC rails - if they do then I would have to run each mains lead inside an earthed shield.
Andy
Attachments
Hi Andy,
Looks great.
That's one of the best solution I've seen for active cooling in a case,
it's almost like a vertical cooling tunnel through the case.
Yes Hugh,
when the recording is good, those high res formats (24b) have something extra, it sounds more natural, effortless on my Soekris DAC.
Looks great.
That's one of the best solution I've seen for active cooling in a case,
it's almost like a vertical cooling tunnel through the case.
Yes Hugh,
when the recording is good, those high res formats (24b) have something extra, it sounds more natural, effortless on my Soekris DAC.
Hi Andy,
Wonderful news that your powered up and so close to hearing music from your AN!! It’s been a long road, but you’ve got a slick, stealth Class A amp in your possession.
Congratulations!!
Time to Enjoy the music!!
Ps, be careful not to stall the fans with too much resistance slowing them.
Wonderful news that your powered up and so close to hearing music from your AN!! It’s been a long road, but you’ve got a slick, stealth Class A amp in your possession.
Congratulations!!
Time to Enjoy the music!!
Ps, be careful not to stall the fans with too much resistance slowing them.
Thank you V - but I have to acknowledge your - and other's - help on this thread to get me to where I am.
Time to Enjoy the music!!
Indeed - very soon!
Yes, sure. From the measurements I've done with my 'fan voltage-dropper board', to start with, I will use a 120 ohm res in the winter ... and maybe only a 100 ohm res in the summer.
I'll see how the temps go for a year - and the fan noise - and then maybe increase the res value, for lower noise.
Andy
Thanks, Danny!
Yes, it is a 'vertical cooling tunnel through the case ' as the top of the case is Modushop's fully vented cover.
It would appear that minek's suggestion that "trying to use 2 cpu coolers and hoping to dissipate 4*44W and staying under 90 degrees is very risky proposal " is not in fact true - as, theoretically, I am dissipating 3a x 43v = 130w with each CPU cooler. (Hugh has told me what measurements I need to do, to get an actual figure for heat dissipation).
Andy
to Andyr,
You asked about how to measure the SLB ripple. You are correct that the only valid way is the use of an oscilloscope. And it needs to be done with a load on the SLB comparable or somewhat higher than you expect in normal use. You just want to make sure you don't see what I can best describe as little blips every half power cycle. The ripple waveform before the series pass transistor in the SLB will look like a sawtooth with the tips slightly rounded off. You just want to make sure the low point of the waveform doesn't show up at the output of the SLB.
You asked about how to measure the SLB ripple. You are correct that the only valid way is the use of an oscilloscope. And it needs to be done with a load on the SLB comparable or somewhat higher than you expect in normal use. You just want to make sure you don't see what I can best describe as little blips every half power cycle. The ripple waveform before the series pass transistor in the SLB will look like a sawtooth with the tips slightly rounded off. You just want to make sure the low point of the waveform doesn't show up at the output of the SLB.
Thank you, JH - yes, I certainly do want to know how to measure the SLB ripple.
Surely, the load on the SLB is constant - either current is going to the speakers ... or it's going into the heatsink?
Or are you saying that - given I've built the 4R AN - more current (than the nominal bias of 3a) gets sucked from the SLB if the spkr load is only 2R, rather than 4R?
Sorry, you've lost me here.
I just remember the CRO screen shot that X posted at the beginning of the thread ... it appeared to show the ripple as extremely low (like 1 or 2 mV).
Do I put the +ve lead of my CRO on the +DC rail output of the SLB and the ground lead on the 0v ... and see what the CRO screen shows?
Also, if ripple is low on the CRO screen - does this mean that I will hear no hum from the amp, when I put it into my system?
Andy
Hi again Andy,
Measuring the ripple when your Nirvana is loading the SLB is sufficient.
Yes, the scope probe connects between the output rail and 0V. AC coupled.
If the observed ripple is low then you won't have hum coming from the SLB power supply.
What I was trying to describe speaking about the sawtooth waveform is what is present before the series pass transistor in the SLB and that it is effectively eliminated after the series pass if it is working correctly.
Measuring the ripple when your Nirvana is loading the SLB is sufficient.
Yes, the scope probe connects between the output rail and 0V. AC coupled.
If the observed ripple is low then you won't have hum coming from the SLB power supply.
What I was trying to describe speaking about the sawtooth waveform is what is present before the series pass transistor in the SLB and that it is effectively eliminated after the series pass if it is working correctly.
Andy, that is one full case! I'm glad you are so close, I have not done active cooling as you have here, but those temps are very good, make sure you go out of your way to check any and every way to make sure what you are seeing is valid.
I'm not saying those temps aren't valid, just verify. I thought mine where in that range and spoke with some people here which caused me to take another look and found my temps were a lot higher than I thought originally.
Those transformers shouldn't be so hot as to make a finger uncomfortable. Get a heat reading on them too. I have never build one that got more than just warm to the touch.
Oh ya, and be careful with those probes! I'd really hate to see you smoke something being so close!!
I'm not saying those temps aren't valid, just verify. I thought mine where in that range and spoke with some people here which caused me to take another look and found my temps were a lot higher than I thought originally.
Those transformers shouldn't be so hot as to make a finger uncomfortable. Get a heat reading on them too. I have never build one that got more than just warm to the touch.
Oh ya, and be careful with those probes! I'd really hate to see you smoke something being so close!!
Last edited:
Very nice work, Andy! Congrats! 
Your temperatures are rather low but maybe it’s not registering on the correct spot or proper emissivity surface. What is your current flowing through the source resistors?
I use Kapton tape over the aluminum to get a good temp.
Truth be told, white paper masking tape is just as good with an emissivity if 0.90 also. It has the problem of leaving nasty tape residue. Black vinyl electrical tape works too. Also nasty residue.
Liquid white out correction fluid is also excellent.
Take a temp on the aluminum surface with any one of those applied and report back what you get.
My MOSFET clamp bar temps are circa 65C to 75C. My heatsink fins on the CPU cooler are 40C to 50C. All greater than 32C.
The NTC-8D operates hot circa 150C so will burn you if you touch it. That’s by design and how it provides a 0.5ohm drop when hot. About 8ohms when cold. It’s bullet proof so don’t worry about damaging it unless you draw more than 7A or so.
Your temperatures are rather low but maybe it’s not registering on the correct spot or proper emissivity surface. What is your current flowing through the source resistors?
I use Kapton tape over the aluminum to get a good temp.
Truth be told, white paper masking tape is just as good with an emissivity if 0.90 also. It has the problem of leaving nasty tape residue. Black vinyl electrical tape works too. Also nasty residue.
Liquid white out correction fluid is also excellent.
Take a temp on the aluminum surface with any one of those applied and report back what you get.
My MOSFET clamp bar temps are circa 65C to 75C. My heatsink fins on the CPU cooler are 40C to 50C. All greater than 32C.
The NTC-8D operates hot circa 150C so will burn you if you touch it. That’s by design and how it provides a 0.5ohm drop when hot. About 8ohms when cold. It’s bullet proof so don’t worry about damaging it unless you draw more than 7A or so.