As far as I can tell, the OPA1656 is only available as SOT8 SMT. But they have DIP8 adapter boards for them so you can pop them in a socket meant for swapping opamps. Otherwise it will require desoldering and resoldering. Not terribly hard with SOT8. I have 4 of them on the BTSB designed by Jhofland and just really like how neutral and great sounding they are.
OPA1656
Hi xrk971,
Beautiful plate...Thank you for your feedback and advice on availability and package size.
Greetings
Hi xrk971,
Beautiful plate...Thank you for your feedback and advice on availability and package size.
Greetings
X, there's also a group buy being put together for a 4Ux500mm heatsink chassis; plenty of room:
Ultimate 4U 500mm Chassis - Who is interested?
"Just been speaking with Gianluca (Modushop) about doing a 4U 500mm Chassis.
Please let me know who might be interested and we can get this show on the road.
For 10 it will be 310 Euro
For 25 it will be 250 Euro"
Ultimate 4U 500mm Chassis - Who is interested?
"Just been speaking with Gianluca (Modushop) about doing a 4U 500mm Chassis.
Please let me know who might be interested and we can get this show on the road.
For 10 it will be 310 Euro
For 25 it will be 250 Euro"
Nirvana compared to the ABBB
Hi
I’ve been looking into this between other life stuff, and I apologise for having to ask, am still unclear - what would be a summary of the key differences/ advantages/ disadvantages - better for system/ purpose of this board, versus the ABBB Amplifier?
Me, I want to get boards for two stereo amps. not certain how I’ll use them but for me
the difference in maximum output isn’t a factor.
Ease of construction would be a factor
Higher 2HD would be a plus
Off the top of my head (it’s 3 am) I recall only
The ABBB
- can output c 53 watts, v 39
…
…
Thank you
Hi
I’ve been looking into this between other life stuff, and I apologise for having to ask, am still unclear - what would be a summary of the key differences/ advantages/ disadvantages - better for system/ purpose of this board, versus the ABBB Amplifier?
Me, I want to get boards for two stereo amps. not certain how I’ll use them but for me
the difference in maximum output isn’t a factor.
Ease of construction would be a factor
Higher 2HD would be a plus
Off the top of my head (it’s 3 am) I recall only
The ABBB
- can output c 53 watts, v 39
…
…
Thank you
otto88,
Measurements and impressions from builders indicate that they are both fine sounding amplifiers.
The Alpha Nirvana would be an easier amp to construct. You can run the AN with passive cooling, and for the ABBB you will need active cooling. The power draw for the AN would also be less.
Measurements and impressions from builders indicate that they are both fine sounding amplifiers.
The Alpha Nirvana would be an easier amp to construct. You can run the AN with passive cooling, and for the ABBB you will need active cooling. The power draw for the AN would also be less.
Thanks Zman!
Otto,
The BB Alpha is a higher output Class A but not so efficient and creates a lot of heat, almost 200W at idle from EACH channel.
The Alpha Nirvana is a refined version of the BB circuit, and uses tricks to reduce energy waste. Whilst doing 39W into 8R, not much less than the 53W BB, it dissipates less than half the heat, only 90W each channel.
I can't emphasise how difficult it is to dissipate high levels of heat from Class A amplifiers. You can dissipate 90W from a large passive heatsink; but 200W is not possible without forced cooling. Both amps have very good pcbs, laid out by an Avionics engineer, but my preference (as the designer of the amp) is the Alpha Nirvana.
It delivers the expected Class A sound, and with 39W you can generally use conventional speakers for high sound levels for listening. With more than 92dB/watt/metre 39W is deafening in even a large room.
Cheers,
Hugh
Otto,
The BB Alpha is a higher output Class A but not so efficient and creates a lot of heat, almost 200W at idle from EACH channel.
The Alpha Nirvana is a refined version of the BB circuit, and uses tricks to reduce energy waste. Whilst doing 39W into 8R, not much less than the 53W BB, it dissipates less than half the heat, only 90W each channel.
I can't emphasise how difficult it is to dissipate high levels of heat from Class A amplifiers. You can dissipate 90W from a large passive heatsink; but 200W is not possible without forced cooling. Both amps have very good pcbs, laid out by an Avionics engineer, but my preference (as the designer of the amp) is the Alpha Nirvana.
It delivers the expected Class A sound, and with 39W you can generally use conventional speakers for high sound levels for listening. With more than 92dB/watt/metre 39W is deafening in even a large room.
Cheers,
Hugh
Thank you, Hugh.
Certainly - other things being equal - I’m attracted to the much more energy efficient Nirvana. Otoh, not too far into my intended long-term ownership of either, I expect to have solar on the roof, and a good battery system …
Yes, needing a fan is not ideal, but in the whole picture it’s not that much of an extra complexity.
And yes, an extra 35% watts is almost immaterial (~ just + 1 dB?)
But what is the difference in the main reason we go for any class A? - the 2HD and 3HD
I’d previously noted the THD measurements on the original and early iterations of the (then called Aksa Lender P-mos Hybrid Aleph!) ALPHA
THD
Original .027 typical Aksa, mostly 2nd, 3HD 10x less, not much else
Feb 11 .0069 1.31 amps
Mar 5: .006 2.75 vrms premium components. 13% less
What are the numbers for the earlier BB Alpha, for which boards are available
Much appreciated
Cheers
Last edited:
otto88,
The following post has a measurement for the ALPHA Big Boy with Buttah (Buttah refers to an onboard SLB power supply):
https://www.diyaudio.com/forums/gro...buttah-abbb-52w-class-amp-gb.html#post5821338
The following post has a measurement for the ALPHA Big Boy with Buttah (Buttah refers to an onboard SLB power supply):
https://www.diyaudio.com/forums/gro...buttah-abbb-52w-class-amp-gb.html#post5821338
Hi X,
With assistance from Hugh over the last couple of months, it seems I have fixed the problem I had, of R18 on an SLB smoking. 🙂
It would seem the AN 4R circuit was oscillating; changing the compensation cap (C4/C114) from 15pF to 22pF ... fixed the problem!
Then, I discovered I had another problem. 🙁 I had mounted my AN 4R boards on the back panel and I found:
* when the back panel was laying away from the ends of case sides ... I got a sine wave out, when feeding 1kHz in, from my sig-gen.
* but when I pushed the back panel closed - which meant that many wires between the CPU cooler brackets and the AN boards were pushed together ... I got a square wave on the CRO and heard a buzzing. (It was this square wave which blew my Maggie ribbons a few months ago.)
On advice from Hugh, I disconnected the 6 'chassis earth' wires which go from each board to the chassis earth bolt. In other words:
* 1 from each SLB board
* 1 from each AN board, and
* 1 from each SLB heatsink.
Having done this ... switching on ... I saw sine waves on my CRO screen and heard no buzzing! 🙂
So I'm wondering what function the earth wires which go from the 'chassis earth' spade on each AN board to the chassis earth bolt, do? What will be the problem if I remove these wires? (Of course, I still have the earth pin on the IEC input socket connected to chassis earth!)
Thanks
Andy
With assistance from Hugh over the last couple of months, it seems I have fixed the problem I had, of R18 on an SLB smoking. 🙂
It would seem the AN 4R circuit was oscillating; changing the compensation cap (C4/C114) from 15pF to 22pF ... fixed the problem!
Then, I discovered I had another problem. 🙁 I had mounted my AN 4R boards on the back panel and I found:
* when the back panel was laying away from the ends of case sides ... I got a sine wave out, when feeding 1kHz in, from my sig-gen.
* but when I pushed the back panel closed - which meant that many wires between the CPU cooler brackets and the AN boards were pushed together ... I got a square wave on the CRO and heard a buzzing. (It was this square wave which blew my Maggie ribbons a few months ago.)
On advice from Hugh, I disconnected the 6 'chassis earth' wires which go from each board to the chassis earth bolt. In other words:
* 1 from each SLB board
* 1 from each AN board, and
* 1 from each SLB heatsink.
Having done this ... switching on ... I saw sine waves on my CRO screen and heard no buzzing! 🙂
So I'm wondering what function the earth wires which go from the 'chassis earth' spade on each AN board to the chassis earth bolt, do? What will be the problem if I remove these wires? (Of course, I still have the earth pin on the IEC input socket connected to chassis earth!)
Thanks
Andy
>but when I pushed the back panel closed - which meant that many wires between the CPU cooler brackets and the AN boards were pushed together ...
>I got a square wave on the CRO and heard a buzzing. (It was this square wave which blew my Maggie ribbons a few months ago.)
This idea of connecting Mosfets to the PCB with long wires never sounded good to me... Mosfets are fast devices, and they like to oscillate...
>I got a square wave on the CRO and heard a buzzing. (It was this square wave which blew my Maggie ribbons a few months ago.)
This idea of connecting Mosfets to the PCB with long wires never sounded good to me... Mosfets are fast devices, and they like to oscillate...
The oscillation was not from the wires but fact that Andy’s amp was the first of an untested 4 ohm variant. So with usual debugging tips with Hugh’s assistance this was fixed. I think the oscillation would have happened even with locally hard mounted MOSFETs. The gate stopper and snubber RC circuits are mounted at the MOSFET pins via a small helper PCB called the “snubber board”.
It sounds like you have a mechanical issue. Probably something is not well insulated (bare wire touching something) when pressed together due to lack of space. This is not a good sign even though you have it working with the protective earth grounds disconnected, there may be something else that might short at some other point. Murphy’s Law and all...
The PE ground connections are there for safety in case of a catastrophic short in the transformer and shunts the short to ground to prevent fire/electrocution. The PE grounds also reduce RF/EMI noise.
With that mindset, I guess you'll never have the opportunity to listen to an AN. 🙂
As X said ... 'The oscillation was not from the wires but from the fact that Andy’s amp was the first 4 ohm build'.
The 4R cct needed a cap value change (from the simulated value) to achieve stability in the real world - very simple (and part of the new-amp development process).
Andy
>I guess you'll never have the opportunity to listen to an AN
I guess you are wrong. I did build one, check this thread while ago 🙂
The one on big aluminum slab with CPU fans exposed.
Of course I did not use long wires, and everything worked perfectly without any troubles.
And of course my build was based on slightly modified 4 Ohm version, not 8 Ohm.
C4 at 15pF worked just fine.
>As X said [..]
If X says so, it must be truth, and nothing but the truth.
My bet is 'solar radiation' caused this.. Or 5G networks 🙂
I guess you are wrong. I did build one, check this thread while ago 🙂
The one on big aluminum slab with CPU fans exposed.
Of course I did not use long wires, and everything worked perfectly without any troubles.
And of course my build was based on slightly modified 4 Ohm version, not 8 Ohm.
C4 at 15pF worked just fine.
>As X said [..]
If X says so, it must be truth, and nothing but the truth.
My bet is 'solar radiation' caused this.. Or 5G networks 🙂
Last edited:
Aah, apologies for forgetting your post. 😱
And my 4R version works fine when the back panel is laying backwards (not screwed shut). Which suggests that having remote mosfets is not, per se, a problem - just as all the other AN builders have found.
Andy
As I said - most of oscillation can be fixed, even these caused by long wires.
But the problem needs to be diagnosed, and the solution will come at the cost of the speed of the amp.
Higher compensation (E.g. C14) - lower Slew Rate of the amp.
So far nobody even asked what is AN's slew rate, so I guess it won't matter..
I'm sure it's technically possible to build an amp with output devices connected via 10 foot wires.
Does anybody do it?
I did built some amps, with devices connected by wires, and I KNOW that longer wires are more likely to oscillate than
short wires. That's why X put snubbers on these little boards.
But they only work for a given frequency range, it's not a silver bullet...
But the problem needs to be diagnosed, and the solution will come at the cost of the speed of the amp.
Higher compensation (E.g. C14) - lower Slew Rate of the amp.
So far nobody even asked what is AN's slew rate, so I guess it won't matter..
I'm sure it's technically possible to build an amp with output devices connected via 10 foot wires.
Does anybody do it?
I did built some amps, with devices connected by wires, and I KNOW that longer wires are more likely to oscillate than
short wires. That's why X put snubbers on these little boards.
But they only work for a given frequency range, it's not a silver bullet...
Last edited:
Elvee was recommending them, but I was never able to find out how to select them, or calculate required params....