Partial listening update, partial because one of the speaker terminals I used in the amp was drilled out too big for a banana plug. Damn annoying to find that.
Anyway, I had a short listen to one speaker, and initial impression is very good - absolutely dead silent. Once I get some more terminals I'll give it a propper session.
I was also trying out a Raspberry Pi with combined iqaudio DAC+ music server/player. Had a mild high pitch buzz whenever I played music - only noticeable just as you started the music (in the silent part). Not there when you stopped the music, so no idea what that is.
Anyway, I had a short listen to one speaker, and initial impression is very good - absolutely dead silent. Once I get some more terminals I'll give it a propper session.
I was also trying out a Raspberry Pi with combined iqaudio DAC+ music server/player. Had a mild high pitch buzz whenever I played music - only noticeable just as you started the music (in the silent part). Not there when you stopped the music, so no idea what that is.
DOH!
I'm glad you like it so far.
If the buzz isn't there when the music is stopped, I would look from the media file to the output of the DAC. There's no practical difference for a DAC in producing 0 V or a music signal, so the analog part of the circuit appears to be working.
Tom
Thanks Tom.
Media files are ok as I've ripped them to FLAC and replayed a few through my PC with no buzz. So it's got to be the DAC, or possibly the transport of the file from the USB attached drive to the server/DAC. I've read that some powered USB hubs (which I'm using for the hard disk), can cause issues with the Pi - it might be that (the quality of the hub is pretty poor).
Media files are ok as I've ripped them to FLAC and replayed a few through my PC with no buzz. So it's got to be the DAC, or possibly the transport of the file from the USB attached drive to the server/DAC. I've read that some powered USB hubs (which I'm using for the hard disk), can cause issues with the Pi - it might be that (the quality of the hub is pretty poor).
Well. The writing was on the wall and the time had to come eventually. Yet, it's still a bit bittersweet for me as I shipped the last Parallel-86 boards this afternoon. The Parallel-86 project is now officially in support mode as TI has discontinued the LM4780 and supplies in the northern hemisphere are dwindling.
I will move the circuit description page to the Support section of my website shortly. Please find it there, should you need to reference it.
A heartfelt thank you to everybody who supported the project, in particular to those who bought Parallel-86 boards.
Those of you who need LM4780 ICs: I have ten left in stock. Fresh from Newark (when they had stock). Toss me an email or PM if you're interested in buying.
I hear RS Components in New Zealand still have them in stock, so if you're down under, that may be an option for you.
Looking forward: The Modulus-86 is still going strong and I have every intention to keep it that way. As mentioned previously, I've switched the project BOM to include the TO-99 can package LME49710HA, so that project is full speed ahead.
Thanks,
Tom
I will move the circuit description page to the Support section of my website shortly. Please find it there, should you need to reference it.
A heartfelt thank you to everybody who supported the project, in particular to those who bought Parallel-86 boards.
Those of you who need LM4780 ICs: I have ten left in stock. Fresh from Newark (when they had stock). Toss me an email or PM if you're interested in buying.
I hear RS Components in New Zealand still have them in stock, so if you're down under, that may be an option for you.
Looking forward: The Modulus-86 is still going strong and I have every intention to keep it that way. As mentioned previously, I've switched the project BOM to include the TO-99 can package LME49710HA, so that project is full speed ahead.
Thanks,
Tom
Oh, I'm sure I can come up with something for you. It may be out a bit, but I'll figure something out. I do need something to satisfy the need moar powah! crowd.
2x25 VAC will give you some ±35-36 V rectified. You can power a Modulus-86 with that. I can't guarantee more than ±28 V swing into a 4 Ω load, as the LM3886 isn't spec'ed for more than 7 A of output current, but if you aren't planning to run a 4 Ω load all the way to the rails, it is an option.
I generally recommend a ±28 V supply with the Modulus-86 as it allows the LM3886 to stay within its guaranteed limits rather than its typical spec. ±28 V will also result in 40 W dissipated in the LM3886 under worst case conditions with a 4 Ω load. 40 W is a good design target for that package.
If it's any consolation, I still have four really nice 530 VA 2x32 VAC toroids that I acquired to use with a Doug Self Blameless amp back before Mr. Self collected his Wireless World articles into a book. I had all the parts, PCBs made, ready to go. Just never assembled the bits. Such is life at times.
Now, some 18+ years later (wow! time flies), I think I may finally have a use for those transformers. See my comment about moar powa above.I have the opportunity to borrow a pair of SL Orion on an extended loan. I see many amp channels in my future...
Tom
Last edited:
Transconductance
I'm waiting (but not holding my breath) for Tom to do a transconductance amp version.
I really like his Modulus-86 but I think a transconductance amp might be the dog's bollix for my TAD2001 mid driver.
Mod-86 on the tweeter, bridged parallel-86 for the woofer, TC or SET for the mids.
Come on Tom I know you really want to do a TC amp.
Thanks for all the support
tim
I'm waiting (but not holding my breath) for Tom to do a transconductance amp version.
I really like his Modulus-86 but I think a transconductance amp might be the dog's bollix for my TAD2001 mid driver.
Mod-86 on the tweeter, bridged parallel-86 for the woofer, TC or SET for the mids.
Come on Tom I know you really want to do a TC amp.
Thanks for all the support
tim
Erm why? Whilst Hawksford did show some benefits to current drive, its a difficult thing to do right and you can already build a Pass F1 or Myref.
I'm glad you're not holding your breath. I prefer to not be responsible for your death.
I don't see any technical advantage at all of current drive. In theory, there is no difference between driving a speaker with an ideal voltage source or an ideal current source. The Modulus-86 is pretty darn close to an ideal voltage source.
If there is a compelling technical reason for why I should take up such a design and this reason can support a good business case, I'm all ears. However, as Bill hints to above, do note that I need to produce a circuit which is stable under a wide variety of load conditions and is reproducible by even relative beginners. I tend to take a mass production mindset when designing and testing circuits.
Tom
Last edited:
Erm why?
Erm why?
Because I want to, Bill--you don't always need to be a dick.
There is plenty written about the possible pros and cons of current drive as you all know.
The clincher for me is that Jean Michel Le Cleach advocated and used current amps on his compression drivers and if it was good enough for him to use on his TAD2001s then its good enough for me to try on mine.
It may well be difficult (though I don't think the Pass amp is too complicated or too unstable) but that then would be an ideal project for Tom.
tim
Erm why?
Because I want to, Bill--you don't always need to be a dick.
There is plenty written about the possible pros and cons of current drive as you all know.
The clincher for me is that Jean Michel Le Cleach advocated and used current amps on his compression drivers and if it was good enough for him to use on his TAD2001s then its good enough for me to try on mine.
It may well be difficult (though I don't think the Pass amp is too complicated or too unstable) but that then would be an ideal project for Tom.
tim
I think the potential for current amps have not been thoroughly explored, so development of such takes quite a bit of effort. You really have to be in the "explore the unknown" mode rather than "use existing technology". To only the people dedicated to improve total audio playback quality will step into this part. It is not an easy task and not for the faint hearted or people that need to make a living first.
The MyRef seems to work best when the current/voltage gain matches the voltage/voltage gain. So this means you have a narrow load impedance range for optimum performance. If development is for a more general audience, then you may choose to go for a pure current mode, but then it is not easy to get right, and there are different safety considerations when you are driving the amplifier close to the limits.
The MyRef seems to work best when the current/voltage gain matches the voltage/voltage gain. So this means you have a narrow load impedance range for optimum performance. If development is for a more general audience, then you may choose to go for a pure current mode, but then it is not easy to get right, and there are different safety considerations when you are driving the amplifier close to the limits.
Last edited:
YOU want a transconductance amp. There are at least 3 you can build off designs on this forum. Why should Tom spend a few weeks of his time just to satisfy a single request? Current mode is a niche backwater well served by existing designs. So I repeat the question. WHY do you want Tom to make this amplifier instead of building a Myref or Pass?
I was not being a 'dick', but thank you for bringing insults onto this thread. Kiwis are normally nice and laid back, you are clearly the exception.
The theory I've looked at shows there is, but perhaps I'm out of date.
With an ideal current source drive the amp's output impedance is infinite. So no damping is applied to the drive unit by the amp. With voltage source drive there is damping as the amp's Zout is very small (dominated by cables in the Mod-86 case). In practice this manifests as a change to the LF response.
Happy to be shown what I'm missing.
Try comparing ideal current source to ideal voltage source for a start, or real world to real world.
Note: This is a build thread for a voltage source amplifier. Lets talk about that and leave current source theory to other threads that actually want to discuss it.
I'm not against the concept, however, as others have pointed out, there are significant risks and design challenges of a high-power current source design. I would like to make sure that the reward is worth the risk.
Bill's estimate of "a few weeks" of design time is optimistic. The Modulus-86 Rev. 1.0 took about 6-7 months to design. Rev. 2.0 added another 3-4 months. Granted, I was working full time at the time, but still.
If you have some references, which state that current sources are objectively better, I'm all ears. Feel free to send them to me via PM so we don't plug up this thread. As I said, if there's a business case...
Regarding current vs voltage. Unless I've misunderstood something fundamental about Ohm's Law and Norton and Thevenin equivalents, current drive and voltage drive should yield identical results.
Tom