John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
dvv,
Thanks for the encouragement. Getting great sound is one of my major goals and making a product that looks great and not like everything else with great industrial design is the other goal. You do have to attract the public these days to your product by differentiation. That is something I have learned along the way. When others start to copy your concept then you know you have changed the conversation. I may not agree with all things that Apple does but at the same time others keep trying to copy their concepts, they seem to win most of the time by being the leader driving pure industrial design. We all know the underlying technology isn't all that revolutionary, but they do usually make sure that they get the basic form and function right.

Christophe,
I'll leave the problems of ground planes for the digital and star grounds for the amplifiers and such to the others, I am not qualified to enter that conversation! I actually see some of the most difficult problems as being the power supply or sub-supplies as one of the biggest problems. How you do that and keep it small and then sub divide that power so you have all those different voltages. Perhaps you want 70 volt rails for a power amp, but then you need a 5v source perhaps for a dac or something else. How you do all that cleanly is way over my head when they are all in the same enclosure and within inches of each other. I don't see how you can consider anything but a smps when size is a factor but how that all works I have some small understanding from reading so many posts. it is to me very high level thinking to get that all correct. An external power supply seems to be an easier solution but then you lose the elegance of a single enclosure.

I still remember a system i put in my car many years ago. It was a three way system, two 10" pro-audio Radian bass drivers, two mid-range horn loaded 6 1/2" cones, and two waveguide coupled dome tweeters. This was all in the back shelf under the window. The bass speakers and mid horn were aimed at the window and the tweeters were aim directly at the front seats. There was no other speaker in the car, all behind your head. I had three separate amplifiers and hand made crossovers in the trunk. In the front end I had a Sony unit and a parametric eq from a high end company, not a typical car audio company. Everything sounded great once you got use to the idea the sound was coming from behind you, i couldn't see a great way to have speakers anywhere in the front. to many problems to do that well without cutting up the dash. The only problem I had was a very high frequency problem on the grounds picking up the damned ignition signal from the spark plug wires. The only way to get rid of that noise was to lift some ground wires going to the amps, it worked and since everything was at 12v I wasn't worried about getting shocked from any floating ground. All the low voltage wiring was done with twisted pairs in shielded cabling so i do get the issue with ground loops. This was the extreme in a crazy environment but it does give me an idea of the problems that have to be solved. Getting that correct in a small space with a smps and not having noise problems makes me remember that car audio system. You can't play around with leaving a lifted ground in a 110v consumer audio system so I'll leave that stuff to the experts!
 
How you do all that cleanly is way over my head when they are all in the same enclosure and within inches of each other.
It is not complicated at all to be understood. Sometimes more to find the best solution ;-)
Just consider in your mind each wire or CI track like a resistance (or a more complex RLC circuit at various frequencies).
In a balanced symmetric connection, like we use in professional studios, the audio signal is the difference between the + &- wire. Differential mode. And common mode signals errors (the ground) are rejected.
Not with unbalanced connections.
If you use perfectly ground isolated devices, like when you use separated batteries on each devices, you will have no current across your ground connections. And the reference point (the ground) for your signal will be at the same potential everywhere. It sound often a lot better or cleaner.

When you use AC transformers, some currents flows between AC and the ground, because there is not a perfect isolation between primary and secondary coils in your transfo. C core transfos, or the ones with a shield between primary and secondary coils helps, as well as SMPS with their little transfo.
Using an unique reference point let all those unwanted currents playing as the want, PSU side. But for the signal, you ensure all your devices are referenced to the same voltage ground.
Hoping it helps.
 
dvv,
Thanks for the encouragement. Getting great sound is one of my major goals and making a product that looks great and not like everything else with great industrial design is the other goal. You do have to attract the public these days to your product by differentiation. That is something I have learned along the way. When others start to copy your concept then you know you have changed the conversation. I may not agree with all things that Apple does but at the same time others keep trying to copy their concepts, they seem to win most of the time by being the leader driving pure industrial design. We all know the underlying technology isn't all that revolutionary, but they do usually make sure that they get the basic form and function right.

...

I understand perefctly, Kindhornman, having been there myself. Often enough, something looking good does not have to be more expensive than something else not looking interesting, I think much of it is the mind of the designer. The trick is not to confuse from with function, function still comes first.

I feel sure you realize that and that makes me feel optimistic about the whole project.
 
In a balanced symmetric connection, like we use in professional studios, the audio signal is the difference between the + &- wire. Differential mode. And common mode signals errors (the ground) are rejected.
Not with unbalanced connections.

Transformers can offer excellent common mode rejection even from unbalanced sources. You can achieve it in active form using Bill Whitlock's INGenius devices sold by THAT Corp.

se
 
SE,
I thought you saw my speaker design idea already? I'm going to do another grill as an option for those who don't like the wire form, the same shape with a normal grill cloth. About 10% of the people who have seen this don't like seeing the speakers themselves. I also removed four of the screws around the tweeter, I can't get rid of the four in the center but did the outer four.
 

Attachments

  • black speaker 2.jpg
    black speaker 2.jpg
    80.6 KB · Views: 188
Last edited:
Ah star grounding, different kettle of fish, the topography is very design dependent, star points are used but care has to be taken, placement can be critical (this is on board). A star point for multiple boards is a bit of a no no as the parasitics to the star point and back can add its own problems causing different levels for the reference plane(s).
PE protective earth and mains is where your classical noisy ground loop comes in, so yes I do agree to a certain extent. A big problem with audio is the fact that the signals go from low 20Hz following the path of least resistance to 20KHz where the return current will follow both the path of least resistance but also the paths of least inductance, add some high frequency noise and this will definitely want to follow the path of least inductance so you have a pot pourii of issues.
 
SE,
I thought you saw my speaker design idea already? I'm going to do another grill as an option for those who don't like the wire form, the same shape with a normal grill cloth. About 10% of the people who have seen this don't like seeing the speakers themselves. I also removed four of the screws around the tweeter, I can't get rid of the four in the center but did the outer four.

Nice, my son wants a set.
 
SE,
I thought you saw my speaker design idea already? I'm going to do another grill as an option for those who don't like the wire form, the same shape with a normal grill cloth. About 10% of the people who have seen this don't like seeing the speakers themselves. I also removed four of the screws around the tweeter, I can't get rid of the four in the center but did the outer four.

No, I don't recall seeing that particular design. Very nice! Is that amplifier heatsink I'm seeing on the back?

se
 
A big problem with audio is the fact that the signals go from low 20Hz following the path of least resistance to 20KHz where the return current will follow both the path of least resistance but also the paths of least inductance,
Not a big problem in the audio range. You can hear the evils and cure them. It is at HF that the problem exists, where inductances produce their worst effects and where electronic is not able to deal with them without producing annoying IM.

This said, "both the path of least resistance but also the paths of least inductance" will be the same with star ground between several boards and no other loop.
 
Last edited:
In my view, 0 inductance is a purely academic value, good only for "what if" analysis. It's just as theoretical as regarding loudspekares as pure impedances.

The best one can hope for is what Kindhornman is designing, as an active loudspeaker where the distance needed to connect the power amp output to the loudspeaker driver is like 5 cm, or 2 inches. That's as pure as it gets.
 
diyAudio Member RIP
Joined 2005
dvv,
Yes that will be the final arbitrator is when it is finally powered up. Moat of the time it should just be loafing along, but those times when someone really cranks them up will be the test or I guess the 1/3 power testing.
1/3 power is absurdly brutal, but it's what Atkinson likes to do, so one must be prepared.

There have been a number of products reviewed and tested in SP that failed during the 1/3 power preconditioning---even some touting fancy-tech heatsinks. Some old Harman automotive products had thermal throttleback, but if one wasn't careful sometimes it just made things worse.
 
Status
Not open for further replies.