I went to see the Terricota soldiers there and ended up seeing a lot more.... got vip treatment with the daughter of the cities chief of police as a guide. She said some sites she showed me and my friend had never been seen by westerners before. A city of 10 million and full of stories and beautiful areas. Saw a lot of people playing pool in their front yards - pool tables in the front yard! You got to go there if you are still in China. I brought art back that are from well known artists but who never sell outside of China. What a great country and history. Oooops. back to electronics....
Say what!? can you give that another go? It looks like English but I cant understand.
I have to disagree with you, John.
If we're drawing parallels between cars and playback systems, this is how typical vinyl PB systems would look like.
Apologies to other forum members for multiple posts and OT pictures!
Attachments
I've heard TO-3's sing when driving a resistive load.
It's the emitter wire bond wire vibrating due to Lorentz forces.
I've also encountered punch through with diodes when a point defect is present on the copper anvils used with 2 and 3 inch diameter diodes, it causes high curretn at the point of pressure.
Also, I've seen emitter dip caused by wafer probing.. generally kills the device during BVceo and BVceo(sus) tests..
But for a bonded to substrate chip, the acceleration needed to modulate any parameters of the chip would flow the metal it's mounted to.
If they cut the can gently, it won't necessarily compromise the chip. It's simple enough to use a good glob top on the chip if it isn't there, or more to encapsulate the bonding wires if the chip is already covered.
If they don't cover the aluminum wires or the chip properly, any reflow process is gonna raise havoc. I don't trust any flux with aluminum or chip level copper.
Glob top is pretty much used a lot of places now, especially overmolded chips.
The high harmonic content does make it hearable. Been there, done that. But it's generally the result of conductor movement doe to Lorentz.
Yah, they cause seal failures. As well as the helium high pressure bomb leak check procedures. They "oilcan" and allow helium through a crack, but when pressure is removed, the crack seals back up.
jn
Don't forget that no small contribution to getting those great Kraut cars and the like designed is because of the trench work to create highly polished turds like this
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
For vinyl if you can find a 12 inch single pressing of a song you like usually you will be astonished at how good it will sound. Many times it's like hearing the tune for the first time. This goes for pre-90's stuff; I don't know what kind of digital madness may have been applied after that.
CD Tweaks
Don't get me started (too late...)
As Dir. of Engineering at a large CD/DVD replication plant, I have talked to many people with 'different' ideas of improving optical disc storage or data recovery. I do forensic data recovery and investigations into data integrity, and use several tools which unequivocally show data equivalence...or not. With over 6 billion pits on a 74-minute audio CD, if you decode the data twice and have zero E32 (uncorrectable) errors, it is almost impossible (1x10^11 for CD-DA, 1x10^17 for CD-ROM chance) that any single recovered bit is wrong. There are MANY, MANY ways (close to an infinite number, practically) to recover different data twice, really only one way to end up with the same data twice. For those nit-pickers who seize upon that microscopic chance of an error being propagated despite the error correction, let me say this: in 20 years of making CDs and DVDs, I have never seen it happen otherwise. If you have other experience, I would love to see it substantiated, otherwise it is mere mental masturbation.
Some of the 'idea' regarding data modification or enhancement include: green marker at the periphery, spraying the surface of the disc with a silicone oil suspension, suspending the disc in a magnetic field, recording pure PCM instead of EFM (Eight-Fourteen Modulation), mastering using an atomic clock instead of a TCXO, aerodynamic 'balance sheets' glued to the disc top, gold instead of aluminum metalize...etc.
I have performed analog pit-track signal analysis, data equivalence and TIA testing on replicated and recorded discs and found NONE of the above tweaks to correlate with improved data recovery with good discs and players. Gold metalize is great for longevity and corrosion resistance in a salt-spray environment, but has lower reflectivity per angstrom than does aluminum at 780nM. If any of these other tweaks actually made a difference, there is a MAJOR problem with the disc, player, or both.
I have had 'creative' interactions with analog audio designers who didn't have a clue as to the data storage methodology of a CD-Audio and wished to store the data directly as PCM, believing the EFM caused audible jitter. When I told them that ALL CD playback systems electronics began with a level-slicer and EFM decoder they told me I didn't know what I was talking about, because another replicator had made those discs and they played fine in their players. Great. So someone else lied (overtly or by omission) to them to get their business, something I could never do. I did not get the business.
The most common misconception I have to deal with when talking to audio engineers is that about jitter. Jitter of a CD pit-track is NEVER directly audible. If severe enough on the disc itself, it only leads to data errors, period. An analog pulse train like that optically recovered from a disc has only the one failure mode: jitter. All optical disc errors are defined as jitter (if you define jitter as the timing of the sliced transitions varying from an integral multiple of 'T', where T is the basic clock interval). A chunk of crap on a disc which momentarily obscures the data track merely causes a REALLLY long T. However, all data from the EFM decoder is then loaded into a buffer with a saw-tooth keep-full variable clock-rate, then error corrected, and then reclocked out to the SPDIF/AES/EBU port or other route to a D>A. If there is jitter in the output bitstream, it originates with the clock in the original A>D or D>A, not the disc.
Sorry to rant; I have had to explain this to (mostly) deaf ears hundreds of times over the last 20 years.
Cheers!
Howie
Howard Hoyt
Dir. of Engineering
AMI, LLC
but because it involves music, more happy in my (basically unpaid) role as:
CE - WXYC-FM
UNC Chapel Hill, NC
www.wxyc.org
1st on the internet
Don't get me started (too late...)
As Dir. of Engineering at a large CD/DVD replication plant, I have talked to many people with 'different' ideas of improving optical disc storage or data recovery. I do forensic data recovery and investigations into data integrity, and use several tools which unequivocally show data equivalence...or not. With over 6 billion pits on a 74-minute audio CD, if you decode the data twice and have zero E32 (uncorrectable) errors, it is almost impossible (1x10^11 for CD-DA, 1x10^17 for CD-ROM chance) that any single recovered bit is wrong. There are MANY, MANY ways (close to an infinite number, practically) to recover different data twice, really only one way to end up with the same data twice. For those nit-pickers who seize upon that microscopic chance of an error being propagated despite the error correction, let me say this: in 20 years of making CDs and DVDs, I have never seen it happen otherwise. If you have other experience, I would love to see it substantiated, otherwise it is mere mental masturbation.
Some of the 'idea' regarding data modification or enhancement include: green marker at the periphery, spraying the surface of the disc with a silicone oil suspension, suspending the disc in a magnetic field, recording pure PCM instead of EFM (Eight-Fourteen Modulation), mastering using an atomic clock instead of a TCXO, aerodynamic 'balance sheets' glued to the disc top, gold instead of aluminum metalize...etc.
I have performed analog pit-track signal analysis, data equivalence and TIA testing on replicated and recorded discs and found NONE of the above tweaks to correlate with improved data recovery with good discs and players. Gold metalize is great for longevity and corrosion resistance in a salt-spray environment, but has lower reflectivity per angstrom than does aluminum at 780nM. If any of these other tweaks actually made a difference, there is a MAJOR problem with the disc, player, or both.
I have had 'creative' interactions with analog audio designers who didn't have a clue as to the data storage methodology of a CD-Audio and wished to store the data directly as PCM, believing the EFM caused audible jitter. When I told them that ALL CD playback systems electronics began with a level-slicer and EFM decoder they told me I didn't know what I was talking about, because another replicator had made those discs and they played fine in their players. Great. So someone else lied (overtly or by omission) to them to get their business, something I could never do. I did not get the business.
The most common misconception I have to deal with when talking to audio engineers is that about jitter. Jitter of a CD pit-track is NEVER directly audible. If severe enough on the disc itself, it only leads to data errors, period. An analog pulse train like that optically recovered from a disc has only the one failure mode: jitter. All optical disc errors are defined as jitter (if you define jitter as the timing of the sliced transitions varying from an integral multiple of 'T', where T is the basic clock interval). A chunk of crap on a disc which momentarily obscures the data track merely causes a REALLLY long T. However, all data from the EFM decoder is then loaded into a buffer with a saw-tooth keep-full variable clock-rate, then error corrected, and then reclocked out to the SPDIF/AES/EBU port or other route to a D>A. If there is jitter in the output bitstream, it originates with the clock in the original A>D or D>A, not the disc.
Sorry to rant; I have had to explain this to (mostly) deaf ears hundreds of times over the last 20 years.
Cheers!
Howie
Howard Hoyt
Dir. of Engineering
AMI, LLC
but because it involves music, more happy in my (basically unpaid) role as:
CE - WXYC-FM
UNC Chapel Hill, NC
www.wxyc.org
1st on the internet
We milled down a dip to the tops of the bond loops and got about 250uV Vos delta on an ordinary JFET op-amp with a titanium needle directly scratched across the surface. The real problem is dynamic/static friction effects from the differential thermal expansion (die/encapsulant). This led to polyaimd overcoat for all JFET products
Cheap R&D, I ground down the package on a grinding wheel, put it in a TEK op-amp curve tracer (remember those?) and called over the chief packaging engineer.
When all else fails back to the locker-room. BTW I think you need a Bentley at your age.
- Status
- Not open for further replies.