TotalDAC USB cable/filter - Teardown

[mansr]

Did a reverse image search on Google. Nothing. But I am pretty sure it's a:

https://www.semicon.sanken-ele.co.jp/sk_content/sjpz-e33_ds_en.pdf

Whatever it is, it's not that exact part. Neither the package size nor the markings agree with that datasheet. The logo is that of Onsemi, but I can't easily find anything in their catalogue with those markings, and they don't seem to have a reverse lookup tool like some manufacturers do. It's possible that this is old stock, and Onsemi have changed their marking scheme. Regardless, it's safe to say it's a diode of some kind.

 

[solderdude]

After a bit more digging...
This part makes the most sense:

Onsemiconductor (Fairchild) SMBJ5V0(C)A with marking KE in SMB/DO-214AA housing (no longer produced part)
It is a TVS intended for 5.0V lines with breakdown between 6.4V and 7.0V and is 600W peak (100A) on 10/1000 μs Waveform.

 

[jipihorn]

After a bit more digging...
This part makes the most sense:

Onsemiconductor (Fairchild) SMBJ5V0(C)A with marking KE in SMB/DO-214AA housing (no longer produced part)
It is a TVS intended for 5.0V lines with breakdown between 6.4V and 7.0V and is 600W peak (100A) on 10/1000 μs Waveform.

THAT makes more sense !

J.

 

[mansr]

After a bit more digging...
This part makes the most sense:

Onsemiconductor (Fairchild) SMBJ5V0(C)A with marking KE in SMB/DO-214AA housing (no longer produced part)
It is a TVS intended for 5.0V lines with breakdown between 6.4V and 7.0V and is 600W peak (100A) on 10/1000 μs Waveform.

How did you find that? It does seem to be the right part.

 

[solderdude]

Looked for 'TVS' in the Farnell parts selector.
Typed in 33V but turned out the 5V had KE markings.
Was a Fairchild part probably from after 2017 and now obsolete so not found in the Onsemi guide.
The E33 was probably production line or some other production code.

 

[fastfreddy666]

After a bit more digging...
This part makes the most sense:

Onsemiconductor (Fairchild) SMBJ5V0(C)A with marking KE in SMB/DO-214AA housing (no longer produced part)
It is a TVS intended for 5.0V lines with breakdown between 6.4V and 7.0V and is 600W peak (100A) on 10/1000 μs Waveform.

So we can agree that it's a surface mounted suppressor diode in the form of a rectangular chip
But what is it suppressing? I have no clue. Probably because the device does nothing.

 

[mansr]

So we can agree that it's a surface mounted suppressor diode in the form of a rectangular chip
But what is it suppressing? I have no clue. Probably because the device does nothing.

It's a transient voltage suppressor. Such devices protect against voltage spikes such as from electrostatic discharges.

 

[fastfreddy666]

It's a transient voltage suppressor. Such devices protect against voltage spikes such as from electrostatic discharges.

So i was right. It's a Zener diode.

 

[JonP]

So we can agree that it's a surface mounted suppressor diode in the form of a rectangular chip
But what is it suppressing? I have no clue. Probably because the device does nothing.

Jeeze... everyone gets so critical. Where's the Fun and Learning?? Can't we all get along? Can't we accept something for what it is? :rofl:

Ok, enough o' that... lemme put my electronics pro hat on...

Why? ESD. So you dont blow things out when you plug it and you're charged from the carpet or whatever. And, any spikes on the power lines would be clamped down. Theres probably one or two TVS's (or Zeners, they work too, not as well but cheaper) in your DAC input, it's common design practice. More would be better, and wont hurt.

Ehhh.. yeah a fairly sloppy soldering job, maybe not too bad. That's one of the major reasons to clean the flux, to be able to see the joints better. Theres a spot where it looks like a solder bridge, might be flux. Most of the time flux wont affect the circuit... depending. Costs $$ in cleani g labor, tho... (at the hand soldering level that is) I'd do a lot better, out of personal pride.

Hand soldering... always worse looking than machine reflow. But so what, if its functionally correct? Machine soldering... makes sense if your making many 1000's or millions. Very cheap. Building a couple hundred or less, not so. Even more so including the cable attaching and box assy. You gotta think, there might be only 100 in the world, as a gamble that they sell enough to warrant "actually going into production" with a larger batch. Bigger investment, more $$ to lay down first.

Circuit... seems OK. If the flux was off, and good pics of the traces, easy to reverse engineer. Dont know all those SMT values, of course... an advantage of thru hole parts, even caps and resistors were marked. Overall, as some have noted, seems like a reasonable filter for USB. (If the parts are sane)

BUT... we are talking >$10 parts in a $10 box, and even with a x5 markup to try to support a biz... this thing has no right costing >$100. Mass produced, could possibly do it for $10-20, but do you wanna drop $100k to have 10k made to try and make more than your money back? Hey, there's ASR out there!

Key issues. Does it do something? Have to come up with a filtering test to determine how much of what frequencies it filters. Do you need it? As stated, probably not unless you have an unusually sensitive DAC or unusually noisy power supply or RF neighborhood.

 

[mansr]

Why? ESD. So you dont blow things out when you plug it and you're charged from the carpet or whatever. And, any spikes on the power lines would be clamped down. Theres probably one or two TVS's (or Zeners, they work too, not as well but cheaper) in your DAC input, it's common design practice. More would be better, and wont hurt.

They're unnecessary in this device. None of the other parts are sensitive, and they won't do much for the USB host or device, which (as you say) almost certainly have their own suppressors, especially with the data lines unprotected.

 

[JonP]

Jeeze... everyone gets so critical. Where's the Fun and Learning?? Can't we all get along? Can't we accept something for what it is? :rofl:

I should note I'm talking at many on here, not just picking on you, fast freddy....

 

[JonP]

They're unnecessary in this device. None of the other parts are sensitive, and they won't do much for the USB host or device, which (as you say) almost certainly have their own suppressors, especially with the data lines unprotected.

Well, I imagine they're not trying to protect the filter...

You might be surprised how often they get dropped in commercial equipment. Just because it's in a spec, dosent mean design laziness or (more often) the bean counters win out.

Hey, making 1 million xxx's, we save $.1 leaving the filter parts out, that's $100,000 profit we'd leave on the table. It probably wont get zapped, it's cheap enough we dont have to provide a meaningful warranty.... etc... You get the picture.

Now, it SHOULD be with pricier devices like a DAC of a certain quality level, that you can expect that not to happen.
Scary thing, I've heard the prior argument a lot even at the *Medical Device* level, and had to make the argument for better design against Management arguing otherwise.

 

[solderdude]

So i was right. It's a Zener diode.

No.. its a Transient Voltage Suppressor which has different properties from a zener. Google it.

 

[fastfreddy666]

Jeeze... everyone gets so critical. Where's the Fun and Learning?? Can't we all get along? Can't we accept something for what it is? :rofl:

Ok, enough o' that... lemme put my electronics pro hat on...

Why? ESD. So you dont blow things out when you plug it and you're charged from the carpet or whatever. And, any spikes on the power lines would be clamped down. Theres probably one or two TVS's (or Zeners, they work too, not as well but cheaper) in your DAC input, it's common design practice. More would be better, and wont hurt.

Ehhh.. yeah a fairly sloppy soldering job, maybe not too bad. That's one of the major reasons to clean the flux, to be able to see the joints better. Theres a spot where it looks like a solder bridge, might be flux. Most of the time flux wont affect the circuit... depending. Costs $$ in cleani g labor, tho... (at the hand soldering level that is) I'd do a lot better, out of personal pride.

Hand soldering... always worse looking than machine reflow. But so what, if its functionally correct? Machine soldering... makes sense if your making many 1000's or millions. Very cheap. Building a couple hundred or less, not so. Even more so including the cable attaching and box assy. You gotta think, there might be only 100 in the world, as a gamble that they sell enough to warrant "actually going into production" with a larger batch. Bigger investment, more $$ to lay down first.

Circuit... seems OK. If the flux was off, and good pics of the traces, easy to reverse engineer. Dont know all those SMT values, of course... an advantage of thru hole parts, even caps and resistors were marked. Overall, as some have noted, seems like a reasonable filter for USB. (If the parts are sane)

BUT... we are talking >$10 parts in a $10 box, and even with a x5 markup to try to support a biz... this thing has no right costing >$100. Mass produced, could possibly do it for $10-20, but do you wanna drop $100k to have 10k made to try and make more than your money back? Hey, there's ASR out there!

Key issues. Does it do something? Have to come up with a filtering test to determine how much of what frequencies it filters. Do you need it? As stated, probably not unless you have an unusually sensitive DAC or unusually noisy power supply or RF neighborhood.

I just like to know how this thing works. That's all.


TVS/Zener Theory and Design Considerations by ON semiconductor:

http://www.reallyreallyrandom.com/www/Zener_Theory_and_Design.pdf

PSUs. I understand them

 

[solderdude]

So we can agree that it's a surface mounted suppressor diode in the form of a rectangular chip
But what is it suppressing? I have no clue. Probably because the device does nothing.

When you have done enough EMC/ESD testing (as I have) you will know what it suppresses and they do nothing in normal situations.
In this particular case it is rather nonsensical as there are electrolytic and compound ceramic capacitors in parallel.
These very fast transient suppressors belong directly on the input (and or output) of a device at the point where one can physically touch pins.
It is rather pointless to put these halfway in a cable. The cable itself is an inductor so at the ends of the cable there can still be a +4kV ESD puls of short duration.

Will the TVS do anything for the sound... of course not.
Is it in the proper place to be of any use ? no it is not.
Is it useful to put some extra capacitors on a voltage line somewhere halfway ? I highly doubt it.
Will it improve the 5V ? Maybe a tiny bit. Will that be beneficial ? No of course not because any decently designed DAC that has to rely on the +5V for operation has buffering and regulators doing that.

 

[fastfreddy666]

No.. its a Transient Voltage Suppressor which has different properties from a zener. Google it.

The wikipedia page:

"A transient voltage suppressor or TVS is a general classification of an array of devices that are designed to react to sudden or momentary overvoltage conditions. One such common device used for this purpose is known as the transient voltage suppression diode that is simply a Zener diode designed to protect electronics device against overvoltages."

so the wikipedia page is wrong. Why am i not surprised?

 

[solderdude]

I just like to know how this thing works. That's all.

Its a semiconductor mostly similar but not equal to a zener diode. It has a different construction, can handle very short very high current peaks and is not intended to work as a zener. It is also a lot faster than a zener. In this case that is moot because of the capacitors and the wires is front and behind.

If a TVS and zener diode would be the exact same thing they would not be called TVS but zener.

 

[fastfreddy666]

Its a semiconductor somewhat similar but not equal to a zener diode. It has a different construction, can handle very short very high current peaks and is not intended to work as a zener. It is also a lot faster than a zener. In this case that is moot because of the capacitors and the wires is front and behind.

If a TVS and zener diode would be the exact same thing they would not be called TVS but zener.

TVS/Zener Theory and Design Considerations

http://www.reallyreallyrandom.com/www/Zener_Theory_and_Design.pdf

 

[JonP]

Its a semiconductor mostly similar but not equal to a zener diode. It has a different construction, can handle very short very high current peaks and is not intended to work as a zener. It is also a lot faster than a zener. In this case that is moot because of the capacitors and the wires is front and behind.

If a TVS and zener diode would be the exact same thing they would not be called TVS but zener.

Yep... TVS are optimized for ultra fast turnon, and beefy construction to withstand swallowing the energy of a heavy pulse. That's one of the ratings, how much energy can they absorb/conduct, as well as voltage and speed.

Regular zeners usually run in more static states with everything more predictable, very different use cases.

And yes, solderdude has a good point about the box in the middle of the cable. Theres a case where I'd AGREE with the frickin' bean counters, and drop them from the parts list!

 

[antcollinet]

The wikipedia page:

"A transient voltage suppressor or TVS is a general classification of an array of devices that are designed to react to sudden or momentary overvoltage conditions. One such common device used for this purpose is known as the transient voltage suppression diode that is simply a Zener diode designed to protect electronics device against overvoltages."

so the wikipedia page is wrong. Why am i not surprised?

Careful - we're stepping into the "all mackerel are fish, but not all fish are mackerel" territory here.