Cavalli Audio Liquid Lightning

[johnwmclean]

I've posted this over on HF;

Alex,

Does the LL pose a safety concern?

KG and Stax amps are voltage and current limited at the output jack. Do your designs follow a similar safety feature?

Cheers,

John

I hope we get a response.

[Augsburger]

You probably will (get a response), but it may not be the one you were expecting.    

Edited September 30, 2013 by Augsburger

[johnwmclean]

On Peters request I’ve deleted the post from his review thread and placed it in the Review: Cavalli Liquid Lightning thread.

[jgazal]

 

 

(...) Now, Therefore THE GENERAL ASSEMBLY proclaims THIS UNIVERSAL DECLARATION OF HUMAN RIGHTS as a common standard of achievement for all peoples and all nations, to the end that every individual and every organ of society, keeping this Declaration constantly in mind, shall strive by teaching and education to promote respect for these rights and freedoms and by progressive measures, national and international, to secure their universal and effective recognition and observance, both among the peoples of Member States themselves and among the peoples of territories under their jurisdiction.

(...)

Article 18.

• Everyone has the right to freedom of thought, conscience and religion; this right includes freedom to change his religion or belief, and freedom, either alone or in community with others and in public or private, to manifest his religion or belief in teaching, practice, worship and observance.

^ Top

Article 19.

• Everyone has the right to freedom of opinion and expression; this right includes freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers.

http://www.un.org/en/documents/udhr/index.shtml

 

I do not know much about electronics, but just by looking the inwards of a Stax amplifier from the 90's I guess they had a 4.5 sigma level of failure already at that time (99.9999998% of the products manufactured are statistically expected to be free of defects). And if they did not, they had a distributor’s chain to repair the products.

 

I think an audio amplifier manufacturer has half the way prepared to achieve low level of defect by simple using components that come from six sigma industrial plants (transistors, resistors etc.).

 

I suppose that if you want to keep such low level of failure on your amplifier you need to path the other half way.

 

According to this thread this means to respect the tolerances of each component, to order PCB’s with the right amount of substrate and to design your copper traces not only predicting the low current, but only the thermal requirements of the solder joints, to assure that your employee apply standard soldering technics (I think this is easier to achieve with Japanese workers) etc., everything to reduce failures derived from shorts and cold joints. How much would this add to cost?

 

Stax currently matches the input transistors and add a thermal pad on them. How many man-hour this add to the cost?

 

There is a philosophical discussion.

 

On the one hand, seeking for a 6 sigma level of failure is the dream of an engineer. Those are products that are going to function for decades. Usually this kind of consumers value vintage equipment and components that were done to match engineer’s criteria and not only the costumer’s perceived value.

 

On the other hand, there are companies that deliberate decide to sell low quality products because they know that the costumers wants to change it in months. Those products need a good fascia though. And this sure is expensive. And that business model keeps the economy flowing fast.

 

Although SMD helps to achieve low level of defect within a production line, they really challenge the repair and led to simply discard the product. 

 

Avoiding those precaution costs in your production line may increase your repair cost, mainly with direct sales.

 

But not adopting safety requirements is a more serious argument.

 

I feel sad that some posts have been deleted at the other forum. I usually feel uncomfortable posting here, but I think the hostility keeps the noise down. I for one have done some unwise questions regarding electronics and all the knowledgeable head-casers were always patient to explain the basics. Nevertheless, what I really appreciate here is the freedom of speech. 

 

Sorry for the long post...

 

edited: my poor english...

Edited October 1, 2013 by jgazal

[gurubhai]

If you had been around longer, you would know that this logic is flawed.  Just look at Mikhail's creations which performed their intended function for years and sounded wonderful until they didn't (function anymore).   I am not at all trying to imply that Cavalli's designs are as bad as Mikhail's designs (in fact, I know nothing of either), I am just pointing out the flaw in your logic. 

 

http://www.head-fi.org/t/437708/warning-about-the-singlepower-extreme-owners-please-read

http://www.head-fi.org/t/437344/my-singlepower-supra-experience-why-mine-almost-blew-up-like-a-hand-grenade-and-yours-might-too/15

I have been around long enough to know about Mikhail, esp. since that reference is thrown around in almost every conversation around these parts. Mikhail used under-specced transformers to power his amplifiers which eventually and not unexpectedly failed at a later date, certainly it was not being used as hammond mfg. intended. The issues being discussed here are different as its obvious that Cavalli amp hasn't skimped on the parts, but the problem is more of a manufacturing process. I am still unconvinced that cold joints would last for that length of time working in those high voltage but in any case, I wouldn't be happy with the soldering skills on display either.

If I were Mr.Cavalli and saw those pictures, I would switch to another production house but then may be he already has. I noticed that he has already released a second version of the LL a few months back while we are discussing the older version.

Since John has already popped the question, I would probably wait for Dr. Cavalli's response to hear his side of story before making up my mind as  to the veracity of these claims.

[The Monkey]

 

I feel sad that some posts have been deleted at the other forum. I usually feel uncomfortable posting here, but I think the hostility keeps the noise down. I for one have done some unwise questions regarding electronics and all the knowledgeable head-casers were always patient to explain the basics. Nevertheless, what I really appreciate here is the freedom of speech. 

 

Sorry for the long post...

 

edited: my poor english...

 

 

lol, I like the portion of your post bolded above.  Nothing wrong with your English.  Certainly better than I could do in another language.

[grawk]

I have been around long enough to know about Mikhail, esp. since that reference is thrown around in almost every conversation around these parts. Mikhail used under-specced transformers to power his amplifiers which eventually and not unexpectedly failed at a later date, certainly it was not being used as hammond mfg. intended. The issues being discussed here are different as its obvious that Cavalli amp hasn't skimped on the parts, but the problem is more of a manufacturing process. I am still unconvinced that cold joints would last for that length of time working in those high voltage but in any case, I wouldn't be happy with the soldering skills on display either.

If I were Mr.Cavalli and saw those pictures, I would switch to another production house but then may be he already has. I noticed that he has already released a second version of the LL a few months back while we are discussing the older version.

Since John has already popped the question, I would probably wait for Dr. Cavalli's response to hear his side of story before making up my mind as  to the veracity of these claims.

I think it's pretty clear Cavalli is content letting his forum admin and his shill army handle things for him.

[Nebby]

I think it's pretty clear Cavalli is content letting his forum admin and his shill army handle things for him.

 

I'm still amazed that someone just claimed that the only thing wrong with Mikhail's builds were under-specced transformers. 

[grawk]

It's a brilliant strategy really. If you don't respond to criticism, then your forum admin has grounds to delete criticism.

Compare it with threads where you have an active designer discussing his projects, then you can leave obviously false criticism to stand. That allows a double standard that is completely justifiable.

[shellylh]

I noticed that runeight joined HC yesterday so maybe he will answer John's question in this thread.

Edited October 1, 2013 by shellylh

[gurubhai]

I'm still amazed that someone just claimed that the only thing wrong with Mikhail's builds were under-specced transformers. 

You misunderstand. I was merely pointing out one of the worst offense Mikhail committed. If you read the  Nate's post on that Singlepower thread at HF you would see that he too focused on that as the major issue.

 

I think it's pretty clear Cavalli is content letting his forum admin and his shill army handle things for him.

I actually had a chuckle about how predictable some of you can be. I knew it was time for some personal attacks. If you really believe that I joined this forum four years back just so I could shill one day for Cavalli, then I truly feel sorry for you.

Anyways, I can see that I have outrun my welcome here and would refrain from posting here any further.

[Voltron]

What a shame.

[grawk]

I actually had a chuckle about how predictable some of you can be. I knew it was time for some personal attacks. If you really believe that I joined this forum four years back just so I could shill one day for Cavalli, then I truly feel sorry for you.

Anyways, I can see that I have outrun my welcome here and would refrain from posting here any further.

I didn't specify you as one of the shills, but I don't mind that you included yourself. As to why you joined 4 years ago, who really cares? It's pretty obvious why you're here now.

[Nebby]

nvm. 

Edited October 1, 2013 by Nebby

[GeorgeP]

 

I actually had a chuckle about how predictable some of you can be. I knew it was time for some personal attacks. If you really believe that I joined this forum four years back just so I could shill one day for Cavalli, then I truly feel sorry for you.

Anyways, I can see that I have outrun my welcome here and would refrain from posting here any further.

 

 

He didn't mean you, based on my reading anyway.

[Pars]

I actually had a chuckle about how predictable some of you can be. I knew it was time for some personal attacks. If you really believe that I joined this forum four years back just so I could shill one day for Cavalli, then I truly feel sorry for you.

Anyways, I can see that I have outrun my welcome here and would refrain from posting here any further.

That didn't seem like much of a personal attack... (if any? dunno, not that sensitive about these things)

 

I think we say something like Grow some thicker skin around here. But what do I know.

[Pars]

I have been around long enough to know about Mikhail, esp. since that reference is thrown around in almost every conversation around these parts. Mikhail used under-specced transformers to power his amplifiers which eventually and not unexpectedly failed at a later date, certainly it was not being used as hammond mfg. intended. The issues being discussed here are different as its obvious that Cavalli amp hasn't skimped on the parts, but the problem is more of a manufacturing process. I am still unconvinced that cold joints would last for that length of time working in those high voltage but in any case, I wouldn't be happy with the soldering skills on display either.

If I were Mr.Cavalli and saw those pictures, I would switch to another production house but then may be he already has. I noticed that he has already released a second version of the LL a few months back while we are discussing the older version.

Since John has already popped the question, I would probably wait for Dr. Cavalli's response to hear his side of story before making up my mind as  to the veracity of these claims.

I think there are some concerns/criticisms of the parts used, as well as the design. Input FETs are one that I recall.

[spritzer]

In the ES-1 and ES-2 the transformers were actually way over spec, it's just the rest that was utterly fucked.  The worst were the bias supplies, filament supply, caps run way over the rated voltage and the minor detail of the -110V offset always present.  In this LLmk1 the parts quality isn't great and just enough to do the job is the mantra.  Why use the CMF55 resistors when the CMF60 units are not that more costly and will result in less noise.  Ditto for the higher wattage resistors, using larger ones can only be better.  Same for the input fets, those aren't good parts and J74's can still be sourced.  The output devices make no sense to anybody and Fairchild does make excellent 800V BJT's which are a perfect fit.  Isolated too so that makes mounting easier. 

 

One other point that needs to be dealt with, how does this amp stack up against the other amps out there.  The only comparison review was done by Tyll and lets face it, all the other amps except the BHSE are fucked in one way or another.  727 is far from linear stock, the SRM-007tII/A is very dull, the WES is a quad of single ended amps with a choke on the output trying to push the EL34 at 600V and the A-10 just needs to be experienced to know how lots of distortion sounds.  The rest of the impressions are by the regular army of shills or people with very little exposure to the high end electrostatics.  You may take my impressions anyway you like but the fact that one of these amps is approaching 10 owners now in the span of a few months says something about the sound quality. 

 

I would welcome somebody like Schit into the market.  We have been taking to a number of Chinese manufacturers and there is interest but who knows what will come of it.  Just this morning I got an offer to build for me anything I'd like.  Not interested in that but it would certainly be possible to build a smaller version of the KGSSHV, fully SMD except the output devices. 

 

As for the amp, I've been busy with other projects but the pot in the bias supply is giving me problems.  Bloody Vishay crap so I'll rewire it a bit and it will work just fine.  After that it's time for the amp board, remove the diodes on the power input (those make no sense), I ordered some LSJ74's so if I get two matched pairs I'll use those or 2SJ109 to replace the Fairchild's and probably remove the transistor in the CCS and use a LED instead.  Also some local capacitance on the board is a given.  Where to put large filmcaps though... 

[Jon L]

We have been taking to a number of Chinese manufacturers and there is interest but who knows what will come of it.  Just this morning I got an offer to build for me anything I'd like. 

 

I would like Chinese-made T2 and KGSSHV on eBay on Buy-It-Now for $399.99 please. 

[spritzer]

Sure, why not.    I do think a version of the KGSSHV could be built and sold for about 1200$ even in small numbers and with our high standards. 

 

If I ever get to work off my  project list then this is something I'd consider doing, at least the R&D side of things. 

[Hopstretch]

 

Just this morning I got an offer to build for me anything I'd like.

 

Sexbot with built-in high quality electrostatic amplifier. Or high quality electrostatic amp with built-in sexbot. Whichever you think will have the better sound quality.

[Augsburger]

Stretch, shame on you.  You forgot to include the built in bottle/wine opener in the specs.

[runeight]

Well folks, I didn’t ever think being involved in the audio business would require this of me. I have filtered through this thread and assimilated what appear to be the primary accusations against the LL.

 

I know that the main objective is claimed to “document just how badly built and designed this amp is” as exposing “safety” concerns, not to mention the magnanimous claim that “We post the data for people to make up their own minds” but I rather suspect it is more than that. 

 

I’ll address the concerns that I pulled out, but I should make it clear that this will be my first and only post on this forum. If anyone has concerns about my designs or amplifiers, please contact me through my website.

1. The direct soldered/exposed IEC connections

This is only an issue for someone who wants to open the amp and modify it while it is plugged in.

 

 It is not a safety issue for a normal customer who doesn’t open the box.

1. Also notice the small jumper next to them, that selects between high and low bias.  Not the best idea to run these at high voltage...

Jumpers are floating and not near any low voltage pads or terminals. The voltage between the pins is less than 50V. There is no risk of problem.

1. cold solder joints

EC standards require the use of lead free solder. It frequently requires higher soldering temperature and cools with a duller finish. This is normal.

1. thin board

Thin board is a matter of opinion. The board does not need to be thicker to do its job well.

1. “What is the cause of a poor soldering?
   1. Worker without skill experience;
   2. Thin PCB substrate and thin traces that cannot handle the temperature; or,
   3. Soldering station without temperature control?

The amp uses ROHS compliant unleaded solder and boards. The boards for LL1 were done by a PCB assembly shop capable of and who regularly does automated and hand lead free soldering for many manufacturers around the US and globally.

 

However, anyone who manufactures complex electronics knows that a highly capable supply chain is essential. I have hired and fired 4 PCB assembly shops before finding the one I’m using now to get the best work possible.

 

No LL has reported failure due to soldering issues.

1. The bias supply is utterly fucked and not safe to use.  Then I'll add loopouts to the back and rewire the input section.

The bias supply works as it’s supposed to. There have been no reported failures of this supply.

 

LL2 has loopouts. This was a shortcoming and has been added.

1. Only thing I've done is remove those bloody mosfets which should never have been there and fix the CCS.  So now the output capacitance is cut down to less than 10pf and the huge bump in distortion from the old CCS is gone.

BJTs do have lower capacitances than Mosfets. They also exhibit thermal runaway. Mosfets have advantages, including not being prone to thermal runaway.

1. All those zeners add a huge quantity of noise and as a whole it's just unstable.

Measurements and owners do not support this notion and the amp is quiet.

 

Additionally, the amp is not unstable. It has exhibited no stability issues on any production amp. When built properly the amp biases and works every time.

1. As for the safety of the amp, of course it isn't safe , there are no output resistors

The amp was designed to current limit and shut down if overloaded.  The CE sticker on the back of the amp is for real. The LL1 was certified for entrance into EU countries by Diversified T.E.S.T. Technologies Lab (http://www.dttlab.com). This includes safety and emissions testing. I have the report in case someone in the EU wants to enquire.

1. “the thing gets stinkin hot” , “extended use of stock LLmk1 and probably LLmk2 due to heat and other issues is going to definitely cause reliability issues, and out of warranty repairs are going to be nasty”

Thank you Kevin, this actually brought a smile. This is a little like the pot calling the kettle black.

 

The thermal conditions of the amp have been tested both theoretically and in reality using IR sensors. The transistor junction temperatures are not near any operational or dangerous limit and if the amplifier is used in a well-ventilated area as recommended, it performs perfectly alright.

 

1. You think that the screws on the current source triple and transistors are somehow replaceable by someone other than a magician

My amplifiers are not intended to be serviced by third party members or DIYers. Registered dealers and my assembly house will have no problem returning the normal function of the amplifier. In HV amps if one part fails there is sometimes other damage. In order to return an amplifier to its original condition, if necessary, the entire board will be replaced. I call this customer service.

1. ““So I take it that figure of 0.0009% THD+N isn't exactly accurate?”
2. J176 input fets designed for switching (not linear applications) with an input noise of a minimum of 10nv per root hz and an amplifier gain of 50db is real? when alex could have used the real thing instead (2sj74 or lsj74)?

2SJ74s are not commercially available that I know of except, possibly, from sources who stocked up on them. They are sometimes counterfeits. As a manufacturer I prefer not to depend on any of these sources. Although if Linear releases new LSJ74 (limited release just a few days ago, rumored OEM late 2013 last time I looked) they will be an obvious choice.

 

Measurements were done by an independent lab here in Austin.

1. Mosfets don't belong in electrostatic amplifiers!

So you say Kevin. I have designed this circuit around the Mosfets with their behavior in mind. This amp works very well and this is affirmed by the owners who have communicated with me.

 

Modifications made in LL MKII have further improved its performance while making various other improvements.

1. As for other makers entering the market, would anybody doubt the design and build quality of a Stax amp sold by Schiit or Pete Millet?  The difference is that they know what they are doing, Cavalli has no clue.

Although I have the highest regard for Jason, Mike and Pete, this is pure speculation regarding a theoretical electrostatic amp. Unless you are privy to all of their business dealings and more importantly, their schematics, you don’t even know that there haven’t been problems that have been handled in one way or another, nor whether any current product has a problem that hasn’t manifested itself yet. And you certainly don’t know what an electrostatic amp from any of them would look like (except, of course, it would no doubt be a copy of something you did first).

 

Furthermore, this is a rhetorical trick. Bringing in other people who are respected is a way to give your argument weight that it doesn’t possess. A casual reader might conclude that these guys are somehow aligned with you on these comments about my amplifiers. But, as far as I know, none of them have made such a public statement.

 

Folks, the audio world can be great place to be with lots of very fine people. I have had the great privilege in the last several years of meeting many of them, both online and in person. I hope you all find that to be true too.

[Knuckledragger]

[roadtonowhere08]

 

Hey, fucker, that's my .gif  

 

 

 

 

 

Well folks, I didn’t ever think being involved in the audio business would require this of me. (Really?  Having to address peoples' concerns about your design?  In academic and field science, it's called peer review; in audio... it's taboo) I have filtered through this thread and assimilated what appear to be the primary accusations against the LL.

 

I know that the main objective is claimed to “document just how badly built and designed this amp is” as exposing “safety” concerns, not to mention the magnanimous claim that “We post the data for people to make up their own minds” but I rather suspect it is more than that. (Nahh, it's just that)

 

I’ll address the concerns that I pulled out, but I should make it clear that this will be my first and only post on this forum. If anyone has concerns about my designs or amplifiers, please contact me through my website.

 

1. cold solder joints

EC standards require the use of lead free solder. It frequently requires higher soldering temperature and cools with a duller finish. This is normal. (Bull. Shit.  You skirted the issue of sloppy soldering and went for the dullness card which might drive away the morons but not anyone who has eyes and has seen other ROHS solder jobs)

1. thin board

Thin board is a matter of opinion. The board does not need to be thicker to do its job well. (With a $5,000 price tag, the quality of parts had better fit the bill... and that anemic PCB surely does not)

1. “What is the cause of a poor soldering?
   1. Worker without skill experience;
   2. Thin PCB substrate and thin traces that cannot handle the temperature; or,
   3. Soldering station without temperature control?

The amp uses ROHS compliant unleaded solder and boards. The boards for LL1 were done by a PCB assembly shop capable of and who regularly does automated and hand lead free soldering for many manufacturers around the US and globally.

 

However, anyone who manufactures complex electronics knows that a highly capable supply chain is essential. I have hired and fired 4 PCB assembly shops before finding the one I’m using now to get the best work possible. (You might want to keep looking...)

 

No LL has reported failure due to soldering issues. (That's what you are going with?  There has been no failure, so there is no problem?)

 

Folks, the audio world can be great place to be with lots of very fine people. I have had the great privilege in the last several years of meeting many of them, both online and in person. I hope you all find that to be true too. (Classic kiss off)

 

Edited October 2, 2013 by roadtonowhere08