![]() |
| | #1 |
| Member Join Date: Mar 2008
Posts: 99
| PCB design and high voltages
Hi, I need some help designing a PCB for a tube amp. I would like to know how high voltages relate to traces width and thickness and spacing between them? Can you point to a standard, application note or calculator for this? Thanks. |
| | |
| ...and now, a word from our sponsor: |
| | #2 |
| Senior Member Join Date: May 2008 Location: Los Gatos, CA
Posts: 158
|
Google "high voltage" trace clearance and you can become an expert, which is a good idea if you're doing a board for a tube amp.
|
| | |
| | #3 |
| Lifetime Member Join Date: May 2006
Posts: 1,035
| |
| | |
| | #4 |
| Member Join Date: Mar 2008
Posts: 99
|
Actually I'm not a beginner - I have several tube amp layouts that work without any problems so far. I just wanted to make sure they comply with the standard.
|
| | |
| | #5 |
| Senior Member Join Date: May 2008 Location: Los Gatos, CA
Posts: 158
|
I didn't mean to blow you off. I have a similar problem. It took me a few tries on Google to get good links, but the search I recommended will point you to standards and calculators. If you've already done a few boards, you know it's all in the details. |
| | |
| | #6 |
| Supporting Member Join Date: May 2006 Location: Glasgow, Scotland
Posts: 2,943
| http://www.ce-mag.com/ce-mag.com/arc...uctSafety.html enjoy FWIW, one of my favourite things about tube amps is that they don't require PCBs. I do PCB layout in my day job, in fact I just finished a 6-layer mixed-signal motherboard with 350 SMT parts and am sick of the things.
__________________ "Ohhhh miracle bulb shines feebly" |
| | |
| | #7 | |
| Lifetime Member Join Date: Jan 2007
Posts: 69
| Quote:
Randall Aiken | |
| | |
| | #8 | |
| Lifetime Member Join Date: May 2006
Posts: 1,035
| Quote:
I just recently fought this battle and mostly won it. But it's an issue of far more than PCBs. You're going to want IEC600650, that being the one which relates to musical gear among other things (like refrigerators and other appliances... A copy of that will set you back between $200 and $400. No, they are not given away free. The standards organizations sell them like novels. A safety standard does NOT tell you what to do to be safe. It tells you if you don't do at least this much you are definitely not safe. While that may sound like sopistry, it's an important legal distinction. You cannot and will not get the information you need to pass safety certification (e.g. CSA, CE, CCC, etc) on an internet forum. If you have to meet a standard (in the USA, you do not; you can be sued even if you do: in Canada, you must meet standard; in the EU, you must meet the standard but are not shielded from lawsuits by it; in... in... - it's almost like there's a different law in each country, isn't it?) then you need a quick education. I recommend reading "Practical Guide to the Low Voltage Directive" by Gregg Kervill and "Electrical Product Safety" by Tzimenakis and Holland to start with, buying a copy of the standard, and then hiring a local certification engineer to review if not actually collaborate. And remember - it's not me, or you, or anyone on this forum you have to convince you did the right thing. It's the judge and jury. | |
| | |
| | #9 |
| Supporting Member Join Date: May 2006 Location: Glasgow, Scotland
Posts: 2,943
|
Randall, sounds like fun The flipside of RG's argument is that, since standards don't protect you from litigation, there's no point in bothering about them. As long as your product doesn't kill anyone or set fire to anything with insurance, I'd bet nobody would even notice a complete absence of safety approvals Better still, sell it as a kit, and in court you can argue that it was badly assembled.
__________________ "Ohhhh miracle bulb shines feebly" Last edited by Steve Conner; 07-22-2008 at 11:02 AM. |
| | |
| | #10 | |
| Lifetime Member Join Date: May 2006
Posts: 1,035
| Quote:
The water is murkier for kits. I didn't look into that. I suspect that one had better be in a position to prove that the kit, assembled to the letter of the instructions and including likely mistakes would pass CE muster. It's hard to argue that a PCB with too-small spacings would get better when assembled. In any case, if it's not CE labeled, the person who builds it may not legally sell it to someone else, and perhaps not even give it away. I'm not an EU lawyer, I just read safety standards and associated literature for several months solid. That is NOT enough. + | |
| | |
| | #11 | ||
| Member Join Date: Mar 2008
Posts: 99
| Quote:
Quote:
Then what about a private person selling a DIY effect/amp on eBay? | ||
| | |
| | #12 |
| Supporting Member Join Date: May 2006 Location: Glasgow, Scotland
Posts: 2,943
|
Well, all you really need to do is make sure that your equipment would comply with the CE mark if a lab ever tested it, and keep a low profile. You have to try really hard to make a tube amp fail an EMC test (though if you wanted to, putting a uC in it would be a good place to start!) so it's basically just safety. There are grey areas: for instance, if I bought someone else's homebrew amp on Ebay, am I right to expect a CE mark? If I got tired of one of my homebrew amps and sold it to a friend, and it malfunctioned and burnt their house down, would their insurance company sue me for the damages because it wasn't CE marked? Could they tell from the remains whether the CE mark was fake? If I left the country, would the loss adjusters come after me? And so on. I personally don't sell any of my homebrew equipment that plugs into a wall socket, for this and similar reasons. I come across similar dilemmas all the time at work, but since I work for a limited liability company, it's not so much of a worry.
__________________ "Ohhhh miracle bulb shines feebly" |
| | |
| | #13 | |
| Lifetime Member Join Date: Jan 2007
Posts: 69
| Quote:
Speaking of DSP, has anyone here ever played around with the Freescale Symphony DSP chips? I like the idea that they are using a free Eclipse-based toolset for the IDE...AD and TI charge a small fortune for their development software. | |
| | |
| | #14 |
| Lifetime Member Join Date: Jan 2007
Posts: 69
| |
| | |
| | #15 |
| Lifetime Member Join Date: Jan 2007
Posts: 69
| Unfortunately, it looks like everything is going to either BGA or those tiny QFN leadless packages like most switching regulators come in nowadays. It makes hand-assembly nearly impossible, even for prototypes.
|
| | |
| | #16 | |
| Senior Member Join Date: Jul 2006 Location: Goteborg, Sweden
Posts: 144
| Quote:
Testing labs do what the name says: testing, they're not authorized to chase non-compliant stuff in shops. This is up to Electrical Safety Board to do. They do sample control of whatever crosses EU boarder, currently they have a spy glass on chinese imports. No CE mark means immediate ban. If CE mark is present the importer/distributor or whoever has put the product on the market is asked to present documentation proving compliance in 48hrs. That documentation is called "Technical File" and has to be maintained by manufacturer if he did all compliance certification testing himself or by the accredited test facility that did the testing and issued certificates. The later is the best way to do it. Now, certification testing by accredited lab is NOT mandatory. Compliance with safety norm is. As manufacturer you are declaring compliance. You are free to declare compliance "by design" but must be able to prove it. As for tubes amps having no problem with EMC compliance - that is not the case. Have seen quite a few that didn't even though there were no digital clocked circuits in them. Undersized PT and heavily distorting amp will feed a lot of garbage back into power line. PT will probably overheat anyway leading to non compliance verdict. Poor grounding scheme will cause the amp to amplify irradiated RF and send it out with mains/input/speaker cable as antenna. Switching spikes from SS rectifiers if not snubbed are RF sources. You can build an amp for your own use AT HOME. Officially you're not allowed to use in public venues, lend or sell it unless you put CE mark on it and can prove compliance. But nobody I know cares. If I receive an amp with 2-prong cord for repair I'm obliged to replace it with earthed 3-prong. If customer refuses I'm obliged to cut the mains cord off. What the customer does after he leaves my shop is not my concern. I involve an accredited lab from early stages of design. This one is run by senior engineers who are tube geeks themselves, they just love having "something fun and real" to certify as a welcome interruption to all chinese coffee grinders and such. RoHS is a totally different barrel of bile. | |
| | |
| | #17 | ||||
| Lifetime Member Join Date: May 2006
Posts: 1,035
| Quote:
1. "... make sure that your equipment would comply..." The ONLY way to do that is to have a certified lab test it, unless you yourself are an experienced compliance tester. It's a little like saying "all you need to do to successfully remove a brain tumor is know all that surgery stuff so you would be licensed as an MD and pass the necessary exams and so on". Worse yet, some of the issues are ambiguous, and whether you would pass can turn on the interpretation of a single word in a phrase of the standard. I was proud of the fact that I pointed out some issues in the standard to the reviewers I hired. But they in turn pointed out items that needed changed even though I had been through the standard with the design documents line-by-line-by-line multiple times. Quote:
Quote:
In the day when European testing was nationally mandated, approval by a compliance lab meant that you WERE by legal definition, safe, and hence not liable. Now, the standard is a de-facto minimum. If you do not at least comply, you could easily be argued as being inattentive to safety minimums. But if someone gets hurt or killed, you are still liable, even if you met the standard. However, if you can prove you met the minimums, you cannot be said to be reckless and irresponsible. What the current system does is perform responsibility splitting. If you have a certified lab do a report and you put on your CE mark, then you can tell people who bring claims to go argue with the lab. It's a version of the old adage that if there are more than two signatures on the form, the blame can never be placed accurately. Quote:
| ||||
| | |
| | #18 |
| Lifetime Member Join Date: May 2006
Posts: 1,035
| That turns out to be a "bookkeeping tax". All you have to do to pass ROHS is to be able to show that all the parts and materials you put into your amps were represented to you as ROHS conforming. That then lets you off the hook in favor of pursuing the much fatter target of a part manufacturer committing fraud.
|
| | |
| | #19 | |
| Supporting Member Join Date: May 2006 Location: Glasgow, Scotland
Posts: 2,943
| Quote:
I'm looking forward to the Line 6 DSP stompbox DDK coming out though! I think that's the same Freescale Symphony DSP that RG mentioned, though I always think of those as the Motorola 56k series. I've written stuff for the ADI 21xx and TI 67xx DSPs before, and I know a guy who does amazing audio things with the Microchip dsPICs. All I'm trying to say is that the whole RoHS/CE thing is a nightmare. I'm not debating this. I'm just wondering what it means to a mom and pop shop. I guess they have two options: close up, or go underground and hope they don't come onto anyone's radar. I took the option of working as an EE in industry and leaving amp building as a hobby. It means I end up designing boring industrial electronics instead of groovy boutique amps, but it also means someone else is paying for the compliance testing, and someone else getting sued if it goes wrong, which suits me just fine.
__________________ "Ohhhh miracle bulb shines feebly" | |
| | |
| | #20 | |
| Lifetime Member Join Date: Jan 2007
Posts: 69
| Quote:
Randall Aiken | |
| | |
| | #21 | |
| Member Join Date: Mar 2008
Posts: 99
| Quote:
| |
| | |
| | #22 | |
| Lifetime Member Join Date: May 2006
Posts: 1,035
| Quote:
However, by and large in the USA we get the government we deserve. We have inherited a simple, effective way to get rid of the weasels. That we do not use it is a tragedy. There is a reason HOR terms are two years. What kills me is that the whole congress does not turn over every few years. I find it amazing that people can be unaware of these issues; and never connect up the dots that having a government in your shorts in any way is a Bad Thing. There is a book that's worth reading. It's called "A Parliament of Whores" by P.J. O'Rourke. | |
| | |
| | #23 |
| Senior Member Join Date: Jul 2006 Location: Goteborg, Sweden
Posts: 144
|
From PCB creepage distances to government in your shorts. Hilarious, you guys are ....creepy Anyway, there was a case of one company reporting a competitor for RoHS non-compliance. Both got investigated and guess which one was found non-compliant? Here in Sweden the Electrical Safety Board has it's own inspectors who pick products randomly or act upon a report of a documented electrical mishap. One distinct difference between EU Safety Boards and UL in US is that UL is a profit driven Inc. Complaining that your cat died when you attempted to dry its fur in a micro does not count for anything other than having yourself declared an idiot and doing time for cruelty to animals. The only state agency that will investigate on anonymous tip is the Tax Office. Corruption is a funny thing, definition varies with geography. We don't care about our politicians sex life unless they get involved with minors or porn. Their job is in danger if they use a government issued credit card to buy toys for their kids or do cut and paste tricks with job related expense receipts. An interesting thing: a product liability insurance policy with EU insurance company states clearly it does not cover USA. For that you need additional policy that costs 5x more. Thanks to great American tradition of suing everybody for anything. What if all these lawyer fees were put to getting stuff properly tested and certified? BTW. CE mark also implies that the product is accompanied by a user manual including detailed safety instructions. Mine has 5 pages of them. Warranty & product registration card must be signed by customer and returned by registered mail. If the statement: "I have read and understood safety instructions" is not initialed all bets are off. Last edited by Alex/Tubewonder; 07-23-2008 at 11:31 AM. |
| | |
| | #24 |
| Senior Member Join Date: May 2006 Location: Lansing, Michigan, USA
Posts: 10,254
|
No argument from this American. I don't care about politicians' sex lives either. Unfortunately I am a minority. Over here if we can point a finger and find someone to blame for something, we consider the job done. We don't really try to change anything, we jist want that AHA! moment, and then we forget it. It is a fine old American tradition swapping stories about who sued whom for how silly a thing. "Did you hear about the guy who...?" I am sure they are really the exception, but we hear about incidents like the fellow who dropped a car battery on his foot as he was trying to install it himself, and successfully sued the battery manufacturer. In some parts of the world, if you bring suit and lose, then YOU are responsible for the other party's legal expenses. I like that plan, that would stop a lot of frivolous suits. |
| | |
| | #25 |
| Supporting Member Join Date: May 2006 Location: Glasgow, Scotland
Posts: 2,943
|
Anyone know how to calculate creepage and clearance for one of these? http://www.neighborhoodvalues.com/nv.../misc/35kc.htm Alex: The Scandinavians are so sensible, they could make even the most messed-up political system work fine. I've never visited Sweden, but I've been to Norway and Denmark on business and loved it there. I'd consider emigrating if it weren't for the price of beer
__________________ "Ohhhh miracle bulb shines feebly" |
| | |
| | #26 |
| Lifetime Member Join Date: May 2006
Posts: 1,035
|
Yeah, but you folks in Sweden are **sane**. Not so in the USA.
|
| | |
| | #27 | |
| Lifetime Member Join Date: May 2006
Posts: 1,035
| Quote:
Of course there are other sections of the standards that might not let that one pass today. | |
| | |
| | #28 |
| Senior Member Join Date: Jul 2006 Location: Goteborg, Sweden
Posts: 144
| We do have our share of people endowed with "alternative intelligence" in every walk of life. It's not all nice and cosy over here even though it may look this way from across the pond.
|
| | |
| | #29 | |
| Senior Member Join Date: Sep 2007
Posts: 124
| Quote:
| |
| | |
| | #30 |
| Junior Member Join Date: Jul 2008
Posts: 1
|
Hi guys.. I'm an absolute newcomer to this forum.... 55 yrs EE in instrumentation and telemetry.. lifelong "audiophool" and snake oil sceptic. Dumped in here because of a guitar freak son.. ( otherwise he's a good lad... [QUOTE=Alex/Tubewonder;67572] Complaining that your cat died when you attempted to dry its fur in a micro does not count for anything other than having yourself declared an idiot and doing time for cruelty to animals. [/QUOTE Even if this and other stories like this may be "designed", I've often wondered how the US legal system came into this "I'll sue you and you'll sue me" situation... I recently read a report that this legal lunacy is costing the US society something like 5.000.000.000.000 $, all together.......? Hm....... But, to the case.. The CE and UL stuff is of course applicable to all aspects of "homegrown" electronics, not only musical related. I started loking into this some years ago, but dropped out along the way. Most of my passtime audio work has been with loudspeakers, although I've made my fair share of amps, but I've never sold anything to other than fairly close friends. Now this questions is again relevant, 'cause of my sons interest in guitars, as there's the chance that whatever I may build may be traded for something else, in the eternal quest for "the holy grail of tone.." Alex.. do you happen to have any links to relevant documents regarding CE and safety...? For the rest of you - the old pre CE swedish and Norwegian regulations very mostly identical, and for most a certificate from the old swedish SEMKO ( official swedish testlab) was automatically accepted by the norwegian NEMKO, and vice versa....Norway and Sweden were known to have some of the strictest regulations for electrical safety in the world... Halgeir Wold Norway |
| | |
| | #31 |
| Senior Member Join Date: Jun 2008
Posts: 319
|
Or do repairs....
|
| | |
| | #32 |
| Junior Member Join Date: Sep 2008
Posts: 7
| PCB Madness and Litigation
First: Back in the 60's and before,people innovated without a "Big Brother" fear.There was a very common AM radio circuit called the "All American 5"... 5 tubes,filaments wired in series,no power supply transformer.The radio's had every knob,screw,etc. insulated or made out of insulated material. I'm positive some folks died as the result of stupidity (washing the radio when it was on)...but for the most part,they were good products,and inexpensive.That's why some tubes had 50 volt filaments (50L6 comes to mind).... I think MOST companies went to task to make products as good and dependable as they could, balanced with safety concerns.There were, however, professional service technicians who fried their test equipment working on those "Hot Chassis" products.I remember the isolation transformers coming onto the market.... Secondly: PCB's and tube amps don't mix well. I've had Guitar amps that had to be "RE-Wired" with real wire, not traces, because the heat,mechanical strain,and voltages were a bit much for circuit boards. Look at the classic Tube amps...Fenders,Marshalls, Vox,among many. Now look at "Modern" Tube amps with PCB's....chinese - built junk. Not worth much money,even when new. Who wants them? Maybe as the basis of a "Modded" amp....putting in all the non-pcb stuff,to make them more reliable. New circuits:I'm working on my brother's Line 6 floorboard....no luck getting out of the manufacturer what the parts in ANYTHING are.In fact - They actually told him "We don't make the products; we design them". So:Unless you are a "Factory authorized service center", you can't find out what the IC's are (house marked).There will be precious little modding of their products! Do you think that equipment will increase in value over time? And unless you have a few Million $ to spend on design and marketing,you won't come close to any of the Emulator's. The "Throw away society". People actually bought products in the past that were meant to be fixed. Vintage amps and modern high - value amps use point - to -point wiring,period. They will always hold their value. >>>> End of Rant <<<< |
| | |
| | #33 |
| Supporting Member Join Date: May 2006 Location: Texas USA
Posts: 914
|
I designed PCB's back in the days before PC computer drafting. Bishop Graphics published a book that had design rules for clearance vs voltage and line width vs current. (Naturally, copper thickness and plating was taken into account.) The line width specs told you what the trace could take but often the voltage drop was more than you would like. To get the voltage drop you had to go to a physics book and calculate the resistance of the copper trace to control voltage drop. UL also had specs that called for additional clearance around traces that connected to the incoming power line and clearance to ground. There was also a UL spec having to do with line width next to the edge of a PCB because there was a tendency for those to start a fire. I'll see if I can find some of those specs.
|
| | |
| | #34 | |
| Member Join Date: Mar 2008
Posts: 99
| Quote:
I recently ordered out of curiousity some RoHS compliant PCBs from the local PCB shop and tried some RoHS solder along. To tell you the truth it's a BIG pain in the ass to solder these. The solder melts at higher temperature and it takes longer for it to spread over the pads. Below is an online calculator I found concerning traces, current, resistance etc.: http://circuitcalculator.com/wordpre...th-calculator/ | |
| | |
| | #35 |
| Junior Member Join Date: Sep 2008
Posts: 7
| PCB's
Wow,Loudthud... What a headache! I imagine PCB design is still a pain...inductance between traces,timing circuits going haywire because of stray capacitance,etc. I know Large scale Integration and PCB's are the only economical way for Manufacturers to recoup their investment, in almost all Consumer Electronics. The very idea of "Mega-Scaling" is a necessary evil. I'm just glad for the things that aren't "Proprietary" and can be hacked and modified. Like Guitar amps that are point-to point. |
| | |
![]() |
| Bookmarks |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
| Display Modes | |
| |
Similar Threads | ||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| voltages too high | cminor9 | 5 E 3 | 1 | 04-18-2008 02:07 AM |
| high plate voltages in blues deville | slapbackdaddy | Maintenance, Troubleshooting & Repair | 22 | 09-15-2007 06:43 AM |
| Hi Fi to Champ conversion questions: Voltages too high? | Regis | Conversions | 11 | 04-18-2007 03:20 AM |
| High voltages afer recap!? | Richard | Maintenance, Troubleshooting & Repair | 14 | 03-20-2007 06:03 PM |
| Measuring voltages in a very high impedance circuit | JC@ | Theory & Design | 4 | 07-04-2006 02:48 AM |