I`m curious : What is a "virgin teflon standoff", and does it have anything to do with the "teflon Don"? I don`t see much SMD, here, at the extreme edge of the known Universe, but I can see that the single sided approach would make tracing a circuit much easier,...... ...........................................................................Don C.
PS : Perhaps the "double sided approach" would be more suitable for Schitzophernics - a few of whome it has been my pleasure to meet. ----- Original Message ----- From: <[EMAIL PROTECTED]> To: <[email protected]> Sent: Tuesday, June 03, 2008 1:53 PM Subject: time-nuts Digest, Vol 47, Issue 5 > Send time-nuts mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > or, via email, send a message with subject or body 'help' to > [EMAIL PROTECTED] > > You can reach the person managing the list at > [EMAIL PROTECTED] > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of time-nuts digest..." > > > Today's Topics: > > 1. Re: PCB design questions thread II (John Day) > 2. Re: PCB design questions thread II (Bruce Griffiths) > 3. Re: PCB design questions thread II (Matthew Smith) > 4. Re: PCB design questions thread II (Bob Paddock) > 5. Re: PCB design questions thread II (Matthew Smith) > 6. Re: PCB design questions thread II (Bob Paddock) > 7. Re: PCB design questions thread II (Chuck Harris) > 8. Re: PCB design questions thread II (Bob Paddock) > 9. Re: PCB design questions thread II (Keith Payea) > 10. Re: PCB design questions (Didier Juges) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Mon, 02 Jun 2008 20:20:14 -0400 > From: John Day <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: Discussion of precise time and frequency measurement > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset="us-ascii"; format=flowed > > At 08:04 PM 6/2/2008, you wrote: >>Personally I *hate* turning boards over and clipping leads. >> >>And I'd much rather layout SMD than T/H. > > A man after my own heart. 0.5mm and even 0.4mm pin pitch is fine, QFN > is doable, but takes patience sadly BGA is a bit beyond the pale for > me right now until I get some more gear. Until I started the job I > have now I hadn't done any PTH in nearly 15 years - and now I know > why! And now I know why we have technicians to assemble prototypes, > but none of them can outdo me for speed and accuracy on an 0.5mm PQFP > FPGA. > > If you are going to do a lot of SMT work by hand, then a good stereo > microscope and a Metcal MX500 series iron are almost indispensable. > > John > > > > > ------------------------------ > > Message: 2 > Date: Tue, 03 Jun 2008 12:20:26 +1200 > From: Bruce Griffiths <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: Discussion of precise time and frequency measurement > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > christopher hoover wrote: >> John Miles wrote: >> >>> For one-off PCBs, I've had good luck with www.batchpcb.com . >>> >> >> I agree. I've used them once and have been happy with the >> results, you just can't be in any hurry. Unfortunately, >> there's no indication of how long it will take a priori. >> >> >>> SMD is not hard to work with by hand, down to 0603 or thereabouts >>> depending on eyesight and/or equipment. I find it easier to deal >>> with than through-hole, frankly. >>> >> >> I agree with the first part, but I actually find SMD to be easier >> than T/H, if you stick with 0603 and larger for the passives. I >> find SMD not only easier but quite bit a faster. >> >> SMD IC packages are easy once you get the hang of them, but a >> microscope is needed for inspecting and fixing fine pitched parts. >> (I think this is a bit of surprise to some folks.) >> >> Personally I *hate* turning boards over and clipping leads. >> >> And I'd much rather layout SMD than T/H. >> >> -ch >> > However for low frequency work the low thermal mass of resistors and > opamps in smt packages can be problematic. > Also SMT packages are more sensitive to board deflections and vibration. > > Using guard rings with some SMT parts is difficult to impossible. > You no longer have the option of directly connecting a leakage sensitive > lead to a virgin teflon standoff. > > Mixers and phase detectors with dc and low frequency isolated grounds > for the IF and RF ports dont appear to be available in SMT packages. > > How do you cope with SMT parts (eg high frequency ADCs) with metal > thermal transfer /ground connections under the package itself? > > Bruce > > > > ------------------------------ > > Message: 3 > Date: Tue, 03 Jun 2008 10:04:44 +0930 > From: Matthew Smith <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: Discussion of precise time and frequency measurement > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset=UTF-8; format=flowed > > Quoth Bruce Griffiths at 2008-06-03 09:50... > >> How do you cope with SMT parts (eg high frequency ADCs) with metal >> thermal transfer /ground connections under the package itself? > > Haven't done it myself, but interested to hear others experiences. I'm > guessing that this would be a job for solder paste and a toaster oven - > or high-tech equivalent. > > > -- > Matthew Smith > Smiffytech - Technology Consulting & Web Application Development > Business: http://www.smiffytech.com/ > Personal: http://www.smiffysplace.com/ > LinkedIn: http://www.linkedin.com/in/smiffy > > > > ------------------------------ > > Message: 4 > Date: Mon, 2 Jun 2008 20:34:43 -0400 > From: Bob Paddock <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: Discussion of precise time and frequency measurement > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset="iso-8859-1" > > On Monday 02 June 2008 02:31:18 pm Patrick wrote: > >> I have wanted to fabricate my own PCBs for several years now but I have >> never made an attempt. I am set up here to do silk screening and I have >> ovens and a hot-air soldering iron. Has anyone else tried to fabricate >> their own boards or is the price of farming the work out just so low now? > > I've played with doing them at work. Found it better to just farm them > out > to the prototype house. It is tough to get any real consistency from week > to > week. > >> If anyone has farmed out work, could you please feedback as to the entry >> level costs and if possible, some suggested companies? > > Cheapest place I've come across for easy boards with 10mil or large lines > is this one, but I've not used them yet (probably next month I will): > > http://www.futurlec.com/PCBService.shtml > >> > 1. Surface mount or through hole? I don't have a re-flow oven (or even >> > a >> > hot air soldering system), so my inclination is to use through hole > > SMT is actually easier if you have a good magnifying system. I use > macro-lenses on the wife's Cam-Corder: > > http://www.designer-iii.com/Solder/ Looks better than the picture of the > screen shows in reality. > > http://www.micromark.com/ has the type of tools that you need, like > insulated > and cross-tweezers (squeeze them to open them, the reverse of normal > tweezers) etc. > > Pick up a small convection oven at Wal-Mart or such place. > > If your making several get a Stencil: > > http://www.smtstencil.com/ > http://www.customlasercutting.com/ > >> Will using through hole cause me grief? > > In the long term, yes. A lot of older TH parts have not been updated to > ROHS > (Lead Free) and they never will be, they will be discontinued at some > point. > Virtually all new parts are SMT. > >> > 2. How many layers? In an ideal world with money no object, if I >> > understand the current art correctly, I think I'd probably aim for a >> > five >> > layer > > I assume that is a typo? You can not have an odd number of layers. > In this current 3D reality each layer has two sides. :-) > >> > board with Vcc, Digital Ground and Power Ground being separate >> > internal planes, and trace routing on the top and bottom of the board >> > with as few vias between top and bottom as possible. Does that sound >> > right? > > Separate ground planes, or split ground planes are a bad idea in this > context. > Always think about where the return current is going to flow. > Read the Analog Devices seminar notes on the subject. > > http://www.analog.com/analog_root/static/library/technicalSeminarSeries.html > is the link to the books but you can find the chapters on line. > > Read 7a & 7b if nothing else: > > http://www.analog.com/en/DCcList/0,3090,1073%255F%255F961,00.html > > > -- > http://www.wearablesmartsensors.com/ > http://www.softwaresafety.net/ http://www.designer-iii.com/ > http://www.unusualresearch.com/ > > > > ------------------------------ > > Message: 5 > Date: Tue, 03 Jun 2008 10:08:03 +0930 > From: Matthew Smith <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: Discussion of precise time and frequency measurement > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset=UTF-8; format=flowed > > Quoth Bob Paddock at 2008-06-03 10:04... > >> Cheapest place I've come across for easy boards with 10mil or large lines >> is this one, but I've not used them yet (probably next month I will): >> >> http://www.futurlec.com/PCBService.shtml > > I was looking at these - they actually do overseas shipments for > near-domestic freight prices. Great for those of us living in countries > where board houses require arm, leg and first-born for even small/simple > jobs. > > -- > Matthew Smith > Smiffytech - Technology Consulting & Web Application Development > Business: http://www.smiffytech.com/ > Personal: http://www.smiffysplace.com/ > LinkedIn: http://www.linkedin.com/in/smiffy > > > > ------------------------------ > > Message: 6 > Date: Mon, 2 Jun 2008 20:39:02 -0400 > From: Bob Paddock <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: Discussion of precise time and frequency measurement > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset="iso-8859-1" > > On Monday 02 June 2008 04:53:17 pm [EMAIL PROTECTED] wrote: > >> Not the cheapest, but great for "professional" proto's when quality >> trumps >> cost (above 1GHz, one source FR4 is totally different from another >> sources >> FR4...) > > Anyone have suggestions for Metal Core Protype Boards? > Used in high power LED applications. > > I know of this one, are there others?: > http://www.protoexpress.com/content/speciality.jsp > > http://www.cif.fr/new/produits_aff.php3?cat=1&scat=3&sscat=89&p=211 > > > > ------------------------------ > > Message: 7 > Date: Mon, 02 Jun 2008 21:05:09 -0400 > From: Chuck Harris <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: Discussion of precise time and frequency measurement > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Bob Paddock wrote: > >>>> 2. How many layers? In an ideal world with money no object, if I >>>> understand the current art correctly, I think I'd probably aim for a >>>> five >>>> layer >> >> I assume that is a typo? You can not have an odd number of layers. >> In this current 3D reality each layer has two sides. :-) > > > Sure you can. I have done 3 layer, 5 layer, and 7 layer boards. > > Having done those, I will advise you not to do 3 layer, it warps. But > everything else is easily do able. > > Multilayer boards are done by bonding together combinations of > 1 side, 2 side, and 0 sided laminate. For instance: > > 4 layer can be made by: > > 2 side | 0 side | 2 side > 1 side | 2 side | 1 side > 1 side | 1 side | 2 side > > ... > > 5 layer can be made by: > > 2 side | 1 side | 1 side | 1 side > 2 side | 0 side | 2 side | 1 side > 2 side | 1 side | 0 side | 2 side > > ... > > 6 layer can be made by: > > 2 side | 0 side | 2 side | 0 side | 2 side > ... > > The only down side to odd numbers of layers is the supply house > has to stock 1 sided laminate. > > -Chuck Harris > > > > ------------------------------ > > Message: 8 > Date: Mon, 2 Jun 2008 21:24:21 -0400 > From: Bob Paddock <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: Discussion of precise time and frequency measurement > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset="iso-8859-1" > > > >> How do you cope with SMT parts (eg high frequency ADCs) with metal >> thermal transfer /ground connections under the package itself? > > "How to succeed the first time with ultra-small QFN packages" > http://www.wirelessnetdesignline.com/showArticle.jhtml?articleID=202800018 > > > > > ------------------------------ > > Message: 9 > Date: Mon, 2 Jun 2008 18:36:46 -0700 > From: "Keith Payea" <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions thread II > To: "'Discussion of precise time and frequency measurement'" > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset="us-ascii" > > A really good way to handle the pad on the bottom of the part is to put a > hole through the board right there. You usually need a bunch of vias to > tie > the pad to the ground plane on the bottom of the board anyway. Put one > hole > large enough for your soldering iron tip to reach in and touch the > underside > of the part. When installing the part, solder the regular pads first, > then > flip the board over and solder the central pad. > > This is sort of self limiting too. If you apply too much heat for too > long, > the part desolders itself and falls off! > > Adding the extra hole is recommended by some of the vendors. I've used it > sucessfully with 2.4GHz band radios and it works great. > > Keith > > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Bob Paddock > Sent: Monday, June 02, 2008 6:24 PM > To: Discussion of precise time and frequency measurement > Subject: Re: [time-nuts] PCB design questions thread II > > > >> How do you cope with SMT parts (eg high frequency ADCs) with metal >> thermal transfer /ground connections under the package itself? > > "How to succeed the first time with ultra-small QFN packages" > http://www.wirelessnetdesignline.com/showArticle.jhtml?articleID=202800018 > > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > > > > > ------------------------------ > > Message: 10 > Date: Mon, 2 Jun 2008 20:52:56 -0500 > From: "Didier Juges" <[EMAIL PROTECTED]> > Subject: Re: [time-nuts] PCB design questions > To: "'Discussion of precise time and frequency measurement'" > <[email protected]> > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset="windows-1250" > > > >> -----Original Message----- >> From: [EMAIL PROTECTED] >> [mailto:[EMAIL PROTECTED] On Behalf Of David C. Partridge >> Sent: Monday, June 02, 2008 1:20 PM >> To: 'Discussion of precise time and frequency measurement' >> Subject: [time-nuts] PCB design questions >> >> I've been working on the design for a frequency divider to >> complement the Thunderbolt I recently bought from TVB (thank >> you Tom, it's working very well as far as I can tell, though >> of course I've no other standard to compare against). >> > > Dave, > > I just made one using a single microcontroller chip soldered on a > Toolstick > from Silabs. The Toolstick is a small PWB with a small 8051 class > microcontroller in surface mount package. The Toolstick costs $10 and you > need a $18 programmer to flash it. I run the uC using the 10 MHz as > external > clock and it has (for now) a 1Hz output. > > The toolstick I used is similar to the one described on this page: > > http://www.ko4bb.com/Test_Equipment/AFSignalGenerator/SigGen.php > > The software is just a couple of for() loops to kill time, a few 2 and 3 > cycles instructions to adjust the timing perfectly (cannot do NOP from C > with the free compiler, but you can program in assembly if you need to > feel > some pain) and one output is twiddled at the right time. The processor is > using the 10 MHz as clock. A single chip is all that is required, the chip > I > actually used is a C8051F530-IT, which costs about $4 on the Silabs web > site > (qty 1) and comes in a 20 pin TSSOP package (surface mount, but solderable > by hand). You really only need to connect 4 pins for that job, supply > voltage and ground, 10 MHz input and 1Hz output. Silabs has a couple of > parts that are available in DIP package, notably the C8051F330D (refered > to > as the 330-GP) which is roughly the same core as the 530. If you buy the > bare chip, you will have to buy the programmer separately, so if you just > want to build a few simple projects, you are better off buying the > toolsticks. > > You might want to add a D flip-flop to clean up the edge, even though I do > not know the jitter of the chip itself (I have planned to test it as well > as > I can, but there are more projects in the pipe at the moment than time to > address them) > > I am working on a version with both 1 kHz and 1 Hz outputs. > > There are a few sine-to-square converters (clock shapers) on Bruce's page > at > > > http://www.ko4bb.com/~bruce/ > > Didier KO4BB > > Internal Virus Database is out-of-date. > Checked by AVG. > Version: 7.5.519 / Virus Database: 269.23.11/1422 - Release Date: 5/8/2008 > 5:24 PM > > > > > > ------------------------------ > > _______________________________________________ > time-nuts mailing list > [email protected] > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > > End of time-nuts Digest, Vol 47, Issue 5 > **************************************** _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
