Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-07-02 Thread Philip Pemberton
On 29/06/10 19:26, Sébastien Bourdeauducq wrote: Soldering wire to 16 pins on a QFP sure sounds like a lot of fun, but with some patience I should be able to make it... Use a Roadrunner wiring pen loaded with the finest wire, pre-tin the ends of the wire, and apply lots of flux (preferably RM

Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-07-02 Thread Adam Wang
Hi, >> Anyway, maybe the simplest solution is to let Adam do it on 1-2 boards as >> he >> has all the tools and experience. Just in order to validate the video >> input Just finished 1 pcs, well...rework this like a doctor doing a surgical operation without hand-shaking. hehe.. will rework anothe

Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-07-01 Thread Bengt Sjölén
what cable are you using? Enameled wire scavenged from a transformer... which isn't the best (bends too easily during soldering). I was also planning to use wirewrap wire but could not find any in the (bad) electronics shops nearby. I would have to order it. enameled wire is actually what

Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-07-01 Thread Sébastien Bourdeauducq
On Thursday 01 July 2010 00:22:45 Bengt Sjölén wrote: > what cable are you using? Enameled wire scavenged from a transformer... which isn't the best (bends too easily during soldering). I was also planning to use wirewrap wire but could not find any in the (bad) electronics shops nearby. I would

Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-06-30 Thread Adam Wang
Hi Sebastien, I'll rework my two pcs here and send to you firstly. My concerns will only be the noise reduction. Since those P[7:0] traces are located at L3 and well surrounded by ground shieldings to eliminate digital noise. This reworks mean that a short distance of them will be exposed. Well, I

Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-06-30 Thread Bengt Sjölén
yeah that's a tricky one especially if you have little or no pad sticking out underneath which is often the case. and very soon you will start loosing pads too. what cable are you using? i normally use single core thin wire wrap wire to patch these kind of things. i did patch one 0.5 mm pit

Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-06-30 Thread Sébastien Bourdeauducq
On Wednesday 30 June 2010 05:44:10 Adam Wang wrote: > Directly solder wires from P[15:8] to P[7:0] on pins should be ok. Well that's harder than I thought and I've been quite unsuccessful at it :( Soldering wires to QFPs is not like soldering QFPs on PCB pads... short circuits you struggle to rem

Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-06-29 Thread Adam Wang
Hi, > Since routing of P[7:0] is went through L3 inside pcb, so I suggest no > need to cut traces on P[7:0] or directly soldering. That's a little > kind of hard. Take apart the one on pcb first then diretly use a new > U21, "raise up" those pins P[7:0] and P[15:8] on IC body first. Then > hand so

Re: [Milkymist-devel] PCB problem with pixel port of video in

2010-06-29 Thread Adam Wang
Hi, On Wed, Jun 30, 2010 at 2:26 AM, Sébastien Bourdeauducq wrote: > Hi, > > There is a stupid PCB problem with the video in pixel port: we connected > P[7:0], but the video decoder outputs on P[15:8] when in 8-bit mode. Nobody > spotted this one :( Stupid. Me either. > Since P[7:0] are high-im

[Milkymist-devel] PCB problem with pixel port of video in

2010-06-29 Thread Sébastien Bourdeauducq
Hi, There is a stupid PCB problem with the video in pixel port: we connected P[7:0], but the video decoder outputs on P[15:8] when in 8-bit mode. Nobody spotted this one :( Besides that, the decoder circuit appears to be working: the crystal oscillator is stable and clean (checked on scope), t