Carlos Posted March 21, 2021 Share Posted March 21, 2021 This post is for Sid, I may have a solution to your noise (emi) problem with the ribbon cable. What about single shielded aircraft wire. That should do the trick. Something like this should work https://www.steinair.com/product/22-ga-single-conductor-shielded/. You can either use 22 gauge or 20 gauge. Should be enough to drive the 12k units trouble free. just a thought Quote Link to comment Share on other sites More sharing options...
Sid Genetry Solar Posted March 21, 2021 Share Posted March 21, 2021 That's exactly what I was going to try next 😉. Debating between having a Chinese cable manufacturer make 70-100pcs sample cables, or me hand-crimp a test line first...but that's what I'll try. One single shielded wire for each MOS signal pair. Quote Link to comment Share on other sites More sharing options...
Carlos Posted March 21, 2021 Author Share Posted March 21, 2021 I would hand crimp a test batch first. Once everything works and tests fine then do a small batch from China. This solution should work as long as the shielded cable is of good quality. I know for the DIY bms project that shielded CAT 7 twisted pair solved the drop packets issue. Your issue is with noise and power as the ribbon cable is too thin and unshielded to deliver the required power/signal needed for the 12kw inverter. As for the 6k version a very short ribbon cable with beads will do the trick. I am very interested to see your design on the new powerboard that will replace the pj design. You can use the OZinverter powerboard as a reference design. Quote Link to comment Share on other sites More sharing options...
The Blind Wolf Posted March 21, 2021 Share Posted March 21, 2021 That is what Tesla uses on thier high voltage line, a shielded power wire. Didn't think about that till after this 😛 Quote Link to comment Share on other sites More sharing options...
InPhase Posted March 21, 2021 Share Posted March 21, 2021 How many wires in the ribbon cable, 10? Quote Link to comment Share on other sites More sharing options...
Carlos Posted March 21, 2021 Author Share Posted March 21, 2021 Yes I believe it’s 10. Sid would know for sure. 1 Quote Link to comment Share on other sites More sharing options...
TheButcher Posted March 21, 2021 Share Posted March 21, 2021 Perhaps screened ribbon would do the trick instead of separate coaxes for the shorter run? There's lots of variants, from foil to copper braided available from Mouser etc. If you are old enough cast your mind back to the shielded ribbon Apple used on the DISK II. Quote Link to comment Share on other sites More sharing options...
Sid Genetry Solar Posted March 21, 2021 Share Posted March 21, 2021 I actually got started on an Apple //e (after the pocket TRS-80)...but I don't really recall the cable on the double (5.25" floppy) disk drive being anything fancy. So the separate coaxes is going to be the best solution--part of the issue with the PJ design is that the H_MOS gate and source wires are crossed and intertwined in the worst possible way--not to mention 28-30AWG being far too small to handle 4A gate drivers. A ferrite over the whole ribbon does nothing...and wrapping foil around the entire ribbon similarly does absolutely nothing. At least right now, I believe that the FET drive lines need to be separately shielded. Yeah, twisted pair Cat 5-7 cable is an idea, BUT the pairs aren't separately shielded, nor are they thick enough. 3 hours ago, Carlos said: I am very interested to see your design on the new powerboard that will replace the pj design. You can use the OZinverter powerboard as a reference design Not sure the Ozinverter mainboard has much to offer...they all seem to use the IR2110 for MOS drive, which isn't strong enough to encounter the issues we are 😉. There were a few posts on TheBackShed about people using 10A+ driver chips (TC4451)...but I didn't see any conclusive results out of them. I wouldn't be doing much significantly different from the PJ board, except for routing the gate/source wires as signal pairs, repinning the connector, as well as upgrading the design slightly. Quote Link to comment Share on other sites More sharing options...
Carlos Posted March 21, 2021 Author Share Posted March 21, 2021 I was talking about the powerboard of the OZinverter. It was based on the PJ design. This one has your fix with the diodes. Quote Link to comment Share on other sites More sharing options...
Carlos Posted March 21, 2021 Author Share Posted March 21, 2021 But I am sure your design is going to be a lot better.😄 Quote Link to comment Share on other sites More sharing options...
Sid Genetry Solar Posted March 21, 2021 Share Posted March 21, 2021 Ha, not like I sent you that drawing awhile back 😉.... Quote Link to comment Share on other sites More sharing options...
TheButcher Posted March 21, 2021 Share Posted March 21, 2021 1 hour ago, Sid Genetry Solar said: I actually got started on an Apple //e (after the pocket TRS-80)...but I don't really recall the cable on the double (5.25" floppy) disk drive being anything fancy. It might have been more strict RFI regulations here. The Apples I had in for repair back then all had pretty substantial ribbon to the floppy drive with copper braid embedded in it. Quote Link to comment Share on other sites More sharing options...
Carlos Posted March 21, 2021 Author Share Posted March 21, 2021 That project was on hold. I will wait for your new design. 1 Quote Link to comment Share on other sites More sharing options...
Waterman Posted March 23, 2021 Share Posted March 23, 2021 On 3/21/2021 at 7:33 AM, Carlos said: This post is for Sid, I may have a solution to your noise (emi) problem with the ribbon cable. What about single shielded aircraft wire. That should do the trick. Something like this should work https://www.steinair.com/product/22-ga-single-conductor-shielded/. You can either use 22 gauge or 20 gauge. Should be enough to drive the 12k units trouble free. just a thought I would think that the https://www.steinair.com/product/22-ga-2-conductor-shielded/ would work as don't they really need two wires? Quote Link to comment Share on other sites More sharing options...
Waterman Posted March 23, 2021 Share Posted March 23, 2021 Would this cable work? https://www.showmecables.com/2-conductor-18-awg-stranded-shielded-pvc-gray-500-ft-west-penn-part-293 It is 18AWG with an Al wrapper and a bare conductor in contact with the shielding wrap. Quote Link to comment Share on other sites More sharing options...
Sid Genetry Solar Posted March 24, 2021 Share Posted March 24, 2021 On 3/22/2021 at 10:32 PM, Waterman said: I would think that the https://www.steinair.com/product/22-ga-2-conductor-shielded/ would work as don't they really need two wires? So technically, I'll probably end up mixing single-conductor shielded with 2-conductor shielded on the connectors. I've ordered some single-conductor shielded wire for a preliminary test. The signals go together, and the FET source wire is connected to a high amperage point--so technically, it doesn't need to be a 2-conductor shielded cable. I really don't care too much what gets induced on the wires...as long as the gate-source pairs move together. Differential EMI seems to be the issue we're up against. 1 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.