From ef1d245c73ff7cc37baed87c60646e762b24ae6c Mon Sep 17 00:00:00 2001 From: "Rich O." Date: Thu, 19 Nov 2026 22:30:01 -0500 Subject: [PATCH] docs: SB-LoRa rev 2 connector: 3M MCS-068SMA1-TR30 (female straddle receptacle, eBay surplus, rows not staggered) replaces the AliExpress part as the J1 candidate, which is dropped for $39 shipping and a 2+ week wait; the J1 pattern becomes opposed F.Cu/B.Cu rows and the board thickness follows the connector's row gap --- hardware/sb-lora/REV2.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/hardware/sb-lora/REV2.md b/hardware/sb-lora/REV2.md index e280e54..67fd0d2 100644 --- a/hardware/sb-lora/REV2.md +++ b/hardware/sb-lora/REV2.md @@ -68,9 +68,9 @@ A 16550-style UART on csSlot1 replaces rev 1's SC16IS740 SPI bridge and its latc ## Open -- **Connector: the Hirose IC1HA order was cancelled (2026-11-17)**, so rev 2 drops the straddle footprint for a single-sided SMT PC Card receptacle (both rows of tails on the top side), harvested or stocked. Candidate: a generic 68-pin female PCMCIA connector sold on AliExpress for CF-to-PC Card adapters (https://www.aliexpress.com/item/32843057045.html, about $8.50 for 5). The listing photos show it's the female side, with both rows of tails flat on the top, but there's no drawing, so the land pattern has to be measured from a sample. The photos show an AMP marking with no part number, while the listing calls it "Jiansu" model "BST", so the photo may be of a different part from what ships. Measure the parts that arrive, check any later batch against them, and buy enough for every board in one order in case the stock is pulls or old stock. Hirose IC1K-68RD-1.27SF(71) (LCSC C17285563) is listed as "Connector and Ejector", which reads as the host-side socket; it isn't confirmed either way. Ruled out: TE 5535657 (the host-side header), Hirose IC1F-68RD-1.27SF and IC9 (obsolete), Amphenol 10014744-007T and 71299-050CALF (out of production), and Amphenol 7111S1615A01LF (a SIM card tray). "Amphenol 1C1K" never turned up as a real part number. The Springboard uses the same 68-pin PCMCIA contacts (Dev Guide §5, §7.1), and the keying is in the module's plastic case, not the connector. Needed from a sample: markings, tail pitch, row spacing, tail distance from the mating face, and board surface height relative to the contact centre line. +- **Connector: the Hirose IC1HA order was cancelled (2026-11-17).** Candidate: 3M MCS-068SMA1-TR30, a 68-contact female PCMCIA receptacle for memory cards ("MEMPAK"), straddle mount, sold on eBay as 3M/Robinson Helicopter surplus, so probably genuine new-old-stock. The listing photos show the top and bottom rows of tails directly above each other, with no stagger. There's no datasheet (3M's MCH-068 parts are the host-side headers), so the land pattern has to be measured from a sample. Buy enough for every board in one order. Ruled out: a generic AliExpress 68-pin female part (single-sided, but $39 shipping and over 2 weeks, and its photos show an AMP-marked part while it's sold as "Jiansu BST"), TE 5535657 (the host-side header), Hirose IC1K-68RD-1.27SF(71) (LCSC lists it as "Connector and Ejector", which reads as the host-side socket), Hirose IC1F-68RD-1.27SF and IC9 (obsolete), Amphenol 10014744-007T and 71299-050CALF (out of production), and Amphenol 7111S1615A01LF (a SIM card tray). The Springboard uses the same 68-pin PCMCIA contacts (Dev Guide §5, §7.1), and the keying is in the module's plastic case, not the connector. Needed from a sample: markings, the gap between the two rows (the board thickness it takes), tail pitch, tail distance from the mating face, tail width and length, and the body size and any pegs. - **Case.** The keying is in the module case, so rev 2 needs a case (printed, or a harvested Springboard shell) whose key and latch match the slot, with the board held at the connector's height. Dry-fit a bare harvested connector in the slot to check the contacts seat fully. -- **Board thickness.** The straddle connector forces 1.6 mm; a top-mount one may allow 0.8-1.0 mm. +- **Board thickness.** A straddle connector sets it: the board is whatever fits between the two rows. The Hirose IC1HA took 1.6 mm, but PC Card internal boards are often about 0.8 mm, so measure the 3M part's gap before assuming 1.6 mm. - **LiPo size.** The aim is at least a few days per charge. Estimate, to be replaced by a measurement: - The XIAO and SX1262 listen all the time (MeshCore companion doesn't sleep the radio): roughly 5-6 mA for the SX1262 in receive with boosted gain, plus 1-3 mA for the nRF52840 with the UART open. Transmits at 22 dBm draw about 120 mA but are short at SF10/BW500 and rare for a companion. Call it 7-10 mA on average. - 3 days at 10 mA is 720 mAh; with margin for usable capacity and ageing, about 1000 mAh. 500 mAh gives about 2 days. @@ -87,10 +87,10 @@ A 16550-style UART on csSlot1 replaces rev 1's SC16IS740 SPI bridge and its latc - **Renders:** `rev2/renders/`. Before ordering: -1. **J1.** Get samples of the AliExpress candidate (above) and measure one with calipers. Then replace the `J1_*` constants at the top of `build_board.py` from its drawing: pitch, row stagger, each row's distance from the edge, and pad size. Keep the 5.5 mm strip along the edge free (`J1_KEEP`), and check whether the connector body needs the board edge moved. Then run `make.sh` and `make_fab.py` again. The current pattern is a placeholder: two staggered rows of 0.5 x 1.6 mm pads, 1.3 and 3.5 mm in from the edge. TE 5535657 is not a candidate: it's the host-side header. +1. **J1.** Get samples of the 3M MCS-068SMA1 (above) and measure one with calipers. The straddle needs a new pattern in `build_board.py`: one row of pads on F.Cu and one on B.Cu, directly opposite each other, in place of the current two staggered rows on top. Set the `J1_*` constants from the sample: pitch, each row's distance from the edge, and pad size. Keep the 5.5 mm strip along the edge free (`J1_KEEP`) on both sides, and check whether the connector body needs the board edge moved. Then run `make.sh` and `make_fab.py` again. The current pattern is a placeholder: two staggered rows of 0.5 x 1.6 mm pads on the top, 1.3 and 3.5 mm in from the edge. 2. **XIAO and Wio footprints.** Check them against the real boards with calipers (print the top copper at 1:1). 3. **Assembly.** In JLC's placement preview, check the rotation of U1 (pin 1 at the top left, beside the triangle), U2, U3, U5 and Y1. -4. **Board thickness.** It stays 1.6 mm unless the chosen connector wants a thinner board. +4. **Board thickness.** Match it to the 3M part's gap between the rows. If that's thinner than 1.6 mm, change `SetBoardThickness` in `build_board.py` and the thickness in the JLC order to suit. 5. Tag the commit you order from as `hw-v0.2`. ## Carried over from rev 1