SHS3321 Firmware rollback

2 Likes

Do you still need help? Iā€™ll be happy to offer advice if you need it still.

I get the connector but on Thursday, so Iā€™ll make my cable then a d get back on here for any questions- thanks!!

1 Like

Just how many of your local Currency Units did it take just out of curiosity?

I really donā€™t want to say- yā€™all gonna make fun of me.

But the pick it was 40$ and the connectors were 10 (bought a bunch because why not)

Thatā€™s not T bag, if I wanted to have a PICKIT in about the same time as I can get the connector it was gonna be ~150 Dollarydoos (~120 Freedom Dollars at the time) and a week and a half wait. The slow boat from china PICKIT I ended up with took 2.5 months to arrive but was only 15 Freedom Dollars.

1 Like

Well - Houston, we have a problemā€¦

The problem is, the key industry lost a customer today. Thanks all who got me the help I needed.
The firmware patch went perfectly and I didnā€™t lose any of the codes I had previously created.

Woo welcome to the future!

4 Likes

Great news mate! Lmk if you ran into any speed bumps I need to add to that wiki!
Test_Pass++!!

2 Likes

None it was perfect

2 Likes

I might need to do this,

Curious if I could recore the lock to comply with my current bitting, might be able to get away with it in my apt if Iā€™m using the same key

2 Likes

@Eriequiet The SHS3320 uses a KIK cylinder Iā€™m pretty sure. So you should just be able to swap the tail piece between that and your current lock. I

ll post some photos of the inside of my spair 3321 when I get Home so you can compare.

Edit: this is the 3321, which has a blanking plate rather than the key barrel face plate, but the internal void fits a KIK cylinder!

2 Likes

I really gotta figure out how those 4 layer PCB antennas with the copper plane covering the top and bottom work. The ACR122U has the same design.

@TamablePumpkin, can you add me as a Contributor to your Github repo? I ran into an issue that I was able to solve, and want to add a bit to the wiki. Username on Github is the same as here, thanks!

1 Like

Iā€™d rather you submit an issue on github than me give you elevated privileges at this time. I know thatā€™s not a great solution however Iā€™ll work on a method for pull requests to be made to the wiki as well as the code base

1 Like

The TownSteel ESmart 5000 uses a Schlage 5 pin keyway and is easy to rekey. Itā€™s harder to swap the cord out as a 6 pin core is fairly common in SC1 locks (Itā€™s an SC4 with only 5 pins in place) but the SC1 core is shorter. I have no issues with it and ring/tags.

About to go through this proces. Just need to wait on the PicKit. That said Iā€™m trying to understand all of this stuff and was looking for some info. Currently the lock ā€œunflashedā€ will read the chip but wont accept it as a tag. Can anyone explain why exactly? is it becuase the NExT serial number is too long? Different protocol? I know I need to roll back the firmware, just not entirely sure what that does and why it makes the device compatible.

I answered this in the Discord, so Iā€™ll transplant it here:

"So just brief context:

  • First there was LF (125kHz) with lots of incompatible competing standards
  • Then NXP created the MIFARE structure at HF (13.56MHz) which uses a 4 byte NUID which was too small so now theyā€™re recycling IDs. It has memory sectors broken up into blocks. They can still be read by the ISO 14443a compliant readers, but they need some MIFARE specific commands to function completely
  • Now there are tons of different NFC tags that comply to ISO 14, but many of the ones by NXP (like the NTAG216 in the NExT) use a 7 byte UID and a memory structure (usually with 1 continuous sector) broken up into pages, which respond to generic ISO 14 commands

Theyā€™re fundamentally different, but also Samsung specifically didnā€™t want it to work, so it doesnā€™t. They want you to buy cards from them like razor blades"

1 Like

Without reverse engineering the software itā€™s hard to know for sure what exactly they changed, but given that we can roll back the firmware of the current hardware revision I suspect that version 1 had a bug that didnā€™t check the length of the UUID.

Now that Iā€™m writing this I should probably check that using a 7 byte UUID card doesnā€™t over flow and corrupt the next ID location.

Any new users should avoid using sequential ID locations until thatā€™s been tested.

2 Likes

https://www.amazon.com/gp/product/B08RMQP6YP/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

are these the correct connectors?

edit: posting a pic as well so none needs to click the link
image

1 Like

Considering that the hardware has the ability to read the chip and the software denies accepting the code (in the current, unflashed version) I think this points to the simple solution that Samsung specifically only wants their chips or chip numbers to work with the lock, to get more people to buy their specific products.

I could see them arguing about something from a safety/security perspective, but if the software used to read any chip, and now only reads Samsung chips, thatā€™s a pretty good ā€œwhyā€ it seems.