Page 2 of 2

Re: TF530 memory options

Posted: 07 Feb 2019 12:20
by terriblefire
Not 1uF caps. X7R multilayered 1uF caps. There is a huge difference. It’s basically the same as putting 7 caps in the same spot with a cumulative capacitance of 1uF

Re: TF530 memory options

Posted: 08 Mar 2020 11:05
by mkl
Fake SRAMs.

Aliexpress or was it eBay, where I bought 5 pcs AS7C4096B-12TCN.
B-suffixed part number does not appear on Alliance website, so one can guess the marking is fake. On the underside of the tsop pacakge, it reads GS74108AGP, which is a 3.3V chip. That marking could be fake too, but at least these srams appear to "work", even if at wrong voltage 5V.

AS7C4096A-12TCN, datecode 0928. These chips appear unused, pins are not shiny, genuine looking marking. I bought 50pcs from chinese seller who are at Alibaba. I got working KM41C4100AS-6 (and later XC9572XL) parts from them previously. But these don't work. Even CHIP ram stops working with these, like with AS7C4096A-TIN from utsource https://github.com/terriblefire/tf530/w ... -TF530-Q&A

Pictures added:

IMG_20200308_123745_copy_800x1048.jpg
IMG_20200308_123844_copy_800x1139.jpg

One chip from another 5 piece lot from ebay or aliexpress has same date code. Old marking can be barely seen, Samsung K6R4....

IMG_20200308_162055.jpg

Re: TF530 memory options

Posted: 06 Apr 2020 21:48
by mkl
Fakes are probably made from 16-bit wide tsop44 srams.
Pin 6 in 16 bit srams is chip select. (address line in 8 bit tsop44 srams)
Bypassing the address line to cpld and connecting pin 6 on srams to vcc makes the board be able to test chip ram ok again.

20200406_230639-1-800x366.jpg

Re: TF530 memory options

Posted: 10 Apr 2021 02:00
by Eduardo L M
Sorry to resurect this thread, but did you ever reach a conclusion about the B variation? Did it really work on the long term? Thanks for any info on this!

Eduardo

Re: TF530 memory options

Posted: 25 May 2023 17:20
by PierreGTT
Hello,

Seems I am facing same kind of issue about fake AS7C4096A-12TCN I got from aliexpress.

I am using this kind of chip in a personnal design. When soldered, accessing other Mem (Flash) on the same bus cause issues even some overheating of the AS7C4096A. So I put a scope on what is referred as pin #37 (/OE) and I was able to see sometimes this pin behaves as an output which is contradictory with its pure input definition. This let me into wondering if the pinout was really that of AS7C4096A.

Then I came up to this thread which conforted me into I got fake AS74096A. The marking looks fully genuine. So if it is a 16bits version sold as an 8 bits one, it is AS7C4098A? Checked for '6' digit which seems genuine.

Don't know why there are doing those tricks, seller most likely is not aware of selling fake chips.
There are too much differences in pinouts between 16 and 8 bits version, cannot be rewired.

So I ordered AS7C4096A-12TIN from RS components. Hope I will got a genuine one!

Pierre