This is the world edition Blackberry 8830 from Dec 2007 (So just after the iPhone had launched and a just off the peak of Blackberry popularity.)
Some teardowns go better than others, this one as you shall see is definitely one of the others 🙂
The label inside the battery compartment indicates the unit was assembled in Mexico.
The front and back cover can then also be removed (After finding and removing two more screws in the base).
You are then left with a frame that supports the PCB and a self contained unit of the base of the keyboard and the display.
In the right picture you can see a thin white coaxial wire that connects to a small board that must be containing or connecting to the antenna (Or antennas as this probably has two antenna.)
Here is the small board on the frame, it appears to have just a handful of passives, the mini RF connector and two gold plated contacts that contact back to the main PCB.
Here is the back of the small board and the plated area on the frame, it is possible that the antenna is built into the frame.
Pulling this all apart exposes the single PCB.
Here is where my troubles started. Look at all the metal cans on this board! Normally the cans (Faraday shields) are thin pieces of tin that are tack soldered in a few places. I find the easiest way to remove them is to simply cut them off with wire snips. These are thicker tin and are soldered to the board all around the base! My soldering iron with a wicking copper strip did nothing. They are too close together to attack them with the dremel. In the end I blasted them with a heat gun – and it took a lot of heat to remove them pretty badly damaging the board in several places. I think repairing one of these boards would be a nightmare – but I guess these weren’t designed to be repaired.
Here is the front of the board with the cans removed
The battery connection is very top right, below that (middle right) you can see the vibration oscillator motor that is set in the board. The area beside that (top right) contains the RF Power Amplifiers in this section you have
- Skyworks SKY-77328 PAM for Quad band GSM/GPRS
- Avago ACPM 7813 RF Amp CDMA 824-849MHz
- Avago ACPM 7833 RF Amp CDMA 1850-1910MHz
- Unknown chip labeled M3512
Together with a couple of metal packages containing SAW filters.
To the left of this in the badly damaged part of the PCB is the
- Qualcomm RFR6500 RF receiver and downconverter.
Below this is the main application processor the
- Qualcomm MSM6550 (application processor)
- Qualcomm PM6650 (Power Manager)
- Spansion 98WS512PE NOR Flash
- Unknown 16 pin device labelled BOW TI
The four small canned areas down the left appear to contain mostly passive devices and I don’t know their functions.
Here is the rear of the board
There is one die in the top left canned area
- Qualcomm RTR6350 the RF transmitter
This is together with a couple of SAW filters
So basically the Blackberry 8830 implements the Qualcomm MSM6500 chipset, here is the block diagram of this chipset showing the functions on each chip
Since this was a CDMA phone (The world edition was very misleading, as I recall it didn’t work in all countries) it is perhaps not that surprising to see a Qualcomm chipset.
I looked at a 2005 Samsung phone containing an earlier version of Qualcomm’s CDMA chipset (MSM6050, L6000 (LNA), RFR6000, RFT6100) earlier this year, so I thought I would compare the RF die.
Getting chips off of the Blackberry PCB proved very difficult. Normally I hit the chips with 20-40 seconds with the hot air gun, the solder melts and the chips can be slid off. I don’t know if was related to the heat received removing the cans, or just the board type, but these chips did not want to budge. I tried cutting them out of the PCB with a dremmel and even after decapsulation in hot acid, the chip would not come off what was left of the glass fibre board.
Anyway I did manage to successfully decap the RFR6500 RF receiver and a good looking die it is!
The RFR6500 was touted as the first single chip implementation of cellular diversity in a press release in 2005. Antenna Diversity is a technique of using two different antenna to always capture the best wireless signal. Ensuring a good connection improves download speed for data.
Other than the addition of antenna diversity the RFR6500 now has the LNA integrated (The RFR6000 was used with a separate L6000 LNA chip) the wireless functionality is the same so I thought it would be interesting to compare the chips.
Here is the die marks from the RFR6500 – codename Firebird, designed and released in 2005.
The designer initials are now gone – it looks like a PHB has gone enough is enough and put a stop to it 🙂
And here is the die mark for the RFR6100 -which was designed in 2004 just a year earlier.
Here are the RFR6100 and L6000 die photos (From the Samsung APH-660)
Given these contain the same functions as what is on the RFR6500 I was expecting to see a lot of re-use of design blocks. Looking them over I can only see one clear piece of re-use. This circuit block is used in RFR6500 and RFR6100.
I am sure there are likely other blocks re-used but none that are obvious, which is quite surprising. IT looks like even the LNA was completely re-designed as I cannot see any re-use of the L6000 circuits.