A couple people will be building the Bus Blaster v1: R (emails below) and swilson (PCB in transit)
(Just a side question: L1 and L2 look like ferrite beads. If you don't mind me asking, what's their impedance? I can't find it on the schematic.)
I just use any ferrite, here's the common ones we use in the lab:
http://dangerousprototypes.com/docs/Par ... rrite_Bead (http://dangerousprototypes.com/docs/Partlist#Ferrite_Bead)
Sorry to keep bothering you. Is the X1 crystal a 12Mhz ABM7-12.000MHZ-D2Y-T? It's listed on the schematic and parts list as CRYSTAL_212M_, and I don't know whether to interpret that as 212Mhz, 21Mhz, etc. I couldn't find it on the common parts list you provided.
Just wanted to make sure I don't order the wrong parts and let the board go to waste.
It is a new SMD crystal (12MHz) that Seeed provided (low cost to them). I believe the one you linked will work, or at least you should be able to wedge it in there :)
Just wanted to let you know that the board arrived today and to ask a (Last one, I promise!) question.
Would you happen to have any tips on repairing a board's lifted pads or how to avoid that in the future? (I had soldered everything except the last buffer chip when things started going downhill XD)
I'm sorry, I always trash boards when I lift pads. The best way to avoid it is an adjustable iron, I haven't lifted one since then (maybe unconnected SSOP pads, but those don't count).
Maybe you're luck: Only IC4, IC5 and IC6 are needed for most JTAG stuff. In fact, you can get by with IC4 and IC6 probably for most minor stuff. IC8 isn't ever used at all in any configuration.
[quote author="ian"]
Maybe you're luck: Only IC4, IC5 and IC6 are needed for most JTAG stuff. In fact, you can get by with IC4 and IC6 probably for most minor stuff. IC8 isn't ever used at all in any configuration.
[/quote]
Now if only IC4 weren't the one I completely messed up. XD
Do you want another PCB?
If it's not too much trouble and you have one lying around...
I'll be more careful this time.
Also, if anyone else is trying to desolder the FT2232H without a hot-air rework station, candles work surprisingly well. (Just make sure you wrap aluminum foil on every other part of the board.)
No problem. Please PM me your address again, I don't keep any addresses on file for privacy reasons.
I got a PM that there was progress, good to hear :) I noticed that the link on candles has been redirected to a spamware/spyware place, so I removed it.
Hi Folks,
I have one of the prototype (white) boards, and started building it this weekend. I'll post progress here, if it's appropriate.
-Crawford
Great to hear, please keep us updated.
OK, so my Bus Blaster is built. This is a 'white board' prototype board.
No magic smoke released on the power up test, so, I plugged it into the shop PC (win7 x86), to see if enumerates. Well, it doesn't, and I vaguely recall that I need to program the raw FT2232 chip with something (MPROG, IIRC). There are some parameters that need to be put into the chip.
Question 1: So what are the parameters, please?
Question 2: IC6,7&8 are covered with solder mask on my board, so I didn't populate them. Do I need to do so, and if so, how?
Thanks!
Crawford
did you get "unknown device" when plugging it in? if not the default mode of the ft2232 requires drivers from ftdi or compatible drivers, it should enumerate after installing them.
Question 1: So what are the parameters, please?
The FT2232H should start up as two COM ports (with the FTDI driver) by default. Windows should see a device, ask for a driver, etc.
If it doesn't work, check the soldering, the 12K bias resistor (R1), the crystal, USB jack, power supply (IC3, L1, L2). Those are the things that really matter on the board.
Question 2: IC6,7&8 are covered with solder mask on my board, so I didn't populate them. Do I need to do so, and if so, how?
IC6-8 are the input buffers for the JTAG signals. The minimum signals required to use JTAG are TDI, TDO, TCK, TMS. At minimum, IC7 should be populated to use JTAG. Some devices also have a bi-directional reset, which is also on IC7. IC8 isn't needed for any current functions, IC6 is for RTCK and another reset, which are less common signals:
http://dangerousprototypes.com/docs/Bus ... n#Hardware (http://dangerousprototypes.com/docs/Bus_Blaster_v1_design#Hardware)
Probably the best way to mount them is to scrape the mask away from the pins, tin them with solder, then solder on the buffer. The board isn't very useful without the input buffer, so it's probably at least work a shot.
Hi Ian
The FT2232H board is working, that was a hard bit of work, the pitch of this ic is very narrow,...
Tokk me about 6 hours of coursing and desperation but finally it enumerated !!!!
[10689.473661] usb 1-2: FTDI USB Serial Device converter now attached to ttyUSB0
[10689.473856] ftdi_sio 1-2:1.1: FTDI USB Serial Device converter detected
[10689.473901] usb 1-2: Detected FT2232H
[10689.473906] usb 1-2: Number of endpoints 2
[10689.473910] usb 1-2: Endpoint 1 MaxPacketSize 512
[10689.473914] usb 1-2: Endpoint 2 MaxPacketSize 512
[10689.473917] usb 1-2: Setting MaxPacketSize 512
[10689.474127] usb 1-2: FTDI USB Serial Device converter now attached to ttyUSB1
There was a short between pin1 and 2 I simply could not make out with my strongest lenses, without ohm meter I would have never found it.I even unsoldered the chip and reflowed it again.
Anyhow it is working now and that is what counts.
The jtag board for the guru plug embedded linux system uses the ft2232h I hope I can reverse engineer it and use your board as backup.
This is an email I got about building the FT2232H breakout. It's not the Bus Blaster, but it will be used the same way so I'm posting it here to contribute to the conversation.
The FT2232H footprint is a little tight for hand soldering. We made a new version for future prototypes with longer legs. I've done 5 or 6 of them in the last month and it makes me nutty :) I use head magnifier and jewelers loupe.
I'm just getting 'unrecognized USB device' and it's not enumerating.
At first I thought that it might have been the loading caps on the oscillator. I had no 27pF caps so used 22pF. Later I remembered my basic electronics and soldered 5pF in parallel (on top of actually) the 22pF's to get 27pF...
Still no go. I printed out the schematic on 11x17 inch paper so I can read it and start checking the pins of the FT2232. I did walk around it with the ohm-meter for shorts earlier, but guess I need to go deeper.
Small correction - i have the flash rom, but didn't install it, but did install all of the caps and resistors for it. At this point it would be easier to put in the flash chip than take the other things out. Maybe I'll try that first.
@Ian thanks for the tips of what to look at first.
-Crawford
some times pull ups can confuse devices if not populated. if they are on the data lines, they will read back 0xff or all logic high. worth checking the ft2232h data sheet errata.
The 22pFs should be ok, and you don't need to the EEPROM to get it to enumerate.
If you have a magnifier or good eyes - try using a needle to move the pins ont he USB jack. The footprint is quite tight (we're using a new one int he future for easier soldering), it keeps giving me problems on lots of PCBs. You can use the needle to test the FT2232H pins too.
i used to have all these problems before i had access to a stereoscope, good quality magnification can save a lot of time. not its safer than soldering with a jewelers loupe!
It's alive!
So, some more time with the magnifying glasses, and checking around the ft2232 with the needle probes (check the pins *and* the lands) showed that 1.8V was not getting to all of the pins. Re-soldered all the pins just to be sure, checked for shorts, and volts (all there now!).
Plugged it and an it enumerated! FTDI drivers loaded on win7, and it looks good (COM4 & 5 now present).
(I did solder in the eeprom just for fun, know it's not needed).
My Target Present LED fades in and out on occasion with nothing plugged in (I think I just guessed at the SOT23 transistor for that circuit).
Next I need to get the ic7 buffer soldered in ...
-Crawford
Thanks for the update, I'm glad it's working for you.
For the LED transistor, I used a BC818xx (BC81816 or BC81840), but as long as the pinout is the same it shouldn't matter too much.
For the Target Present transistor I used a SOT23 version of the 2N3904, which is pretty generic.
I may give it a shot with OpenOCD or some other JTAG tool today.
Cheers,
Crawford
So, I was beginning to despair of this prototype version of Bus Blaster ever fully working. None of the several JTAG-compatible devices would be detected in any tool I tried, not even my OpenBench Logic Sniffer.
But, I started buzzing out the JTAG connections and buffers, and guess what? The TDI pin and the TMS pin are reversed, that is, connected to the wrong pins of the FT2232 ! Grr! Arrgh! (Note that I am not bitching about a board I got for free on PCB Sunday, just the time it took me to find this).
After a simple swap of the TDI/TMS wires, the OpenBench, the AT91SAM7S board, the Coolrunner x-board, they all detect now.
So, this perhaps begs the question: Is there a simple wire-back test (not unlike wiring Tx to Rx in rs-232) to test out the connectivity of the board? There are 2 ports on it... (I know this is probably overcome by events with the CPLD version).
-Crawford
I'm really sorry about that, I was aware of this issue and thought it was listed in the Free PCB Sunday post. My epiphany on this issue arrived here, after a week of on-and-off debugging:
http://dangerousprototypes.com/forum/in ... 8#msg13688 (http://dangerousprototypes.com/forum/index.php?topic=403.msg13688#msg13688)
Is there a simple wire-back test (not unlike wiring Tx to Rx in rs-232) to test out the connectivity of the board? There are 2 ports on it... (I know this is probably overcome by events with the CPLD version).
As far as I know there is no simple loop-back test for JTAG. In the factory we have been testing against actual JTAG targets. The secondary port on v1 is unused and unexposed.
On v2 we can load a special CPLD logic that tests connectivity, but that is still a bit in the future too.