>
> You can queue up the pin change commands and send them over the USB as a
> bunch - it's not as fast as using the accelerated SPI modes, but at
> least you don't have a separate USB transaction for each change of the
> pin state.
>
It's on the reads, where you have to sequence the bit input with the
clock edge outputs, that it gets very slow. There may be ways of
optimising this, but I haven't found any that make much of a difference.
Paul Burke
Reply by David Brown●July 30, 20072007-07-30
Paul Burke wrote:
> Thomas Magma wrote:
>> While I wait... does anyone have any experience with their FT2232D
>> chip? We are looking to use it to do some FIFO dumps and some bit
>> banging. How are the API's to deal with?
>
>
> The 2232 is good in as far as it goes, which is SPI, I2C and JTAG.
> Anything other than that, beware. Although the data sheet includes the
> sentence, "The MPSSE interface is designed to be flexible so that it can
> be configured to allow any synchronous serial protocol (industry
> standard or proprietary) to be interfaced ...", in point of fact it is
> not flexible enough to accommodate many of the protocols used for serial
> devices like ADCs. In particular, any protocol that requires the clock
> to be high in the idle state, with data changing on the falling edge,
> has to be manually bit- banged, which is very slow indeed given the 1ms
> USB time frame.
>
> Paul Burke
You can queue up the pin change commands and send them over the USB as a
bunch - it's not as fast as using the accelerated SPI modes, but at
least you don't have a separate USB transaction for each change of the
pin state.
Reply by Uwe Bonnes●July 24, 20072007-07-24
Thomas Magma <somewhere@overtherainbow.com> wrote:
...
> The documentation for the DLP2232M says:
> "There are also several new modes which can be enabled in the external
> EEPROM, or by using DLL driver commands. These include Synchronous Bit-Bang
> Mode, a CPU-Style FIFO Interface Mode, a Multi-Protocol Synchronous Serial
> Engine Interface Mode, MCU Host Bus Emulation Mode, and Fast Opto-Isolated
> Serial Interface Mode."
> I have the eval board on the way and will soon find out for myself. I have
> also asked the question to tech support at FTDI.
There is no way to activate FIFO mode without the EEPROM
--
Uwe Bonnes bon@elektron.ikp.physik.tu-darmstadt.de
Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
Reply by Thomas Magma●July 24, 20072007-07-24
>>I'm curious as to why the datasheet for the evaluation board DLP2232M
>>states: "There are also several new modes which can be enabled in the
>>external EEPROM, or by using DLL driver commands." Would this not imply
>>that you can change modes by commands only and therefore do not require a
>>EEPROM?
>
> NO, if either of the ports needs to be in FIFO mode (245 mode) it must
> have
> an external EEPROM.
>
> The extra modes are how many different ways the Port A serial port can be
> configured for standard serial/JTAG/SPI/I2C....
>
Hi Paul,
The documentation for the DLP2232M says:
"There are also several new modes which can be enabled in the external
EEPROM, or by using DLL driver commands. These include Synchronous Bit-Bang
Mode, a CPU-Style FIFO Interface Mode, a Multi-Protocol Synchronous Serial
Engine Interface Mode, MCU Host Bus Emulation Mode, and Fast Opto-Isolated
Serial Interface Mode."
I have the eval board on the way and will soon find out for myself. I have
also asked the question to tech support at FTDI.
Thomas
Reply by ●July 24, 20072007-07-24
On Monday, in article <r09pi.253$fJ5.50@pd7urf1no>
somewhere@overtherainbow.com "Thomas Magma" wrote:
>>
>>> I think that they should have just put the EEPROM right on
>>> the chip. Or even have volatile configuration registers on
>>> the chip that you can configure via their driver command set.
>>> It's expensive to inventory and control new components.
>>
>> On embedded world 2007 the people at the FTDI booth told my that they plan
>> to redo the FT2232 with embedded EEPROM, unique hardware ID and two MPSSE
>> channels. I hope they keep promise.
>>
>> And let's hope they get mode setting right.
>>
>
>I'm curious as to why the datasheet for the evaluation board DLP2232M
>states: "There are also several new modes which can be enabled in the
>external EEPROM, or by using DLL driver commands." Would this not imply
>that you can change modes by commands only and therefore do not require a
>EEPROM?
NO, if either of the ports needs to be in FIFO mode (245 mode) it must have
an external EEPROM.
The extra modes are how many different ways the Port A serial port can be
configured for standard serial/JTAG/SPI/I2C....
I still think they should have pins for basic serial/parallel for less
inventory and manufacturing steps of programming several devices on board.
If they could get the driver and device to tell you what chip was at the
other end of the link, let alone have driver config and version numbers
reported in a better way would help.
--
Paul Carpenter | paul@pcserviceselectronics.co.uk
<http://www.pcserviceselectronics.co.uk/> PC Services
<http://www.gnuh8.org.uk/> GNU H8 & mailing list info
<http://www.badweb.org.uk/> For those web sites you hate
Reply by Jim Granville●July 23, 20072007-07-23
Uwe Bonnes wrote:
> Thomas Magma <somewhere@overtherainbow.com> wrote:
>
>
>>>Yes, unfortunately serial EEPROM and 3 x resistors. If you want
>>>2 x serial no EEPROM required, I personally wished that two pins
>>>had been used for selecting each port as serial/parallel.
>>>
>>>I'll add a note on my web page.
>>>
>
>
>>I think that they should have just put the EEPROM right on
>>the chip. Or even have volatile configuration registers on
>>the chip that you can configure via their driver command set.
>>It's expensive to inventory and control new components.
>
>
> On embedded world 2007 the people at the FTDI booth told my that they plan
> to redo the FT2232 with embedded EEPROM, unique hardware ID and two MPSSE
> channels. I hope they keep promise.
The Silabs CP21xx series have on board EEPROM. It also has
a PLL on board, so saves a Xtal.
Not sure about how the other finer points compare.
-jg
Reply by Thomas Magma●July 23, 20072007-07-23
>
>> I think that they should have just put the EEPROM right on
>> the chip. Or even have volatile configuration registers on
>> the chip that you can configure via their driver command set.
>> It's expensive to inventory and control new components.
>
> On embedded world 2007 the people at the FTDI booth told my that they plan
> to redo the FT2232 with embedded EEPROM, unique hardware ID and two MPSSE
> channels. I hope they keep promise.
>
> And let's hope they get mode setting right.
>
I'm curious as to why the datasheet for the evaluation board DLP2232M
states: "There are also several new modes which can be enabled in the
external EEPROM, or by using DLL driver commands." Would this not imply
that you can change modes by commands only and therefore do not require a
EEPROM?
Thomas
Reply by Uwe Bonnes●July 23, 20072007-07-23
Thomas Magma <somewhere@overtherainbow.com> wrote:
> > Yes, unfortunately serial EEPROM and 3 x resistors. If you want
> > 2 x serial no EEPROM required, I personally wished that two pins
> > had been used for selecting each port as serial/parallel.
> >
> > I'll add a note on my web page.
> >
> I think that they should have just put the EEPROM right on
> the chip. Or even have volatile configuration registers on
> the chip that you can configure via their driver command set.
> It's expensive to inventory and control new components.
On embedded world 2007 the people at the FTDI booth told my that they plan
to redo the FT2232 with embedded EEPROM, unique hardware ID and two MPSSE
channels. I hope they keep promise.
And let's hope they get mode setting right.
--
Uwe Bonnes bon@elektron.ikp.physik.tu-darmstadt.de
Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
Reply by Thomas Magma●July 23, 20072007-07-23
> Yes, unfortunately serial EEPROM and 3 x resistors. If you want 2 x serial
> no EEPROM required, I personally wished that two pins had been used for
> selecting each port as serial/parallel.
>
> I'll add a note on my web page.
>
I think that they should have just put the EEPROM right on the chip. Or even
have volatile configuration registers on the chip that you can configure via
their driver command set. It's expensive to inventory and control new
components.
Thomas
Reply by ●July 23, 20072007-07-23
On Monday, in article
<8rl8a3tcvte8drqa2v4pnerpocns7m7er2@4ax.com>
mike@whitewing.co.uk "Mike Harrison" wrote:
>On Sun, 22 Jul 2007 13:33:21 +0100 (BST), paul$@pcserviceselectronics.co.uk
> (Paul Carpenter) wrote:
>
>>On Friday, in article <iY5oi.133185$NV3.125271@pd7urf2no>
>> somewhere@overtherainbow.com "Thomas Magma" wrote:
>>
>>>Well the website is sort of working, at less than 1K Bytes per second it is
>>>very painful.
>>>
>>>Just looking at the FT2232D datasheet it looks like it might need an
>>>external EEPROM. Is this true?
>>
>>Yes, unfortunately serial EEPROM and 3 x resistors. If you want 2 x serial
>>no EEPROM required, I personally wished that two pins had been used for
>>selecting each port as serial/parallel.
>>
>>I'll add a note on my web page.
>>
>>>All I want to do is transfer half a MByte of
>>>data from a 8bit FIFO (< 1 second) (from a laptop) and do some bit banging
>>>for control.
>>
>>You may be at the limit of the device as these are 12Mbps (High speed
>>NOT full speed) devices. So 12Mbps transfer rate is for BOTH ports
>>combined.
>
>12M is full speed, 480M is high speed - 'full' referred to the maximum speed
> at the time of the 1.1
>standard. Confusing I know!
That is what I meant to put, you are right. I confusingly got them wrong
way round.
>>>I will get their evaluation board (DLP-2232M) and find out for
>>>myself but I thought someone might be able to give me a heads-up. Should I
>>>be looking into the Vinculum VNC1L that has the EEPROM built in? I would
>>>like to minimize the component count in my design.
>
>VNC1L is a totally different beast - it's a host controller
>
>Have you looked at the FT245R?
If he needs a FIFO port and bit banging at same time FT2232D or other
chip is better. Having done something similar hence my example page.
The thing you have to watch is how many USB 2.0 devices are actually
USB 1.1 that works with USB 2.0.