Skip to main content

Messages

This section allows you to view all Messages made by this member. Note that you can only see Messages made in areas you currently have access to.

Messages - Gridstop

1
Bus Pirate Support / Re: What is the current recommendation v3 or v4?
Personally I think the move to STM 32 is really, really exciting. I tried a few times to get into PIC, but I'm really liking eclipse & openocd & arm-none-eabi toolchain. It's been great.

Frankly, I don't see a problem with keeping a FTDI chip, the new FT2--X chips are quite a bit cheaper than the classic 232RL used in the v3. (Unless you've already switched to an X in the newest versions?) With the STM32's having DMA for the host side interface, that doesn't seem to have much penalty vs integrated USB, really. The only thing missing is the programmable peripheral pin? Or is there something like that some of the STM32 chips?

Now I realize I'm making an assumption when you say 'STM' and not 'STM32' specifically.
2
Bus Blaster JTAG debugger / JTAG circulate/echo operation?
I'm investigating using my bus blaster to replace an old parallel-port JTAG setup. (It's for an old proprietary ASIC)

For the most part, everything seems fine, the original authors of the test software were nice enough to carefully contain everything in three main routines:
scan_ir (just what you think it is)
scan_dr (also what you think it is)

But then there is a 'circulate_dr' function, that's a scan_dr read-only, but it's setup so just before each clock, it echos the input to the output. I guess that would allow the jtag to cycle the entire chain through and put all the original data back in the shift registers when reading. This doesn't seem to be standard, from what I can tell, but it's used repeatedly for this device.

The FT2232 doesn't do this echo on a normal JTAG_Read (using FTDI's FTCJTAG.dll). Perhaps there's a way to get it to do it directly via the D2XX.dlls and lower level MPSSE stuff?

I could easily fake this with the CPLD, by having a GPIO set a mux that sources TDI from either the FT2232 or an echo of the TDO input. That's one thing awsome about this setup. But I'm wondering if anyone else has seen anything like this or if it can be done directly in the FT2232. At the end of the day, it may be easiest to go with the XDS100, which has the same capability and would be easy to sell to management, I suppose. I'm just curious if anyone has ever seen anything like this.

Thanks!
3
General discussion / Any guess why PICkit freaks out when I power-cycle target?
My pickit 3 is having serious problems when I power up the target after it's already connected to USB (no big deal), but that includes even power-cycling the target while the pickit is connected.

That is to say, if the pickit is connected to the PC and the target while the target is off, and then I turn the target on, the pickit will no longer even communicate with the PC until I disconnect & reconnect the USB cable. Note that I am not using the pickit to power the target.

Start with pickit connected to PC and target, target's supplies are OFF. First I try to connect in MPLAB X:
Code: [Select]
*****************************************************

Connecting to MPLAB PICkit 3...
Firmware Suite Version.....01.28.92
Firmware type..............dsPIC33F/24F/24H

Target device was not found. You must connect to a target device to use PICkit 3.


Totally normal, of course. Now I turn on the target's supplies without touching the pickit 3, and then try again:

Code: [Select]
*****************************************************
Connection Failed.


Note that's the same message you get if the pickit is not responding, not if the target isn't responding.

Now leaving the target power on, I disconnect & reconnect the USB cable, and it works:
Code: [Select]
*****************************************************

Connecting to MPLAB PICkit 3...
Firmware Suite Version.....01.28.92
Firmware type..............dsPIC33F/24F/24H

Target detected
Device ID Revision = 3003

The following memory area(s) will be programmed:
program memory: start address = 0x0, end address = 0x3ff
configuration memory

Programming...
Programming/Verify complete

Running
Target Halted


And lastly, stop the debug session, do a power-cycle on the target, and try again:

Code: [Select]
*****************************************************
Connection Failed.


This makes the pickit 3 pretty much 100% worthless for in-circuit debugging if the target uses its own supplies. Any guesses on what is causing this or how to reset the damn thing without unplugging & replugging the USB every single time?

Note: The target device is a pretty simple TI ua7833 linear reg, and has a totally floating supply with a low inter-winding capacitance power transformer. I might try wiring target ground to earth ground, which is OK for development, even though ultimately the circuit must float. Right now the only ground connection to the circuit is through the pickit 3, but again given the power transformer I'm using, it shouldn't be a problem.
4
Project development, ideas, and suggestions / Re: Power supply teardown / Upgrade (suggestions needed)
Mainly I'm talking about the HP analog supplies (no digital circuitry/controls at all). Most of them have a pretty standard set of terminals along the back that configure the voltage & current feedback circuitry. With a couple DACs and ADCs they can be converted to digital control & readout fairly simply. What I found is most of the older even simple non-programmable digital supplies tend to be very low wattage or very expensive.

At this point I will probably just target some kind of serial protocol that can either go over USB via any rs232<->usb adapter, or via bluetooth bridge if so desired. And leave that stuff until later.

For micros I am indeed thinking I'll stick with atmel for now and if there's enough interest someone wants to redo it with pics, so be it.

For PCB my main reason for switching to diptrace is pcb size limits of eagle are causing me problems, not in this project but in a few others. Diptrace also has just a much better feel than eagle so I'm liking it so far. I would go to KiCAD but it's just a nightmare still.

Thanks for the input. I'll take your advice on waiting until I have more finished (maybe at least have PCB's in hand) to start putting real info out.
5
Project development, ideas, and suggestions / Power supply teardown / Upgrade (suggestions needed)
Hi all,

After watching the power supply threads here and elsewhere not really turn into anything yet, I'm going to go through the process of taking an older HP power supply (usually various kinds are available on ebay for less than a hundred bucks), doing a tear down and repair (the one I got is 230v, needs to be changed to 110, has some other minor issues). I'm hoping to document the process as well.

After getting it up and running, I'm also going to convert it over to full digital control. The end result will be some modular PCB's that could be used with almost any older HP/Agilent power supply (they all use virtually identical feedback/control circuits), and then daisy chained together to a master controller that will interface either via Bluetooth directly to an android device, or via usb to a PC (or r-pi) and then to a Linux or android front end via tcp/ip.

So I'm just interested in feedback as far as:

1) Should I just do a blog with pictures/text, or is it worth doing some videos? Combination of them? There will probably be long gaps in the process as I'm busy with school and/or getting PCB's made and such.
2) Is it worth trying to document software/firmware design process and/or PCB layout? I'm planning on switching from eagle to diptrace so I'll be kind of a newbie with the software, so that might be pretty ugly to watch.
3) I'd like to experiment with the android front end (I've done tons of java, I just need to learn the new UI layout stuff), is it worth making it work on a phone or just stick with tablet designs (probably a Nexus 7 as reference device).
4) Is it worth exploring Bluetooth SPP? It seems generally sketchy making it work, but then if it works on the standard nexus devices that's probably good enough. This avoids all the usb nonsense as well. The alternative is connecting to a PC or r-pi (probably using MCP2200 or FT232 for vid/pid's sake) and then using tcp/ip after that, which is fine but just adds more layers.
5) I'm most comfortable with atmel/avr, though I'd like to get into PIC too, but I feel that might be just one too many things to learn & deal with. Just wondering if people had strong opinions on the matter.

My plan is to make all the designs & software open/free, of course.
6
Open Bench Logic Sniffer / Re: open logic sniffer vs bus pirate
My quick advice is this:
If you know what protocols are involved, that is, what kind of digital signals are present, and the bus pirate understands them, that's the tool to use.
If you don't know exactly what protocol is being used or it is simply something the bus pirate doesn't understand, obviously you need the obls.

My use is usually I'll look at something first with OBLS, and if I can figure out that a serial protocol is involved, and what sort it is, then Is witch to bus pirate if appropriate.
7
Client software / Re: Jawi's Logic Sniffer client software - releases
Hey jawi, I finally got a chance to test out b3.

Everything looks really nice so far. I did encounter one bug that was driving me crazy until I got out a voltmeter and checked manually...

There seems to be a rendering glitch that a single DC signal (3.3v on channel 0) is drawn as a continuous low instead of continuous high. Here's the contents of the save file:

;Size: 1
;Rate: 50000000
;Channels: 8
;EnabledChannels: 255
;Compressed: true
;AbsoluteLength: 24575
;CursorEnabled: true
00000001@0

which pretty clearly shows it did record a '1' for the entire trace on channel 0, but draws the trace low along with channel 1,2,3,etc...

Keep up the amazing work.
8
Project logs / Re: Bus Blaster Free PCB build
Sorry, I guess I should have been more clear. I meant like the digikey/mouser part numbers. Last time I peeked at it the cpld was only available from digikey, and the usb or something was only from mouser, I was trying to find maybe everything from digikey but then I got involved in some other project. I'll have to dig it out again I just don't have time until finals are over.
9
Project logs / Re: Bus Blaster Free PCB build
hmmms, did you happen to save your partslist? I have a v3 board sitting here on the bench but I've just been too lazy to go through and pick all the parts out.
11
Bus Blaster JTAG debugger / Re: Bus Blaster V3?
ian, I saw there was both a v3 and v4 in the pcb bin, I still have a coupon left, so I was wondering what would be more helpful to you as far as testers go? It sounds like the v3 is just a slight PCB spin but the v4 is a whole new design... I don't have an immediate need for one, though eventually I will be doing some ARM jtag and possibly xsvf work because we'll be doing some FPGA/CPLD work in school next year.
12
Project logs / Re: avr usb breakout & RS-422 "IR" blasters
I used an SI8410 isolator and a separate SN75176 receiver, and it's about $2 for both in small quantities, so definitely more than the attiny25. Still could be quite a bit cheaper than special interfaces if you have the board space. If generating the isolated power is a sticking issue, it's quite possible to do with a full h-bridge. http://www.maxim-ic.com/app-notes/index.mvp/id/4553

You could use an h-bridge IC driven by complementary oscillator outputs on your avr, like the ZXMHC3A01N8TC (just looked for 5 seconds on mouser so don't hold me to that) and it's only like 60 cents, a couple diodes, and a cap.

Communication is unidirectional so a single channel isolator is sufficient for me, and cable lengths are fairly short so I actually power the receiver and isolator via the cat5 itself because the current draw is extremely low.
13
Project logs / avr usb breakout & RS-422 "IR" blasters


So there's a few things here:

Bottom right is an at90usb162 breakout I had made at seeed. I pretty much stole the Teensy's reset circuit, so I don't know that I should put up the gerbers. But I added a few important features for me, such as an AP2141 current-limited load switch which lets me put unlimited capacitance on any daughter cards without issue. I also added extra ESD protection ic's on the bottom (just special zener 4-packs). There's space for a 3.3v regulator, not used for this design.

The other board can be populated two ways. The one attached to the usb breakout is a simple 3x RS422 transmitter. The other boards are isolated RS422 receivers and attiny25's used as IR blasters. Right now I'm not actually doing 38khz modulation, because these boards are going to be hard-wired internal to the equipment being controlled. An open-collector transistor output will pull-down the IR module's output (which is also open-collector) to fake an IR signal.

The receiver boards could be easily changed to daisy-chain compatible with a very simple re-spin, it just wasn't an issue for me, I intended to use three separate cables since my controlled equipment is not in easily daisy-chained locations anyway.

After moving to almost all SMD parts I can say that 1206 parts are actually a joy to work with, and 0805's are acceptable. 0603's are quite doable but a real pain, so I only intend to use them where absolutely necessary. I had no problems with the IC's, using a hoof tip on my metcal, they were very very easy.


I've already tested transmitting codes from the PC via virtual serial port, and it's working OK so far. Haven't hooked it into the equipment yet, that should happen this week. I also need to wire up a 3x4 keypad to the micro.

I'm hoping now that the open USB stack for PIC is coming together I'll eventually switch and make it into a PIC18F breakout.
14
Bus Pirate Support / Re: BusPirate V3.5 free PCB built!
Any advice on part #'s for the resistor networks, tantalum caps, or usb port?

I want to build a 3.5 or a 4, and most of the parts should be easy to pick, but I wasn't sure about an exact match for these.

The v4 of the BP has a crystal that in the partlist just says "Seeed 12mhz crystal"... :/
15
USB Infrared Toy / Non-USB IRtoy?
I haven't looked at the code, but from what I understand, the IRtoy's USB controller just creates an internal virtual serial port, right? If one were to move its interface to the real serial port on the device (and then possibly chose a non-usb micro) and give each IRtoy an address using jumpers, then it should be possible to place a large number of them on a RS-485 network.

I'm building a small RS-422 (star, not daisy-chained) network of attiny25's used as IR blasters (though they are hardwired with open-collectors, so they produce IR codes, but have no actual IR LEDs). We're using it to control multiple cheap component-type video switches to save some guys buying a massive video mixing panel.