Electronics > Projects, Designs, and Technical Stuff
A Universal Programmer
(1/2) > >>
colejohnson66:
After seeing the two(?) "universal" programmer videos from Dave (TL866 and Wellon) and how they only can switch between GND, VCC, and VPP (and possibly a VPP2), I thought about building a "true" universal programmer. It won't be as great as the DataI/O Unisite, obviously, but it will have a true DAC per pin based system.

Currently, I've been working on an octal pin driver board based around the MAX528+ (octal DAC capable of -5V to 5V). The output of each DAC is ran through an opamp (MC33174) to quadruple the voltage to -20V to 20V. Each of the eight pins on the board will be switchable between the DAC/opamp output ("Vpp") and ground. I'm also using two 74HC259 (1:8 demuxer with latched outputs) to decode the selection for each pin (bottom left of the schematic).

As for the main processor that would control all six (6*8=48 pins) driver boards, I'm thinking of the Adafruit Metro M4 Grand Central (unless people suggest better options). I was thinking I could use the M4 as an interpreter for a control script that could be sent over serial (over USB) from a desktop program.

I'm also trying to keep the cost down as even $20 per pin driver board would cost $120 for all six. I haven't calculated the BOM for this schematic, but any ideas on how to reduce it further would be accepted.

And my reason for posting is this: to post updates, but also to get feedback as I'm a beginner. I'm unsure of my choice of MOSFETs among other things. Basically, I want to know if my circuit is good, and if not, what I should change. I've attached my schematic.
FrankBuss:
The P-channel MOSFET won't work, gate voltage must be higher:



There are high-side FET drivers for it.

But I don't know much about analog electronics. For me it always helps, if I test it the individual parts on a breadboard first. And maybe would be nice to have some short circuit protection as well, or even an adjustable max current per pin.
Ian.M:
Some old parallel EPROMs needed 25V Vpp (26.5V abs max).   I am not aware of any that needed a larger negative bias rail than -12V.  I therefore suggest offsetting your amplified DAC outputs to the range +27V to -13V.  Your pin drivers need to be quite robust - e.g an Intel D2716 needs up to 100mA of Vcc and up to 30mA from Vpp during programming.

IMHO a DAC per pin approach is excessively expensive.  *MOST* of the pins only need to be capable of being driven to a Vcc dependent logic level, or read back, with an adjustable threshold.   Lets assume the max Vcc it wil ever need is 7V, and give it 16 outputs and 24 I/Os that can handle that (to allow for up to 16 address and 16 data lines, with the spare 8 I/Os usable for control lines or additional address lines).  Then it would only need ONE of your full function pin driver boards to handle the limited number of power rail and programming pulse pin positions for each target socket. 

A 96 way DIN41612 connector would allow customisation of driver to pin mapping for up to 48 ZIF socket pins or allow different socket boards to be connected.  Seven of the eight fully programmable supply pins would go through the 96 way connector, as one would be needed as the comparator bias voltage for the input thresholds, and it needs a ground pin.  The DIN41612 connector on the programmer should itself be socketed so it can be easily replaced when worn out.  Locate it next to the 48 way DIL ZIF socket so personality modules can carry a label that indicates the pin 1 position to use.   
jaromir:

--- Quote from: Ian.M on October 07, 2019, 08:27:29 am ---Some old parallel EPROMs needed 25V Vpp (26.5V abs max).   I am not aware of any that needed a larger negative bias rail than -12V.

--- End quote ---

I agree with your analysis, just nitpicking over the quoted bit:
Old i1702 and similar EPROMs did have wider voltage swing (47V) during programming, but I believe it's reasonable to omit those from support list. Also, early EAROMs, like ER1400 (used in some 80's high-class multimeters) needed -35V (or thereabouts) against substrate, but this perhaps may be workarounded by creatively moving ground potential to higher voltage.
Ian.M:
Thanks for the EAROM bias supply note.  Perhaps one of the DAC channels should go to a dedicated 0 to -40V negative bias supply generator.

Moving ground would be a PITA, you'd either need a full DAC/ADC per pin approach or would need the DAC board to be fully isolated, with one of its channels connected to system ground.
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod