Yeh you are right about that. RF kicks butt. The HV5522 chip is driven with a serial link so I can slow it way down for my application but if I'm draining 208 volts at 100 ma per channel I may have to split the cables and re-weave them in a twisted pair configuration. In my first application I'll only be draining 3-5 volts at 20 ma per channel but at higher powers cross-talk could mess things up. I'd like to do a layout with this chip and a Propeller QFP-44 on the same board. Having serial control of 32-channels like this would make a fresh alternative for the usual noob question on here: " how do I control eight 24 volt DC coil relays?"
I love the concept. I've got so many of the older RAM chips I'd like to be able to use them the way you tried.
I could have used jumpers to get the connections down to just two wider ribbon cables by using the wider side pad arrays. I did not see a way to solder on just two lines of single row pin headers and still push them directly into a 0.1" pitch breadboard. The way Schmart board laid this out makes the row a double. I sent copies of the photos to both Schmart board and the manufacturer of the chip. I'm such a chronic whiner that Schmart board has started to produce small boards that fit more of the SMT chips we like. I'm not sure they've hit the nail on the head yet w.r.t. our basic needs. The chip MFG could not offer an easy connection solution for me to experiment with this chip on bread boards.
Duane: While the DRAM chips run extremely fast in a pc, this is not the case with the prop. I presume they are 3V3 logic chips (not all memory on SIMM cards are 3V3)?
Since you are using the prop, your access will be much slower than 333MHz-1333MHz or whatever they are up to now. At much lower speeds, the cables will be less of an issue. Did you try placing a large cap (say 100uF between the power rails) on the memory pcb? Also, try much slower access to ensure the cable length is not an issue.
Attached is a photo of my attempt to interface with a DIMM SDRAM module.
It didn't work. I thought I had all the refresh rates correct but I could never read back data I attempted to write to the chip.
I read that long wire can cause all sorts of problems in circuits. I hope your chip worked correctly with the longish wires connected.
Duane
I've made EDO DRAM work with much uglier wiring There is code floating around the forums for a latched interface.
Series 1K resistors should be used to keep from over-driving propeller with 5V devices ... not an issue with TTL usually, but CMOS can drive higher than 3.3V.
The advantage of EDO (and other "synchronous" devices) over SRAM is that bytes can be read without toggling address bits
What are you trying to do? Get me to try it again? The capacitor is a good idea. It looks like there is a cap under each chip but I'll add an extra (if I try this again).
I hate to admit it but there is a good chance the software is the problem. This was my second PASM project after flashing a few LEDs. I was waiting for Steve to finish his SDRAM boards so I thought I'd give SDRAM a try myself. I didn't try anything fancy like hidden refreshes. I'm pretty sure I had the refresh rate fast enough (I found a datasheet for the chips on the internet).
The chips are 5V. I'm pretty sure I have resistors on the Prop input lines. I realized after I soldered all the connections that I should have used pins 0 through 7 for data. I used the low pin numbers for address lines.
Comments
Attached is a photo of my attempt to interface with a DIMM SDRAM module.
It didn't work. I thought I had all the refresh rates correct but I could never read back data I attempted to write to the chip.
I read that long wire can cause all sorts of problems in circuits. I hope your chip worked correctly with the longish wires connected.
Duane
I love the concept. I've got so many of the older RAM chips I'd like to be able to use them the way you tried.
http://forums.parallax.com/attachment.php?attachmentid=82715&d=1309984974
Since you are using the prop, your access will be much slower than 333MHz-1333MHz or whatever they are up to now. At much lower speeds, the cables will be less of an issue. Did you try placing a large cap (say 100uF between the power rails) on the memory pcb? Also, try much slower access to ensure the cable length is not an issue.
Series 1K resistors should be used to keep from over-driving propeller with 5V devices ... not an issue with TTL usually, but CMOS can drive higher than 3.3V.
The advantage of EDO (and other "synchronous" devices) over SRAM is that bytes can be read without toggling address bits
What are you trying to do? Get me to try it again? The capacitor is a good idea. It looks like there is a cap under each chip but I'll add an extra (if I try this again).
I hate to admit it but there is a good chance the software is the problem. This was my second PASM project after flashing a few LEDs. I was waiting for Steve to finish his SDRAM boards so I thought I'd give SDRAM a try myself. I didn't try anything fancy like hidden refreshes. I'm pretty sure I had the refresh rate fast enough (I found a datasheet for the chips on the internet).
The chips are 5V. I'm pretty sure I have resistors on the Prop input lines. I realized after I soldered all the connections that I should have used pins 0 through 7 for data. I used the low pin numbers for address lines.
Dang it! Now I'll have to try again.
@yarisboy, Sorry for hijacking your thread.
Duane