Sorry, I wasn't clear in my description. I was talking about a regular USB to DB25 cable, not a TPDD cable.
On Wed, Jul 17, 2019, at 12:04 PM, Brian K. White wrote: > A tpdd cable needs to have a 2x4 female pin header with ttl signals on > one end, and a db25 male with rs232 on the other end. > Unless you wanted to make a tpdd cable that only connected to modern pcs? > > -- > bkw > > On 7/17/19 2:25 PM, Kurt McCullum wrote: > > The old serial cables do. But the new FTDI USB cables have a standard > > USB end with the FTDI + Max232 chips in it and the other end is the > > RJ45. Here is an example. > > > > https://www.ebay.com/itm/USB-to-RJ45-For-Cisco-USB-Console-Cable-FTDI-744664241835-Y2C9/263834776764?epid=19015691857&hash=item3d6dc734bc:g:b9cAAOSw6jJb8oOZ > > > > The do have DSR/DTR and all other signals. I've been making these for > > my own use and find them to be almost perfect. If I could just find > > one at the same price with a round black cable I would consider it > > 'perfect' > > > > Kurt > > > > On Wed, Jul 17, 2019, at 11:20 AM, Brian K. White wrote: > >> Don't the cisco cables have a rj45 on one end, and db9 on the other? > >> Don't you need a dupont connector for the tpdd, and a 9-25 adapter for > >> the M100? > >> And do they actually include dsr/dtr? > >> > >> Myself I am hoping to no longer use the db25 pcb. I just made a new > >> version just for the heck of making a more optimized version of it. I > >> think the most optimal is with the pcb and transistors on the tpdd end, > >> because it takes care of the 2x4 plug for the tpdd in a much more > >> convenient way than crimping and assembling a dupont connector. And the > >> cable and db25 comes from any generic factory molded modem cable. The > >> latest version I added a place to zip-tie the cable end to the pcb, and > >> I think that will be the final most optimal version. > >> > >> But my first samples haven't arrived yet so it's still a theory. > >> > >> It might actually be worth making a little 3d-printed housing for the > >> tpdd-end, to fill up that opening in the case just like the original > >> cable does. Not just for polarity, but also to provide strain relief, > >> where the case opening is what holds the cable in place instead of the > >> pins soldered on the pcb inside the tpdd. But that's honestly not a > >> major worry and I haven't actually tried to design anything. > >> > >> -- > >> bkw > >> > >> > >> On 7/17/19 1:57 PM, Kurt McCullum wrote: > >> > Brian, > >> > > >> > I appreciate the insights. My first reaction to seeing your new board > >> > was that this was a good solution. But after reading through your > >> > post, I changed my mind. It's probably still much easier (and cheaper) > >> > to buy the FTDI cable designed for CISCO routers, clip off the RJ45 > >> > end and then solder on the DB25. Those cables have the proper levels > >> > and it's easy enough to solder the DB connector. The only thing I > >> > don't like is the flat, light blue, cable. > >> > > >> > Kurt > >> > > >> > On Wed, Jul 17, 2019, at 10:49 AM, Brian K. White wrote: > >> >> Kinda sorta maybe possibly. I had thought about using one of the > >> >> variants of max232, instead of the 3 transistor hack. But in the end, > >> >> the 3 transistors is probably the cheapest and simplest circuit to > >> make > >> >> a tpdd work, and this new circuit with the part that Rick Shear > >> >> identified, really does seem to be just like the original cable. I > >> >> haven't used a scope to compare the actual signals, but I have a scope > >> >> and an original cable so I could. I think I'll do that. > >> >> > >> >> It was a while ago so I might have some of this wrong, but what I > >> think > >> >> I remember was most of those ttl to rs232 level-shifter chips > >> don't have > >> >> 3 channels, and so you can't handle all 3 lines that the three > >> >> transistors handle, you wouldn't get DSR/DTR. And most of them > >> also need > >> >> auxiliary caps for the charge pump to parasite power from the rs232 > >> >> lines which do not actually deliver any power to speak of. There are > >> >> special ones that have their own caps built in, and there are special > >> >> ones with more channels, or you could use multiple regular ones. > >> So, you > >> >> might only be trading 3 transistors for a max232 and two caps, or > >> a big > >> >> expensive special max232 etc. > >> >> > >> >> I still wouldn't mind tryng it just for the heck of it. The result > >> would > >> >> be a cable that should be more technically correct, and so and more > >> >> reliable, when used with other random generic serial ports besides the > >> >> one on a model 100-200, yet still work perfectly on a model 100-200 > >> >> (which has pullup resistors, which the original tpdd cable circuit > >> banks > >> >> on). > >> >> > >> >> But I don't think the result would be any cheaper or simpler to make. > >> >> > >> >> -- > >> >> bkw > >> >> > >> >> > >> >> On 7/17/19 10:59 AM, Kurt McCullum wrote: > >> >> > Brian, > >> >> > > >> >> > I noticed you have a new DB25 PCB on Osh Park. I suspect the whole > >> >> > TPDD cable was intended to take a TTL level floppy drive and > >> bring it > >> >> > to a standard RS-232 level. I'm wondering if that same PCB could be > >> >> > used with one of the inexpensive TTL level FTDI USB to serial > >> >> > converters to create a PC to Model-T cable? The only problem I > >> see is > >> >> > that the TTL level FTDI cables are all 6 wire and don't seem to have > >> >> > the DTR wire. But I suspect that could be hard wired some how. > >> >> > > >> >> > Kurt > >> >> > > >> >> > On Wed, Jul 10, 2019, at 11:28 PM, Brian K. White wrote: > >> >> >> I finally did the version of this that I had envisioned. > >> >> >> https://oshpark.com/shared_projects/Sk27mJwq > >> >> >> > >> >> > > >> >> > >> >> > >> > > >> > >> > > > >
