On Jan 20, 2014, at 4:13 PM, Al Kossow <aek at bitsavers.org> wrote: If you can?t find a data sheet, there?s at least one place that has programming information (though not pinouts): the DEC Pro 300 series technical manual (available on bitsavers.org). The PRO 350/380 uses the 7201 USART for the comm port, and register information is found in Volume 1 starting at pages 5-97 to 5-112. I used that to write a working drivers (sync as well as async) so that information is pretty accurate and complete. paul

I just looked in my old NEC databooks and the upd7201A is upd7201AC in a plastic package and upd7210AD in a ceramic package. The older data book for upd7201 mentions the two package types but fails to give detailed part number package suffixes. by inference you may have the earlier 7201 in plastic Dave Caroline On 20/01/2014, Holm Tiffe <holm at freibergnet.de> wrote:

[Z80-SIO .vs. 7201] The bus itnerface signals are rather different too. Everything I've seen suggests that the final letter describes the package -- 'C' is a plastic package, 'D' is a ceramic package. I have a 7201D on an HP150 CPU board, and this is a ceramic package. So the 7201 data sheet will cover it. I am not sure abotu thignsl ike temperature range (but I don't think that will bother you). -tony

[7201 .vs. Z80-SIO) OK, there are 8 data liens and 2 address lines, and a chip select. But the actual cotnrol lines are quire different. The 7201 hads Rd/ and Wr/. The Z80-SIO, like most (all?) Z80 peripheral chips makes do with Rd/,. It also takes IORQ/ and a clock of the same freqeuncy as the CPU and if there is no read at the apporpriate time, it assumes it's a write cycle. Dido the intterupt acknowledge. In my experience the Z80 peripheral chips are a right pain to use with anything other htan a Z80. The 7201 can be linked to just about any CPU. -tony

True to a point. But I found it a lot easier to use a 6522 in a Z80 system than to use a Z80-SIO in a 6809 machine. -tony

True, but the 6500 and 6800 buses are very similar. Interfacing a 6522 to a 6809 or a 6850 toa 6502 is trivial. What';s the 6828? I suspect the 6829 would be very hard to use with anything other than a 6809 (it.s the MMU for the 6809). Not a common chip, but I have seen it. -tony

Right. It's not in my Motorola 8 bit data book (from the mid 1980s), so I guess it was discontinued by then (the 6829 MMU is i nthere). I've never seen a 6828 in anything either. In gneral, I think it's best ot use the interrupt controlelr nad DMA controlelr designed for the famil of CPUs you're using. The time it's necesasry to use chip for ma differnet family is when you want a particualr I/O functio nwhich does nto exist i nhe family you're msotly using. I menitond the 6522, there is really no Intel equivlent to that, for exmaple. -tony

ISTR that the MC6828 was bipolar, not MOS, and that it was a gate array. It probably was dropped earlier than other MC6800 family members due to the specific gate array being discontinued, and there probably wasn't enough demand to justify development of a replacement. On Jan 27, 2014 3:55 PM, "Chuck Guzis" <cclist at sydex.com> wrote:

[MC6828 vectored priority interrupt encoder] That's it. I couldn't remembed the number. I think all of the rest of the MC850x bipolar gate array series were various polynomial/CRC/ECC circuits, except for one that was a 4K DRAM refresh controller. They were all based on the same array, except for one of the polynomial generators that was full-custom, and apparently put in that series only due to having similar functionality.

Tony Duell wrote: Where are your problems? The SIO supports IMHO the Reset of the IEI-IEO Chain trough a bit in a register, that's easier as on the other chips. I never tried this, but I see no big problem here. Please clearifiy this a little.. Regards, Holm -- Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe, Freiberger Stra?e 42, 09600 Obersch?na, USt-Id: DE253710583 www.tsht.de, info at tsht.de, Fax +49 3731 74200, Mobil: 0172 8790 741

Tony Duell wrote: Maybe not trivial but it is doable. Sometimes you have to build an Sequencer to get the timings to fit. I have other examples of using Z80 Peripherals on other Processors, CTC on a Z8 is one.. In this case the CTC is memory mapped. Want a schematic? IORQ just distinguishes between memory and IO Access, it is possible to do memory mapping with those chips. I've from the former GDR, Germany from behind the iron curtain. The Semiconductor plants there build Z8000, Z80 and Z8 so the z80 Peripherals was what we could get and we had to use. There where until the end of the 80's no Z8000 peripherals available besides the MMU. Later the SCC and the CIO where introduced. There where 8086 Computers build with the Z80 SIO in them and they used the SIO even as console in a mostly PDP11 compatible Computer, the K1620/1630 Systems. (which was a big flaw, since no Software could run w/o modification). There are other things that where good meant but bad at the end in this machine, eg. there where some bus controllers on the cards, each programmable to its IO adresses and Vectors and that programming has to be done before the machine is compatible at all. After that, the conterollers for floppy and disk used the same register adresses as the original HW. Again, you have to boot only OS's which where patched to do this. (BTW, the CPU was an 8 Bit Slice in NMOS, a asynchron processor w/o a clock). Here are pictures http://www.robotrontechnik.de/index.htm?/html/computer/k1600.htm We still searching for a bootable media for this machine. It is the slowest PDP11 ever build. Here is a picture of the CPU: http://de.wikipedia.org/wiki/Datei:K_1630_CPU.jpg You can see the very slow control store build from 2708's. Mostly I saw some 8255s used in Z80 computers, bit at least last year I've repaired a controller with an 6809 and a Zilog SCC in there. Don't ask why they used this chip, it seems to me that speciall the UARTs from other families wheren't "that good" at all. Regards, Holm -- Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe, Freiberger Stra?e 42, 09600 Obersch?na, USt-Id: DE253710583 www.tsht.de, info at tsht.de, Fax +49 3731 74200, Mobil: 0172 8790 741

Tony Duell wrote: I wouldn't say that using Z80 chips on other CPUs are problematic, the only thing you have take care about is the RETI Instruction that is decoded from the Chips to reset the interrupt Chain. I've experimented with an Z80-CTC on a Z8001 CPU which has such a different Bus. The Problem was easy to solve with only a GAL16V8. Otherwise the 7201 isn't that easy to use for processors other than the 8086 family since it want's 3 Int ACK Pulses. Regards, Holm -- Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe, Freiberger Stra?e 42, 09600 Obersch?na, USt-Id: DE253710583 www.tsht.de, info at tsht.de, Fax +49 3731 74200, Mobil: 0172 8790 741

I think you mean 7201, not 7210 here. The 7210 is a GPIB interface chip. Many, many years ago I bought a 7210 to add the GPIB prot to an FTS-88 machine (a strange 8088 box runnign CP/M-86 (FTOS-C) from 8" floppies) -tony