My RS-232 / 6551 Design

[cbmeeks]

I *THINK* I've finished my design in regards to my serial port. I plan on using a 65C51 (and work around the bug) or use an old-school Rockwell version (which I have found works in my other design).

I also want a real serial port that could communicate with my other legacy computers like my Apple II or C64. So I'm hoping I have implemented all of this correctly. Especially the hardware handshaking/etc.

I'm certainly interested in the opinions and advice of you guys.

Thanks!

[GARTHWILSON]

Why are TX and RX going through the line driver and receiver but the control lines going to the same connector are not?

[cbmeeks]

Hmm. Good question.

I thought the control lines went directly to the 6551.
So the DTR/DSR and RTS/CTS also get translated?

[GARTHWILSON]

cbmeeks wrote:

Hmm. Good question.

I thought the control lines went directly to the 6551.
So the DTR/DSR and RTS/CTS also get translated?

Yes, everything. It might be good to go through my RS-232 primer.

[cbmeeks]

Ah, obvious fubar on my part. For some reason, I was thinking those are just plain TTL levels. I'll rework that and try again.

Thanks!

[GARTHWILSON]

The popular MAX232 has two line drivers and two line receivers, commonly used for TX, RX, RTS, and CTS. If you also want DSR and DTR, you'll need an extra line driver and receiver pair. If you already have ±12V (or even ±9V—it's not critical) on the board for other things anyway, you can save board space by using something like the 16-pin MC145406 triple line driver and receiver which does not need any charge-pump capacitors.

[cbmeeks]

Ah! OK. That makes more sense. Most examples I have seen really don't utilize DSR/DTR and sometimes not even RTS/CTS.
I couldn't figure out how to get all of those pins into the MAX232 which is one reason why I thought those were TTL.
I've learned more about RS-232 today than I have in the last 10 years. LOL

To be honest, I only went with the MAX232 because I have some. I will investigate the MC145406 chip as a replacement.

Thanks again.

[BigDumbDinosaur]

cbmeeks wrote:

Hmm. Good question.

I thought the control lines went directly to the 6551.
So the DTR/DSR and RTS/CTS also get translated?

Yes. If you use a MAX238 you will be able to accommodate the control lines in a single device. If you refer to my white paper on interfacing the 28L91 I show a suitable DTE-to-DCE hookup on page 17.

28l91_interfacing.pdf

Interfacing the 28L91

(374.21 KiB) Downloaded 172 times

[BigDumbDinosaur]

GARTHWILSON wrote:

If you also want DSR and DTR, you'll need an extra line driver and receiver pair.

That requirement may be implemented with a MAX238.

[cbmeeks]

It looks like the MC145406 is obsolete on Mouser.

What about this one?

https://www.mouser.com/ProductDetail/595-SN75C1406N/

Pretty cheap and has 3 drivers/receivers.

[BigDumbDinosaur]

cbmeeks wrote:

Ah! OK. That makes more sense. Most examples I have seen really don't utilize DSR/DTR and sometimes not even RTS/CTS.

Use of CTS/RTS handshaking is de rigueur above 9600bps in most cases in which handshaking is required. XON/XOFF handshaking becomes unreliable above that speed.

Quote:

To be honest, I only went with the MAX232 because I have some. I will investigate the MC145406 chip as a replacement.

I recommend the MAX238. The MAX238 uses the same charge pump caps as the MAX232. My POC V1.2 unit has two MAX238s to handle the four serial channels.

[cbmeeks]

Thanks BDD.

I will look into the MAX238 and your document above.

EDIT
Forgot to mention that it would be great to only have a 5V supply but I could possibly put in a 9/12V and step down to 5V. But it seems the MAX238 doesn't need 9/12V!

So that seems to be a perfect match. I'm a little puzzled on why there are more drivers than receivers on the chip.

[BigDumbDinosaur]

cbmeeks wrote:

Forgot to mention that it would be great to only have a 5V supply but I could possibly put in a 9/12V and step down to 5V. But it seems the MAX238 doesn't need 9/12V!

When the MAX232 was originally released in the mid-1980s it was considered revolutionary in the TIA-232 world because it didn't require multiple voltage sources like the then-ubiquitous 1488 and 1489 devices.

Quote:

So that seems to be a perfect match. I'm a little puzzled on why there are more drivers than receivers on the chip.

Huh? The MAX238 has four TIA-232 outputs and four TIA-232 inputs.

[GARTHWILSON]

I just went through and verified or fixed all the links in my RS-232 primer. (I can't believe Maxim changed their whole domain name, making millions of links across the internet, including many in my own pages, dead. I'm sure they have since figured out that was a dumb idea.) And darn! I see the MC145406 is out of production. Jameco still has them, but their being out of production would explain why Jameco's price is so high ($4.95US). I might get a bunch to for my own inventory anyway.

[GARTHWILSON]

I like to have the higher ± voltages on the board anyway for things like:

• the negative voltage required for the backplane of many high-contrast supertwist LCDs
• my D/A converter
• the op amp for the A/D and D/A converters
• my analog signal-conditioning plug-in modules
• the speaker amplifier
• and in this case, having the triple line drivers and receivers that take less board space because they have fewer pins and don't require the charge-pump capacitors that the MAX23_ use

Also, it is often better to have local voltage regulation on the board. If you have a 5V regulator off-board and bring the 5V in through a cord, the resistance of the cord and connectors may result in having significantly less than 5V on your 5V parts, depending on your current draw. This is especially true of tiny USB connectors and thin, cheap USB cables, as I found out on a work project in 2018. If you have a higher voltage coming into the board, then regulate down to 5V, you'll have a better setup.