RS232 Issues

457 views
Skip to first unread message

Gracana

unread,
Jan 21, 2017, 10:40:20 PM1/21/17
to RC2014-Z80
Are there any gotchas involving MAX232 installation on the Serial I/O board?

I've had my system running, connected with an FTDI cable. Everything has been good so far. I wanted to switch to RS232 so I can use my video terminal as an interface, and it's just not working. I've been troubleshooting with my PC as well, and on either device I can see data from the RC2014, but it doesn't see anything coming back.

Take a look at this photo. In blue, an RS232 signal from my PC. In yellow, a signal from the RC2014. Why so low?

When I probe the signals between the ACIA and the MAX232, I see only data from the RC2014, and nothing from the PC.

When I probe the signals between the PC and the MAX232, I see only data from the RC2014, and nothing from the PC.

If I disconnect the serial cable from the RC2014 and probe it, I get the lovely output shown in the photo.

Any ideas?

Regards,
Gracana

Gracana

unread,
Jan 21, 2017, 10:41:20 PM1/21/17
to RC2014-Z80
Forgot the photo. Here it is.
rs232.jpg

Scott Lawrence

unread,
Jan 21, 2017, 10:50:52 PM1/21/17
to rc201...@googlegroups.com
Make sure that you're not using both the FTDI interface to your PC along with the MAX232 chip, since both driving the lines would produce bad results.

-s

On Sat, Jan 21, 2017 at 10:41 PM, Gracana <kb1...@gmail.com> wrote:
Forgot the photo. Here it is.

--
You received this message because you are subscribed to the Google Groups "RC2014-Z80" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rc2014-z80+unsubscribe@googlegroups.com.
To post to this group, send email to rc201...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/rc2014-z80/c6bfa2e0-7e3f-40c7-95a9-ffac377dfb66%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Scott Lawrence
yor...@gmail.com

Mild Lee Interested

unread,
Jan 21, 2017, 11:26:35 PM1/21/17
to RC2014-Z80
Dumb question, but are you using a null modem or a straight through cable? 
Maybe the MAX232 is dead?  I've killed a few through poor handling over the years...

Gracana

unread,
Jan 21, 2017, 11:31:52 PM1/21/17
to RC2014-Z80
On Saturday, January 21, 2017 at 10:50:52 PM UTC-5, Scott Lawrence wrote:
Make sure that you're not using both the FTDI interface to your PC along with the MAX232 chip, since both driving the lines would produce bad results.

Good thought. But I'm definitely not doing that.

Gracana

unread,
Jan 21, 2017, 11:36:18 PM1/21/17
to RC2014-Z80
On Saturday, January 21, 2017 at 11:26:35 PM UTC-5, Mild Lee Interested wrote:
Dumb question, but are you using a null modem or a straight through cable? 

Null modem cable.
 
Maybe the MAX232 is dead?  I've killed a few through poor handling over the years...

I have two and both exhibit the same problem, but that doesn't mean it's not the chip.

Just now I tried replacing the 'lytics with 0.1uF ceramics. TI's datasheet says that's ok. Now I get characters echoing back to the PC, but no output from the RC2014.

I tried the FTDI cable again, with the MAX232 removed, just to check that I haven't killed the board or anything. Nope.. It's fine.

Mild Lee Interested

unread,
Jan 22, 2017, 5:08:51 AM1/22/17
to RC2014-Z80
I'm having similar problems, but I unfortunately don't have a scope suitable for fault finding.
The probable cause for me is the RS232 - USB converter I'm using (which has a Prolific chipset).  These converters are the bane of my existance and notorious for not working correctly (or at all) with some serial equipment.  Could this be your problem also, or are you using a machine with a true RS232 serial port?
At work I keep several laptops with native serial ports running XP simply because modern hardware and OS suck at talking to legacy equipment :(

Mild Lee Interested

unread,
Jan 22, 2017, 8:56:40 AM1/22/17
to RC2014-Z80
Here's another stupid thought:  did you use a male or female D9 connector on the RC2014 Serial IO module?
I assumed that as a DTE device, the RC2014 should have a male D9 connector, but I later realised the pin out is for a female D9.  This causes a big mess if you then try to use a standard null modem cable, as pin 5 becomes pin 1, etc. 

With a female D9 fitted to the RC2014 and using a standard F-F null modem cable plus gender changer, everything works out.

Because I fitted a male D9 to the RC2014, I ended up having to build my own null modem as follows:

PC end (F)                RC2014 end (F)
Pin 2(Rx) ---------------- Pin 3(Tx)
Pin 3(Tx) ---------------- Pin 4(Rx)
Pin 5(Gnd) ------------- Pin 1(Gnd)
Pin 8(CTS) ------------- Pin 8(RTS)

Works fine now.

On Sunday, 22 January 2017 12:36:18 UTC+8, Gracana wrote:

Gracana

unread,
Jan 22, 2017, 9:31:57 AM1/22/17
to RC2014-Z80
On Sunday, January 22, 2017 at 8:56:40 AM UTC-5, Mild Lee Interested wrote:
Here's another stupid thought:  did you use a male or female D9 connector on the RC2014 Serial IO module?
I assumed that as a DTE device, the RC2014 should have a male D9 connector, but I later realised the pin out is for a female D9.  This causes a big mess if you then try to use a standard null modem cable, as pin 5 becomes pin 1, etc.

Oh! Yeah, I made the same assumption. That must be my problem. I'll see if I can make an adapter.

Gracana

unread,
Jan 22, 2017, 9:42:01 AM1/22/17
to RC2014-Z80
Well, I swapped things around. Now I get unprintable characters... But it's starting to work. Maybe I should go back to the 1uF caps.

Mild Lee Interested

unread,
Jan 22, 2017, 11:48:20 AM1/22/17
to RC2014-Z80
You must have 1uF caps for the charge pump on the MAX232, but yes, they can be ceramics.  0.1uF ceramics way too small, or was that a typo in your post?  I didn't pick up on this on your earlier post, sorry.
I guess you could also double check your com port settings, perhaps?
Failing all that, giving problematic PCB's a good bath in isopropal alcohol gets rid of any number of gremlins, I've found!

Jon Elliott

unread,
Jan 22, 2017, 1:11:01 PM1/22/17
to RC2014-Z80
The photo on RC2014.co.uk shows a male connector...

http://rc2014.co.uk/wp-content/uploads/2014/10/SerialIO_Tindie.jpg

...proper confused! so I need to rewire my standard F to F cable...?

Gracana

unread,
Jan 22, 2017, 3:55:15 PM1/22/17
to RC2014-Z80
On Sunday, January 22, 2017 at 11:48:20 AM UTC-5, Mild Lee Interested wrote:
You must have 1uF caps for the charge pump on the MAX232, but yes, they can be ceramics.  0.1uF ceramics way too small, or was that a typo in your post?  I didn't pick up on this on your earlier post, sorry.
I guess you could also double check your com port settings, perhaps?
Failing all that, giving problematic PCB's a good bath in isopropal alcohol gets rid of any number of gremlins, I've found!

 Oh, the datasheet says the MAX202 (not 232) can operate on 1uF capacitors. I wondered why they said that but wrote 1uF everywhere else.

An alcohol bath might be a good idea. I'll see if I have some.

Jon Elliott

unread,
Jan 22, 2017, 4:08:29 PM1/22/17
to RC2014-Z80
Failing that a toothbrush and washing up liquid on an unpopulated board.

Gracana

unread,
Jan 22, 2017, 4:28:38 PM1/22/17
to RC2014-Z80
Hooray! It was the wiring. With a fresh head this morning I took a look again, and with the wires bodged appropriately, it works. Thanks everyone for your input!


20170122_162713.jpg

Gracana

unread,
Jan 22, 2017, 9:11:09 PM1/22/17
to RC2014-Z80
There, now it hooks up to the VT nicely!
20170122_210509.jpg
20170122_210529.jpg

Mild Lee Interested

unread,
Jan 23, 2017, 2:09:33 AM1/23/17
to RC2014-Z80
As am I...

The schematic shows ground on pin 5 (as it should be) but it's only pin 5 if you use a female D9.  Ground is the pad on the back of the board closest to the top RHS corner.  On a male D9, this is pin 1.

Because of the way they are internally connected, it won't do you any good to just put a F-F gender changer on the male socket - pin 1 will still be pin 1, not pin 5.

On the other hand, putting a female D9 on the board then using a M-M gender change will work just fine - you end up with the standard pin out of a DTE device.

The photo is, unfortunately, wrong.  Are we really the first ones to notice this??  Or have we tied ourselves in knots and there's really nothing wrong with it...

It's no real drama to make a customised adapter to get the pin out right again, but I think I'll just replace the male D9 with a female one and have done. 


On Monday, 23 January 2017 02:11:01 UTC+8, Jon Elliott wrote:

Jon Elliott

unread,
Jan 23, 2017, 9:02:27 AM1/23/17
to RC2014-Z80
Ok, mine partly works now - I've reversed the wiring, and swapped the RX/TX, sending and receiving both data and status from the module, but not receiving keypresses from the PC. Need to look at this later when I've got more time - work to do and all.

Good work chaps!

James Damewood

unread,
Jan 26, 2017, 4:29:57 AM1/26/17
to RC2014-Z80
I am having the same problem and pin 5 (GND) is open.  Also noticed the pads for the DB-9 connector was covered with solder mask. Removed solder mask surface to expose pads.  But later realized Pin 9 on MAX232 was not connected to the correct pin.  Will need to remove connector and bodge wire from MAX232 and to DB-9 male or female connector.

James

Mild Lee Interested

unread,
Jan 26, 2017, 7:34:58 PM1/26/17
to RC2014-Z80

Hi James, if you put a male DB-9 on the serial IO board, then yes pin 5 of that connector is open.  The layout is on the board is for a female DB-9. Have a look at the attached images and you should see what I mean.  Black image is female and grey is male.  The pins are opposite. If you put a male connector on the board, pin 1 is ground. 
Pin 9 on the MAX232 is the TTL level RX pin.  It's not connected to the DB-9 connector at all.  According to the schematic, it's connected to the Rx line on the back plane, pin 2 on the MC86B50 and the tx pin on the FTDI header via a resistor.  Is the schmatic wrong?

Spencer Owen

unread,
Jan 31, 2017, 8:16:48 AM1/31/17
to rc201...@googlegroups.com
On 22 January 2017 at 20:55, Gracana <kb1...@gmail.com> wrote:

An alcohol bath might be a good idea. I'll see if I have some.

I find a glass of Merlot works best, with some relaxing music and candles for mood lighting.

Spencer 

Spencer Owen

unread,
Jan 31, 2017, 8:21:54 AM1/31/17
to rc201...@googlegroups.com
I think you're right, the DB9 should be a female connector.  I have to admit, however, that I don't use the RS232 output myself, and although there have been quite a few of the RS232 kits sold, this hasn't come up before.

I think that somewhere in the RS232 standard it explicitly states that whatever cable or connector you have available will be the opposite to what you need.  I know that I tend to have DB25 connectors when I need DB9, male when I need female, and null modem cables when I need straight ones.  When I did a lot more work with RS232 stuff Inevitably I ended up making my own cables up for each thing.

Spencer

--
You received this message because you are subscribed to the Google Groups "RC2014-Z80" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rc2014-z80+unsubscribe@googlegroups.com.
To post to this group, send email to rc201...@googlegroups.com.

Gracana

unread,
Feb 1, 2017, 7:08:40 PM2/1/17
to RC2014-Z80
I think that somewhere in the RS232 standard it explicitly states that whatever cable or connector you have available will be the opposite to what you need.  I know that I tend to have DB25 connectors when I need DB9, male when I need female, and null modem cables when I need straight ones.  When I did a lot more work with RS232 stuff Inevitably I ended up making my own cables up for each thing.

 That definitely seems to be how it goes. After this I bought a bunch of shells and connectors and whatnot, so now I'm prepared. Maybe.

Mild Lee Interested

unread,
Feb 2, 2017, 7:22:10 AM2/2/17
to RC2014-Z80

I think that somewhere in the RS232 standard it explicitly states that whatever cable or connector you have available will be the opposite to what you need.  I know that I tend to have DB25 connectors when I need DB9, male when I need female, and null modem cables when I need straight ones.  When I did a lot more work with RS232 stuff Inevitably I ended up making my own cables up for each thing.

Yes! It also states that the only pins that will be used consistantly for the intended purpose are 2, 3 and 5 (except when they are not, of course).  Pin 2 male on a DB9 will always be Tx (except when it's Rx) and the voltage level will always be the exact opposite of what you expect.  If you try to use recursive thinking to defeat this last part ("well I think it's going to be +15V, so therefore I'll assume it's -15V") some sort of of macro level Schrodinger's Cat effect ensures you always get it wrong.
And then you have to get the hardware handshaking to work.  My favorite are the vendors who make their equipment so that it dies expensively if you get the connecting cable pin out incorrect (looking ar you, Allen Bradley...)
Message has been deleted

James Damewood

unread,
Feb 6, 2017, 2:20:28 AM2/6/17
to RC2014-Z80
Here is a follow up.  I completed to bodge wire for the serial PCB by cutting tracings for a male DB9 connector.  Cut trace to pin 1 moved to pin 5(GND) too any ground point.  I used pin U1 pin 15 .  Pin 3 (TX) was already in the correct position. Cut trace to pin 4 and jumped to pin 2(RD). 

I am not seeing any hand shaking with pin 7(RTS) after moving from pin 8.  It is just staying at 7.4VDC from MAX232.   

Is hardware handshaking being used in the current firmware?  Otherwise the communication with my host PC works fine. 


Thank you,

James
 

 


Here is a follow up if you are using a female to female null modem.  Completed the bodge wire for the serial PCB by cutting tracings for a male DB9 connector.  Cut trace to pin 1 moved to pin 5(GND) too any ground point, using pin U1 pin 15.  Pin 3 (TX) was already in the correct position, but need to jump a wire to U1 pin 7 because I did not pay attention  while cutting.  Cut trace to pin 4 and jumped to pin 2(RD) to U1 pin 8. 
Cut trace on the other side on DB9 male pin 8 and jumper pin 7(RTS) to U1 pin 14.

I am not seeing any hand shaking with pin 7(RTS) after moving from pin 8.  It is just staying at 7.4VDC from MAX232.   

Is hardware handshaking being used in the current firmware?  Otherwise the communication with my host PC works fine. 


Thank you,

James

Joel Maslak

unread,
Apr 24, 2017, 12:59:44 AM4/24/17
to RC2014-Z80
On Monday, February 6, 2017 at 12:20:28 AM UTC-7, James Damewood wrote:
I am not seeing any hand shaking with pin 7(RTS) after moving from pin 8.  It is just staying at 7.4VDC from MAX232.   

Is hardware handshaking being used in the current firmware?  Otherwise the communication with my host PC works fine. 

Having built my RC2014 over the weekend (it works!), I can actually answer this.

Assuming you're using the 32K BASIC ROM, I can confirm it handshakes reasonably well - I can paste long BASIC listings and the like without any keyboard delay.  I get occasional glitches but they very much are the exception.  Note that if anything is accessing the UART directly, it might not properly handshake.

My setup is very simple - a null modem going into a gender changer going into a female socket.  I didn't modify any traces, although I did use the female socket instead of the male socket.

Before I hooked up the MAX232, I had a miserable time trying to get a PL2303 cable working with it (I'll probably pull the MAX232 out and try again when I have a chance to swap in less resistive resistors).  FWIW, my Rigol scope won't decode the TTL serial reliably either (lots of bogus decodes) - so it's not just PL2303.  I spent a couple hours on this issue, thinking I must have built something wrong, despite seeing normal looking signals on all bus lines. I finally decided to just try RS232 instead after seeing that my logic analyzer had no trouble decoding the TTL serial.

Thomas Riesen

unread,
Apr 24, 2017, 5:07:49 AM4/24/17
to RC2014-Z80
BTW, please note there are different MAX232 types available. The cheap MAX232CPE and the MAX232A / MAX202.
With the MAX232CPE I use always 1uF pump-capacitors, the MAX232A / MAX202 runs with 0.1uF ...

Over the time, I had several problems with USB/COM converters, sometimes strange things happen ... if I in doupt
about the circuit and wireing, I connect my circuit to a "real" COM port.

Thomas

Peter Willard

unread,
Jun 6, 2017, 7:52:40 AM6/6/17
to RC2014-Z80
There is definitely a "desired" function of the male vs female connectors.  Long ago, when dinosaur PDP-11's were still living, a MALE connector was a DTE device or "data terminal equipment" and cables "generally" were DCE or Data Communications Equipment. Both the Computer and the Terminal were considered DTE devices... so Computers and Terminals should always have male connectors.  Cables and Modems should be female, since they are communications components.  The only oddball (actually... the a good thing) is when you are wiring a Computer or a terminal to a Modem, you use a straight through cable to the modem devices.  When the modem is not used... you EMULATE the modem connection by creating a "wired" crossover of signals... hence the term... Null Modem Cable.

So, the RC2014 board "should" have had a male connector... and I think that is why some of us really want to put a male connector there...


Scott Lawrence

unread,
Jun 6, 2017, 8:40:42 AM6/6/17
to rc201...@googlegroups.com
I was exposed to VAX & PDP equipment(*) back in college, but I never got into the wiring of the MassBus or any of that behind them then. I had no idea about the gender of the connectors!  Thanks for the info on this!  :D

-s

* I was exposed to VAX & PDP equipment and all I got was this lousy t-shirt

Sent from my fancy-schmancy phone.

On Jun 6, 2017, at 7:52 AM, Peter Willard <petew...@gmail.com> wrote:

There is definitely a "desired" function of the male vs female connectors.  Long ago, when dinosaur PDP-11's were still living, a MALE connector was a DTE device or "data terminal equipment" and cables "generally" were DCE or Data Communications Equipment. Both the Computer and the Terminal were considered DTE devices... so Computers and Terminals should always have male connectors.  Cables and Modems should be female, since they are communications components.  The only oddball (actually... the a good thing) is when you are wiring a Computer or a terminal to a Modem, you use a straight through cable to the modem devices.  When the modem is not used... you EMULATE the modem connection by creating a "wired" crossover of signals... hence the term... Null Modem Cable.

So, the RC2014 board "should" have had a male connector... and I think that is why some of us really want to put a male connector there...


--
You received this message because you are subscribed to the Google Groups "RC2014-Z80" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rc2014-z80+...@googlegroups.com.

To post to this group, send email to rc201...@googlegroups.com.

Peter Willard

unread,
Jun 6, 2017, 10:21:41 AM6/6/17
to RC2014-Z80
Hehe... 19 Year DEC employee, 8 years as a communications technician... (sometimes called  Human Serial Port Switch in the early 80's)   If you have never had to wire up a 8 port loom for a DEC W991 module... you missed out on some real fun, as some of these systems did not have the concept of a bank of DB25's used for serial interfacing.  Imagine one of these with 18 solder lugs on each side of the board.  You had to remember which pins were GND TX RX RTS DTR etc.  And then solder your wires to bring the signals out of the cabinet.





I once earned the comment... "Ask Pete... he can connect anything to anything and have it working in no time"




Reply all
Reply to author
Forward
0 new messages