• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Quadra 700 overclocked at 37.5 MHz

timtiger

Well-known member
the printer was tested working with this Quadra before the overclock?
Thanks for thinking along. I was using the LabelWriter a week ago, the StyleWriter some time in autumn on this machine successfully.

The ImageWriter was not tested with the 700 before I think.

The 'leave without power for a while with no battery' trick has some weight to it, in weird cases like this.
I have cleaned the board last evening and give it later today another try. If it shows no improvement, I take it apart again and lay it by side. I won't be home until Monday anyway.
 

timtiger

Well-known member
After putting everything together again, I was doing some tests:
- Checked HD/FS integrity - both okay
- Updated drivers for StyleWriter
- Doing some plain text prints in 8.1 and 7.0.1 (clean, minimal system) on Image- and StyleWriter
- Setup a direct serial connection between an iBook and the Quadra

1) The ImageWriter is printing - but not what it should (plain text „hello world“ is something formatted)
2) The StyleWriter is not printing - but reacts on some commands
3) A direct serial connection can be setup and transfers data in both directions (Though I only tried ascii text, no control commands)

See some „screenshots“ attached.

As you both guessed - it seems to be a software thingy. But what I don‘t understand is: Why happens it on a fresh, clean, minimal system too? I am still back on 25MHz…
 

Attachments

  • 92FABB33-565E-4256-A039-95BEBD967F44.jpeg
    92FABB33-565E-4256-A039-95BEBD967F44.jpeg
    6.9 MB · Views: 21
  • 0CBDB796-616E-42D2-A2B0-FD895DCC65BF.jpeg
    0CBDB796-616E-42D2-A2B0-FD895DCC65BF.jpeg
    5.2 MB · Views: 17
  • 63711F39-48AE-480C-9CD4-81EFE01B97AC.jpeg
    63711F39-48AE-480C-9CD4-81EFE01B97AC.jpeg
    2.8 MB · Views: 13
  • F7E1F6F7-22DA-4465-BE39-65842C03C7BF.jpeg
    F7E1F6F7-22DA-4465-BE39-65842C03C7BF.jpeg
    1.2 MB · Views: 18
Last edited:

Phipli

Well-known member
Daft question - is there any reason one of the two devices wouldn't be properly earthed? A step-down transformer or power lead with no earth pin?
 

timtiger

Well-known member
Daft question - is there any reason one of the two devices wouldn't be properly earthed? A step-down transformer or power lead with no earth pin?
Both printers are untouched (in the sense of the housing being originally closed) and connected with two different proper power cables - the Quadra, too.

For the connection an original apple cable is being used.
 

Phipli

Well-known member
So, if the modern computer can talk to the Quadra, and the loopback says data and flow control works... it really doesn't feel like there is much wrong with the computer. The only things left is are the voltages right, and is there a fault with the resistor packs that doesn't show with the adapter to the modern computer, or there is a speed related issue.

Have you turned Appletalk and TCP/IP off?
 

timtiger

Well-known member
@Phipli @jeremywork Quick question: If it’s neither software nor hardware - could it be both? In other words: Could the ROM be affected because different routines for printing and serial tasks are existing? Maybe I damaged a connection from the clock to the ROM or so…
 

Phipli

Well-known member
@Phipli @jeremywork Quick question: If it’s neither software nor hardware - could it be both? In other words: Could the ROM be affected because different routines for printing and serial tasks are existing? Maybe I damaged a connection from the clock to the ROM or so…
If the ROM is broken, the checksum fails and the computer doesn't boot. If the clock was broken, the processor wouldn't run.

Weirdly, I don't think your problem is because of your overclock.

I'm also wondering if there is nothing wrong with the Q700 and the problem is elsewhere.
 

jeremywork

Well-known member
Looking at those prints, if the overclock weren’t in question at all, I would first guess a similar but technically incompatible printer driver or printing extension is being used.

Just what it *looks* like. I know you’ve installed it fresh from a source, and there really aren’t too many ways to go wrong with Macintosh printers… maybe the printer extension is a different language version than the SSW? That might be grasping at straws.
 

lobust

Well-known member
I have some thoughts on this, that may be something or nothing.

The old mac serial ports are RS422, which is a four cable differential serial connection. However, there are three additional lines in apple's implementation, two for some kind of handshaking and clock sync, and one that is called a "General purpose input".

1) Simple terminal emulation will probably only use the four basic TxD± / RxD± lines using software flow control, so is unlikely to indicate any potential faults on the remaining three lines that the printers may or may not use.

2) It's not uncommon for differential signals to sort of kind of work if either side of the differential pair fails. For example, RxD LOW is 0V on both RxD- and RxD+, while RxD HIGH is (usually, for older equipment) 5V on RxD+ and -5V on RxD-, resulting in a differential voltage of 10V seen by the line receiver. If for example the RxD- line driver fails during RxD HIGH, you will have some floating voltage around 0V on that line, but RxD+ will still be dutifully pumping out its 5V, which results in some low-but-non-zero differential voltage that may or may not trigger the RxD line receiver.

I would lean towards 1) as being likely, RxD and TxD are working, but the printers use hardware flow control to increase communication bandwidth, and one or both of those lines is not working properly resulting in incomplete and corrupted instructions being received by the printer. Of course this is speculation, but it seems likely to me.

Do any serial port sniffer applications exist for classic mac os?
 

cheesestraws

Well-known member
printers use hardware flow control to increase communication bandwidth, and one or both of those lines is not working properly

Good thought. OP, you should be able to test this by turning hardware flow control on on your terminal emulator during the two-computer test...
 

Phipli

Well-known member
Good thought. OP, you should be able to test this by turning hardware flow control on on your terminal emulator during the two-computer test...
They said that snooper confirmed flow control in the loopback I thought? I suspect I misunderstood.
 

timtiger

Well-known member
Snooper tested Handshake and different transfer rates… It’s worth to test flow control with Kermit…

Will do this on sunday - Thank you all!
 

lobust

Well-known member
They said that snooper confirmed flow control in the loopback I thought? I suspect I misunderstood.

Oops, I missed that!

I looked at Branchus' snooper loopback pinout and it is specifically testing TxD/RxD- and handshake lines. TxD/RxD+ are not connected. That is probably a non-issue as those are the two that would normally be connected when hooked up to an RS232 device (which the IW at least appears to be).

I haven't done a loopback test in Snooper before (no test cables and no need so far) - does it have a signal display or does it just run with no feedback and issue a pass/fail?

I have a fair amount of experience troubleshooting serial connections, but not with vintage macs unfortunately.

Edit: The original ImageWriter is single-ended rs232, but OP said IW II, which does use rs422 differential. So I'd suggest that the snooper loopback test is not proof that this is working properly. Supporting that theory, OP used an iBook as the terminal emulator, which means a USB-Serial adapter, which means rs232 not rs422, which also does not prove full function of the serial ports...

source: http://mirrors.apple2.org.za/ground.icaen.uiowa.edu/MiscInfo/Cables/Pinouts.txt
 
Last edited:

timtiger

Well-known member
does it just run with no feedback and issue a pass/fail?
It runs with no feedback and tells only pass/fail afterwards.
So I'd suggest that the snooper loopback test is not proof that this is working properly. Supporting that theory, OP used an iBook as the terminal emulator, which means a USB-Serial adapter, which means rs232 not rs422, which also does not prove full function of the serial ports...
Check. To test the rs422 port to check full function, could I connect an SE/30 directly to the Q700 by using Kermit? If so, which cable could I use?
 

lobust

Well-known member
It runs with no feedback and tells only pass/fail afterwards.

Check. To test the rs422 port to check full function, could I connect an SE/30 directly to the Q700 by using Kermit? If so, which cable could I use?

I assume a standard mac 8 pin mini-din serial cable will have all lines connected at both ends, but whether Kermit or any other terminal program on the mac can be made to actually use RS422 I just don't know.

I'd be more interested in learning what your other printers use. If it can be determined that they all use RS422 differential signalling then I'd take that as more evidence that one/both of the TxD/RxD+ signals is not working. I don't know what the transceiver setup is like on the 700, but both ports exhibiting the same behaviour can mean one of a few things:

a) this theory is totally wrong and I am leading you on a wild goose chase
b) they have both failed with same failure mode by pure coincidence
c) the fault lies upstream of the serial port transceivers
d) the 700 somehow uses the same transceiver for both ports
 
Top