2.5 VDC is much too low. Are you sure you used 1 uF caps as shown in the data sheet? Also, those caps should be tantalums for best results—tantalums are polarized, BTW. On my POC unit, I'm seeing about +9.5 VDC coming out of the MAX238 charge pump circuit (pin 11—the MAX238 is two MAX232s in one package). The corresponding pin on the MAX232 is pin 2. You should see at least +8 VDC positive to ground at that pin if everything is right. You should also see the same voltage, but opposite polarity, at pin 6. If you see these voltages but only a couple of volts at TxD you have something loading down TxD (and possibly RxD as well).


You may need to get another one if all else fails.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

The Micro UK101 circuit diagram shows them as 22uF electrolytic capacitors:

http://home.micros.users.btopenworld.co ... uk101.html

That is what I have used. Lee built one too, and he said it works - no mention of changing any of the components:

viewtopic.php?p=13790#13790

..confused.. ..again..

Correct. I actually built it twice, I set fire to the first one, the second time I replaced the decode logic with a single GAL and added another 32K RAM so unused blocks or ROM could be replaced by RAM.

Looking at the layout diagram however I see that the +V capacitor on the MAX232, the one that goes to pin 2, is the wrong way round. I spotted this when I built it and emailed Grant but I see it hasn't been fixed. It is correct in the schematic though. Perhaps both were originally wrong.

Lee.

I get you - the layout that shows the logic & glue wiring is different to the schematic. NMaybe he fixed the schematic but not the layout. Doesn't brook well for the next project (Grant's Jupiter Ace implementation) although there's just a schematic for that, no layout, so I may be in luck...

ARRRGHHHHHH!

That's better.

Nice one, Lee, I'll turn it round tout suite.

Logic probe is in the post, but I lost Oscilloscope number 5 to an auction sniper (again). Must try harder...

Heh, heh, heh.... My time for 6502 pyrotechnics will come I am sure.

Hang on a mo... the schematic shows the -ve end of the capacitor connecting to Vcc - is that correct?

..so..

Vcc -------[ - 22uF + ]------- +V (MAX232 pin 2)

..is correct? Seems counter intuitive to conect -ve on a cap to +ve power supply..?

Yes, because the + side of the capacitor is near 10V. RS-232's minimum drive voltages are ±5V, and recommended is ±12V. The MAX232 takes the 5V and doubles it (which with a little loss gets close to 10V), and also inverts it for the negative drive voltage.

OK - Clock pulses, can't check until I get the logic probe.
RSS is high after reset (it gets pulled low by the reset button). I assume this is what you mean by /RESET?
Rdy, IRQ, NMI are all held high.

After turning the serial line cap round, it is still reading low on the Tx line, but the ACIA is not plugged in as I am still trying to get the NOPGEN to do something. Inserting ACIA makes no difference, Tx and Rx are both showing ~4v, as is the +ve side of that capacitor (on Pin 2 of the MAX232). Does it need Rx / Tx pulses from the 6850 to give 10v on the outputs?

The UK101 kit is wrong. Please review the MAX232 data sheet.


According to the MAX232 data sheet, capacitors C1 through C5 should be 1uF (same for the MAX238).

In a conversation I had with a Maxim engineer sometime ago, I was told that exceeding the recommended charge pump capacitor values produces no beneficial effect and will cause eventual failure of the chip. He described cases where large values, such as the 22uF you mentioned, causing failure within a year of service, due to the very large initial charging current at power-on. He also recommended not using an electrolytic. I used tantalums in the recommended values and achieved excellent performance.

I strongly recommend you stick with Maxim's recommendations. They designed the parts and would be the final authority on their proper application.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

What about BE (bus enable)? It must be pulled high, or else the MPU will tri-state the data and address buses.


Remove the 6850, disconnect anything attached to the EIA-232 port and make your voltage measurements at MAX232 pins 2 and 6, not TxD and RxD.

The MAX's charge pump operation has nothing to do with the ACIA. You should see ~10 VDC positive on pin 2 and ~10 VDC minus on pin 6. If you used those big electrolytics mentioned above, yank 'em out of there and put in 1uF tantalums like MAX recommends. I'm reasonably certain it'll work, assuming the big caps didn't already damage the chip.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

I'm not sure how relevant it is to this discussion but my recollection is
that the original MAX232s ran the charge pump slower and used bigger
caps (without looking it up, my recollection is 10uF was the recomendation)

I thought the faster ones were MAX232A.

He has the Rockwell part, so there's no BE pin.

OK - I'll try it. You're absolutely right, the datasheet says 1uF. My part is a Maxim MAX232CPE which looks like the basic MAX232. I have to assume it is zapped now, as from what you and others have said, it should still work with the 22uF caps, if only for a little while.

No BE pin on mine (or on the schematic / wiring diagram). It's a Rockwell R6502AP. I looked up the Rockwell datasheet and couldn't see BE either. Which pin number is it supposed to be? One of the "NC" pins? I assume that if it's not present I can ignore the requirement to keep it high.

There is a lot of variation in cap values in published circuits using the MAX232CPE. I use 10uF in mine.

The 'bus enable' BE pin is a feature of the WDC 65C02. Its not present on earlier 6502s

_________________
Andrew Jacobs
6502 & PIC Stuff - http://www.obelisk.me.uk/
Cross-Platform 6502/65C02/65816 Macro Assembler - http://www.obelisk.me.uk/dev65/
Open Source Projects - https://github.com/andrew-jacobs

Your recollection was correct. I've used the MAX232 in projects since the chip's inception back in the 1980s. My first application was to take the fake RS-232 output of the C-128 and produce the true RS-232 voltage levels. At that time, the capacitor recommendation was a 10uF electrolytic. A redesign of the chip c. 1991 reduced the capacitor size to 1uF, and a revised data sheet was released highlighting the change. It was some time in 1992 when I spoke with Maxim on whether the old (i.e., large) sizes would still work. I was told to use the 1uF to avoid harming the chip. I have done so ever since.

The reasoning behind using tantalums instead of electrolytics is the former exhibits a lot less leakage, thus making the charge pump more efficient. Electrolytics will work, just not as well.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!