11 replies to this topic

[Fox-1 / mnx]

I never really had any problems with XL's. If there were any they were always related to (memory-)upgrades. But what about that thing called "delay line" ? (U13 in 600XL)

Just came across a 3rd XL (stock) with a broken one. 2 Were 800XL's, and now this 600XL. Is there a way to not use this component by using another way to create the delay? I don't even know what's in it (a coil?).

[Marius1976]

Is it that 'odd' looking IC where some pins are missing?

[candle]

its bunch of coils with some caps
you can check atari 5200 for reference, and atari 800 personality board schematics to get some clue about how it works
freddy uses synchronous clock (8x system clock) and shift register to have the timing well defined and always in sync and good timing proportions regardless of temperature, humidity and other factors analog circuitry may suffer from
its all about phi2 signal being delayed by 1/4 so RAS and CAS signals could fit in inside the same clock cycle
you could also get rid of it and put SRAM chip in place of all that DRAM circuitry
all it takes is RW to WE/OE decoding logic plus cas inhibit connected to one of chip select inputs of 128k SRAM chip, if you want to get whole picture, make some provisions for EXTSEL line too (OR gate non inverted CS input of SRAM chip will suffice)

[Fox-1 / mnx]

Marius1976, on Mon Jan 30, 2012 1:20 PM, said:

Is it that 'odd' looking IC where some pins are missing?

That's the one.

candle, on Mon Jan 30, 2012 1:57 PM, said:

its bunch of coils with some caps

Ah, thats what I thought.

candle, on Mon Jan 30, 2012 1:57 PM, said:

you could also get rid of it and put SRAM chip in place of all that DRAM circuitry
all it takes is RW to WE/OE decoding logic plus cas inhibit connected to one of chip select inputs of 128k SRAM chip,

I was hoping for an easier solution, like replacing that one IC with some other replacement components but it doesn't seem that easy :-) In that case I think the cleanest way is to try to find a replacement of the original.

[Marius1976]

Fox-1 experiences that 'many' of those components seems to fail now or soon.

Is this an expectable fact that this component won't last as long as let's say any LSxxx chip?

Even when it's not used?

[Rybags]

According to the 1064 schematics, it does away with the delay line and feeds the signal through an invertor 4 times to acheive the delay.

[candle]

i wonder why - CAS and RAS are present on PBI connector

[Fox-1 / mnx]

Rybags, on Mon Jan 30, 2012 3:30 PM, said:

According to the 1064 schematics, it does away with the delay line and feeds the signal through an invertor 4 times to acheive the delay.

Interesting. That would translate to about 40nS, I think. I'm rusty but can look it up.


Is below info right?

Standard XL delay line IC has 5 delayed outputs, with O2 as input. Outputs are 25, 190, 260, 315 and 440nS.


How tight are these values? I can imagine 25nS is tight but how far can I drift away from 315 and 440nS?

[Rybags]

Not sure... the standard propogation delay on LS devices seems to be anything 10-30 ns. 4 x 30 gives 120 which is getting near to 1/4 a cycle.

[tf_hh]

Fox-1 / mnx, on Mon Jan 30, 2012 12:41 PM, said:

Just came across a 3rd XL (stock) with a broken one.  2 Were 800XL's, and now this 600XL.  Is there a way to not use this component by using another way to create the delay?  I don't even know what's in it (a coil?).

You have three defective delay lines? Or the memory expansion won´t work correctly?

[Fox-1 / mnx]

tf_hh, on Tue Jan 31, 2012 2:44 AM, said:

You have three defective delay lines? Or the memory expansion won´t work correctly?

No memory expansions involved. All 3 XL's are stock (2 x 800XL, 1 x 600XL).

1st died 2 (3?) years ago. Got a replacement from someone (from Guus?).

2nd died last year. Found a replacement that came with some stuff I bought from someone.

3rd died last week. Or better, it was put away in working condition years ago but now it doesn't work anymore.


Of course it's all old stuff so one can expect problems over time but it's always the same issue. Apart from an odd RAM and one particular capacitor in XLF's it all keeps running without problems. The fact that it all still works also has a side effect. I have close to none Atari-specific spare parts! That's why I want to solve this little issue by using more common parts (standard delay line IC's) but did some research at the meanwhile and I don't think this will lead to a good solution. Delay lines seem to be much more expensive than I thought they would be and there doesn't seem to be one that covers the whole ranch as used in an XL (5 delayed outputs from 25nS to 440nS) so I'll have to use either 2 of them (2 x 8 delayed outputs) or using a programmable one that's even more expensive.

I guess I'll just have to live with it and put it away in non-working condition. Ordering such a part from Best won't do as I can just buy a whole XL for less money over here.

[Rybags]

I wonder could something alternative be rigged up - maybe feeding the 3.59 MHz clock direct could give more options?