[Coco] CoCo 3 to RBG...

[gene heskett]

On Wednesday, July 20, 2011 09:27:13 PM Tim Fadden did opine:

> On 7/20/2011 2:37 PM, Steven Hirsch wrote:
> > On Wed, 20 Jul 2011, gene heskett wrote:
> >> On Wednesday, July 20, 2011 04:52:43 PM Steven Hirsch did opine:
> >>> On Wed, 20 Jul 2011, gene heskett wrote:
> >>>> On Wednesday, July 20, 2011 01:00:14 PM Steve Batson did opine:
> >>>>> Steve,
> >>>>> 
> >>>>> Can you provide details on building your 2 chip circuit to get it
> >>>>> working?
> >>>>> 
> >>>>> I received mine this week, built cables to Map Coco 3 RGB signals
> >>>>> to the on board 8 pin connector as they describe. The board
> >>>>> powered fine, I can access it's menu and functions, but no Coco
> >>>>> output. Just a green screen, but I don't think it's Coco output
> >>>>> though. Coco works fine with composite  out to TV.
> >>>>> 
> >>>>> Any help will be appreciated!  :)
> >>>> 
> >>>> This is going to be best troubleshot with a decent scope, only that
> >>>> can give you a definitive answer.
> >>>> 
> >>>> Oh Wait, you said the 8 pin connector?  Humm, I may be half a
> >>>> bubble off here, but ISTR reading in the excruciatingly fine print
> >>>> (I had to drag out an old projector lens & use it for a magnifying
> >>>> glass to read that teeny little booklet) that the 8 pin connector
> >>>> only handles 31khz & up sources, so I had not considered doing
> >>>> anything but stuffing the coco's signals into the adjacent DB15,
> >>>> which claims to work down to 15 khz, and which means one must pay
> >>>> attention to the pinout of the 10 conductor cable from the coco as
> >>>> to where it goes in the DB15.
> >>> 
> >>> Where in the manual does it say that?  The one I'm looking at claims
> >>> that 15Khz. video can be applied to P3 (The large pins on the left),
> >>> P11 (8-pin SIL) or P12 (HD15).
> >> 
> >> Did you not get the xx20 version of that board?  Mine came with dual
> >> db15
> >> outs on the back edge.  And my booklet text says P3, P10, & P11 for
> >> the 15
> >> khz input.
> > 
> > I have an older revision.  But, I think I misread your paragraph
> > above.  I thought you were claiming that ONLY the DB15 accepted
> > 15.75Khz. video. Ain't the English language great?  I agree that all
> > three are intended to sync at that freq.  But, the big pins on P3 do
> > not have anything brought out for H+V sync.
> > 
> >>> Since P3 appears to bring out only the
> >>> composite sync input I never tried it.  I can tell you that H+V on
> >>> P11 never worked for me.  I made the (perhaps erroneous) assumption
> >>> that the H and V pins on the HD15 were connected to the same
> >>> circuit nodes as H+V on P11.  But you know what they say about
> >>> "assume" :-).
> >> 
> >> Yes, its been applied to me on numerous occasions.
> >> 
> >> A digital meter in ohms mode can tell that story.
> > 
> > I'll check after dinner this evening.
> > 
> >>> I've never been able to get mine to work with H+V sync on a 15Khz.
> >>> signal. I wrote to their tech support and recall being told that it
> >>> wasn't intended to work with H+V at that horz. freq - only composite
> >>> sync.
> >> 
> >> That's a genuine class A bummer if by composite, they also mean fully
> >> interlaced.
> > 
> > No nothing that draconian.  It simply needs a dumb two-chip sync
> > combiner. I built one up on a proto-board and it worked just fine.
> > I'll send you the GIF of the schematic in private e-mail.
> > 
> >> Taking the coco's H & V, and combining them does not a
> >> composite signal make.  In std Never Twice Same Color, the V Sync is
> >> not a
> >> solid 3 line long signal, it is 'serrated' by returning to black at
> >> 2x the
> >> h-rate, such that there is a 4.7 u-sec hole, which returns to the
> >> sync output such that each serration end, is co-incident with the
> >> falling edge
> >> of the H-sync for every other serration.  That is why, years ago, you
> >> could
> >> adjust the vertical hold to make the picture roll down, and the
> >> vertical sync bar you could see had a square tail on the left end of
> >> the bar from the middle to the right of the screen.  The one you
> >> couldn't see at the left end would have been identical.  Its
> >> actually slightly more complex than that because for 3 lines ahead
> >> of V-sync, the H sync pulse was reduced
> >> to half its normal time to reduce any dc offsets created by the
> >> apparent doubling of the repetition rate.  Ditto for 3 lines after
> >> the V pulse for
> >> exactly the same reason.  This is what forms the arrow's 'tail' you
> >> could
> >> see.
> >> 
> >> Making composite by Nanding the two signals substitutes a solid bar 3
> >> H lines long, which contains no H sync info, and causes the hooking
> >> to the left or right at the top of the picture because the H sync is
> >> momentarily
> >> lost.  That might well be the only thing we can do in which case
> >> there is a
> >> bit of false advertising involved.
> > 
> > Wow.  Now I know who to ask about video particulars!
> > 
> >> However I am still confused by the apparent differences in our
> >> booklets, and now wonder which is correct.  Or are you miss reading
> >> the PCB because
> >> the photos do not call the connectors out by their P3, P10 or P11
> >> designations?
> > 
> > No, I think we were saying the same thing.
> > 
> > Steve
> 
> I also received one of these boards. (yesterday) Please share the two
> chip combiner circuit with me also.  I'm pretty sure I could build it,
> but not bright enough to design it!  :-)
> 
> Thanks
> 
> Tim Fadden
> 

You may want to hang on for a day maybe.  I've been doing some poking 
around in my machine with my scope and found that while the gime's output 
is right side up, sync at logic 0, video active time in between pulses is 
logic 1.  BBuuuuttt, they send it on to the video header by way of 
buffering in IC15, under the keyboard ribbon, which inverts it.  Its 
entirely possible that all we need to do is re-invert it so we have 
negative sync.  Because of the top of the pix sync shake when combined, I 
would much rather it work with separate drives.

Steve, if you have an in path to these folks, can you ask what polarity of 
sync it wants on the 8 pin connector?

In my checking the Vcc being given to the gime on pin 35 in my machine, 
it's 4.79-4.81 volts, which is borderline low for the gime.  Going back to 
where my 5 volt input from the PSU hits the 5 volt line to everything at a 
point adjacent to the SALT chip, I find a nice fat 5.02 volts.  WTH???  
Then by accident almost, I note that slightly to the rear and right of the 
gime it says FB5 on the board screen print, which is in series with the 
gime's Vcc pin 35.  That is a Ferrite Bead, and should have only milli-ohms 
of resistance, causing maybe a millivolt loss at most.

Huh?  Power down, measure its r because it sure looks like a resistor to 
me, 2.2 ohms, which at the gimes draw, accounts for most of the loss.  Hunt 
around and finally on another board, find a wire standing there with a 
Ferrite Bead on it.  Warm up the iron and extract it, add a small blob of 
GC's silver bearing solder to each end of FB5 where it goes into the board, 
then bridge it with this bead.  Bingo, 4.99 volts at the gimes pin 35, and 
a lot brighter display.  Crisper too I think.

But however, I have a more pressing problem in that I seem to have lost the 
heads or something in /d0, my token 40 tpi drive.  I have had several power 
bumps that my UPS's gobbled up in the last week, but when I thought to go 
check the coco, mainly because the LED's on my rsw-232 sniffer had gone 
out, and found the coco sitting there, spinning /d0 in place but totally 
crashed.  And likely had been since the bump the day before.  The disk in 
it shows no sign of a well polished or worn track and I can run  mb.dw3 on 
it but get errors I wish the script was smarter about reporting, but I'll 
have to put some echo's in the script to see what is going on.  I may, no 
certainty, have one more 40 track drive if the heads on that one are ~30~.

If not, how is your supply of those at Cloud 9 north, Mark M.?

This tells me I had better bring in an old 350 watt UPS I fitted with an 
external, much bigger battery a year back and see if I can rig it for the 
coco, cuz if a power bump of 1/4 second can raise this much hell, I had 
better be installing some insurance.

Cheers, gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Landru! Guide us!
		-- A Beta 3-oid, "The Return of the Archons", stardate 
3157.4