Source for 68A09 CPU or replacement?

dropde...@gmail.com

unread,

Aug 31, 2011, 4:35:59 PM8/31/11

to

I am trying to repair a friend's vectrex that is dead as a doornail,
except for a white dot, is there any source for 68A09 CPU (not bob
roberts) or is there another replacement I can use it it's place
(assume 1mhz 6809 is too slow, 2mhz 68B09, would it need a different
clock crystal (what value?)

anything else to watch out for?

thanks

Mittens

unread,

Aug 31, 2011, 5:00:47 PM8/31/11

to

On Wed, 31 Aug 2011 21:35:59 +0100, uncle...@jerk.org
<dropde...@gmail.com> wrote:

> I am trying to repair a friend's vectrex that is dead as a doornail,
> except for a white dot, is there any source for 68A09 CPU (not bob
> roberts) or is there another replacement I can use it it's place
> (assume 1mhz 6809 is too slow, 2mhz 68B09, would it need a different
> clock crystal (what value?)

MC68B09 (Motorola) or HD68B09 (Hitachi) will be a straight swap. No need
to change any crystals. Just make sure you don't get an 'E' suffix as the
pinout is slightly different. Ebay has several of them at the moment.

> anything else to watch out for?

Are you sure it's just the CPU that's toasted, or if the CPU is broken at
all? Sounds like it might be something simple like a power supply rail has
gone down. You shouldn't see a dot if the analog circuitry is working
right. It should automatically blank if it gets no signals from the CPU
after a very short while to protect the CRT from getting burnt.

If the CPU is dead - has a 5V supply but the built in oscillator isn't
running - then chances are some of the other logic will have been killed
by whatever mishap caused the CPU to fry.

dropde...@gmail.com

unread,

Aug 31, 2011, 10:04:36 PM8/31/11

to

On Aug 31, 5:00 pm, Mittens <mittens@_nospam_hush.ai> wrote:
> On Wed, 31 Aug 2011 21:35:59 +0100, unclefre...@jerk.org

>
> <dropdeads...@gmail.com> wrote:
> > I am trying to repair a friend's vectrex that is dead as a doornail,
> > except for a white dot, is there any source for 68A09 CPU (not bob
> > roberts) or is there another replacement I can use it it's place
> > (assume 1mhz 6809 is too slow, 2mhz 68B09, would it need a different
> > clock crystal (what value?)
>
> MC68B09 (Motorola) or HD68B09 (Hitachi) will be a straight swap. No need
> to change any crystals. Just make sure you don't get an 'E' suffix as the
> pinout is slightly different. Ebay has several of them at the moment.
>
> > anything else to watch out for?
>
> Are you sure it's just the CPU that's toasted, or if the CPU is broken at
> all? Sounds like it might be something simple like a power supply rail has
> gone down. You shouldn't see a dot if the analog circuitry is working
> right. It should automatically blank if it gets no signals from the CPU
> after a very short while to protect the CRT from getting burnt.

I am not sure, I was just going by reading old posts and other stuff
online.
I have replaced the CPU with a socket,have to find a cpu to populate
it with now.

> If the CPU is dead - has a 5V supply but the built in oscillator isn't
> running - then chances are some of the other logic will have been killed
> by whatever mishap caused the CPU to fry.

well that's not good. I did a cap kit on it as well, I guess next step
is put it
together and turn it on see what happens after I get a CPU.

dropde...@gmail.com

unread,

Sep 6, 2011, 9:11:06 PM9/6/11

to

On Aug 31, 5:00 pm, Mittens <mittens@_nospam_hush.ai> wrote:

> On Wed, 31 Aug 2011 21:35:59 +0100, unclefre...@jerk.org

>
> MC68B09 (Motorola) or HD68B09 (Hitachi) will be a straight swap. No need
> to change any crystals. Just make sure you don't get an 'E' suffix as the
> pinout is slightly different. Ebay has several of them at the moment.

besides the "E" suffix what else should I watch for? "P" seems ok how
about "C"?

Mittens

unread,

Sep 7, 2011, 4:02:26 AM9/7/11

to

Yes, 'P' apparently denotes plastic package for Motorola 6809s. I have a
68B09P here and it has the same pin out. I have not seen a 'C' suffix, but
I believe it might refer to extended temperature range for military use
and might be followed by an additional 'P'. MC68A09CP, MC6809C, etc. There
might also be 'S' for ceramic (not seen one of those either). Either way,
if it doesn't have the 'E' anywhere hear the the end it /should/ just work.

If it will save you a lot of aggravation procuring it, you could in theory
get an 'E' to work by piggybacking a canned oscillator. The 'E' parts lose
the built in oscillator circuit but gain multi bus mastering capability
for multi processor systems, which obviously the Vectrex wouldn't use.
There would be some effort involved in working out which pins need to be
rerouted, but most of the bus signals are on the same pins AFAIK.

i...@millenniumrobotics.com

unread,

Sep 7, 2011, 7:18:41 PM9/7/11

to

On Sep 7, 2:11 am, "unclefre...@jerk.org" <dropdeads...@gmail.com>
wrote:

well the micrprocessor type in my Vectrex consoles is the Motorola
MC68A09P.
I saw one on ebay for sale now:
http://www.ebay.co.uk/itm/MC68A09P-Motorola-CPU-40-Pin-Vintage-Collectible-NEW-/120487695385?pt=LH_DefaultDomain_0&hash=item1c0da05419

there are also on ebay the same model number chips but made by
FairChild and Hitachi.

VectrexMad!
http://www.vectrex.co.uk

dropde...@gmail.com

unread,

Sep 7, 2011, 11:35:52 PM9/7/11

to

On Sep 7, 7:18 pm, i...@millenniumrobotics.com wrote:
> On Sep 7, 2:11 am, "unclefre...@jerk.org" <dropdeads...@gmail.com>
> wrote:
>
> > On Aug 31, 5:00 pm, Mittens <mittens@_nospam_hush.ai> wrote:
>
> > > On Wed, 31 Aug 2011 21:35:59 +0100, unclefre...@jerk.org
>
> > > MC68B09 (Motorola) or HD68B09 (Hitachi) will be a straight swap. No need
> > > to change any crystals. Just make sure you don't get an 'E' suffix as the
> > > pinout is slightly different. Ebay has several of them at the moment.
>
> > besides the "E" suffix what else should I watch for? "P" seems ok how
> > about "C"?
>
> well the micrprocessor type in my Vectrex consoles is the Motorola
> MC68A09P.

> I saw one on ebay for sale now:http://www.ebay.co.uk/itm/MC68A09P-Motorola-CPU-40-Pin-Vintage-Collec...

>
> there are also on ebay the same model number chips but made by
> FairChild and Hitachi.

too expensive. My budget is $50 and those cpus are $40ppd alone. I've
already spent $20+ on capkit.

dropde...@gmail.com

unread,

Sep 15, 2011, 5:21:48 PM9/15/11

to

[slice]

so I sourced some used MC68B09P cpus (10 for $24, some of them have to
work right?) until this shows up can I
test the unit with a plain 6809 or will that fry it in 2 seconds? I'm
getting antsy haha.

Mittens

unread,

Sep 16, 2011, 5:29:44 AM9/16/11

to

No, it won't fry. It just might not work properly. Or at all. And you
won't be able to rule out whether the old processor was duff or something
else is wrong.

dropde...@gmail.com

unread,

Oct 11, 2011, 10:30:52 AM10/11/11

to

On Sep 16, 5:29 am, Mittens <mittens@_nospam_hush.ai> wrote:
> On Thu, 15 Sep 2011 22:21:48 +0100, unclefre...@jerk.org

>
> <dropdeads...@gmail.com> wrote:
> > [slice]
>
> > so I sourced some used MC68B09P cpus (10 for $24, some of them have to
> > work right?) until this shows up can I
> > test the unit with a plain 6809 or will that fry it in 2 seconds? I'm
> > getting antsy haha.
>
> No, it won't fry. It just might not work properly. Or at all. And you
> won't be able to rule out whether the old processor was duff or something
> else is wrong.

ok been taking my time/slacking on this, I replaced all the
electrolytic caps and the cpu and I still got nothing when turned on.
I removed the logic board and put it in a known working vectrex
chassis tested DC at logic board connector have 5v, -5v and 12v there
and nothing, tested a couple cpus in the socket ... so what does this
leave? RAM? Crystal? I don't think blown 6522 would do this as far as
I could tell from reading.

Any additional tips? (with links if applicable!)

Mittens

unread,

Oct 11, 2011, 2:33:02 PM10/11/11

to

On Tue, 11 Oct 2011 15:30:52 +0100, uncle...@jerk.org
<dropde...@gmail.com> wrote:

> ok been taking my time/slacking on this, I replaced all the
> electrolytic caps and the cpu and I still got nothing when turned on.
> I removed the logic board and put it in a known working vectrex
> chassis tested DC at logic board connector have 5v, -5v and 12v there
> and nothing, tested a couple cpus in the socket ... so what does this
> leave? RAM? Crystal? I don't think blown 6522 would do this as far as
> I could tell from reading.
>
> Any additional tips? (with links if applicable!)

I never heard of a crystal failing, but that's not to say it couldn't
happen. Do you have a scope? Probing the crystal will at least tell you if
it's oscillating. If it's not and the CPUs you tried are good, then it or
either of the two load caps connected to it might be dead.

If it is oscillating then check all the IRQs, Halt and Reset pins are
deasserted (up at +5V).

If they are, sorry to say probably any of the other chips could be
suspect. I would start with the most complex, 6522, AY-3, then RAM, then
work my way down. All the I/O goes through the 6522, so a dead one of
those will mean no sounds or picture. At least they are relatively easy to
come by. AY-3's are getting rare and expensive. Address decoding is done
in a bunch of TTL chips, so any one of those might stop it working but
they are generally pretty reliable.

dropde...@gmail.com

unread,

Oct 11, 2011, 5:40:16 PM10/11/11

to

On Oct 11, 2:33 pm, Mittens <mittens@_nospam_hush.ai> wrote:
> On Tue, 11 Oct 2011 15:30:52 +0100, unclefre...@jerk.org

I do have a scope it hasn't been calibrated since april 1970 (lol) but
for this
purpose if it turns on I think it's good enough. I have a logic probe
would that
be a proper tool in CMOS mode to test pins on the CPU and other chips?
Or just go with voltage levels on multimeter?

Mittens

unread,

Oct 11, 2011, 6:08:11 PM10/11/11

to

On Tue, 11 Oct 2011 22:40:16 +0100, uncle...@jerk.org
<dropde...@gmail.com> wrote:

>
> I do have a scope it hasn't been calibrated since april 1970 (lol) but
> for this
> purpose if it turns on I think it's good enough. I have a logic probe
> would that
> be a proper tool in CMOS mode to test pins on the CPU and other chips?
> Or just go with voltage levels on multimeter?

I have only ever used a multimeter, but either will do I guess. (Or the
scope in free running 'auto' mode.)

Good luck.

dropde...@gmail.com

unread,

Oct 11, 2011, 8:08:13 PM10/11/11

to

On Oct 11, 6:08 pm, Mittens <mittens@_nospam_hush.ai> wrote:
> On Tue, 11 Oct 2011 22:40:16 +0100, unclefre...@jerk.org

>
> <dropdeads...@gmail.com> wrote:
>
> > I do have a scope it hasn't been calibrated since april 1970 (lol) but
> > for this
> > purpose if it turns on I think it's good enough. I have a logic probe
> > would that
> > be a proper tool in CMOS mode to test pins on the CPU and other chips?
> > Or just go with voltage levels on multimeter?
>
> I have only ever used a multimeter, but either will do I guess. (Or the
> scope in free running 'auto' mode.)
>
> Good luck.

well I get 5V on VOM, but no pulse on the crystal as far as I could
tell, I tried multiple voltages and time scales, flat all the way. I
tested the scope by hooking up the output of a CD walkman and I saw
some action doing that so I am fairly certain I'm doing it right,
tested every combination of the 2 leads and case ground on the XTAL.

Do these look right as a replacement? different shape than current one
and bob roberts had Crystals and Clock Oscillators in slightly
different categories so I'm not 100% positive

http://www.ebay.com/itm/5-x-6Mhz-High-Precision-Crystal-Clock-Oscillator-/300530157959?pt=LH_DefaultDomain_0&hash=item45f8fe4987

Mittens

unread,

Oct 12, 2011, 3:28:09 AM10/12/11

to

On Wed, 12 Oct 2011 01:08:13 +0100, uncle...@jerk.org
<dropde...@gmail.com> wrote:

> well I get 5V on VOM, but no pulse on the crystal as far as I could
> tell, I tried multiple voltages and time scales, flat all the way. I
> tested the scope by hooking up the output of a CD walkman and I saw
> some action doing that so I am fairly certain I'm doing it right,
> tested every combination of the 2 leads and case ground on the XTAL.
>
> Do these look right as a replacement? different shape than current one
> and bob roberts had Crystals and Clock Oscillators in slightly
> different categories so I'm not 100% positive
>
> http://www.ebay.com/itm/5-x-6Mhz-High-Precision-Crystal-Clock-Oscillator-/300530157959?pt=LH_DefaultDomain_0&hash=item45f8fe4987

Yep, they should be ok. Before you do though, what exactly do you read on
pins 1, 7, 37 of the 6809. Expect 0V, 5V, >2.5V respectively. If you've
got anything different see if you can fix that first before buying a new
crystal. If the oscillator is running I would expect to see about 1V on
pins 38 and 39 on a multimeter.

dropde...@gmail.com

unread,

Oct 12, 2011, 10:53:31 AM10/12/11

to

On Oct 12, 3:28 am, Mittens <mittens@_nospam_hush.ai> wrote:
> On Wed, 12 Oct 2011 01:08:13 +0100, unclefre...@jerk.org

>
> <dropdeads...@gmail.com> wrote:
> > well I get 5V on VOM, but no pulse on the crystal as far as I could
> > tell, I tried multiple voltages and time scales, flat all the way. I
> > tested the scope by hooking up the output of a CD walkman and I saw
> > some action doing that so I am fairly certain I'm doing it right,
> > tested every combination of the 2 leads and case ground on the XTAL.
>
> > Do these look right as a replacement? different shape than current one
> > and bob roberts had Crystals and Clock Oscillators in slightly
> > different categories so I'm not 100% positive
>

> >http://www.ebay.com/itm/5-x-6Mhz-High-Precision-Crystal-Clock-Oscilla...

>
> Yep, they should be ok. Before you do though, what exactly do you read on
> pins 1, 7, 37 of the 6809. Expect 0V, 5V, >2.5V respectively. If you've
> got anything different see if you can fix that first before buying a new
> crystal. If the oscillator is running I would expect to see about 1V on
> pins 38 and 39 on a multimeter.

1=0v
7=5v
37=5v
38=5v
39=0.89v

the 37 and 38 are high I guess, when you say "fix that first" am I
looking
at resisters and caps then?

I already ordered the crystals found another ebay dealer with
3 for $1.39 postpaid so no great loss if I don't need.

Mittens

unread,

Oct 13, 2011, 6:42:11 AM10/13/11

to

On Wed, 12 Oct 2011 15:53:31 +0100, uncle...@jerk.org
<dropde...@gmail.com> wrote:

> 1=0v
> 7=5v
> 37=5v
> 38=5v
> 39=0.89v
>
> the 37 and 38 are high I guess, when you say "fix that first" am I
> looking
> at resisters and caps then?

They are all fine I reckon. 1 and 7 are the power pins and 37 is nReset.
AFAIK those are the only ones which should have any effect on the free
running of the oscillator.

38 and 39 are the oscillator pins, and the fact one is the inverse of the
other suggests the CPU is trying to do the right thing, so check
continuity between those pins and the crystal legs and for shorts to the
+5V rail or ground. If all is good, then it's definitely worth swapping
out the crystal. Also check the ceramic caps C210 and C222 aren't short
circuit.

Mittens

unread,

Oct 13, 2011, 6:47:04 AM10/13/11

to

On Thu, 13 Oct 2011 11:42:11 +0100, Mittens <mittens@_nospam_hush.ai>
wrote:

> Also check the ceramic caps C210 and C222 aren't short circuit.

I'm using this schematic as a reference: http://imgur.com/by5NQ

dropde...@gmail.com

unread,

Oct 13, 2011, 12:07:07 PM10/13/11

to

On Oct 13, 6:42 am, Mittens <mittens@_nospam_hush.ai> wrote:
> On Wed, 12 Oct 2011 15:53:31 +0100, unclefre...@jerk.org

I have continuity between each pin and one of the legs of the crystal,
I have continuity from outside leg of C210
and C222 to the grounding wire between boards so that looks good, from
power connector pins no shorts to ground
waiting for a crystal then, I looked at some junk arcade boards I have
to see if I could find a 6.000 Mhz crystal but no dice. I don't want
to fiddle with NTSC or 8+Mhz crystals to test this.

dropde...@gmail.com

unread,

Oct 13, 2011, 12:09:20 PM10/13/11

to

thanks

one other thing, I am not hearing the familiar vectrex buzz, would
that be
there if the oscillator wasn't working?

Mittens

unread,

Oct 13, 2011, 1:56:00 PM10/13/11

to

No. The buzz is there due to activity in the VIA and analog circuitry and
if the CPU stops writing to it it all just goes into a stable state.

Mittens

unread,

Oct 13, 2011, 1:59:12 PM10/13/11

to

> I have continuity between each pin and one of the legs of the crystal,
> I have continuity from outside leg of C210
> and C222 to the grounding wire between boards so that looks good, from
> power connector pins no shorts to ground
> waiting for a crystal then, I looked at some junk arcade boards I have
> to see if I could find a 6.000 Mhz crystal but no dice. I don't want
> to fiddle with NTSC or 8+Mhz crystals to test this.

8MHz probably wouldn't hurt anything. You might also find an NTSC crystal
oscillates at some sub-harmonic if the 6809 can't keep up.

dropde...@gmail.com

unread,

Doubtful though I think seeing as there is evidently some life in this one.

> Photos will take a bit if you would still like to see them.

Sure, it may help.

dropde...@gmail.com

unread,

Nov 1, 2011, 10:23:24 PM11/1/11

to

On Nov 1, 7:30 pm, Mittens <mittens@_nospam_hush.ai> wrote:
> On Tue, 01 Nov 2011 20:44:53 -0000, unclefre...@jerk.org

>
> <dropdeads...@gmail.com> wrote:
> >> I'm running out of ideas I'm afraid. I can't think why if the 6809 is a
> >> good one, you are seeing a signal on Q but not E.
>
> >> How about removing R216 temporarily and seeing what's on pin 34 then?
> >> Just wondering if something else might be driving that signal too.
>
> >> Have you got photos top and bottom of your board as it is at the moment?
>
> > I desoldered the side of r216 closer to pins 20-21 on 6809 and tested
> > from lifted side there pins 34+35 on cpu, I see what could be a 6mhz
> > wave (little fuzzy on my crappy scope) on both pins I could not say if
> > they are.
>
> Ok, that sounds promising. So now you have a signal on pin 34 (E) with
> that resistor removed? That suggests you have a short to ground or +5V on
> the other side of it.

ok I measured resistance from here to the 4 pin power connector and I
get readings from 550-120 ohms, no shorts, does this sound reasonable?

Mittens

unread,

Nov 3, 2011, 6:08:08 PM11/3/11

to

On Wed, 02 Nov 2011 02:23:24 -0000, uncle...@jerk.org
<dropde...@gmail.com> wrote:

>> Ok, that sounds promising. So now you have a signal on pin 34 (E) with
>> that resistor removed? That suggests you have a short to ground or +5V
>> on the other side of it.
>
> ok I measured resistance from here to the 4 pin power connector and I
> get readings from 550-120 ohms, no shorts, does this sound reasonable?

Not really, no. What are you measuring it with? DMM or a moving coil meter?

It there is a bad IC though, I guess this narrows it down to either IC202
(74LS00, quad NAND) or the 6522.

dropde...@gmail.com

unread,

Nov 4, 2011, 3:26:50 PM11/4/11

to

On Nov 3, 6:08 pm, Mittens <mittens@_nospam_hush.ai> wrote:
> On Wed, 02 Nov 2011 02:23:24 -0000, unclefre...@jerk.org

>
> <dropdeads...@gmail.com> wrote:
> >> Ok, that sounds promising. So now you have a signal on pin 34 (E) with
> >> that resistor removed? That suggests you have a short to ground or +5V
> >> on the other side of it.
>
> > ok I measured resistance from here to the 4 pin power connector and I
> > get readings from 550-120 ohms, no shorts, does this sound reasonable?
>
> Not really, no. What are you measuring it with? DMM or a moving coil meter?
>
> It there is a bad IC though, I guess this narrows it down to either IC202
> (74LS00, quad NAND) or the 6522.

DMM, that should be 1200ohms not 120 btw. HOWEVER I went back and
didn't use continuity mode and the readings are more like 18M for
right most pin and 120k-190k ohms for the rest, so I misspoke. does
THAT sound ok then?

ok I have read a dead 6522 will present itself with no picture but
sound still
works,which this has neither. Does it depend how it's shot maybe? I
dunno...

bradley...@gmail.com

unread,

Apr 1, 2017, 3:38:01 AM4/1/17

to

On Thursday, September 1, 2011 at 6:35:59 AM UTC+10, uncle...@jerk.org wrote:
> I am trying to repair a friend's vectrex that is dead as a doornail,
> except for a white dot, is there any source for 68A09 CPU (not bob
> roberts) or is there another replacement I can use it it's place
> (assume 1mhz 6809 is too slow, 2mhz 68B09, would it need a different
> clock crystal (what value?)
>
> anything else to watch out for?
>
> thanks

Hello There

I am very interested to know if you ever fixed this? I have the same problem and I am very keen to repair it. I have a good quality scope and decent electronics knowledge. I was reading through this conversation and you were all very helpful and knowledgeable, wondering if I could get some help if any of you are around.

Regards
Brad

Computer Nerd Kev

unread,

Apr 1, 2017, 8:48:21 PM4/1/17

to

There are a lot of potential causes, it means that the Logic board
circuitry isn't generating a picture signal - which is the key
purpose for most of it, so there's a lot that could go wrong. Unless
you already have reason to already suspect the 68A09, I would begin
by checking the power supply voltages on the logic board.

If the sound works, then it's likely a problem in the area of the
circuitry that generates the final picture signals for the CRT,
rather than the CPU and processing circuitry.

There are some notes here on the Vector Gaming Forums:
http://vectorgaming.proboards.com/thread/1469/vectrex-white-dot-problem-fixed

In case you haven't already got it:
http://computernerdkev.heliohost.org/vecadapt/Vectrex_service_manual.pdf

--
__ __
#_ < |\| |< _#

bradley...@gmail.com

unread,

Apr 2, 2017, 3:52:15 AM4/2/17

to

Thankyou for your quick response, looks like I have found the right place to find some help.

Thankyou for that Forum link, I have done a lot of internet searching and had not found that one, with the testing I have done I feel the problem had to be a chip between the cart connector and Sound/ I/O Chip. In this discussion two people with identical symptoms found the 6522A VIA to be at fault, So it looks like I might need to get one of these unless I can do further testing to see if its at fault?

Here is the testing I have done so far:

*CRT displays white dot and no sound when first turned on. However my fault has an interesting twist, when wriggling a cart in the cart slot and pressing reset many times I can get the white dot to move, with enough patience perhaps after 15 mins of doing this, It boots! shows startup screen and sound will work.

*However first time I did this the sound was slow and the buttons on the controller did not work, so via the schematic I was confident this was the sound chip, so I replaced it and when it booted again after tirelessly wriggling a cart and pressing reset the sound and controller problems were fixed.

This is not the only problems I have, when I can get it to boot, the vectored image is very muddled, Here is a video I posted to Youtube before I replaced the sound chip:

https://www.youtube.com/watch?v=iwjD1T6spz4

*I cant stress enough that it isn't easy to get it to boot, and I believe anyone with the white dot no sound problem may also be able to get it to boot by wriggling a cart and resetting but haven't had the patience :)

I however believe the screen issue is a separate fault in the analogue section IC's or could also be integrator caps?

I have done lots of other testing, power supply voltages was the first thing I checked and these were all good on both boards.

One interesting thing I would like to note is when it is showing white dot no sound the bus between the VIA and DAC is all 5V, This proves to me that nothing is driving the DAC so any of the chips between the Cart connector slot and Sound I/O Chip are probably at fault?

Let me know what you think.

Cecil Casey

unread,

Apr 2, 2017, 1:22:19 PM4/2/17

to

Hi Bradley,

I sounds like the bad old capacitor rot. Vectrex are known for having this problem.

The first clue is the 15 minute boot. The multiple reboots are minimally reforming the capacitor's in the power supply so that the ac ripple is below the threshold where the CPU can actually boot. If you use your o-scope on the power rails you will see that they are off voltage (likely) and have a lot of noise on the rails. Use the service manual above to find the test points.

A good second clue is that when the machine does boot up the sound is fine. Or at least plays. This could not happen if the CPU and all the main chips around it were not working. So you can skip looking for new chips.

Thirdly lets look at your video, specifically when the gunship is zooming into the play field. You can clearly see the vectors being drawn, but none of them are closing the loop to correctly draw the ship. The vectrex uses the R/C time constant of the capacitor charge to draw the correct slope of the line from any point to some other point. So if your caps are dried up and failing, or your voltages are not stable, your time constant has gotten all screwed up and nothing on the screen will ever be right. You could try the alignment process and use the test cart, but it will never be right with the caps shot.

Doing a quick search turns up a site that has a cap kit for sale. https://console5.com/store/vectrex-cap-kit.html
They have a link on this page for the integrator cap kit as well.
https://console5.com/store/vectrex-integrator-cap-kit.html

Good idea to do these while you have everything torn apart. You will need to do the alignment once everything has been replaced.

Their Wiki has some nice work and data for you on the replacement locations. I have no experence with this company so good luck and do further searches for more cap information.

-Cecil

Computer Nerd Kev

unread,

Apr 2, 2017, 6:32:23 PM4/2/17

to

bradley...@gmail.com wrote:
>
> Thankyou for your quick response, looks like I have found the right place
> to find some help.
>
> Thankyou for that Forum link, I have done a lot of internet searching and
> had not found that one, with the testing I have done I feel the problem
> had to be a chip between the cart connector and Sound/ I/O Chip. In this
> discussion two people with identical symptoms found the 6522A VIA to be
> at fault, So it looks like I might need to get one of these unless I
> can do further testing to see if its at fault?

As already suggested, you might want to check for power supply ripple
using your oscilloscope, if it's a few hundred millivolts, that could
be the problem. Perhaps the ripple settles down after a time as described
already, and resetting the system after that point allows it to work.

> Here is the testing I have done so far:
>
> *CRT displays white dot and no sound when first turned on. However my
> fault has an interesting twist, when wriggling a cart in the cart slot
> and pressing reset many times I can get the white dot to move, with
> enough patience perhaps after 15 mins of doing this, It boots! shows
> startup screen and sound will work.
>
> *However first time I did this the sound was slow and the buttons on
> the controller did not work, so via the schematic I was confident this
> was the sound chip, so I replaced it and when it booted again after
> tirelessly wriggling a cart and pressing reset the sound and controller
> problems were fixed.

The sound chip may have been damaged by excessive voltage from the power
supply, though this would suggest the voltage regulator ICs at fault
rather than the power supply capacitors. These can also cause ripple
when fail, although more in the form of intermittent voltage spikes.

> This is not the only problems I have, when I can get it to boot, the
> vectored image is very muddled, Here is a video I posted to Youtube
> before I replaced the sound chip:
>
> https://www.youtube.com/watch?v=iwjD1T6spz4
>
> *I cant stress enough that it isn't easy to get it to boot, and I
> believe anyone with the white dot no sound problem may also be able
> to get it to boot by wriggling a cart and resetting but haven't had
> the patience :)
>
>
> I however believe the screen issue is a separate fault in the analogue
> section IC's or could also be integrator caps?

I'll look at the video later, but from the description in the other
post, the capacitors there do seem most likely at fault.

> I have done lots of other testing, power supply voltages was the first
> thing I checked and these were all good on both boards.
>
> One interesting thing I would like to note is when it is showing white
> dot no sound the bus between the VIA and DAC is all 5V, This proves to
> me that nothing is driving the DAC so any of the chips between the Cart
> connector slot and Sound I/O Chip are probably at fault?
>
> Let me know what you think.

Have you had a close look at the capacitors to see whether any are
bulging, leaking, or emitting a bad smell? They can fail without doing
this though.

If you're confident of the power supply, then a logic chip is indeed
most likely still at fault. However the loss of more than one logic
chip suggests in itself a power supply issue.

Bradley Fechner

unread,

Apr 3, 2017, 8:25:03 AM4/3/17

to

On Thursday, September 1, 2011 at 6:35:59 AM UTC+10, uncle...@jerk.org wrote:
> I am trying to repair a friend's vectrex that is dead as a doornail,
> except for a white dot, is there any source for 68A09 CPU (not bob
> roberts) or is there another replacement I can use it it's place
> (assume 1mhz 6809 is too slow, 2mhz 68B09, would it need a different
> clock crystal (what value?)
>
> anything else to watch out for?
>
> thanks

Thankyou both for your reply's

I think the voltages are alright, there is little to no noise on the 5V lines, a bit more on the 13v line, a 100mv saw shaped ripple, but when I was able to boot the Vectrex, this ripple did not change.

There is definitely no visible problems with any of the electrolytic caps, this was the very first thing I looked for when I opened it up. I worked as an apprentice for while at an AV workshop, Bulging caps were the first things they looked at every time opening a tv :)

I don't want to rule out caps, as I know they quite possibly could be bad, but would like to be more certain if possible before I buy them, I would have to round them up as the kit would be to expensive to have posted from the US.

Any other ideas? I am tempted to buy a 6522A Via IC, seeing this was faulty in two similar cases.

Computer Nerd Kev

unread,

Apr 3, 2017, 5:34:28 PM4/3/17

to

One other possibility might be that there's a dry solder joint on the
board and your wriggling of the cartridge in the slot causes the board
to flex and the joint to make contact again. You'd want to inspect the
joints on the board visually and reflow any that look like they might
not be properly bonding with the component pins, feeding in some new
solder at the same time.

If the logic board components themselves are the only suspects left,
the only way to confirm a particular failure short of swapping chips
would be to analyse the signals on the board from when boots. So
you'd want to check that the CPU was toggling its address lines, the
ROM was sending data back, the I/O chips were receiving/sending I/O,
etc. You'd ideally want a logic analyser that could capture and
display all the signals at once. A possible poor-man's alternative
could be to rig the CPU clock up so that it can be manually advanced,
then step through the boot-up routine checking all the chip inputs
and outputs one by one using a logic probe, oscilloscope, or
multimeter.

You might find the chip swapping approach easier.

Cecil Casey

unread,

Apr 4, 2017, 10:45:56 AM4/4/17

to

Sorry, I guess I messed up, the caps on the Vectrex do not swell or bulge like cheep modern caps, they dry out and loose the means to store a charge. A tube of toilet paper has as good a capacitance value once that happens. ESR goes to heck, and filtering drops off. Like I said it is a common vectrex issue, check any old vectrex archive to find many issues just like yours.

Your machine can boot, it can draw, it can make music. It's not the chips. It is only the chips when you never get anything in one of those classes.

If you want to know for sure plug in a controller and start the game once it boots, even if you can't see it does it start? If it can then the PIA is fine, but then sounds and drawing would not work with out the PIA either.

Do both cap kits and the alignment before you mess with the IC chips.

-C

Bradley Fechner

unread,

Apr 5, 2017, 7:36:22 AM4/5/17

to

Hello Cecil

I understand your absolutely right, I know electrolytic caps are notorious for going out of tolerance with age. So I know they could still be at fault, however at the same time I am also not disregarding chips as they still can be at fault aswell, the sound chip was bad, but it was still able to create sound.

Something I would like to clear up with my fault, when I get it to boot, its only that time, as soon as I press reset its straight back to a white dot, no sound...dead and its just as hard to get it to start again. When I wriggle a card in the slot it makes the dot move in what I would call a holding pattern, as if it freezes, sometimes I can make out dashed lines in the squiggles as if it has attempted starting the vectrex welcome screen and froze.

Some Further things I have done:

*Replaced all electrolytic caps on the logic board with spares I have on hand

*I moved Integrator caps around, (If only one was bad, maybe the screen issue would change) Which it did, the screen is now almost back to the way it should be.

* Re-flowed new solder on ALL the points on the logic board, I thought I should do this while I had the logic board out... again, thanks Computer Nerd Kev

Hello Computer Nerd Kev
Thats a very interesting idea, to step the CPU and observe the I/O's, not as easy as simply swaping chips, but would be cheaper and a fun interesting test. I built a simple computer from an old book made from 60 TTL chips that has the ability to step the cpu like this. I might give it some thought

Brad

Computer Nerd Kev

unread,

Apr 6, 2017, 7:10:49 PM4/6/17

to

Bradley Fechner <bradley...@gmail.com> wrote:
>
> Hello Cecil
>
> I understand your absolutely right, I know electrolytic caps
> are notorious for going out of tolerance with age. So I know
> they could still be at fault, however at the same time I am
> also not disregarding chips as they still can be at fault
> aswell, the sound chip was bad, but it was still able to
> create sound.
>
> Something I would like to clear up with my fault, when I get
> it to boot, its only that time, as soon as I press reset
> its straight back to a white dot, no sound...dead and its
> just as hard to get it to start again. When I wriggle a card
> in the slot it makes the dot move in what I would call a holding
> pattern, as if it freezes, sometimes I can make out dashed lines
> in the squiggles as if it has attempted starting the vectrex
> welcome screen and froze.

Seems more likely a fault with the ROM or CPU causing the CPU to
lock up, as the I/O is apparantly doing its job in that situation.

Maybe wriggling the cartridge and thereby varying the load on the
data and address lines is causing either a ROM (Data) or CPU
(address) output to start to come back to life. Or it might be
the load on the 5V power line being connected and disconnected
with the ROM IC in the cartridge. Just guesses.

> Some Further things I have done:
>
> *Replaced all electrolytic caps on the logic board with spares
> I have on hand
>
> *I moved Integrator caps around, (If only one was bad, maybe
> the screen issue would change) Which it did, the screen is
> now almost back to the way it should be.

That's good news.

> * Re-flowed new solder on ALL the points on the logic board,
> I thought I should do this while I had the logic board out...
> again, thanks Computer Nerd Kev
>
>
> Hello Computer Nerd Kev
> Thats a very interesting idea, to step the CPU and observe the
> I/O's, not as easy as simply swaping chips, but would be cheaper
> and a fun interesting test. I built a simple computer from an
> old book made from 60 TTL chips that has the ability to step
> the cpu like this. I might give it some thought

Impressive. For this you'd only need a simple debouncing circuit
on a push button. Ahh, I just checked the 6809 datasheet and it
states a minimum clock frequency of 400KHz. Perhaps still worth
trying, some other CPU designs are specified to operate at such
low frequencies (then again, you might not know what to blame
if there's a visible problem). Pin 38 (EXTAL) is connected to
the external clock signal, while pin 39 (XTAL) is connected
to ground according to the datasheet.