flex cable

[+dhe]

This weekend while moving cards about in a pbox, I lost a flex cable.

   Yes - I always power down and wait 10 minutes (minimum) before inserting or extracting a card.

   Yes - I try to make sure the card is perfectly aligned up and down, before inserting them.

   Yes - I touch the metal of the pbox to make sure I'm fully un-staticed.

 

I found out by going into e/a with just the flex and ti-32k - no memory expansion error.

 

I moved that to a different pbox and the error followed.

A different flex cable with the same 32K unit worked.

 

Bud Mills and Chris both mentioned the ls245 as having a particularly hard life.

 

Are the schematics for the flex cable out there?

 

I did look at TI's:

Peripheral Expansion System - Theory of Operation and Technical Training Manual

-and-

TI--99/4 Home Computer Troubleshooting Guide

and found nothing helpful.

[+OLD CS1]

16 minutes ago, dhe said:

Are the schematics for the flex cable out there?

Yes, I have referred to them many times.  Look at sheets six and seven, pages 64 and 65, of the "TI-99/4A Console and Peripheral Expansion System Technical Data" manual.

[+dhe]

Excellent - Thanks.

[+Ksarul]

The schematics are also on Mainbyte's site.

[+dhe]

I checked my toy box. I have a handful of 74LS245 and 244's.

I didn't have any 20 pin sockets.

I searched ebay for 20 pin Machine Screw Sockets.

Only came up with two vendors, one in China and one in Poland.

I have 10 Polish sockets on the way.

[+OLD CS1]

3 minutes ago, dhe said:

I have 10 Polish sockets on the way.

There is a joke in this somewhere. I can feel it in my bones.

[+Ksarul]

53 minutes ago, dhe said:

I checked my toy box. I have a handful of 74LS245 and 244's.

I didn't have any 20 pin sockets.

I searched ebay for 20 pin Machine Screw Sockets.

Only came up with two vendors, one in China and one in Poland.

I have 10 Polish sockets on the way.

You forgot the two most likely local sources for parts: Unicorn Electronics and me. I always have lots of sockets on hand for the various projects I'm working on. I actually have 500 of the 20-pin sockets on their way from a reliable Chinese source at the moment (they are somewhere in the outbound Chinese customs process as of yesterday). Bonus with this source is that all of the metal parts are gold plated, so I never have to worry about ROHS/Legacy soldering issues with them. Their prices are pretty good too, although they can be a bit slow to ship. I always get my items though. . .so I can live with a bit of slow, and I program the standard delays with them into my order frequency. I have probably ordered 10,000 sockets from them over the last ten years.

[+helocast]

44 minutes ago, Ksarul said:

You forgot the two most likely local sources for parts: Unicorn Electronics and me. I always have lots of sockets on hand for the various projects I'm working on. I actually have 500 of the 20-pin sockets on their way from a reliable Chinese source at the moment (they are somewhere in the outbound Chinese customs process as of yesterday). Bonus with this source is that all of the metal parts are gold plated, so I never have to worry about ROHS/Legacy soldering issues with them. Their prices are pretty good too, although they can be a bit slow to ship. I always get my items though. . .so I can live with a bit of slow, and I program the standard delays with them into my order frequency. I have probably ordered 10,000 sockets from them over the last ten years.

Yeah, I keep stuff like this on subscription order better than my Keurig pods ...

[+Ksarul]

3 hours ago, helocast said:

Yeah, I keep stuff like this on subscription order better than my Keurig pods ...

I've bought from them before too. As I noted, the source I use gets sockets with all metal parts gold plated, not just the socket inserts. Of course, they also cost about double what the ones you ordered cost. I also buy the sockets like yours from time-to-time, especially when the price is right (like when I found 3,000 of the 16-pin sockets for a price that could not be passed up earlier this year--I think it was about 8 cents a socket).

[+dhe]

I didn't realize what a pain it would be to work with a board, with 2 pounds of cable permanently attached. So I very loosely tied the cable up.

 

 

I used to photo with a Nikon SLR, but have gotten lazy and use my phone. Since many of the photo's I try to take our up close, I said - self, wouldn't it be nice to have a macro lens again? Sure enough they make them for phones!

 

This photo is using the iphones aelock - it's ok, but not good.

 

 

This one is with the macro lens.

 

 

Now, here is my question for you guru's. I've seen plenty of photo's of bulging/leaking radial caps. But I've never seen photo's of bulging axial caps. Do these caps look like they are bulging and need replaced?

 

They are 22ufx25V - C8/C9 and roughly 6.38mm x 12.20mm.

 

Does anyone know of a good source for these?

 

 

 

[GDMike]

When in doubt - perhaps,

change out those caps.

But they look just like mine,

And so far they are just fine.

So if I guess, there normal.. but I'm just a poet and didn't know it. I'll just run em..

[Shift838]

14 hours ago, dhe said:

I didn't realize what a pain it would be to work with a board, with 2 pounds of cable permanently attached. So I very loosely tied the cable up.

 

 

I used to photo with a Nikon SLR, but have gotten lazy and use my phone. Since many of the photo's I try to take our up close, I said - self, wouldn't it be nice to have a macro lens again? Sure enough they make them for phones!

 

This photo is using the iphones aelock - it's ok, but not good.

 

 

This one is with the macro lens.

 

 

 

Now, here is my question for you guru's. I've seen plenty of photo's of bulging/leaking radial caps. But I've never seen photo's of bulging axial caps. Do these caps look like they are bulging and need replaced?

 

They are 22ufx25V - C8/C9 and roughly 6.38mm x 12.20mm.

 

Does anyone know of a good source for these?

 

 

 

 

I bought mine from mouser.  I stepped up to 35v caps.

 

 

[+dhe]

@Shift838 said checking the pining is time consuming, I thought he meant as in check pin 1 on the paw to pin 1 on the solder side of the double row of connectors on the back of the interface board. TI Fooled me, there aren't associated pins sticking out the back on header on the solder side.

 

Instead I decided to check resistors on the interface board, below you will find my test, and I believe I'm in the clear there, also a few notes on the capacitors.

 

As mentioned before, Bud said - "If you swap cards very often, the LS245 in the elephant paw will eventually fail".

A google of "Do ls245 fail often" - brings up the answer yes.

 

I think I will try replacing the LS245 on the interface side... If that fails I will try replacing the LS245 in the paw.

 

Tons of oxidation removed from fingers.

L245 - Octal Bus Transceivers (tri-state).
L244 - Octal Buffer with line driver.

R1=46.3 o
R2=46.6 o
R3=46.7 o
R4=46.6 o
R5=46.1 o
R6=46.7 o
R7=46.0 o
R8=46.0 0

R9=.465   ko
R10=.464  ko
R11=.463  ko
R12=10.07 ko  <- R12 is not marked on the silk screen

R13=.466 ko
R14=.463 ko
R15=.468 ko

R16=.990  ko
R17=46     o
R18=99.3   o
R19=10.07 ko

Gold Values + or - 5% so on 47 ohms would be 45-47-49
.464 ko = 464 o ~= 470 ohms.
 

Schematic States
  R1-R8 = 47 ohms
  R9-R15 = 470 ohms

 

Caps C8, C9
  Axial 22uf x 25V
  6.38mm thick and 12.20mm long.
 

 

[+Lee Stewart]

3 hours ago, dhe said:

R1=46.3 o
R2=46.6 o

R1=46.3 Ω
R2=46.6 Ω

Dan, you can use Alt+234(on the number keypad) to get ‘Ω’ for “ohm”.  Also,

3 hours ago, dhe said:

Caps C8, C9
  Axial 22uf x 25V

Caps C8, C9
  Axial 22µf x 25V

you can use Alt+230 to get ‘µ’ for “micro”.

 

...lee

[Shift838]

7 hours ago, dhe said:

@Shift838 said checking the pining is time consuming, I thought he meant as in check pin 1 on the paw to pin 1 on the solder side of the double row of connectors on the back of the interface board. TI Fooled me, there aren't associated pins sticking out the back on header on the solder side.

 

Instead I decided to check resistors on the interface board, below you will find my test, and I believe I'm in the clear there, also a few notes on the capacitors.

 

 

The correct resistor values should be:

 

Flex Interface PEB Card:

 

R1 - R15, R17 = 47 Ω

R16 = 1K Ω

R18 = 100 Ω

R19 = 10K Ω

 

Flex Interface Elephant foot:

 

All resistors are 47Ω

 

I think if you are reading the resistor numbers correctly your R12 is not correct.  It should be 47Ω on the PEB card.  If R12 is indeed 10K, then swap it out for a 47Ω.

 

Edited December 30, 2021 by Shift838

[+dhe]

Ok.. Another day of errors....

 

I decided to go with Bud Mill's and try replacing the 245 in the paw first.

 

And here is what I found:

 

 

Can you guess where the 40 year old double sided sticky tape is covering?

Yea... The 245...

 

If you followed my errors and tribulations on the horizon, you know I have experience removing double sided sticky tape.

 

Praise @arcadeshopper his advice to buy a Hakko has NEVER been regretted.

 

Desoldered, board cleaned up, and lets drop in my imported from Poland socket!

 

 

Apparently there is another standard for width between pin row that is 10.1 vs 7.62!!

 

When I lived in Indianapolis and did a bit of electronics, I was really spoiled rotten - about 30 Radio Shack with ok to great electronics knowledgeable employees, and THREE electronics supply houses - I could go in to any of them, show them a picture or bring in a part and they would grab what I need from the back (even if I didn't know what I needed) and bring it to the counter....

 

This time, I double checked the dimensions and put in an order with Unicorn Electronics, I prefer them, and usually your going to get parts compatible with 70/80/90's electronics, but the $30 minimum can be a killer, when you need two .20 capacitors...

 

O' and all the resistors checked out on the paw.

 

 

[Stuart]

That would be the point where I cut the plastic between the two rows and install as two SIL connectors rather than one DIL connector. 

[+OLD CS1]

On 12/30/2021 at 3:28 PM, Shift838 said:

The correct resistor values should be:

 

Flex Interface PEB Card:

 

R1 - R15, R17 = 47 Ω

R16 = 1K Ω

R18 = 100 Ω

R19 = 10K Ω

 

Flex Interface Elephant foot:

 

All resistors are 47Ω

 

I think if you are reading the resistor numbers correctly your R12 is not correct.  It should be 47Ω on the PEB card.  If R12 is indeed 10K, then swap it out for a 47Ω.

 

I want to take a moment to recognize the effort put into this post using real "ohm" symbols.

[+mizapf]

105-key PC keyboard with Linux: Ω = Shift-AltGr-q

(since those key combinations are typically not labeled, it's a bit like Easter egg searching)

[+dhe]

Here it is June. I finally got in the correct sockets and TI branded chips to continue work.

 

I replaced the chip in the paw. Tested.

Result: No change - 32K not seen.

 

I replaced all four chips on the interface card side. Tested.

Result: No change.

 

I'm not very good at electronics, but I try to go slow and methodical.

    First I tested all chips in the TL866.

    I checked as best I could each pin to it's next destination.

    I also checked to make sure none of the pins was shorted, do to a solder bridge.

 

I was frankly, pretty amazed, after all the messing around I did, it didn't either work, or broke worse?!

 

Fortunately, a kind soul who's much better at this sort of thing is going to take a look at my work.

[+Ksarul]

On 12/31/2021 at 4:25 PM, dhe said:

Desoldered, board cleaned up, and lets drop in my imported from Poland socket!

 

 

Apparently there is another standard for width between pin row that is 10.1 vs 7.62!!

There are .300 sockets--the most common type for discrete logic chips, .400 sockets--much less common but the type you have here, .600 sockets--mostly used for RAM, ROM, and larger logic chips, and random sockets that are even wider for really special chips like the TMS9900. Just one chip that is in TI-99 hardware uses the .400 sockets: the Hex-Bus controller chip.

[+dhe]

Name used with permission. The kind soul who took pity upon me was @Shift838.

 

He plugged it in to his first system and it worked.

He tried his second system and it also ... worked.

 

So all of my board butchering didn't total destroy it.

 

@Shift838 - took his meter and went to each point on the cable - there he ended up with some really high resistance. He said some of the pins looked rusted.

 

He reflowed allow the solder joints again and retested resistance - all of them - then showed as straight through.

In addition, he saw a number of capacitors who's pins stuck out farther then he was comfortable with and he trimmed those back.

He felt under the right condition they may cause a short.

 

He then tested the unit with copies from TIPI <> IDE. All worked.

 

I mentioned this on the TI Zoom meeting, that I'd never heard of equipment self destructing. @arcadeshopper said it's fairly common. One of the first things recommended on many old stand up arcade machines is to re-flow all the solder joints and then start troubleshooting.

 

When I received the card I put it in a PBOX and used Corcomp's PDM to test the 32K several times.

 

Thank you to everyone that helped me bring this old card back from the dead!