Superbrain #1 : The ratty one

Talk about non-Acorn classic computers/hardware/software here (including retro consoles)
User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Superbrain #1 : The ratty one

Postby jonb » Mon Jul 03, 2017 2:32 pm

Hi

The subject of this thread is one of the three Superbrains I collected from The Centre for Computing History in Cambridge. I call it "The ratty one" because it has lots of grime on the keyboard and case, and the case itself is in need of a respray.

It's a standard Superbrain with 64K RAM and twin DS/DD floppies so in that regard it is QD spec like mine. It has an ominous label attached to its power lead: "TRIPS MAINS", which turned out to be caused by a dodgy power switch with a metal body that was shorting to earth intermittently. I pulled it and fitted a connector block temporarily to enable testing to continue.

On power up, it attempted to boot (drive A: was accessed), then died. I suspected the power supply had failed (it was clicking) and pulled the motherboard to try on mine. Hmm, power supply on mine is clicking now! Must be a fault on the board.. and so I checked for shorts and sure enough, +12v and GND are shorted. This will be tricky to isolate.

I can't see any bits of metal lying around. Maybe be a failed IC. As far as I can see, the only ones needing 12v are the 4116 DRAM and serial line level transducers.

So my first question is this: Is this (12v short to GND) a common failure mode of the DRAM?

Welcome to the ride.... :twisted:

JonB

User avatar
hoglet
Posts: 6352
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Superbrain #1 : The ratty one

Postby hoglet » Mon Jul 03, 2017 3:34 pm

jonb wrote:So my first question is this: Is this (12v short to GND) a common failure mode of the DRAM?

I think it's much more likely to be a tantalum capacitor (if there are any) than the DRAM.

Can you post a decent picture of the board?

User avatar
sirmorris
Posts: 701
Joined: Wed Feb 11, 2009 12:18 pm
Location: oxfordshire uk
Contact:

Re: Superbrain #1 : The ratty one

Postby sirmorris » Mon Jul 03, 2017 4:17 pm

Welcome to the ride.... :twisted:


Can't wait! I love following a good bug hunt. Not so keen when I have to perform one tho ;)

bear
Posts: 19
Joined: Sat Jul 16, 2016 10:04 pm
Location: Seattle, USA
Contact:

Re: Superbrain #1 : The ratty one

Postby bear » Mon Jul 03, 2017 5:19 pm

hoglet wrote:I think it's much more likely to be a tantalum capacitor (if there are any) than the DRAM.


+1 for that

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Mon Jul 03, 2017 5:39 pm

This is my motherboard's RAM.

IMG_0862.JPG


The blue beads.. I checked across a few of their legs and they are closed circuit.

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Mon Jul 03, 2017 5:45 pm

Let's take a closer look...

IMG_0924.JPG
Tantamount to tantalum?


Says 104 Z5E on one side, R 948 on the other.

104 = 100nF
Z5E = Min temp 10 degrees, max temp 85 degrees, tolerance +-4.7%

Can't see any polarisation marks and they appear to be oriented randomly.

dp11
Posts: 670
Joined: Sun Aug 12, 2012 8:47 pm

Re: Superbrain #1 : The ratty one

Postby dp11 » Mon Jul 03, 2017 6:03 pm

All of them would be in parallel so measuring one is the same as measuring them all. If you have a low voltage but high current psu e.g. 1 to 2v . You could connect this to the 12 rail and find which cap gets hot.

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Mon Jul 03, 2017 6:15 pm

Then I noticed these black ones that look like diodes. Polarised, 20v, 10uF and the one with its leg in the air is closed circuit.

IMG_0925.JPG
Bah, tantalum.

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Mon Jul 03, 2017 6:18 pm

So let's give it a whirl now, having checked the 12v rail is not shorted anymore.

IMG_0926.JPG
Master, it lives!!!


Looks like a stuck P happening there. No surprise given the filth under the keyboard (you should see my desk after turning the board over, YUCK!).

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Mon Jul 03, 2017 6:48 pm

Well, I have pulled the P key and found it is not actually closing properly. Pulled it apart and cleaned it, all good. However the P key still repeats even with the key missing! We are left with some possibilities:

  • A short between KR3600 keyboard encoder pins 24 and 36 (this is the combination that indicates the P is pressed)
  • Odd KR3600 failure mode.

There's no short, so I am thinking the KR3600 is playing up.

KR3600 circuit.JPG
Keyboard circuit.

User avatar
hoglet
Posts: 6352
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Superbrain #1 : The ratty one

Postby hoglet » Mon Jul 03, 2017 7:06 pm

jonb wrote:Well, I have pulled the P key and found it is not actually closing properly. Pulled it apart and cleaned it, all good. However the P key still repeats even with the key missing! We are left with some possibilities:

  • A short between KR3600 keyboard encoder pins 24 and 36 (this is the combination that indicates the P is pressed)
  • Odd KR3600 failure mode.

There's no short, so I am thinking the KR3600 is playing up.

What happens if you actually press some of the other keys?

It might be worth a probe around Z81 (74HC74) just to make sure that seems to be working correctly.

Dave

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Mon Jul 03, 2017 8:52 pm

Hi Dave

Pressing other keys has no effect.
Occasionally there will be other random keys repeating, not just P.
Will check the '74 tomorrow.

Cheers
JonB

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Tue Jul 04, 2017 8:26 am

'74 seems OK.

I took the keyboard encoder out and put in a spare - an AY-5-3600-PRO, not the right keymap but will tell me if the encoder is the problem. It is... so now I will build an encoder board http://www.vcfed.org/forum/showthread.p ... ter-boards and try again.

Getting close!

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Tue Jul 04, 2017 9:59 am

OK, checkpoint.

Fitted a 40 way IC socket, built a KR3600 adapter board and whacked it in. The adapter has scavenged parts from my own Superbrain - an AY-5-3600-PRO and EEPROM containing the right keyboard map for the Superbrain.

IMG_0928.JPG
KR3600 adapter board in situ


Now it is booting and there are no spurious characters.

IMG_0929.JPG
Boot and CP/M Prompt


However (there's always a "but..."), I can see that quite a few of the keys are not working. Testing them all gives a fail list:

Main keyboard
  • Row 1: ESC 1 3 5 8 9 - = BS BREAK
  • Row 2: Y I [
  • Row 3: CRTL J K ;
  • Row 4: L-RESET X . / RIGHT-SHIFT R-RESET
  • Row 5: SPACE-BAR
Numeric pad
  • .

I've checked the failed keys for a pattern that corresponds with the way the keys are connected to the encoder that might indicate a broken track, but I don't see any. Because of the way these keys work, hitting them repeatedly doesn't tend to clean the contacts, so I will have to remove each one, dismantle and scrape the contacts clean, then reassemble / refit. As I have not yet been under all the key caps I expect there will be lots of the usual keyboard crud to remove, and of course the key caps themselves will need a clean as they are exceptionally filthy.

This will take a while... :lol:

User avatar
1024MAK
Posts: 6668
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...

Re: Superbrain #1 : The ratty one

Postby 1024MAK » Tue Jul 04, 2017 11:07 am

jonb wrote:Let's take a closer look...

Image

Says 104 Z5E on one side, R 948 on the other.

104 = 100nF
Z5E = Min temp 10 degrees, max temp 85 degrees, tolerance +-4.7%

Can't see any polarisation marks and they appear to be oriented randomly.

From the picture, these look like ceramic type capacitors. Hence no polarisation.

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

User avatar
1024MAK
Posts: 6668
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...

Re: Superbrain #1 : The ratty one

Postby 1024MAK » Tue Jul 04, 2017 11:12 am

jonb wrote:Then I noticed these black ones that look like diodes. Polarised, 20v, 10uF and the one with its leg in the air is closed circuit.

Image

I'll admitt, I'm not 100% sure what this type is. They are however NOT ceramic. But I have come across them before. Can't remember in what though :roll:

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

User avatar
1024MAK
Posts: 6668
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...

Re: Superbrain #1 : The ratty one

Postby 1024MAK » Tue Jul 04, 2017 11:22 am

Oh, BTW, 4116 DRAM can develop internal faults which either result in the chip drawing a much higher current from the +12V supply, or leak the +12V out on one of the TTL level pins :twisted:

In the ZX Spectrum and ZX80/ZX81/TS1000 16k RAM pack, they take out the DC/DC converter/inverter circuit due to overload rather than becoming a simple short circuit.

In the Memotech MTX machines (where they form the video RAM), it appears they just get a lot hotter than normal.

Keep going Jon, we know you love getting covered in crud as you dive in deep, up to your neck in it :lol:

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Tue Jul 04, 2017 2:09 pm

Ok Mark

IMG_0931.JPG
Eurghhh....

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Tue Jul 04, 2017 2:10 pm

Let's get closer..

IMG_0931.JPG
Shudder...

:-&

User avatar
vanpeebles
Posts: 369
Joined: Wed Nov 28, 2012 10:01 am
Location: UK
Contact:

Re: Superbrain #1 : The ratty one

Postby vanpeebles » Tue Jul 04, 2017 2:13 pm

Ah, skankamank!

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Tue Jul 04, 2017 2:58 pm

But with a stiff brush, vacuum cleaner and a bit of meths, we can improve matters.

IMG_0934.JPG
Much better


The keys were washed (in a mixture of hot water, washing up liquid and sterilising powder) but it has only made a small improvement. They're badly yellowed and the grime on them is very tenacious. At a guess I'd say the key cap surfaces are suffering from UV damage and have microscopic crazing, which is trapping the grime.

IMG_0935.JPG
Retr0brite, anyone?

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Tue Jul 04, 2017 3:42 pm

INTERLUDE: Repairing CMC key switches.

This brief guide applies to repairs to BBC Master key switches as well (they are the same as the Superbrain, SMT type 4).

IMG_0938.JPG
Mark the solder side of the board with the key cap values. Makes it easier to find a specific key for desoldering.

IMG_0940.JPG
Using a Duratool Z915 desoldering station to heat and suck solder off the switch pins.

IMG_0942.JPG
Use an iron and screwdriver to heat and move the pins away from the sides of the via. This'll help them come out. Check there is space around the pins. Preferably they should be loose in the via.

IMG_0943.JPG
Flip the board over and use an IC extractor to pinch the clips holding the key in place, while levering the body from the cutout in the metal frame. Be gentle. If it is binding, maybe the pins are still soldered slightly.

IMG_0944.JPG
It's out.

IMG_0945.JPG
Should have done this before it was out! Check continuity with a meter. This one was not working at all, as expected.

IMG_0946.JPG
Use a craft knife to prise the clips open, and gently detach the base of the switch.

IMG_0947.JPG
When apart it looks like this.

IMG_0948.JPG
You can see where the contacts are, at the tip if these lugs. They appear to be gold coated, and ridged where they make contact. Use the craft knife to scrape at the contact ridges on both lugs, then reassemble and test.
Last edited by jonb on Wed Jul 05, 2017 3:25 pm, edited 3 times in total.

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Tue Jul 04, 2017 3:55 pm

IMG_0949.JPG
When refitting the switch, check the orientation and carefully line the pins up with the vias. Do not press the switch body fully home until you can see the pins through the vias. Then solder up the pins and flip the board over.
(18.92 KiB) Not downloaded yet


IMG_0950.JPG
As we have replaced the I key, I can do a DIR on the disk contents now!


Only another 24 switches to go. :(

That said, I can do it in batch now because the key values are all marked on the underside of the board.

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Tue Jul 04, 2017 4:37 pm

Top Tip : Treat the space bar as a special key. It often has a stronger spring in order to lift the space bar mechanism, which is heavier than a standard key top, so make sure you do not swap its switch for one of the others.

User avatar
vanpeebles
Posts: 369
Joined: Wed Nov 28, 2012 10:01 am
Location: UK
Contact:

Re: Superbrain #1 : The ratty one

Postby vanpeebles » Tue Jul 04, 2017 9:00 pm

I wonder if meths would help with my soldering, it might take the edge off. :lol:

User avatar
daveejhitchins
Posts: 3617
Joined: Wed Jun 13, 2012 5:23 pm
Location: Newton Aycliffe, County Durham

Re: Superbrain #1 : The ratty one

Postby daveejhitchins » Wed Jul 05, 2017 9:21 am

A way I've found, for removing key switches, is: after desoldering use a small coin to apply pressure to the pins and then just release the clips one at a time on the top side - they just pop out!

Dave H :D
Parts: UM6502CE, GAL22V10D, GAL16V8D, AS6C62256A, TC514400AZ, WD1772, R6522, TMS27C512, AT28C256
Products: ARA II, ABR, ATI, AP6, MGC, AP5 . . .
For a price list, contact me at: Retro Hardware AT dave ej hitchins DOT plus DOT com

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Wed Jul 05, 2017 3:20 pm

The problem is that one of the pins is thin and could be damaged by pushing with a coin.

On the Superbrain, it was often folded over prior to soldering so you are looking at potentially snapping it off, once you'd unfolded it.

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Thu Jul 06, 2017 9:25 am

After refitting all the (washed but not Retr0brited) key caps and spending a fiddly hour or so trying to get the space bar to not stick, we find ourselves at this juncture:

IMG_0952.JPG
The keys look white-ish here, but they are actually yellow.


The next steps will be to:
  • Test the serial ports, run a memory test and other main board testing.
  • Replace the (assumed)tantalum caps near the memory that are smoothing the 12v lines. I'm going to use radial electrolytics here.
  • Check the power supply (remember, this board is fitted to my own Superbrain; its case, drives and PSU are awaiting attention).
  • Test and service the floppy drives.
  • Test the CRT and repair (if necessary).
  • Find and fit a new mains switch, wire it properly (at the moment, the switch operates on the 115v line, after the step-down transformer, which is powered all the time when the thing is plugged in).
  • Replace the power cord as its cores have illegal colouring (white, blue, green instead of brown, blue, green/yellow)
  • Possibly Retr0Brite the key caps and respray the cabinet (at the discretion of CCH)

User avatar
jonb
Posts: 2041
Joined: Sat May 21, 2011 12:42 pm
Location: South Coast of England

Re: Superbrain #1 : The ratty one

Postby jonb » Sat Jul 08, 2017 11:31 am

Update time again, folks!

IMG_0953.JPG
Twin Tandon TM100-2 drives removed and partially disassembled. To service, I cleaned out the spiders webs and other debris, cleaned the read/write heads, lubricated the main bearing and head carriage and gave the front panels a bit of a clean.

IMG_0954.JPG
Unfortunately, one of the sensor pins broke loose...

IMG_0955.JPG
...so it's out with the finer of my soldering iron tips.

IMG_0956.JPG
Never say "never". The lead is easily bodged back into place, with patience and a needle thin tip.

InkedIMG_0959_LI.jpg
Both drives have the motor spindown mod applied. Look closely at the 8th tab on the edge connector. This is the MOTOR_ON signal which the Superbrain hardwires to "ON". There is a bit of Sellotape covering it, because we need to isolate the drive. The other part of this mod can be seen on the option block - a link between the right hand most sides, made of a small bent staple. The effect of this is to start the motor when the drive is selected. There's another mod required - a resistor swap on the main board - that gives a slightly longer delay to allow the motor to spin up before attempting to access the data.

IMG_0960.JPG
This picture shows the Superbrain running with its own PSU, CRT and drives. Everything is good, apart from the fan (it's failed) and the MOTOR_OFF for Drive B which looks to be a ROM bug. Seems like it is not deselecting the drive after access, but when you switch to A:, B: goes off immediately, so the mod itself is OK.

IMG_0961.JPG
The PSU and CRT boards haven't been touched as they are working, but you can see the connector block I fitted after removing the broken power switch.

IMG_0962.JPG
Speaking of which, the replacement switch is on the right. It's exactly the same size and spec as the original. I need to fit it.

User avatar
1024MAK
Posts: 6668
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...

Re: Superbrain #1 : The ratty one

Postby 1024MAK » Sat Jul 08, 2017 11:58 am

Ahh, tag-strip terminals. I had to show that I could solder wires to a tag-strip in order to make through as part of the selection of potential candidates for a job in 1987. Of the 34 that applied, I was one of three to get a job :D

Good work there Jon =D> =D> =D>

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...


Return to “other vintage computer hardware, software and games”

Who is online

Users browsing this forum: No registered users and 4 guests