ECONET WOES

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

ECONET WOES

Postby torrind » Thu Apr 12, 2018 9:10 pm

Hi All,

I'm after some help please! Do bear with me if I begin to ramble :)

I recently upgraded both of my model B's to ECONET and as a result both now display station numbers and 'NO CLOCK' error. I have now built a clock based on the Simon (waiting for Friday) version 2 clock project with a board Kindly supplied to me by Liam.

For BEEB #1 when I activate the clock and press N+BREAK I get the station number only - Success! as it recognises the clock attached!
However, for BEEB #2, its still telling me I have no clock attached! I have checked my soldering and re-touched all points - I've checked the sockets with a continuity tester and unplugged, checked and reseated the IC's but to no avail.

I've read on various websites that other issues on the board could affect this type of issue?? (which is interesting, as I've long suspected issues with the board even though it appears to work fine!

Does anyone have any initial thoughts as to what I can test? I don't have a scope and my technical ability does fluctuate somewhat!

Are there any general test programs I can use to interrogate the board / RAM / ROMS / etc etc?

Kind Regards,

Darren

User avatar
lcww1
Posts: 120
Joined: Wed Mar 15, 2017 11:16 pm
Location: Out There
Contact:

Re: ECONET WOES

Postby lcww1 » Thu Apr 12, 2018 10:04 pm

torrind wrote:Are there any general test programs I can use to interrogate the board / RAM / ROMS / etc etc?


I can't suggest anything specific, but here are a few bits and pieces from this great forum that might help you make a start on getting to the bottom of the problem:

tricky's TestROM - http://www.stardot.org.uk/forums/viewtopic.php?f=3&t=13973&hilit=bbc+test+rom+tricky

the *. fault finding index - http://www.stardot.org.uk/forums/viewtopic.php?f=3&t=2190

MartinB's Memory Test Tool - http://www.stardot.org.uk/forums/viewtopic.php?f=3&t=14809&hilit=Martinb+ram+test

and the BBC Microcomputer Service Manual - http://www.8bs.com/othrdnld/manuals/essentials/BBCBServiceManual.zip

I hope these help......

johnkenyon
Posts: 139
Joined: Wed Jul 20, 2011 2:21 pm
Location: Coventry
Contact:

Re: ECONET WOES

Postby johnkenyon » Thu Apr 12, 2018 11:12 pm

Refer to the BBC B circuit diagram at http://mdfs.net/Info/Comp/BBC/Circuits/BBC/bbc.gif top left hand corner.

Recheck the Econet installation see: https://www.beebmaster.co.uk/BeebShop/D ... 20v2-5.pdf

Double check all resistor values - R38, R39, R44, R45, R47, R48, R50, R51, R52
Check that C17 is the correct value (Issue 1,2 &3 boards)
Check that IC94 (LM319), IC87 (74LS123) and IC89 (68B54) are all seated correctly.
Ensure that link S2 is broken.

If socketed, swap IC95 and IC94 (both LM319), and see if the fault clears or is replaced by Line Jammed.
If your problem clears then swap them back, to verify that your original IC94 is duff.
If your no clock is replaced by Line Jammed, then the original IC94 is duff.

If that doesn't work then use a 'scope with the circuit diagram to verify that the clock signal is reaching the 68B54 (pins 4&5), and pin 1 of IC87.
Then check the state of the 68B54 DCD input (pin 27), should be low when the clock is connected, and high when disconnected.
Finally check IC91 - however I reckon that a fault there would give a Line Jammed error.

If you don't have a scope, try swapping components with the working Beeb - IC94, IC87, and IC89.

/John

torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

Re: ECONET WOES

Postby torrind » Fri Apr 13, 2018 10:41 pm

Hi Liam / John,

Firstly, thank you both for your collective responses!

OK, so I have made some progress :D

John - After following your very detailed instructions, I found out that IC87 was the culprit. The problem was further compounded by the fact that it was soldered directly to the board and I've spent all evening removing it, adding a socket and then plugging in a spare IC - success! BEEB #2 now see's the clock - whooo hoo!

However, its not all good news!

Once I've connected up the BEEBS to the clock and entered the *NET command on both, I get the following errors:

On BEEB #1 if I issue *. I get "Line Jammed" (I've swapped over IC91 and IC94 with BEEB #2 with no success)
On BEEB #2 If I issue *. I get "Net Error"

As always, thanks for the support and advice! I'm feeling positive on the progress so far...

Darren

User avatar
IanS
Posts: 545
Joined: Mon Aug 31, 2009 6:02 pm
Contact:

Re: ECONET WOES

Postby IanS » Fri Apr 13, 2018 11:24 pm

Is the setup two beebs, two cables and a clock? Do you have any terminator or line-biasing components - viewtopic.php?f=3&t=12580&p=161029#p161029

torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

Re: ECONET WOES

Postby torrind » Sat Apr 14, 2018 7:13 am

Hi Ian,

Yes correct. I have my two BEEBs connected directly to the clock with no line biasing or terminators.

If I need to add line biasing components as per the schematic, how is it added to the existing data pins on the clock? Is it added in parallel for example?

Thanks

Darren

User avatar
IanS
Posts: 545
Joined: Mon Aug 31, 2009 6:02 pm
Contact:

Re: ECONET WOES

Postby IanS » Sat Apr 14, 2018 1:36 pm

torrind wrote:Yes correct. I have my two BEEBs connected directly to the clock with no line biasing or terminators.

If I need to add line biasing components as per the schematic, how is it added to the existing data pins on the clock? Is it added in parallel for example?

Yes, in parallel, no need to cut any tracks or cables. The easiest way is to probably add the extra resistors onto the clock pcb, as you have +5 available there.

Or create a passive terminiator, which can plug in anywhere - https://www.retro-kit.co.uk/page.cfm/co ... rminators/
(for a small network, only one will be sufficient to bias the data lines)

User avatar
lcww1
Posts: 120
Joined: Wed Mar 15, 2017 11:16 pm
Location: Out There
Contact:

Re: ECONET WOES

Postby lcww1 » Sat Apr 14, 2018 4:01 pm

I’ve just done a quick test using my simoni clock built on the same pcb - I connected an iss4 BBC B and a Master 128 directly to the clock, and both machines picked up the clock signal straightaway. So, unless there’s some line biasing from the Econet module in the Master, this setup should be equivalent to Darren’s, should it not? This would make line biasing issues less likely to be the issue, perhaps.

Have you tried swapping the MC68B54 ADLC chips between the Beebs, to see if that causes the problem to switch from one machine to the other?

User avatar
IanS
Posts: 545
Joined: Mon Aug 31, 2009 6:02 pm
Contact:

Re: ECONET WOES

Postby IanS » Sat Apr 14, 2018 6:30 pm

"line jammed" is usually an indication of missing terminators (or line-biasing). Every Econet should have them, so it's just one less thing to worry about.

User avatar
lcww1
Posts: 120
Joined: Wed Mar 15, 2017 11:16 pm
Location: Out There
Contact:

Re: ECONET WOES

Postby lcww1 » Sat Apr 14, 2018 6:44 pm

oops! I'd somehow missed all the developments above - just ignore me :oops:

anyhow, great progress Darren........

torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

Re: ECONET WOES

Postby torrind » Sun Apr 15, 2018 8:11 pm

Thanks for the confirmation chaps.

I'll source the components for the line biasing module add them to the clock and report back....

stay tuned folks.....

<cue loony toons music>

User avatar
1024MAK
Posts: 7192
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: ECONET WOES

Postby 1024MAK » Sun Apr 15, 2018 9:31 pm

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

torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

Re: ECONET WOES

Postby torrind » Mon Apr 16, 2018 3:46 pm

Haha...



Quick question whilst I await the line biasing components.....

Would the issue be solely down to line biasing if one BEEB says "Line Jammed" and the other says "Net Error"?

I would have assumed they would both say "Line Jammed"


Just a thought.....

Darren

User avatar
lcww1
Posts: 120
Joined: Wed Mar 15, 2017 11:16 pm
Location: Out There
Contact:

Re: ECONET WOES

Postby lcww1 » Tue Apr 17, 2018 8:25 pm

torrind wrote:Would the issue be solely down to line biasing if one BEEB says "Line Jammed" and the other says "Net Error"?


I don’t know the answer to this question Darren, but I have a nagging suspicion that there may be more to this problem than just the line biasing. Some of my suspicion is down to me wondering just how sensitive small/short Econets are to the presence or absence of line termination. For some reason, my ever changing Econet has so far never given rise to any line errors, despite the fact that I’ve been fairly cavalier about whether or not I bother plugging in terminators as I’ve altered the network.

A couple of other, possibly irrelevant, suggestions - have you tried changing the clock settings at all? I wonder if slowing things down would change anything?

Lastly, I came across this handy Econet troubleshooting guide from BeebMaster http://www.beebmaster.co.uk/Econet/TestNet.html - there are a few suggestions there that might be worth a try while you’re waiting on the terminators.....

Anyhow , keep us posted :)

torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

Re: ECONET WOES

Postby torrind » Tue Apr 17, 2018 8:53 pm

Thanks Liam,

Yes this is what I suspect too..

Based upon my use of both BBEBs, BEEB #2 (net error) is the one I feel something is "off" with - but although it seems to function correctly on the surface, I can't help having doubts about it.....

I have tried changing the clock settings to no avail and I have also built Steve's Econet clock v1.0 to see if this made any difference but sadly no!

A couple of questions for you though.....

1. If you have a single 'working' BEEB connected to a clock and you issue *NET and *. - what should happen? - I know very little of the commands you use.
2. I'm guessing the 5 pin DIN cables are "straight through" Pin 1 to Pin 1 etc etc?


Thanks,

Darren

User avatar
lcww1
Posts: 120
Joined: Wed Mar 15, 2017 11:16 pm
Location: Out There
Contact:

Re: ECONET WOES

Postby lcww1 » Tue Apr 17, 2018 11:53 pm

torrind wrote:A couple of questions for you though.....

1. If you have a single 'working' BEEB connected to a clock and you issue *NET and *. - what should happen? - I know very little of the commands you use.
2. I'm guessing the 5 pin DIN cables are "straight through" Pin 1 to Pin 1 etc etc?


Ok, a good excuse for a quick bit of late night Econet tinkering, if indeed any excuse is needed .....

Q1 - when I set things up that way I get a not listening error after a longish pause - the Master with ANFS expects to find a fileserver at station 254 - the Beeb with NFS just says not listening........

Q2 Yes, the Econet cables are straight through, just as you describe

Whilst I was testing the Master and BBC B connected via the clock (with no terminators), I had a rummage around the discs on my Gotek drive (7 segment display only, so I’ve long since lost track of what’s where - just like BITD with piles of unlabelled/mislabelled floppies :D) and found the following program called NetTest - dunno where this came from - either somewhere on BeebMaster’s site or from JGH’s MDFS - anyhow, this little ditty will send a message from one station to another and confirm that communication has been established - this program worked perfectly on my “two beebs both directly connected to the clock” setup - here it is:

2668CE4F-B291-45F7-9F80-60EB1A6D2885.jpeg
NetTest


too late for typing - Yawn!

User avatar
lcww1
Posts: 120
Joined: Wed Mar 15, 2017 11:16 pm
Location: Out There
Contact:

Re: ECONET WOES

Postby lcww1 » Wed Apr 18, 2018 6:39 am

One more thing to note - once you’ve got your Econet up and running, it would make sense to set one machine up as a fileserver - there’s a succinct summary of the options here: https://www.heyrick.co.uk/econet/intro/servers.html - and you can find more details, lists of Econet related * commands, and the necessary software here: https://www.beebmaster.co.uk/Econet/SetFile.html and on successive pages.

You’ll need a coprocessor for a level 2 fileserver, but as you probably know you can use a raspberry pi zero for this these days, see here http://stardot.org.uk/forums/viewtopic.php?t=11325 - you can get a very nice, and inexpensive, level shifter board for interfacing the pi to a Beeb here: http://www.sundby.com/index.php/project/pitubedirect/. So, few things to be thinking about :D

torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

Re: ECONET WOES

Postby torrind » Wed Apr 18, 2018 8:12 pm

I've now built most of the line bias circuit - just waiting on two 56k resistors in the post - hopefully tomorrow, then I can test. (I'll be honest - I'm not hopeful of resolution! - but hey, I'd better be upbeat about it: ...... IT WONT WORK! :lol: :lol: )

Yeah both BEEBs have a PiTubeDirect copro added - they are great little devices.

Thanks for the links and additional info - very useful.

P.S. I'm thinking of buying an A5000 - What are your thoughts? I'm guessing I can also Econet that too? (I'd use it as a fileserver! If I ever get past this ECONET thingumylarky!!)

Stay Tuned......

Darren

User avatar
lcww1
Posts: 120
Joined: Wed Mar 15, 2017 11:16 pm
Location: Out There
Contact:

Re: ECONET WOES

Postby lcww1 » Wed Apr 18, 2018 10:14 pm

Keep going Darren! :D
You'll get it fixed.........

re A5000 - these are my favourite Acorn 32bit machines - I have one working A5000 that I use as a level 4 Econet fileserver (the A5000 takes the same Econet modules as a Beeb Master - though it's best to use one of the low profile types, if you can, to avoid the module clashing with podule 2). This A5000 also has an Access+ Ethernet podule fitted, which makes transferring files to and from my Acorn computers and my more modern computers very easy (there's a great RISC OS NFS client called Sunfish that runs on RO 3.1 http://www.cp15.org/internet/). I've also got 3 other A5000 pcbs in various states of disrepair - perhaps one day I'll even manage to get one (or more) of them fixed.......

torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

Re: ECONET WOES

Postby torrind » Thu Apr 19, 2018 9:03 pm

Well not so good news!

Firstly, I missed out on the purchase of the A5000 and secondly, my completed line bias circuit didn't work (see pic)

I still get a line jammed error on one BEEB and Net Error on the other. My next task would be to build a similar circuit as a passive terminator using a 10uf capacitor - although I'm doubtful it will work.

Although each BEEB can be assigned a station number and they both detect the presence of a clock signal, I'm obviously missing something fundamental here, but I'm at a loss as to what?
Attachments
image1.jpeg

User avatar
IanS
Posts: 545
Joined: Mon Aug 31, 2009 6:02 pm
Contact:

Re: ECONET WOES

Postby IanS » Thu Apr 19, 2018 9:06 pm

Can you post a picture of where you've attached the wires from the biasing circuit?

torrind
Posts: 45
Joined: Fri Dec 01, 2017 4:57 pm
Location: Bristol
Contact:

Re: ECONET WOES

Postby torrind » Thu Apr 19, 2018 9:53 pm

Hi Ian,

Here is a pic of the whole thing...

image1.jpeg


Ive checked the pin outs using a continuity tester as they appear labelled up wrongly on the SV1 and SV2 links..... (btw, I've for completeness moved the pins to the other side of ground (pin 2) just to check, but with the same results).


Darren