The Sinclair QL Forum

Hi Ruptor,

there looks to be a problem somewhere. Nasta's post that I responded to appears to have vanished!

I had to delete one of my posts this morning as I had a duplicate, but it was definitely my own that I deleted. I've seen the SQL error occasionally too, and I suspect that this might be causing the duplicates. I see to remember the error being "SQL Server has gone away".


Cheers,
Norm.

NormanDunbar wrote: I had to delete one of my posts this morning as I had a duplicate, but it was definitely my own that I deleted. I've seen the SQL error occasionally too, and I suspect that this might be causing the duplicates. I see to remember the error being "SQL Server has gone away".

I've experienced similar issues during my last few posts (mainly related to editing a post I've submitted a few minutes before) and seen the "SQL Error" page as well. Re-loading the post, however, showed my edit had gone through and even created a duplicate entry (which I deleted).

NormanDunbar wrote: I had to delete one of my posts this morning as I had a duplicate, but it was definitely my own that I deleted. I've seen the SQL error occasionally too, and I suspect that this might be causing the duplicates. I see to remember the error being "SQL Server has gone away".

I've experienced similar issues during my last few posts (mainly related to editing a post I've submitted a few minutes before) and seen the "SQL Error" page as well. Re-loading the post, however, showed my edit had gone through and even created a duplicate entry (which I deleted).

Now i just wanted to save this post. See what happened:

I have raised an email with the admins (using the email address on the error page) and DM'd Rob Heaton with similar details. Call me Mr Spam if you like!

Hopefully, things will be better soon.


Cheers,
Norm.

Back to my keyboard although this evening I wish I wasn't. The PCB is secured under the frame and belling out the tracks I found I had missed one cut so spent hours trying to cut it before moving on to cut the back side tracks. Even the latter took a lot longer than expected because I forgot the metal frame wasn't long enough to put all the keys on the top row and there is an extra unconnected key next to the return key otherwise a hole would exist. The extra keys scrambled my brain trying to remember where they were when counting the keys to leave some key connections uncut. Now the return key seems to be getting stuck and it is a devil of a job to get the cap off let alone get the switch out. It is a weird one because the hall effect chip was near to the PCB than all the others so it should make less contact with the slider. I will come back to that at the end if the keyboard works. I thought I might finish the PCB wiring by now but I haven't even started.

Ruptor wrote:

Derek_Stewart wrote:Is it possible to re-instate him, as I do not like to see people being banned for a singular view.

If he was the one slagging off the dead Sinclair on a pro Sinclair forum it does question his logic and where his brain was at.

I have been updated with the reasons for the ban, so it looks like no come back.

I do not understand why he has access to the QL Chat...

Ruptor wrote:Back to my keyboard although this evening I wish I wasn't. The PCB is secured under the frame and belling out the tracks I found I had missed one cut so spent hours trying to cut it before moving on to cut the back side tracks. Even the latter took a lot longer than expected because I forgot the metal frame wasn't long enough to put all the keys on the top row and there is an extra unconnected key next to the return key otherwise a hole would exist. The extra keys scrambled my brain trying to remember where they were when counting the keys to leave some key connections uncut. Now the return key seems to be getting stuck and it is a devil of a job to get the cap off let alone get the switch out. It is a weird one because the hall effect chip was near to the PCB than all the others so it should make less contact with the slider. I will come back to that at the end if the keyboard works. I thought I might finish the PCB wiring by now but I haven't even started.

Sounds great, will you show the result of your work.

I looked o the Cherry Switch website of the Chrery Switch life, they indicate 50 Million keypresses, not sure how ong that is?!

Derek_Stewart wrote:

Ruptor wrote:I looked o the Cherry Switch website of the Chrery Switch life, they indicate 50 Million keypresses, not sure how ong that is?!

I think that works out to 137 years for one key pressed a 1000 times a day so I think they will see us out.
This job is a lot of wiring made difficult by the old solder being a different melting point to lead free stuff so dry joints are real easy to make. The better method that in hind sight would have been more useful to my next QL build is to use the existing keyboard matrix with a PIC each end of 4 wires on a serial link and key transposer in the QL end. I persuaded myself out of that route during hacking the PCB to move the blue keys from above the top row to form the QL function key on the left because I cut all the tracks but it was still the easiest route I think. Now I have 24 wires from keyboard to QL that might cause a problem electrically and are unwieldy to say the least. I can still implement the serial method since I have another hall effect keyboard it is just double the work but I suppose It could still be applied to this rewired keyboard without key transposition. Still 48 wires to go and hacking some breadboard with edge connectors to fit the QL sockets so back to the grindstone.

Here is my hacked hall effect switch keyboard with shaved prototype PBC plugs for the membrane to main QL board connectors on the left. I used old telephone cable for keyboard to QL connections and have to pick up 5V from the QL board to supply the hall effect switches. I don't have a specification for the switches so does anybody know if the QL 10 K pull down resistors need to be 1K like on the original donor keyboard circuit for the hall effect switches to work?
Now I have to figure out the best way to test it so it doesn't blow up my QL.

I wanted more information about the hall effect switches and found these two resources that I couldn't find last time I looked.https://deskthority.net/wiki/Micro_Switch_SD_Series
https://telcontar.net/KBK/Micro_Switch/SD#output
It looks like a hall effect switch is not necessarily a switch although the three terminal one does seem like a momentary type. Mine is R4A3E with 40 and 4R on the left & right of sensor with the R in grey but there are '?' all over the sites on the bits I want to know. My impression is the switches are quite high current and originally they were driven by a 74154 with 1K resistor pull downs in series so don't look directly compatible with the QL 3K & 10K set up.
I checked the QL output signals that are all the same but the first line is different looking like it doesn't pull down low enough. When I tested the fixed QL with a wire short in place of the membrane I had the impression it wasn't very consistent but put it down to my wire poking because all the lines did give characters. So the first line needs investigating but the others would show if my keyboard works or not.
Confident the keyboard wouldn't damage the QL I plugged it in and it doesn't work. Checking the drive signals with it unplugged they are the same so no damage caused.
Things don't look promising because the switches originally worked with 5mA and the 8049 is limited to <2mA so needs buffering.
Any suggestions? Time for a cup of tea.