Yes, I thought it was an XC2064 but when I put my magnifier on the pic it was just a blur.
The print was done on a centronics dot matrix with standard computer paper. I had my own paper with my logo and details printed as I used to write custom software for many ICL customers. (still have a box of my paper stacked in the garage, but sadly no printer to print on it) I also used to put my software printouts on microfiche (another dead system). At one of my customers, I used to log all their invoices/credit notes/picking slips, statements, product listing, and all supporting transactions monthly to microfiche. This way they didn’t have to keep paper records. This was 1980+.
Still have my microfiche but have no idea if anyone has a reader these days. Remember the auto parts shops used microfiche for all their car manuals and parts lists, with pricing. Those days are gone!
The print was done on a centronics dot matrix with standard computer paper.
It looks like it folds out in both directions.
Yes, it is 4 pages of 15"x11" fanfold paper taped together to form a 2x2 sheets, printed with a dot matrix printer in graphics mode.
There was an option in the software to print it smaller that almost fit on a single sheet - just slightly longer but the width was ok.
The XC2018 only had 100 cells in a 10x10 arangement. It allowed me to design the pcb without knowing the full bus specification of the ICL Mini which I had to reverse engineer. There were only two short wires and a single track cut needed to my original pcb design so I never redesigned it for production. Not bad considering I designed the pcb before I could reverse engineer the bus - I used my pcb to snoop the bus first.
Here is a pic of a pair of ICL System 25's. Between them is a 65MB removable HDD with a fixed 130MB underneath. The one on the right is in the pic with the cover removed. My boards are on the right with the cables coming out of them (the smaller RJ45 cables) - there are 4 pcbs each supporting 10 RS232 remote terminals / PCs.
Turns out I failed to read the datasheet for the FT231X chip and assumed that I/O could still be powered at 5V, like with the FT232R. Turns out you can't. This may explain some strange issues I was having...
New boards on the way though. Hoping to have a proved design soon.
The power chip and 9dof chips gave me trouble, but was able to make it work with hot air gun after a several tries...
Next, need to put together the other version that has simple linear regulators instead of this fancy power chip.
What is the power chip ?
Linear regulators are feasible for 3v3, but it's a bigger thermal challenge to get decent 1.8V power without a switcher.
In trying to avoid the harder-to-use SMPS choices, I'm about to trial a TLV62585PDRL 2.5V-5.5V input, 3-A high efficiency step-down converter in SOT563 package paired with an over-rated inductor to keep heat sources to a minimum. TI look to have recently dropped the price on that part, now ~ 16c/1k, and it has a PGOOD out as a bonus.
I think the power chip is TPS650532RGER (one in that series anyway).
I just wish I was better at getting to solder right the first time... Need to work on that...
The EVE2 interface for Newhaven's EVE2 displays works. Knew it would, but good to check. It's the main purpose of this board...
Also, the RGBX led works (tested with Jon's code in obex). I'm sure the demo looks better with 24 leds instead of just 1
I noticed the VGA display has some noise on it with the Parallax VGA breakout attached.
It's probably from the power supply chip's switching regulator.
Next time, I'll route power from one of the LDOs to certain pin headers to fix that.
Right now, the LDOs aren't attached to anything...
Found two ways to fix noisy VGA. One is to add two 10 uF caps the VIO pins at the header.
Second is to change the power supply mode pin from Power Save Mode to Forced PWM mode.
Still, if VGA is intended purpose for these pins, probably better to connect VIO to one of the LDOs...
Found two ways to fix noisy VGA. One is to add two 10 uF caps the VIO pins at the header.
Second is to change the power supply mode pin from Power Save Mode to Forced PWM mode.
Still, if VGA is intended purpose for these pins, probably better to connect VIO to one of the LDOs...
Interesting data point. Is that PCB 2 or 4 layers ? ( I think it is 2L, top side placement only ?)
I'm leaning more to 4L default on small boards, for thermal and decoupling/supply impedance reasons, and I'm also pondering if making the spread caps larger than 100nF would help the DACs.
Here’s another version that can use either LDOs for power or switching regulators.
Shown here with 3-pin regular for 3.3 V and a tiny SMT module for 1.8 V.
The 1.8 V thing looks like three little caps next to the HDMI connector...
Comments
XC3064A-7 looks like a match for the later model.
The print was done on a centronics dot matrix with standard computer paper. I had my own paper with my logo and details printed as I used to write custom software for many ICL customers. (still have a box of my paper stacked in the garage, but sadly no printer to print on it) I also used to put my software printouts on microfiche (another dead system). At one of my customers, I used to log all their invoices/credit notes/picking slips, statements, product listing, and all supporting transactions monthly to microfiche. This way they didn’t have to keep paper records. This was 1980+.
Still have my microfiche but have no idea if anyone has a reader these days. Remember the auto parts shops used microfiche for all their car manuals and parts lists, with pricing. Those days are gone!
Yes, it is 4 pages of 15"x11" fanfold paper taped together to form a 2x2 sheets, printed with a dot matrix printer in graphics mode.
There was an option in the software to print it smaller that almost fit on a single sheet - just slightly longer but the width was ok.
The XC2018 only had 100 cells in a 10x10 arangement. It allowed me to design the pcb without knowing the full bus specification of the ICL Mini which I had to reverse engineer. There were only two short wires and a single track cut needed to my original pcb design so I never redesigned it for production. Not bad considering I designed the pcb before I could reverse engineer the bus - I used my pcb to snoop the bus first.
Here is a pic of a pair of ICL System 25's. Between them is a 65MB removable HDD with a fixed 130MB underneath. The one on the right is in the pic with the cover removed. My boards are on the right with the cables coming out of them (the smaller RJ45 cables) - there are 4 pcbs each supporting 10 RS232 remote terminals / PCs.
I am about to design my board. Which company do you recommend to place the p2 on the board.
Thanks
Not too hard when you have a stencil...
Only company I’ve used to make boards for me is screaming circuits.
They are good. But probably cost a lot ...
New boards on the way though. Hoping to have a proved design soon.
That took forever...
The power chip and 9dof chips gave me trouble, but was able to make it work with hot air gun after a several tries...
Next, need to put together the other version that has simple linear regulators instead of this fancy power chip.
I posted the Eagle source files and also gerbers to top post.
What are the buffer ics on the two USB-A connections?
Same as Eval board I think...
Linear regulators are feasible for 3v3, but it's a bigger thermal challenge to get decent 1.8V power without a switcher.
In trying to avoid the harder-to-use SMPS choices, I'm about to trial a TLV62585PDRL 2.5V-5.5V input, 3-A high efficiency step-down converter in SOT563 package paired with an over-rated inductor to keep heat sources to a minimum. TI look to have recently dropped the price on that part, now ~ 16c/1k, and it has a PGOOD out as a bonus.
I just wish I was better at getting to solder right the first time... Need to work on that...
But, I see it's because I put in a HyperFlash chip instead of HyperRam chip...
Back to the hot air gun...
Also, the RGBX led works (tested with Jon's code in obex). I'm sure the demo looks better with 24 leds instead of just 1
It's probably from the power supply chip's switching regulator.
Next time, I'll route power from one of the LDOs to certain pin headers to fix that.
Right now, the LDOs aren't attached to anything...
Second is to change the power supply mode pin from Power Save Mode to Forced PWM mode.
Still, if VGA is intended purpose for these pins, probably better to connect VIO to one of the LDOs...
Interesting data point. Is that PCB 2 or 4 layers ? ( I think it is 2L, top side placement only ?)
I'm leaning more to 4L default on small boards, for thermal and decoupling/supply impedance reasons, and I'm also pondering if making the spread caps larger than 100nF would help the DACs.
Shown here with 3-pin regular for 3.3 V and a tiny SMT module for 1.8 V.
The 1.8 V thing looks like three little caps next to the HDMI connector...
1.8 V says 1W
So ~500 mA I guess