Sinclair ZX80 / ZX81 / Z88 Forums

I realize that I've been creating a lot of posts around questions and debugging of my zx81 laptop project. Rather than that I thought it would be better etiquette to make one thread documenting my progress, questions, frustrations, and solutions.

Today's progress was around the battery pack.
I took apart an old Sony NP-F960 (six 18650 cells) I had laying around.
It was originally wired for 7.2v (max 8.4) and ran the zx81 and laptop for about 3 hours on a full charge.

We rewired it for 10.8v (max 12.6) and hooked it up to a new BMS
It's been running for 3 hours already and has only dropped 1.5v (of the 5.4 available)

Even with a very uneven discharge curve I expect to get 6-8 hours.

For a spare parts battery pack, I'm pretty happy.

Thanks for sharing this with us.
I already have a question:
You expect twice the runtime with the reconfigurated cells. I don't understand why. The ZX81 is very happy with 7.2V. More voltage means more heat being created and so less efficiency. Had there been no BMS before and now with BMS you expect better charging?
About the Zeddy: did you change anything there to draw less power?
My ZX81 now has CMOS CPU, 32k CMOS RAM, CMOS ROM and last not least a vLA81 ULA replacement.
I have to measure the current it draws, but I expect it to be significantly less than original.
Kind regards Paul

Paul wrote: ↑Sun Apr 26, 2020 10:45 am Thanks for sharing this with us.
I already have a question:
You expect twice the runtime with the reconfigurated cells. I don't understand why. The ZX81 is very happy with 7.2V. More voltage means more heat being created and so less efficiency. Had there been no BMS before and now with BMS you expect better charging?

Mostly it's about the batteries being able to more fully discharge.
There are six 18650 cells. Each nominally 3.6v. Fully charged they are 4.2v - They can safely discharge down to 2.3v

With 2 cells in series that doubles the voltage and gives a rage of 4.6v - 8.4v (nominal 7.2v)
The zx81 and TFT monitor both start shutting down just under 7v. Less than half the total discharge.

With 3 cells in series it triples the voltage and gives a range of 6.9v - 12.6v (nominal 10.8v)

This means that the low voltage matches nearly perfectly with the low voltage needs of the system. The BMS will shut down output to protect from over discharge. Meaning the battery pack can be fully drained.

I reconfigured the battery pack from 2 serial, 3 parallel to 3 serial 2 parallel (also switching out the BMS for an appropriate one) to achieve this.

Paul wrote: ↑Sun Apr 26, 2020 10:45 am
About the Zeddy: did you change anything there to draw less power?

The only real change here was replacing the 7508 with an equivalent switching regulator. Because the 7508 uses heat to dissipate excess voltage it is very inefficient at higher voltages. Switching regulators are more expensive and add noise to the signal, but *much* more efficient. This also removed the need for the big heatsink.

The zx81 seems to handle the extra noise well. And if it was a problem it could be taken care of by a few bypass capacitors at the regulator.

Cheers,
Matthew

Thank for pointing that out.
Of course that makes sense.
What kind of a BMS do you use?
And what kind of a TFT? Mine requires more than 8V to start working(12V nominale).

Paul wrote: ↑Sun Apr 26, 2020 5:49 pm Thank for pointing that out.
Of course that makes sense.
What kind of a BMS do you use?
And what kind of a TFT? Mine requires more than 8V to start working(12V nominale).

The TFT is a 7" 1024x600 (I had trouble with lower resolutions)
https://www.amazon.com/gp/aw/d/B07NQ3R ... sin_title
I haven't tested how low it will start. I'll do that with my bench supply and report back. But once it's up, it stays on until just a bit bellow 7v

Here is the BMS I'm using
https://www.amazon.com/gp/aw/d/B07YKCM ... sin_title

And right around 6 hours is the answer for how long the new battery pack lasts


That's double the old battery configuration.
It ran for 3 hours last night and 3 hours today (with a little counting program in SLOW mode to add CPU load)

The BMS shut it down at 8v under load. Not bad at all.

I also just proved to myself (with the aid of my bench supply) that the BMS handles charging in parallel to the zx81 running with a 12v 2A power supply.

That means I'll be able to make a low voltage warning that tells me to plug in. I've made one of these before out of an LM3914 for the 12v battery on my 1980 Westphalia van
I'll just have to adjust resistors to get the correct range

Tonight was spent 3d modeling and test printing an enclosure that holds the zx81, battery, and display.
(BTW: Fusion 360 is free for hobbyists/non-commercial work)

Design constraints:
I want to keep the general lines of the zx81
I love the retro futuristic look of the case and want to continue the same feel.
It needs to be structurally strong enough to hold the parts (battery pack, display, etc.) but not too bulky. Extra material adds a lot of print time.
I want the zx81 to be "drop in" so I can use it as a portable or take it out and still have a regular working zx81.
That means keeping all connectors stock.
It has to be printed in multiple pieces. My printer doesn't have a large enough volume to do it all in one go.

This is proving to be a really fun problem.
I'm excited to wake up tomorrow morning to see how well my test print fits around the zx81

I haven't updated this in awhile.
I've been doing modeling of the zx81 and test prints of the enclosure.

These don't show it well (because I took quick photos to answer another question and the zx81 isn't totally in place) however it fits quite well. Easy to pop in and out. But snug enough not to fall out if I hold it upside down.

Next is to extend the back to make a place for the battery, charging electronics, and a mount for the 7" LCD

Thank for the update. I have an old ZX81 PCB not doing anything. This thread has opened up a couple of options for me.

Cheers.