Re: [M100] REXCPM battery
On 9/22/23 18:09, Brian White wrote: > Took some measurements > > This is with the BAT54C diodes. > > > While not installed in 100: > BATT 2.99v > RAM_RST 2.91v 7uA > SRAM VCC 2.76v > > > While installed in 100, power on, wall power, original tandy 6v adapter: > RAM_RST 4.85v > BATT, installed 3.0v > BATT, not installed 3.65v 0.5uA > That was sent from my phone just to get the numbers down. What I GET from it is, > While not installed in 100: > BATT 2.99v > RAM_RST 2.91v 7uA > SRAM VCC 2.76v * Draws 7uA while being powered from the batteries alone. * Almost no voltage drop from the battery to RAM_RST. * Still very low voltage drop all the way to the actual vcc pin on the sram (vs just at the cap), so a simple battery life estimate based on just nominal mAh capacity and uA drain is correct with no unaccounted further drops. > While installed in 100, power on, wall power, original tandy 6v adapter: > RAM_RST 4.85v > BATT+, batts installed 3.0v > BATT+, batts not installed 3.65v 0.5uA * 0.5uA reverse leakage through the BAT54C * trying to pull the cells up not only over their nominal 3.0v but even over their actual measured brand new initial 3.2v. And the datasheet for a CR1025 cell says "Max Rev Charg: 1 microampere" So, in nice cool under 25c conditions there isn't a problem, but that 0.5uA goes up over 1uA in temperatures that are still possible human environment ambient. The risk is only while powered on, not while say, sitting in a hot car not being used. Then again, unattended running 24/7 has always been a use case for these machines, like famously the EME Online Weather Logger setup. So it could actually be on and running in a pretty hot environment. So I guess stick with the plan to call for BAV170 or other low reverse leakage diode. Another thing I looked at, which also in the end resulted in no change: I read that tantalum caps have leakage current and maybe not ideal for long term battery standby applications. So I removed the tantalum cap and still measured 7.4uA when powered from the battery. So the drain is not from tantalum leakage current. I was hoping for a possible hack to replace the tantalum with a ceramic to trade away some minutes of battery-change time for some more years of battery standby time. But if it's still drawing over 7uA with no cap installed at all, then it's not going to get any better. I can't believe the sram chip is drawing 7uA itself. My REXCPM originally came with a CY62177EV30 whose datasheet says has a standby current of 3uA typical, max 25uA, so IT might possibly draw 7-8uA, but I replaced that with a RMLV3216AGSA to upgrade it from 2M to 4M, and that datasheet says the typical standby current is only 0.6uA, though when you look at the tables vs the summary in both sheets, they actually look like pretty similar min max range. They may both actually only draw about 1uA normally or up to 24-25 in worst case high temp. There is at least a 3.3v voltage regulator and who knows what else besides just the sram being powered while on standby. I thought maybe the 4 batteries was a little unnecessary overkill but I guess it's worth it after all, since even with all 4, the estimated life is still only about 2 years. It would have been nice to go back to 2 cells and use the original big standard header pins (4 big pins still fits where the 3 pins were on the first version). Actually no reason I can't still just make an option for that. So I think this is pretty much it, as good as it's gonna get without picking apart the rexcpm itself to identify some possible improvement to the standby mode, like cutting out the voltage regulator and switching over to a direct connection on standby, or maybe just replacing the regulator with some other model, maybe finding pullups/downs that could be replaced with higher value resistors, etc. But 2 years is good enough to be worth doing I think, and the default hookup method without hacking a qwiic connector onto the rexcpm is easy enough. bkw
[M100] REXCPM battery
Took some measurements This is with the BAT54C diodes. While not installed in 100: BATT 2.99v RAM_RST 2.91v 7uA SRAM VCC 2.76v While installed in 100, power on, wall power, original tandy 6v adapter: RAM_RST 4.85v BATT, installed 3.0v BATT, not installed 3.65v 0.5uA bkw
Re: [M100] rexcpm battery
Yeah, that's one of the disadvantages of using schottkys - they do tend to leak more reverse current. I appreciate the offer to send me a BAV170. I regularly put in orders at Mouser, so I'll just add some to my cart when I do my next order. Very kind to offer, though 🙂 //Ken From: M100 on behalf of Brian K. White Sent: Thursday, September 21, 2023 10:53 PM To: m100@lists.bitchin100.com Subject: Re: [M100] rexcpm battery I've been reading about charging protection for lithium coin cells, and I think we might want to swap out the BAT54C for something with as little reverse leakage as possible. I thought a few uA wouldn't really matter, especially if only during on/active time, and didn't worry about it, but I guess even that causes the cells to die early and only last a few months instead of years. So I think you should not use the BAT54C from the original BOM and use maybe BAV170 instead: https://www.digikey.com/en/products/detail/diodes-incorporated/BAV170-7-F/822819 or any of these: https://www.digikey.com/short/q9bfjch5 This was my other pick: https://www.digikey.com/en/products/detail/central-semiconductor-corp/CMPD6001C-TR-PBFREE/5325257 Lowest reverse leakage but a little higher forward voltage drop and also more expensive. Probably neither difference actually matters much at the voltages and currents involved. Maybe the BAT54C is actually fine as long as the computer isn't turned on for long periods in 90F heat. The reverse leakage graph for the BAT54C at 2v (difference between 5v from the computer and 3v from the battery) at 25C (77F) is only 0.3uA, but could go up around 5uA for 3v (say the batts are almost dead at 2v, so 5-2) at 40C (104F). So... maybe not really a problem 99% of the time. But still. I'm ordering some for myself and if you mail me your address I'll mail you some so you don't have to make a whole digikey order and shipping just for that. No change to the PCB other than the silkscreen. The new part drops in the same place. Sorry about that. -- bkw On 9/20/23 14:19, Ken St. Cyr wrote: > It turned out nice and looks pretty solid (at least in the pictures). I > ordered boards and parts the other day, so I'm just waiting for those to > arrive now. I'll report back and share my experience after getting one > together - > > //Ken > > *From:* M100 on behalf of Brian K. > White > *Sent:* Tuesday, September 19, 2023 10:18 PM > *To:* m100@lists.bitchin100.com > *Subject:* Re: [M100] rexcpm battery > I have repaired my REXCPM and now the github readme includes pics that > show a good way to mount a qwiic connector and where to connect bodge > wires if you want to take on the challenge of such small loose wires > going to such small places. All 4 solder points are at least a > reasonable chip leg, no soldering directly to delicate traces. So the > wires are reasonably robust in the end. You can install & remove the > rexcpm module without straining the wires. I will probably further > secure them with glue sometime later but haven't yet. > > I don't think it's worth it just to get the connector if your original > pins are still intact, even though the connector is more convenient in > the end, but since my original pins rotated and broke their traces > anyway, I had to do a bodge wire repair anyway. And mounting the > connector seems to be solid enough by super-glueing it to the top of a > chip, and by using one of the black versions of the connector which I > think is a material that the glue can stick to easier. The natural > colored ones look like they might be something no glue will stick to. > The black ones feel like they have some glass filler, and can be scuffed > to a rough texture. Seems to be pretty solid so far though I'm not > testing-to-destruction just to find out. :) > single https://www.sparkfun.com/products/14417 > <https://www.sparkfun.com/products/14417> > 10 pack https://www.adafruit.com/product/4208 > <https://www.adafruit.com/product/4208> > > My rexcpm is running and working so I can resume testing battery life. > > I think it's safe to order pcbs if you want now. The dupont breakout > cable is a good solution for keeping the REXCPM side of things simple > and stock, and adding the single wire for gnd is easy, reasonably safe > against blunders, and leaves the REXCPM still compatible with it's > original bus board and cable. So the bus board can keep the qwiic > connector. > > I have tested the 102/200 board now too. good to go. I don't have a > pcbway or gerbers up for that yet but will shortly. > > -- > bkw > > > On 9/19/23 04:29, Brian K. White wrote: >> My boards came
Re: [M100] rexcpm battery
I've been reading about charging protection for lithium coin cells, and I think we might want to swap out the BAT54C for something with as little reverse leakage as possible. I thought a few uA wouldn't really matter, especially if only during on/active time, and didn't worry about it, but I guess even that causes the cells to die early and only last a few months instead of years. So I think you should not use the BAT54C from the original BOM and use maybe BAV170 instead: https://www.digikey.com/en/products/detail/diodes-incorporated/BAV170-7-F/822819 or any of these: https://www.digikey.com/short/q9bfjch5 This was my other pick: https://www.digikey.com/en/products/detail/central-semiconductor-corp/CMPD6001C-TR-PBFREE/5325257 Lowest reverse leakage but a little higher forward voltage drop and also more expensive. Probably neither difference actually matters much at the voltages and currents involved. Maybe the BAT54C is actually fine as long as the computer isn't turned on for long periods in 90F heat. The reverse leakage graph for the BAT54C at 2v (difference between 5v from the computer and 3v from the battery) at 25C (77F) is only 0.3uA, but could go up around 5uA for 3v (say the batts are almost dead at 2v, so 5-2) at 40C (104F). So... maybe not really a problem 99% of the time. But still. I'm ordering some for myself and if you mail me your address I'll mail you some so you don't have to make a whole digikey order and shipping just for that. No change to the PCB other than the silkscreen. The new part drops in the same place. Sorry about that. -- bkw On 9/20/23 14:19, Ken St. Cyr wrote: It turned out nice and looks pretty solid (at least in the pictures). I ordered boards and parts the other day, so I'm just waiting for those to arrive now. I'll report back and share my experience after getting one together - //Ken *From:* M100 on behalf of Brian K. White *Sent:* Tuesday, September 19, 2023 10:18 PM *To:* m100@lists.bitchin100.com *Subject:* Re: [M100] rexcpm battery I have repaired my REXCPM and now the github readme includes pics that show a good way to mount a qwiic connector and where to connect bodge wires if you want to take on the challenge of such small loose wires going to such small places. All 4 solder points are at least a reasonable chip leg, no soldering directly to delicate traces. So the wires are reasonably robust in the end. You can install & remove the rexcpm module without straining the wires. I will probably further secure them with glue sometime later but haven't yet. I don't think it's worth it just to get the connector if your original pins are still intact, even though the connector is more convenient in the end, but since my original pins rotated and broke their traces anyway, I had to do a bodge wire repair anyway. And mounting the connector seems to be solid enough by super-glueing it to the top of a chip, and by using one of the black versions of the connector which I think is a material that the glue can stick to easier. The natural colored ones look like they might be something no glue will stick to. The black ones feel like they have some glass filler, and can be scuffed to a rough texture. Seems to be pretty solid so far though I'm not testing-to-destruction just to find out. :) single https://www.sparkfun.com/products/14417 <https://www.sparkfun.com/products/14417> 10 pack https://www.adafruit.com/product/4208 <https://www.adafruit.com/product/4208> My rexcpm is running and working so I can resume testing battery life. I think it's safe to order pcbs if you want now. The dupont breakout cable is a good solution for keeping the REXCPM side of things simple and stock, and adding the single wire for gnd is easy, reasonably safe against blunders, and leaves the REXCPM still compatible with it's original bus board and cable. So the bus board can keep the qwiic connector. I have tested the 102/200 board now too. good to go. I don't have a pcbway or gerbers up for that yet but will shortly. -- bkw On 9/19/23 04:29, Brian K. White wrote: My boards came in and at least the model 100 one works fine. There's room, but no good way to solder the qwiic connector to the REXCPM, so the most practical way to handle the REXCPM side is just use the pre-made adapter cable that has female dupont sockets, and stick them right onto the original pins. For the GND wire, you can treat it as optional and leave that wire unconnected, or add a single solid-core wire to the big cap for the gnd connection. If you use 23 gauge solid core wire, or maybe a size or so thicker, the wire is stiff enough to hold it's shape and thick enough for a female dupont socket to stick onto it and not fall off. I used 23 gauge wire from some solid core ethernet cable and
Re: [M100] rexcpm battery
On 9/20/23 15:25, Stephen Adolph wrote: Hi, I only want to comment that I think this is very clever and useful! Looks like great work indeed! A point of clarification though. Brian, rexcpm is not usable in Tandy 200. Hah, well ah ok then... ok I just updated everything to remove all mention of model 200. Thank you Cheers Steve On Wednesday, September 20, 2023, Ken St. Cyr <mailto:k...@stcyrfamily.net>> wrote: It turned out nice and looks pretty solid (at least in the pictures). I ordered boards and parts the other day, so I'm just waiting for those to arrive now. I'll report back and share my experience after getting one together - //Ken *From:* M100 mailto:m100-boun...@lists.bitchin100.com>> on behalf of Brian K. White mailto:bw.al...@gmail.com>> *Sent:* Tuesday, September 19, 2023 10:18 PM *To:* m100@lists.bitchin100.com <mailto:m100@lists.bitchin100.com> mailto:m100@lists.bitchin100.com>> *Subject:* Re: [M100] rexcpm battery I have repaired my REXCPM and now the github readme includes pics that show a good way to mount a qwiic connector and where to connect bodge wires if you want to take on the challenge of such small loose wires going to such small places. All 4 solder points are at least a reasonable chip leg, no soldering directly to delicate traces. So the wires are reasonably robust in the end. You can install & remove the rexcpm module without straining the wires. I will probably further secure them with glue sometime later but haven't yet. I don't think it's worth it just to get the connector if your original pins are still intact, even though the connector is more convenient in the end, but since my original pins rotated and broke their traces anyway, I had to do a bodge wire repair anyway. And mounting the connector seems to be solid enough by super-glueing it to the top of a chip, and by using one of the black versions of the connector which I think is a material that the glue can stick to easier. The natural colored ones look like they might be something no glue will stick to. The black ones feel like they have some glass filler, and can be scuffed to a rough texture. Seems to be pretty solid so far though I'm not testing-to-destruction just to find out. :) single https://www.sparkfun.com/products/14417 <https://www.sparkfun.com/products/14417> 10 pack https://www.adafruit.com/product/4208 <https://www.adafruit.com/product/4208> My rexcpm is running and working so I can resume testing battery life. I think it's safe to order pcbs if you want now. The dupont breakout cable is a good solution for keeping the REXCPM side of things simple and stock, and adding the single wire for gnd is easy, reasonably safe against blunders, and leaves the REXCPM still compatible with it's original bus board and cable. So the bus board can keep the qwiic connector. I have tested the 102/200 board now too. good to go. I don't have a pcbway or gerbers up for that yet but will shortly. -- bkw On 9/19/23 04:29, Brian K. White wrote: > My boards came in and at least the model 100 one works fine. > > There's room, but no good way to solder the qwiic connector to the > REXCPM, so the most practical way to handle the REXCPM side is just > use the pre-made adapter cable that has female dupont sockets, and > stick them right onto the original pins. > > For the GND wire, you can treat it as optional and leave that wire > unconnected, or add a single solid-core wire to the big cap for the > gnd connection. If you use 23 gauge solid core wire, or maybe a size > or so thicker, the wire is stiff enough to hold it's shape and thick > enough for a female dupont socket to stick onto it and not fall off. I > used 23 gauge wire from some solid core ethernet cable and that was > about perfect. Other common sources, thermostat wire, doorbell wire. > > Take a 25mm piece of wire, strip 3mm on one end and bend it sharp 90 > degrees, strip 6mm on the other end, poke the wire in between the cap > and the 3 pins and solder the short bent end to the cap. > > Then the 4 female dupont wires go like this: > > black (GND) -> wire on cap > red (/WR) -> closest pin to cap > blue (RAM) -> middle pin > yellow (RAM_RST) -> furthest pin from cap > > > > It turns out you have to be very careful not to strain the original > pins on the REXCPM.. They rotate pretty easily and break the via
Re: [M100] rexcpm battery
Hi, I only want to comment that I think this is very clever and useful! Looks like great work indeed! A point of clarification though. Brian, rexcpm is not usable in Tandy 200. Cheers Steve On Wednesday, September 20, 2023, Ken St. Cyr wrote: > It turned out nice and looks pretty solid (at least in the pictures). I > ordered boards and parts the other day, so I'm just waiting for those to > arrive now. I'll report back and share my experience after getting one > together - > > //Ken > -- > *From:* M100 on behalf of Brian K. > White > *Sent:* Tuesday, September 19, 2023 10:18 PM > *To:* m100@lists.bitchin100.com > *Subject:* Re: [M100] rexcpm battery > > I have repaired my REXCPM and now the github readme includes pics that > show a good way to mount a qwiic connector and where to connect bodge > wires if you want to take on the challenge of such small loose wires > going to such small places. All 4 solder points are at least a > reasonable chip leg, no soldering directly to delicate traces. So the > wires are reasonably robust in the end. You can install & remove the > rexcpm module without straining the wires. I will probably further > secure them with glue sometime later but haven't yet. > > I don't think it's worth it just to get the connector if your original > pins are still intact, even though the connector is more convenient in > the end, but since my original pins rotated and broke their traces > anyway, I had to do a bodge wire repair anyway. And mounting the > connector seems to be solid enough by super-glueing it to the top of a > chip, and by using one of the black versions of the connector which I > think is a material that the glue can stick to easier. The natural > colored ones look like they might be something no glue will stick to. > The black ones feel like they have some glass filler, and can be scuffed > to a rough texture. Seems to be pretty solid so far though I'm not > testing-to-destruction just to find out. :) > single https://www.sparkfun.com/products/14417 > 10 pack https://www.adafruit.com/product/4208 > > My rexcpm is running and working so I can resume testing battery life. > > I think it's safe to order pcbs if you want now. The dupont breakout > cable is a good solution for keeping the REXCPM side of things simple > and stock, and adding the single wire for gnd is easy, reasonably safe > against blunders, and leaves the REXCPM still compatible with it's > original bus board and cable. So the bus board can keep the qwiic > connector. > > I have tested the 102/200 board now too. good to go. I don't have a > pcbway or gerbers up for that yet but will shortly. > > -- > bkw > > > On 9/19/23 04:29, Brian K. White wrote: > > My boards came in and at least the model 100 one works fine. > > > > There's room, but no good way to solder the qwiic connector to the > > REXCPM, so the most practical way to handle the REXCPM side is just > > use the pre-made adapter cable that has female dupont sockets, and > > stick them right onto the original pins. > > > > For the GND wire, you can treat it as optional and leave that wire > > unconnected, or add a single solid-core wire to the big cap for the > > gnd connection. If you use 23 gauge solid core wire, or maybe a size > > or so thicker, the wire is stiff enough to hold it's shape and thick > > enough for a female dupont socket to stick onto it and not fall off. I > > used 23 gauge wire from some solid core ethernet cable and that was > > about perfect. Other common sources, thermostat wire, doorbell wire. > > > > Take a 25mm piece of wire, strip 3mm on one end and bend it sharp 90 > > degrees, strip 6mm on the other end, poke the wire in between the cap > > and the 3 pins and solder the short bent end to the cap. > > > > Then the 4 female dupont wires go like this: > > > > black (GND) -> wire on cap > > red(/WR) -> closest pin to cap > > blue (RAM) -> middle pin > > yellow (RAM_RST) -> furthest pin from cap > > > > > > > > It turns out you have to be very careful not to strain the original > > pins on the REXCPM.. They rotate pretty easily and break the via free > > inside the pcb, and then the traces break. > > > > At one point I was stuffing the gnd wire down from the top in between > > the pin and the cap, and that put sideways strain on the pin. > > > > My REXCPM doesn't work now unless I hold the pins a certain way to > > cause them to make contact. > > > > I have to desolder a few parts to figure out where
Re: [M100] rexcpm battery
It turned out nice and looks pretty solid (at least in the pictures). I ordered boards and parts the other day, so I'm just waiting for those to arrive now. I'll report back and share my experience after getting one together - //Ken From: M100 on behalf of Brian K. White Sent: Tuesday, September 19, 2023 10:18 PM To: m100@lists.bitchin100.com Subject: Re: [M100] rexcpm battery I have repaired my REXCPM and now the github readme includes pics that show a good way to mount a qwiic connector and where to connect bodge wires if you want to take on the challenge of such small loose wires going to such small places. All 4 solder points are at least a reasonable chip leg, no soldering directly to delicate traces. So the wires are reasonably robust in the end. You can install & remove the rexcpm module without straining the wires. I will probably further secure them with glue sometime later but haven't yet. I don't think it's worth it just to get the connector if your original pins are still intact, even though the connector is more convenient in the end, but since my original pins rotated and broke their traces anyway, I had to do a bodge wire repair anyway. And mounting the connector seems to be solid enough by super-glueing it to the top of a chip, and by using one of the black versions of the connector which I think is a material that the glue can stick to easier. The natural colored ones look like they might be something no glue will stick to. The black ones feel like they have some glass filler, and can be scuffed to a rough texture. Seems to be pretty solid so far though I'm not testing-to-destruction just to find out. :) single https://www.sparkfun.com/products/14417 10 pack https://www.adafruit.com/product/4208 My rexcpm is running and working so I can resume testing battery life. I think it's safe to order pcbs if you want now. The dupont breakout cable is a good solution for keeping the REXCPM side of things simple and stock, and adding the single wire for gnd is easy, reasonably safe against blunders, and leaves the REXCPM still compatible with it's original bus board and cable. So the bus board can keep the qwiic connector. I have tested the 102/200 board now too. good to go. I don't have a pcbway or gerbers up for that yet but will shortly. -- bkw On 9/19/23 04:29, Brian K. White wrote: > My boards came in and at least the model 100 one works fine. > > There's room, but no good way to solder the qwiic connector to the > REXCPM, so the most practical way to handle the REXCPM side is just > use the pre-made adapter cable that has female dupont sockets, and > stick them right onto the original pins. > > For the GND wire, you can treat it as optional and leave that wire > unconnected, or add a single solid-core wire to the big cap for the > gnd connection. If you use 23 gauge solid core wire, or maybe a size > or so thicker, the wire is stiff enough to hold it's shape and thick > enough for a female dupont socket to stick onto it and not fall off. I > used 23 gauge wire from some solid core ethernet cable and that was > about perfect. Other common sources, thermostat wire, doorbell wire. > > Take a 25mm piece of wire, strip 3mm on one end and bend it sharp 90 > degrees, strip 6mm on the other end, poke the wire in between the cap > and the 3 pins and solder the short bent end to the cap. > > Then the 4 female dupont wires go like this: > > black (GND) -> wire on cap > red(/WR) -> closest pin to cap > blue (RAM) -> middle pin > yellow (RAM_RST) -> furthest pin from cap > > > > It turns out you have to be very careful not to strain the original > pins on the REXCPM.. They rotate pretty easily and break the via free > inside the pcb, and then the traces break. > > At one point I was stuffing the gnd wire down from the top in between > the pin and the cap, and that put sideways strain on the pin. > > My REXCPM doesn't work now unless I hold the pins a certain way to > cause them to make contact. > > I have to desolder a few parts to figure out where two of the traces > go so I can find where to run bodge wires. Fixing the original traces > right where they broke at the base of the pins is probably too fragile > now that the vias are moving like that. > > Maybe I'll just remove the pins, fix the existing traces where they > broke, and solder flexible wires instead of pins where the pins used > to go, and secure the solder joints with glue or epoxy. Hopefully the > combination of the glue and the flexible wire, and being super careful > from now on, will be enough to keep the traces from breaking again. > > So I think the battery board is fine, and the cable connection at > least for model 100 is practical en
Re: [M100] rexcpm battery
I have repaired my REXCPM and now the github readme includes pics that show a good way to mount a qwiic connector and where to connect bodge wires if you want to take on the challenge of such small loose wires going to such small places. All 4 solder points are at least a reasonable chip leg, no soldering directly to delicate traces. So the wires are reasonably robust in the end. You can install & remove the rexcpm module without straining the wires. I will probably further secure them with glue sometime later but haven't yet. I don't think it's worth it just to get the connector if your original pins are still intact, even though the connector is more convenient in the end, but since my original pins rotated and broke their traces anyway, I had to do a bodge wire repair anyway. And mounting the connector seems to be solid enough by super-glueing it to the top of a chip, and by using one of the black versions of the connector which I think is a material that the glue can stick to easier. The natural colored ones look like they might be something no glue will stick to. The black ones feel like they have some glass filler, and can be scuffed to a rough texture. Seems to be pretty solid so far though I'm not testing-to-destruction just to find out. :) single https://www.sparkfun.com/products/14417 10 pack https://www.adafruit.com/product/4208 My rexcpm is running and working so I can resume testing battery life. I think it's safe to order pcbs if you want now. The dupont breakout cable is a good solution for keeping the REXCPM side of things simple and stock, and adding the single wire for gnd is easy, reasonably safe against blunders, and leaves the REXCPM still compatible with it's original bus board and cable. So the bus board can keep the qwiic connector. I have tested the 102/200 board now too. good to go. I don't have a pcbway or gerbers up for that yet but will shortly. -- bkw On 9/19/23 04:29, Brian K. White wrote: My boards came in and at least the model 100 one works fine. There's room, but no good way to solder the qwiic connector to the REXCPM, so the most practical way to handle the REXCPM side is just use the pre-made adapter cable that has female dupont sockets, and stick them right onto the original pins. For the GND wire, you can treat it as optional and leave that wire unconnected, or add a single solid-core wire to the big cap for the gnd connection. If you use 23 gauge solid core wire, or maybe a size or so thicker, the wire is stiff enough to hold it's shape and thick enough for a female dupont socket to stick onto it and not fall off. I used 23 gauge wire from some solid core ethernet cable and that was about perfect. Other common sources, thermostat wire, doorbell wire. Take a 25mm piece of wire, strip 3mm on one end and bend it sharp 90 degrees, strip 6mm on the other end, poke the wire in between the cap and the 3 pins and solder the short bent end to the cap. Then the 4 female dupont wires go like this: black (GND) -> wire on cap red (/WR) -> closest pin to cap blue (RAM) -> middle pin yellow (RAM_RST) -> furthest pin from cap It turns out you have to be very careful not to strain the original pins on the REXCPM.. They rotate pretty easily and break the via free inside the pcb, and then the traces break. At one point I was stuffing the gnd wire down from the top in between the pin and the cap, and that put sideways strain on the pin. My REXCPM doesn't work now unless I hold the pins a certain way to cause them to make contact. I have to desolder a few parts to figure out where two of the traces go so I can find where to run bodge wires. Fixing the original traces right where they broke at the base of the pins is probably too fragile now that the vias are moving like that. Maybe I'll just remove the pins, fix the existing traces where they broke, and solder flexible wires instead of pins where the pins used to go, and secure the solder joints with glue or epoxy. Hopefully the combination of the glue and the flexible wire, and being super careful from now on, will be enough to keep the traces from breaking again. So I think the battery board is fine, and the cable connection at least for model 100 is practical enough (I'd prefer a matching polarized connector instead of 4 individual wires, but there is just no good way to add that tiny connector that won't be too delicate and just break in no time) but my rexcpm is out of commission and I can't do any further real testing to get real current and voltages etc right now. I did have it fully hooked up and running briefly before I twisted the pins. I can't test the 102/200 board right now, and the pre-made qwiic-dupont cable is only 150mm, not long enough to reach all the way from the board to the rexcpm. You could add short male-female dupont extensions/jumpers, but that's kind of annoying. For the 102/200 board, really may
Re: [M100] rexcpm battery
My boards came in and at least the model 100 one works fine. There's room, but no good way to solder the qwiic connector to the REXCPM, so the most practical way to handle the REXCPM side is just use the pre-made adapter cable that has female dupont sockets, and stick them right onto the original pins. For the GND wire, you can treat it as optional and leave that wire unconnected, or add a single solid-core wire to the big cap for the gnd connection. If you use 23 gauge solid core wire, or maybe a size or so thicker, the wire is stiff enough to hold it's shape and thick enough for a female dupont socket to stick onto it and not fall off. I used 23 gauge wire from some solid core ethernet cable and that was about perfect. Other common sources, thermostat wire, doorbell wire. Take a 25mm piece of wire, strip 3mm on one end and bend it sharp 90 degrees, strip 6mm on the other end, poke the wire in between the cap and the 3 pins and solder the short bent end to the cap. Then the 4 female dupont wires go like this: black (GND) -> wire on cap red(/WR) -> closest pin to cap blue (RAM) -> middle pin yellow (RAM_RST) -> furthest pin from cap It turns out you have to be very careful not to strain the original pins on the REXCPM.. They rotate pretty easily and break the via free inside the pcb, and then the traces break. At one point I was stuffing the gnd wire down from the top in between the pin and the cap, and that put sideways strain on the pin. My REXCPM doesn't work now unless I hold the pins a certain way to cause them to make contact. I have to desolder a few parts to figure out where two of the traces go so I can find where to run bodge wires. Fixing the original traces right where they broke at the base of the pins is probably too fragile now that the vias are moving like that. Maybe I'll just remove the pins, fix the existing traces where they broke, and solder flexible wires instead of pins where the pins used to go, and secure the solder joints with glue or epoxy. Hopefully the combination of the glue and the flexible wire, and being super careful from now on, will be enough to keep the traces from breaking again. So I think the battery board is fine, and the cable connection at least for model 100 is practical enough (I'd prefer a matching polarized connector instead of 4 individual wires, but there is just no good way to add that tiny connector that won't be too delicate and just break in no time) but my rexcpm is out of commission and I can't do any further real testing to get real current and voltages etc right now. I did have it fully hooked up and running briefly before I twisted the pins. I can't test the 102/200 board right now, and the pre-made qwiic-dupont cable is only 150mm, not long enough to reach all the way from the board to the rexcpm. You could add short male-female dupont extensions/jumpers, but that's kind of annoying. For the 102/200 board, really maybe for both boards, maybe it's better to just go down to 2 or 3 cells and full size generic 4-pin headers instead of the qwiic connectors. It does work fine for the 100 version though since there is a pre-made cable that is long enough all in one piece. I've updated the github with bom and pcbway links, and gerber zip for the model 100 version. I say it's safe to order the 100 version at least. The 102/200 version I haven't worked out how best to connect to the REXCPM yet. https://github.com/bkw777/REXCPM_UPS -- bkw On 9/17/23 20:33, Ken St. Cyr wrote: Thanks for the detailed explanation - I totally get it now. I'll wait until you verify the board and post the gerber before ordering a set. I typically have an order or two open with JLCPCB at any given time - for just about every order, it's exactly 10 days from the time I click the purchase button before the blue box shows up in my mailbox, and that's with the cheapest shipping option. I'm not in a hurry to try it out, but if I was, I would probably just mill a prototype board on my CNC. Regarding the 4-wire connector... I'm happy to just solder the ground wire directly to the REXCPM, though I'm sure most normal people would prefer to have it on a connector. //Ken S. *From:* M100 on behalf of Brian K. White *Sent:* Thursday, September 14, 2023 10:16 PM *To:* m100@lists.bitchin100.com *Subject:* Re: [M100] rexcpm battery On 9/14/23 21:27, Brian K. White wrote: But somehow there is power differential between gnd and vmem (or whatever the positive side of the big cap is, I say vmem because obviously it powers the sram) on the rexcpm, and 0.7uA of current flowing along RAM_RST. 7uA not 0.7uA! -- bkw -- bkw
Re: [M100] rexcpm battery
Thanks for the detailed explanation - I totally get it now. I'll wait until you verify the board and post the gerber before ordering a set. I typically have an order or two open with JLCPCB at any given time - for just about every order, it's exactly 10 days from the time I click the purchase button before the blue box shows up in my mailbox, and that's with the cheapest shipping option. I'm not in a hurry to try it out, but if I was, I would probably just mill a prototype board on my CNC. Regarding the 4-wire connector... I'm happy to just solder the ground wire directly to the REXCPM, though I'm sure most normal people would prefer to have it on a connector. //Ken S. From: M100 on behalf of Brian K. White Sent: Thursday, September 14, 2023 10:16 PM To: m100@lists.bitchin100.com Subject: Re: [M100] rexcpm battery On 9/14/23 21:27, Brian K. White wrote: > But somehow there is power differential between gnd and vmem (or > whatever the positive side of the big cap is, I say vmem because > obviously it powers the sram) on the rexcpm, and 0.7uA of current > flowing along RAM_RST. 7uA not 0.7uA! -- bkw
Re: [M100] rexcpm battery
On 9/14/23 21:27, Brian K. White wrote: But somehow there is power differential between gnd and vmem (or whatever the positive side of the big cap is, I say vmem because obviously it powers the sram) on the rexcpm, and 0.7uA of current flowing along RAM_RST. 7uA not 0.7uA! -- bkw
Re: [M100] rexcpm battery
also the updated design is visible here https://github.com/bkw777/REXCPM_UPS The schematic now also documents which pins are which on the rexcpm too just for convenience & reference. You *could* got ahead and export gerbers and order boards right now without waiting. But I have a set on the way and I did pay for DHL, so it will only be about a week. Once I build one and there is no bonehead mistake, I'll update the readme with links to oshpark or pcbway or both, and add gerber zips under releases. I think the board should have ENIG finish so that the battery contacts are gold plated, but that makes the cost go up a lot on pcbway. elecrow and jlc are a lot cheaper for ENIG. oshpark is always ENIG. But the soldermask from jlc is kind of junk. Maybe it's just if getting blue or black and maybe green is better, but definitely black especially is weak and scrapes off easily. I'm still using them though at least for testing because, well that's why it's called prototyping. The boms are already updated with the new parts, including an extra male connector for the rexcpm, though there are no directions or plan or anything for what exactly to do with that connector. It's an SMT part and no obvious non-janky way to hook it up yet. Not even sure it even fits in the space allowed yet (but probably does, because the original connectors are already 2.54mm thick and they do not take up the entire space, so 2.95mm should be ok) There's some other options for the cables too not shown. Digikey has some plain black wire cables made by JST. That doesn't matter for the 100 where the wires are inside a compartment, but for 102 or 200 where it has to run externally and be taped to the bottom or something it might be slightly less ugly. 12" for 102/200 https://www.digikey.com/short/0m7pnpwm 4" for 100 https://www.digikey.com/short/5b7vjjvq or 2" for 100 but which might not be long enough https://www.digikey.com/short/mqh5ncbd Similarly, although the bom has white connectors from digikey. adafruit sells some black connectors: https://www.adafruit.com/product/4208 -- bkw On 9/14/23 21:26, Brian K. White wrote: Yes to most of your understanding, not yet on the 4-cell test. I just ordered the pcb's yesterday so it'll be about a week. Although there is not much to actually test about the updated board. It's just more of the same. It's the same circuit. All it changes is adding more celles and a ground wire. The gnd wire isn't really a new addition because when the boards are installed in the machine, there is an equivalent gnd connection through the normal gnd pins in both sockets. Both boards do have their local GND connected to the GND pins on their own respective sockets, and both sockets are connected to the same GND rail all throughout the 100. So when both boards are installed, there is a common GND connection between the two boards the same as a wire. Nothing about the fitment changes, it's the same pin headers and same battery holders. The pin headers are pretty tall at just over 4mm and the was a real question if that would fit or if it needed a special pin header with shorter shoulders. But that turned out ok. Yiu do need to flush-cut all the pins on top and then ideally touch each one again to melt the shap cut to a smooth dome. But that's easy. The new board is no different. The Qwiic connector is new, but it is 2.95mm, and so that's that, no worry even though I haven't actually got it yet. So nothing much changes either electrically or physically, at least not in terms of wondering if it will work or not. In the mean time, while waiting for the new boards I have had the original board installed in a 100 with the memory power switch turned off, and it's still reading 2.69v at the cap right now after... 4 or 5 days so far? My original email was 4 days ago and I had done over 24 hours of testing at that time. This is good. The leakage thing is this: First, background, the REXCPM lives in a socket that was only designed for a ROM, and so that socket only has minimal bus signals. Just power, address, data, chip-enable, output-enable. It doesn't even have a write-enable signal like an sram would have. What matters in this case is among the things it doesn't have is it has no provision keeping something powered-but-disabled while the machine is turned off. It just has a VCC pin the dies when everything else dies. The system bus has a bunch of other signals, including RAM_RST, which is held *high* the entire time the machine is "off". This is to keep sram disabled and asleep by holding their active-low /CE pins high. Not just for a peripheral connected to the system bus but also all the internal ram is disabled (or enabled) by that same signal. I don't know if this signal was originally intended to be used as a power source vs just a control signal. The internal ram has an actual VBAT power line on their VCC pins, and there is no pin for that
Re: [M100] rexcpm battery
t need mounting tape or who knows what. Maybe I just change the design again to use some other more convenient type of connector that isn't quite so small but still small enough? There are cheap male & female pigtail kits for JST-PH 2.0mm that have a male connactor crimped on to wires. That would be dead easy to solder the male side onto the rexcpm. https://www.amazon.com/dp/B081CRLN8B And I guess the other end could just be wires soldered to the bus board too, so that one kit does everything, but I actually really prefer the original arrangement with only fixed connectors (be they male or female) on the boards and wires only in the cable between. I would rather figure out some way no matter what it takes, to have just fixed connectors on the boards. Maybe I can desolder the original pins but only to slide them up slightly so that they are as high in the air as they can go before hitting the compartment cover, and maybe that leaves enough space between the pins and the pcb to solder the connector to the underside of the pins, and the pins provide the rigid mounting. Then the gnd wire can be a loose bodge wire glued down. Maybe I can add a 4th rigid pin extracted from some other generic header or connector, from the rear of the cap over to the connector, to solder to one of the mechanical mounting solder pads on the side of the connector too, as well as the gnd pin. The rexcpm side of things is now the unknown tricky part but hopefully I figure out some plan that isn't too terribly hacky and is actually doable and solid. -- bkw On 9/14/23 14:53, Ken St. Cyr wrote: Great work! To make sure I’m following - the REXCPM relies on having a ground path through the socket, so the idea is to swap out the 3-wire JST connection for a 4-wire version, which adds the ground pin... And you’re saying the only reason it works outside the unit right now without the ground path is because there’s leakage on the RAM_RST line? I'd like to give it a shot ... have you done a prototype run of the 4x battery board yet? //Ken S *From:* M100 on behalf of Brian K. White *Sent:* Wednesday, September 13, 2023 3:13:53 AM *To:* m100@lists.bitchin100.com *Subject:* Re: [M100] rexcpm battery Let's give THIS a try... https://photos.app.goo.gl/JaSofoK2VotHzVco9 <https://photos.app.goo.gl/JaSofoK2VotHzVco9> The connector is a standard JST-SH, and cables are available pre-made in both 2" for the 100 and 12" for the 102 or 200. There's even an off-the-shelf cable that has dupont sockets on one end, so you could connect to an unmodified rexcpm. So now we have 120 mah gnd connection polarized plug I think we're looking at at least 2 years of shelf life now. -- bkw On 9/12/23 15:05, Brian K. White wrote: I've done a little more testing. The GND path is backfeeding only via the RAM line. Nothing on /WR. (the only possible connections) I am measuring only 6.6uA on the RAM_RST line between the two boards with the boards outside of the 100. That is a little high but actually not too bad. I would hope for under 5uA ideally but 7uA is in the same class. At that rate, nominally the batteries should last over 11 months. And at that low current there shouldn't be much voltage drop across the diode on RAM_RST, and indeed there isn't. I only lose about 0.1v from the battery to RAM_RST. The bulk of the voltage drop is happening along the GND path, and things change a lot when the boards are installed in the 100 and the 100 provides a real GND connection between the two boards. With both boards outside of the 100, I currently measure (the batteries are draining as time goes on so these numbers all change a little each time I check something, so I'll keep saying the battery voltage along with whatever else I'm talking about) 2.99v between BATT- and BATT+ 2.91v between BATT- and RAM_RST (so practically no drop through the diode on the bus board) 2.76v between BATT- and CAP+ (it's not labelled CAP+, I just mean the positive side of the big cap) on the REXCPM (after giving it time to drain away the 5v from the 100) 2.26v between CAP- and CAP+ So there is not actually much drop from BATT+ along RAM_RST to REXCPM VCC, but a 0.5v loss along the GND-via-RAM path. Next, when I provide an actual GND connection between the boards by plugging them in to the 100 with the memory power switch turned off, the current draw actually increases slightly to 7.2uA (after an initial spike probably charging the cap), but the voltage drop decreases. The battery is currently at 2.99v and I get 2.72v at the cap instead of 2.26v While writing this the current along RAM_RST very gradually dropped further to 6.9uA. Maybe it will continue creeping down to settle at the same 6.6-6.7 eventually? This suggests that maybe the batteries would provide a
Re: [M100] rexcpm battery
t need mounting tape or who knows what. Maybe I just change the design again to use some other more convenient type of connector that isn't quite so small but still small enough? There are cheap male & female pigtail kits for JST-PH 2.0mm that have a male connactor crimped on to wires. That would be dead easy to solder the male side onto the rexcpm. https://www.amazon.com/dp/B081CRLN8B And I guess the other end could just be wires soldered to the bus board too, so that one kit does everything, but I actually really prefer the original arrangement with only fixed connectors (be they male or female) on the boards and wires only in the cable between. I would rather figure out some way no matter what it takes, to have just fixed connectors on the boards. Maybe I can desolder the original pins but only to slide them up slightly so that they are as high in the air as they can go before hitting the compartment cover, and maybe that leaves enough space between the pins and the pcb to solder the connector to the underside of the pins, and the pins provide the rigid mounting. Then the gnd wire can be a loose bodge wire glued down. Maybe I can add a 4th rigid pin extracted from some other generic header or connector, from the rear of the cap over to the connector, to solder to one of the mechanical mounting solder pads on the side of the connector too, as well as the gnd pin. The rexcpm side of things is now the unknown tricky part but hopefully I figure out some plan that isn't too terribly hacky and is actually doable and solid. -- bkw On 9/14/23 14:53, Ken St. Cyr wrote: Great work! To make sure I’m following - the REXCPM relies on having a ground path through the socket, so the idea is to swap out the 3-wire JST connection for a 4-wire version, which adds the ground pin... And you’re saying the only reason it works outside the unit right now without the ground path is because there’s leakage on the RAM_RST line? I'd like to give it a shot ... have you done a prototype run of the 4x battery board yet? //Ken S *From:* M100 on behalf of Brian K. White *Sent:* Wednesday, September 13, 2023 3:13:53 AM *To:* m100@lists.bitchin100.com *Subject:* Re: [M100] rexcpm battery Let's give THIS a try... https://photos.app.goo.gl/JaSofoK2VotHzVco9 <https://photos.app.goo.gl/JaSofoK2VotHzVco9> The connector is a standard JST-SH, and cables are available pre-made in both 2" for the 100 and 12" for the 102 or 200. There's even an off-the-shelf cable that has dupont sockets on one end, so you could connect to an unmodified rexcpm. So now we have 120 mah gnd connection polarized plug I think we're looking at at least 2 years of shelf life now. -- bkw On 9/12/23 15:05, Brian K. White wrote: I've done a little more testing. The GND path is backfeeding only via the RAM line. Nothing on /WR. (the only possible connections) I am measuring only 6.6uA on the RAM_RST line between the two boards with the boards outside of the 100. That is a little high but actually not too bad. I would hope for under 5uA ideally but 7uA is in the same class. At that rate, nominally the batteries should last over 11 months. And at that low current there shouldn't be much voltage drop across the diode on RAM_RST, and indeed there isn't. I only lose about 0.1v from the battery to RAM_RST. The bulk of the voltage drop is happening along the GND path, and things change a lot when the boards are installed in the 100 and the 100 provides a real GND connection between the two boards. With both boards outside of the 100, I currently measure (the batteries are draining as time goes on so these numbers all change a little each time I check something, so I'll keep saying the battery voltage along with whatever else I'm talking about) 2.99v between BATT- and BATT+ 2.91v between BATT- and RAM_RST (so practically no drop through the diode on the bus board) 2.76v between BATT- and CAP+ (it's not labelled CAP+, I just mean the positive side of the big cap) on the REXCPM (after giving it time to drain away the 5v from the 100) 2.26v between CAP- and CAP+ So there is not actually much drop from BATT+ along RAM_RST to REXCPM VCC, but a 0.5v loss along the GND-via-RAM path. Next, when I provide an actual GND connection between the boards by plugging them in to the 100 with the memory power switch turned off, the current draw actually increases slightly to 7.2uA (after an initial spike probably charging the cap), but the voltage drop decreases. The battery is currently at 2.99v and I get 2.72v at the cap instead of 2.26v While writing this the current along RAM_RST very gradually dropped further to 6.9uA. Maybe it will continue creeping down to settle at the same 6.6-6.7 eventually? This suggests that maybe the batteries would provide a
Re: [M100] rexcpm battery
Great work! To make sure I’m following - the REXCPM relies on having a ground path through the socket, so the idea is to swap out the 3-wire JST connection for a 4-wire version, which adds the ground pin... And you’re saying the only reason it works outside the unit right now without the ground path is because there’s leakage on the RAM_RST line? I'd like to give it a shot ... have you done a prototype run of the 4x battery board yet? //Ken S From: M100 on behalf of Brian K. White Sent: Wednesday, September 13, 2023 3:13:53 AM To: m100@lists.bitchin100.com Subject: Re: [M100] rexcpm battery Let's give THIS a try... https://photos.app.goo.gl/JaSofoK2VotHzVco9 The connector is a standard JST-SH, and cables are available pre-made in both 2" for the 100 and 12" for the 102 or 200. There's even an off-the-shelf cable that has dupont sockets on one end, so you could connect to an unmodified rexcpm. So now we have 120 mah gnd connection polarized plug I think we're looking at at least 2 years of shelf life now. -- bkw On 9/12/23 15:05, Brian K. White wrote: > I've done a little more testing. > > The GND path is backfeeding only via the RAM line. Nothing on /WR. > (the only possible connections) > > I am measuring only 6.6uA on the RAM_RST line between the two boards > with the boards outside of the 100. > > That is a little high but actually not too bad. I would hope for under > 5uA ideally but 7uA is in the same class. > > At that rate, nominally the batteries should last over 11 months. And > at that low current there shouldn't be much voltage drop across the > diode on RAM_RST, and indeed there isn't. I only lose about 0.1v from > the battery to RAM_RST. > > The bulk of the voltage drop is happening along the GND path, and > things change a lot when the boards are installed in the 100 and the > 100 provides a real GND connection between the two boards. > > With both boards outside of the 100, I currently measure > (the batteries are draining as time goes on so these numbers all > change a little each time I check something, so I'll keep saying the > battery voltage along with whatever else I'm talking about) > > 2.99v between BATT- and BATT+ > 2.91v between BATT- and RAM_RST (so practically no drop through the > diode on the bus board) > 2.76v between BATT- and CAP+ (it's not labelled CAP+, I just mean the > positive side of the big cap) on the REXCPM (after giving it time to > drain away the 5v from the 100) > 2.26v between CAP- and CAP+ > > So there is not actually much drop from BATT+ along RAM_RST to REXCPM > VCC, but a 0.5v loss along the GND-via-RAM path. > > Next, when I provide an actual GND connection between the boards by > plugging them in to the 100 with the memory power switch turned off, > the current draw actually increases slightly to 7.2uA (after an > initial spike probably charging the cap), but the voltage drop > decreases. The battery is currently at 2.99v and I get 2.72v at the > cap instead of 2.26v > > While writing this the current along RAM_RST very gradually dropped > further to 6.9uA. Maybe it will continue creeping down to settle at > the same 6.6-6.7 eventually? > > This suggests that maybe the batteries would provide a worthwhile > shelf-life extension after all, at least while installed in the > machine which provides a real GND connection. > > 60mAh at 8uA works out to just over 10 months. (according to > https://www.digikey.com/en/resources/conversion-calculators/conversion-calculator-battery-life) > > That's worth doing I think. Especially when added to the 100's own > shelf-life, that should work out to maybe 1.5 years total. > > I still have to check the actual current drain at the battery. My > current measurements above were just on the RAM_RST line from the bus > board to the rexcpm, not at the battery itself. It's possible the > battery is seeing more total current than just that through additional > paths like reverse leakage through the schottkey from BATT+ to VBUS, > and through RAM_RST into the 100. > > Anyway it looks a lot more promising than I thought at first. > > > As for getting that same long shelf-life outside of the 100, one idea > might be to mod the REXCPM to use a 4-pin JST-SH connector and use the > same connector on the bus board, and a pre-made cable assembly. > > Those are very small connectors and would be difficult to crimp the > female connectors on the cable by hand, but they sell those pre-made > in suitable lengths. And the male connectors are only 2.95mm tall and > 7mm wide (really only 6mm but the flanges on the cable connector are > 7mm) and that fits easily on both the rexcpm and the
Re: [M100] rexcpm battery
Let's give THIS a try... https://photos.app.goo.gl/JaSofoK2VotHzVco9 The connector is a standard JST-SH, and cables are available pre-made in both 2" for the 100 and 12" for the 102 or 200. There's even an off-the-shelf cable that has dupont sockets on one end, so you could connect to an unmodified rexcpm. So now we have 120 mah gnd connection polarized plug I think we're looking at at least 2 years of shelf life now. -- bkw On 9/12/23 15:05, Brian K. White wrote: I've done a little more testing. The GND path is backfeeding only via the RAM line. Nothing on /WR. (the only possible connections) I am measuring only 6.6uA on the RAM_RST line between the two boards with the boards outside of the 100. That is a little high but actually not too bad. I would hope for under 5uA ideally but 7uA is in the same class. At that rate, nominally the batteries should last over 11 months. And at that low current there shouldn't be much voltage drop across the diode on RAM_RST, and indeed there isn't. I only lose about 0.1v from the battery to RAM_RST. The bulk of the voltage drop is happening along the GND path, and things change a lot when the boards are installed in the 100 and the 100 provides a real GND connection between the two boards. With both boards outside of the 100, I currently measure (the batteries are draining as time goes on so these numbers all change a little each time I check something, so I'll keep saying the battery voltage along with whatever else I'm talking about) 2.99v between BATT- and BATT+ 2.91v between BATT- and RAM_RST (so practically no drop through the diode on the bus board) 2.76v between BATT- and CAP+ (it's not labelled CAP+, I just mean the positive side of the big cap) on the REXCPM (after giving it time to drain away the 5v from the 100) 2.26v between CAP- and CAP+ So there is not actually much drop from BATT+ along RAM_RST to REXCPM VCC, but a 0.5v loss along the GND-via-RAM path. Next, when I provide an actual GND connection between the boards by plugging them in to the 100 with the memory power switch turned off, the current draw actually increases slightly to 7.2uA (after an initial spike probably charging the cap), but the voltage drop decreases. The battery is currently at 2.99v and I get 2.72v at the cap instead of 2.26v While writing this the current along RAM_RST very gradually dropped further to 6.9uA. Maybe it will continue creeping down to settle at the same 6.6-6.7 eventually? This suggests that maybe the batteries would provide a worthwhile shelf-life extension after all, at least while installed in the machine which provides a real GND connection. 60mAh at 8uA works out to just over 10 months. (according to https://www.digikey.com/en/resources/conversion-calculators/conversion-calculator-battery-life) That's worth doing I think. Especially when added to the 100's own shelf-life, that should work out to maybe 1.5 years total. I still have to check the actual current drain at the battery. My current measurements above were just on the RAM_RST line from the bus board to the rexcpm, not at the battery itself. It's possible the battery is seeing more total current than just that through additional paths like reverse leakage through the schottkey from BATT+ to VBUS, and through RAM_RST into the 100. Anyway it looks a lot more promising than I thought at first. As for getting that same long shelf-life outside of the 100, one idea might be to mod the REXCPM to use a 4-pin JST-SH connector and use the same connector on the bus board, and a pre-made cable assembly. Those are very small connectors and would be difficult to crimp the female connectors on the cable by hand, but they sell those pre-made in suitable lengths. And the male connectors are only 2.95mm tall and 7mm wide (really only 6mm but the flanges on the cable connector are 7mm) and that fits easily on both the rexcpm and the bus board. On the rexcpm side it would have to be bodge wired to the 3 existing pins and the 4th pin to the negative side of the big cap, then secured with glue or mounting foam tape. On the bus board it's nothing. The connector footprint fits right where the existing 3-pin header is now. https://mou.sr/3K9FwRe https://mou.sr/44Pcako or https://www.digikey.com/short/7hb4cm4q https://www.digikey.com/short/wzj8ppjb And as an added bonus, the new cable is fully polarity keyed on both ends too. -- bkw
Re: [M100] rexcpm battery
I've done a little more testing. The GND path is backfeeding only via the RAM line. Nothing on /WR. (the only possible connections) I am measuring only 6.6uA on the RAM_RST line between the two boards with the boards outside of the 100. That is a little high but actually not too bad. I would hope for under 5uA ideally but 7uA is in the same class. At that rate, nominally the batteries should last over 11 months. And at that low current there shouldn't be much voltage drop across the diode on RAM_RST, and indeed there isn't. I only lose about 0.1v from the battery to RAM_RST. The bulk of the voltage drop is happening along the GND path, and things change a lot when the boards are installed in the 100 and the 100 provides a real GND connection between the two boards. With both boards outside of the 100, I currently measure (the batteries are draining as time goes on so these numbers all change a little each time I check something, so I'll keep saying the battery voltage along with whatever else I'm talking about) 2.99v between BATT- and BATT+ 2.91v between BATT- and RAM_RST (so practically no drop through the diode on the bus board) 2.76v between BATT- and CAP+ (it's not labelled CAP+, I just mean the positive side of the big cap) on the REXCPM (after giving it time to drain away the 5v from the 100) 2.26v between CAP- and CAP+ So there is not actually much drop from BATT+ along RAM_RST to REXCPM VCC, but a 0.5v loss along the GND-via-RAM path. Next, when I provide an actual GND connection between the boards by plugging them in to the 100 with the memory power switch turned off, the current draw actually increases slightly to 7.2uA (after an initial spike probably charging the cap), but the voltage drop decreases. The battery is currently at 2.99v and I get 2.72v at the cap instead of 2.26v While writing this the current along RAM_RST very gradually dropped further to 6.9uA. Maybe it will continue creeping down to settle at the same 6.6-6.7 eventually? This suggests that maybe the batteries would provide a worthwhile shelf-life extension after all, at least while installed in the machine which provides a real GND connection. 60mAh at 8uA works out to just over 10 months. (according to https://www.digikey.com/en/resources/conversion-calculators/conversion-calculator-battery-life) That's worth doing I think. Especially when added to the 100's own shelf-life, that should work out to maybe 1.5 years total. I still have to check the actual current drain at the battery. My current measurements above were just on the RAM_RST line from the bus board to the rexcpm, not at the battery itself. It's possible the battery is seeing more total current than just that through additional paths like reverse leakage through the schottkey from BATT+ to VBUS, and through RAM_RST into the 100. Anyway it looks a lot more promising than I thought at first. As for getting that same long shelf-life outside of the 100, one idea might be to mod the REXCPM to use a 4-pin JST-SH connector and use the same connector on the bus board, and a pre-made cable assembly. Those are very small connectors and would be difficult to crimp the female connectors on the cable by hand, but they sell those pre-made in suitable lengths. And the male connectors are only 2.95mm tall and 7mm wide (really only 6mm but the flanges on the cable connector are 7mm) and that fits easily on both the rexcpm and the bus board. On the rexcpm side it would have to be bodge wired to the 3 existing pins and the 4th pin to the negative side of the big cap, then secured with glue or mounting foam tape. On the bus board it's nothing. The connector footprint fits right where the existing 3-pin header is now. https://mou.sr/3K9FwRe https://mou.sr/44Pcako or https://www.digikey.com/short/7hb4cm4q https://www.digikey.com/short/wzj8ppjb And as an added bonus, the new cable is fully polarity keyed on both ends too. -- bkw On 9/11/23 15:47, Ken St. Cyr wrote: I love the idea, and would be keen to build one for myself if it could indeed extend the shelf life of my REXCPM memory to at least a few weeks. I tend to swap between my REXCPM and a standard REX, so even just having something to keep my REXCPM memory active for a few days while it’s out of the machine would save some trouble. I was thinking about doing a quick backup and restore using a TL866, but haven’t gotten around to tinkering with it. It doesn’t sound like this would work well with that big cap needing to be kept charged. It’d be great if there was a way to take it out of circuit… maybe a jumper that you could remove if you have the REXCPM UPS hooked up //Ken *From: *M100 on behalf of Brian K. White *Date: *Monday, September 11, 2023 at 2:28 AM *To: *m...@bitchin100.com *Subject: *[M100] rexcpm battery I thought I made the slickest thing. https://github.com/bkw77
Re: [M100] rexcpm battery
On 9/11/23 15:47, Ken St. Cyr wrote: I love the idea, and would be keen to build one for myself if it could indeed extend the shelf life of my REXCPM memory to at least a few weeks. I tend to swap between my REXCPM and a standard REX, so even just having something to keep my REXCPM memory active for a few days while it’s out of the machine would save some trouble. I was thinking about doing a quick backup and restore using a TL866, but haven’t gotten around to tinkering with it. It doesn’t sound like this would work well with that big cap needing to be kept charged. It’d be great if there was a way to take it out of circuit… maybe a jumper that you could remove if you have the REXCPM UPS hooked up I think the big cap is only a good thing. It's not draining the batteries, it's only helping them. It's providing the first 15 minutes of life for free, and then not hurting after that. It also provides up to 15 minutes of battery-change grace period, depending on the starting level. If you briefly power-on the main computer before changing the batteries, you'll get the full 15 minutes at least (a few more before the ram actually loses data, but 15 is safe). If you start with almost dead batteries and simply remove them, you still may get a whole minute or more depending on just how dead they were. Even 15 seconds is a lot if you prepare and have the new batteries unpackaged and ready to go first. The excess drain might possibly be because of the unintended ground path backfeeding though what are supposed to only be signal handling input or output gates in other components, instead of having an actual GND connection from the REXCPM to the battery negative. It could even be causing some ICs to no be disabled by their active-low /CE pins maybe. Or it might be the REXCPM is keeping more stuff alive than the sram, or maybe there are pullups or pulldowns that have lower than necessary values burning current at all times. Or, maybe my replacement sram chip burns more than the original*. These questions don't really have to be mysteries, they can all be answered by just looking and testing, I just haven't done that yet. * When I got my REXCPM, Steve didn't have any 4M chips, so I got a 2M unit and replaced the chip myself, and I don't remember off the top of my head exactly what it's quiescent current was. The "over 1 year" estimate is just based on the typical value for almost any sram since the early 90's of 5 uA or less. It just turned me on to get removable batteries into the available space right on the board and still be able to close the compartment cover. It does require flush-cutting all the solder posts and the battery holder tabs. But to me that is more straightforward, convenient, and repeatable than finnicky hack or otherwise specialized pin construction. Normally for "fake DIP legs" on a pcb intended to go into a dip socket, you want the pins to be as thin as possible to mimick an actual dip leg, so as not to stretch out flat leaf-style sockets. And often also want the pcb to be as low profile as possible with little or no shoulder/insulator between the contact leg and the pcb, to allow the most room for the components on top. But in this case the socket has machined round pin holes, and commodity machined round pin headers are perfectly fine in them, and the huge 4mm shoulder height is ok if all the parts can fit on the bottom of the pcb and leave the top clean and flat. A separate keeper board will be trivial so I'll add that to the repo shortly. So, even if we have to just live with the inefficient sleeping instead of solving it, a couple AA's or something will be cheap, easy, and last a long time. That will be better than nothing and simple to do. //Ken *From: *M100 on behalf of Brian K. White *Date: *Monday, September 11, 2023 at 2:28 AM *To: *m...@bitchin100.com *Subject: *[M100] rexcpm battery I thought I made the slickest thing. https://github.com/bkw777/REXCPM_UPS <https://github.com/bkw777/REXCPM_UPS> https://photos.app.goo.gl/i87E4wzimexCR3wL6 <https://photos.app.goo.gl/i87E4wzimexCR3wL6> I was able to get 60mah of battery onto the system bus interface board for REXCPM, and theoretically that should be able to keep the sram memory for at least a year, but it looks like it will only last about 2 possibly 3 days. So plan B is a separate "keeper". A separate thing with a much larger battery that you connect to the rexcpm when not in the computer. But that's only half of what I wanted. What I was really hoping for though was to have the battery built-in, so that when the 100 batteries die sitting on the shelf, the on-board battery is still there for at least another year. Both of these boards do work as merely ordinary rexcpm bus adapters, whether the batteries are installed or not. The 102/200 board also a
Re: [M100] rexcpm battery
I love the idea, and would be keen to build one for myself if it could indeed extend the shelf life of my REXCPM memory to at least a few weeks. I tend to swap between my REXCPM and a standard REX, so even just having something to keep my REXCPM memory active for a few days while it’s out of the machine would save some trouble. I was thinking about doing a quick backup and restore using a TL866, but haven’t gotten around to tinkering with it. It doesn’t sound like this would work well with that big cap needing to be kept charged. It’d be great if there was a way to take it out of circuit… maybe a jumper that you could remove if you have the REXCPM UPS hooked up //Ken From: M100 on behalf of Brian K. White Date: Monday, September 11, 2023 at 2:28 AM To: m...@bitchin100.com Subject: [M100] rexcpm battery I thought I made the slickest thing. https://github.com/bkw777/REXCPM_UPS https://photos.app.goo.gl/i87E4wzimexCR3wL6 I was able to get 60mah of battery onto the system bus interface board for REXCPM, and theoretically that should be able to keep the sram memory for at least a year, but it looks like it will only last about 2 possibly 3 days. So plan B is a separate "keeper". A separate thing with a much larger battery that you connect to the rexcpm when not in the computer. But that's only half of what I wanted. What I was really hoping for though was to have the battery built-in, so that when the 100 batteries die sitting on the shelf, the on-board battery is still there for at least another year. Both of these boards do work as merely ordinary rexcpm bus adapters, whether the batteries are installed or not. The 102/200 board also allows one to use a REXCPM on a 102 or 200 without soldering any wires or opening the case, although it does mean using an external wire which is not exactly robust. -- bkw
[M100] rexcpm battery
I thought I made the slickest thing. https://github.com/bkw777/REXCPM_UPS https://photos.app.goo.gl/i87E4wzimexCR3wL6 I was able to get 60mah of battery onto the system bus interface board for REXCPM, and theoretically that should be able to keep the sram memory for at least a year, but it looks like it will only last about 2 possibly 3 days. So plan B is a separate "keeper". A separate thing with a much larger battery that you connect to the rexcpm when not in the computer. But that's only half of what I wanted. What I was really hoping for though was to have the battery built-in, so that when the 100 batteries die sitting on the shelf, the on-board battery is still there for at least another year. Both of these boards do work as merely ordinary rexcpm bus adapters, whether the batteries are installed or not. The 102/200 board also allows one to use a REXCPM on a 102 or 200 without soldering any wires or opening the case, although it does mean using an external wire which is not exactly robust. -- bkw
