A multi power source capable power supply for PSION Organiser II (all family) devices with power failure memory saver. The design reduces risk of internal device memory loss either due to external power interruption or due to Lithium 9 Volt surrogate battery discharge characteristics (limited if any warning). By providing sufficient standby power for longer term internal memory retention and specifically not operation. Permitting hours / days for recharge of main power battery or months of power down whilst preserving the internal RAM contents. The design (should) fits a classic PSION Organiser II Top Slot Case with minor modification for the USB connection (See all notes).
All the files are required for a complete assembly.
- The default repository format is STEP ( ISO 10303 ) due to its high fidelity.
{ STL files are surface geometry as opposed to solid model representations, often containing export processing artifacts }.
- The default repository format for PCB files is GERBER .
FreeCAD may prove suitable for viewing, handling and, if desired modifying STEP files.
KiCad may prove suitable for viewing GERBER files.
Models might be viewed on line using the following links. Noting display of any mesh lines is a feature of the renderer and are not contained in the source STEP model. It is strongly recommended to use the source STEP file where ever possible, as any export process will introduce artifacts.
The design presents a multi source capable power system for the Top Slot of a PSION Organiser II family of devices. Along with power source capability from the classic Barrel jack 10.4 Volt power supply and a USB-C connection, a memory protection feature is included to protect the devices internal RAM memory. In the event of power loss / removal, internal RAM memory would be maintained for up to 12 months together with time keeping functionality.
The PCB dimensions (38.0 x 43.6 mm) are ideally suited to low cost manufacture. The 4 layer PCB design provides low cost of implementation. Whilst the PCB space claim is intended to be suitable for a classic Top Slot Case with minor modification (See Considerations) a potential custom case is demonstrated in the repository. Construction is intentionally a combination of surface mount and through hole PCB for implementation flexibility.
The through hole technology allows straight forward assembly of Barrel Jack and memory protection feature. Whilst the more advanced surface mount features are required for the USB-C power source. The design will operate as a memory saver device without the surface mount components, though this design might be preferred in such circumstances.
A minimum configuration comprises:
- a 4 layer PCB
- 1N4148 diode
- 2 battery clips
- 2 x 8 way Right angle header (~8 mm engagement length)
- 2 off CR2032 batteries or similar
Device power is prioritised to select from the various sources in the following order of priority:
- Classic Barrel Jack
- USB-C
- Internal 9 Volt source
- Memory protection system
A USB-C connection is provided using basic 5 Volt capability only to provide maximum compatibility and obviate endpoint power management systems. A high quality buck boost power system with additional over voltage protection. Provides an additional feature to protect against potential USB-C Power Delivery over voltage scenarios.
A classic Barrel Jack (5.5 mm outer, 2.0 mm pin) connector for the ~10.4 Volt 175 milli Amp centre positive pin power supply is located to be compatible with the classic Top Slot Case. The Barrel Jack connections are brought to 2 header pins on the left of the PCB to permit alternate power sources to be readily accommodated. (This is the same connector used on the classic Arduino Uno and Mega boards).
A low voltage 'secondary' power source is held on the PCB to allow for longer term retention when the main power supply or internal PP3 battery become depleted. Switching is automatic via diodes and the purposefully lower secondary battery level restricts usage to notification and memory backup only.
Details on testing and backup battery performance are located in the Top Slot Battery Saver Basic repository.
The device is transparent to the Organiser's Operating system and requires no software to operate.
It is preferable to avoid connecting the unit with no power source available e.g. absence of 9 Volt internal battery or power sourced via the units power connections i.e. 10.4 Volt Barrel Jack connection or USB-C connection. So as to avoid unnecessarily discharging the backup batteries. By providing a main power source on connection of the unit. The internal capacitors of the Organiser are 'pre-charged' from the main power source rather than the backup battery system, thus ensuring maximum life and thus protection.
Use of a battery bank to power the Organiser will require a 'Keep-Alive' device (effectively a 555 timer plus low value resistance). The power drawn by the Organisers normal use is far too low for battery banks typical auto shut down features.
Remove batteries from unused devices. Also see here.
All measurements are taken at a nominal temperature of 21 C against unit 4w2xf_0003.
Typical voltage output levels from the devices boost converter with an input voltage of 5.13 Volts nominal was measured :
Load | V Boost | Vb | I(Vb) mA |
---|---|---|---|
No Load | 10.086 | 9.789 | 0.000 |
Standby | 10.085 | 9.654 | 0.000019 |
Nominal | 10.083 | 9.378 | 0.004 |
Full Load | 10.141 | 9.355 | 0.10142 |
Conditions
- No main 9 Volt battery fitted
- Standby refers to unit powered off in sleep mode
- Nominal refers to a program running an updating counter on the display at the maximum iteration rate
- Full load is measured against a passive load frame
- Measurements taken after 1 hr unless otherwise noted
- Measurement system characteristics - Input resistance 10 MOhm, 0.1 mV resolution
To simulate a defective or over load scenario (test performed with no Organiser device connected). The Vb line was connected to zero volts for a period of 5 seconds with 10 second intervals over 5 cycles. No degradation was observed and regulation voltage was measured at 10.084 Volts 30 seconds following the last cycle.
Note: Longer short circuit intervals typically (should) cause USB upstream (e.g. polyfuse) protection to activate.
The unit was tested for a minimum of 1 hour in both standby and operational modes against the following power sources:
Type | Model | Manufacurer | Harness / Connection | Comment |
---|---|---|---|---|
PowerPort III | A2033 | Anker | 1 m cable (Anker) | |
KOPPLA Charger | E1408 | Ikea | 1 m cable (Anker) | (3 port USB Charger) |
QI charger & USB | E1401 | Ikea | 1 m cable (Anker) | |
Charger | S005BBB0500100 | Samsung | 1 m cable (Anker) | 5 Volt I.T.E Power Supply (5 V, 1 A) |
Mobile phone | Galaxy S23 | Samsung | 1 m cable (Anker) | |
Mobile phone | Galaxy S20 | Samsung | 1 m cable (Anker) | |
Personal Computer | Generic | Asus | 3 m USB ext. cable (Amazon Basics) | |
Personal Computer | Generic | Asus | 1 m cable (Anker) | |
Portable computer | DELL | DELL | 1 m cable (Anker) |
- Over 100 connector mating cycles were performed of the USB-C connector.
- Connector pull testing aligned to the connectors principle axis was performed to ensure the connector robustly retained on the PCB.
No degradation observed in soldering and attachment of the USB C connector to the PCB.
Note: It is contra indicated to pull a cable from the connector(s) without applying a counter acting force to the Top Slot case.
Connector was tested in both orientations to ensure power supplied regardless of USB C connector orientation
Performance of the voltage via the barrel jack is exactly as per the classic power supply configuration due to use of the same 1N4001 diode.
The forward voltage drop across the 1N4148 diode was measured at 0.4264 Volts whilst the unit in standby mode. This will vary slightly due to clock interrupt activation.
Main power was applied via USB or internal Battery. Content was stored on the devices internal memory and all main power sources removed. A period of 10 minutes was permitted to elapse with no main power source applied. A main power source (USB or Battery) was reconnected and memory contents verified. 32 cycles were tested with no loss of data identifed.
1, 12 and 24 hour main power source removal tests were performed, with no loss of data identifed.
Data verification comprised:
- Time keeping function maintained
- Program storage function maintained (program(s) run following power restoration).
All combinations of incorrect CR2032 battery insertion were tested. No measurable reverse voltage was applied. No backup power was supplied. No degradation identified.
Single battery insertion. No measurable voltage was applied. No degradation identified.
Operation returned on correct insertion of batteries.
Reverse biased leakage current from the on-board battery system into the boost convertor via the S1G diode was measured at ~15 nA @ 21 C. Though less than 100th of the current required for battery backup of the device and in such conditions can be ignored. Where a system is not in use for extended periods this current will over extended periods discharge the on board batteries. Hence, it is recommended to remove batteries from a unit which is in long term storage. See also here concerning batteries.
Thermal evaluation of the device under full load (100 mA) conditions was performed to assess the temperature rise of the regulator and diode of the USB boost voltage regulator system. The input USB voltage was 4.9610 Volts at the PCB (5.250 Volts no load - 3 m USB cable), 10.118 Volts from the boost regulator output with the SVB voltage was 9.332 Volts presented to the load bank. The USB extension lead selected, so as to simulate a non-ideal scenario.
Following 2 hours of load, a maximum thermal rise of 10 C at 25 C room ambient was observed. The image was captured at this point, noting that the metal components surface finish alias temperature indication, as shown by the battery and holder indicating excessively differing temperatures. The image is for indication only, with measurements taken using a 0.8 mm dia. welded bead thermocouple.
As anticipated the regulator and diode are the heat sources, the thermal image shows good heat soak into the PCB and additionally into the USB connector and its over mould.
Voltage regulators output voltage regulation was found to be better than 0.3% between no load and full load. Effective voltage regulation at SVB pin was 0.5%.
Near Field H Probe indication of the Rev 0.1 unit confirmed expectations that under typical operating conditions the units emissions were low / undetectable (i.e. orders of magnitude lower than a subsequently removed co-located Arduino Nano). At full load, the magnetics due to current flow gave rise to a characteristic ~150 MHz trace shown. Principally located at the boost convertor inductor and boost convertor output filter capacitor. A future refined iteration will seek to improve the full load performance by selecting a shielded inductor and additionally targeting the output filters performance during full load. At ~ 5 cm from the device no appreciable H-Field signal was evident. Note the trace recorded with the device connected to a passive load frame.
In some aspects, especially the lower frequency band, the performance was better than an EBL 5400 mWh battery (6F22 9V boost converter type) under similar load conditions and particularly under step load change conditions typical of the type occurring during EPROM programming!
The H Field probe was maintained parallel to the top of PCB for assessment repeatability, this may not represent the highest amplitude. The lower side of the PCB exhibited levels ~ 1/3 of those on the upper side.
Durability testing comprised over 168 hours use under typical user load cycling e.g. standby ~ 85%, programming EPROM's ~ 1%, running programs ~ 14% etc. and was performed following all other testing. No issues were identified.
When running at higher current loads these are outside the efficient normal operating range of the device. The design is intended for typical use, i.e. which does not involve continuous high current loads. The nominal 20 mA load current is more than sufficient for the majority of Organiser use cases. The relatively short period EPROM programming loads can typically be readily accepted with the reduced efficiency of the voltage conversion, which under 100 mA load can fall to as low as 80% (from 90%). Use of high quality capacitors is particularly recommended to accommodate ripple current typically arising from EPROM programming. A typical duty cycle might be of the order of 10%, though 24 hour (room temperature) burn in testing has been completed in order to demonstrate design robustness.
Models or files makes no accommodation for manufacturing tolerances, process or material - see Notes below.
The original male 8 way 2 row top slot header style connector (marked MXS 70224) may have included a custom pin support moulding. Data from a parts list kindly indicated in this hardware forum seemed to confirm this theory. Readily available 8 way 2 row header connectors tend to have smaller pin support mouldings. The effect of using pin headers with smaller moulding is to permit the PCB to displace vertically in the slot guide channel, resulting in poor alignment with the mating female connector and potential insertion difficulty. There are a number of mitigations, such as changing the height of the male header connector in the PCB and adding material to support the top of a smaller male header pin support moulding. The precise accommodation will depend on your selected pin header moulding, also see here.
Connecting any device which has not undergone thorough testing, will lead to irreversible degradation!
See Organiser 2 Hardware forum, though see note below first.
Such a system does not obviate the need for standard good practice backup procedures.
Assessment for any fitness for purpose is entirely the users responsibility. Use is entirely at the users own risk. Any degradation to hardware, data loss or corruption is the responsibility of the user.
All information is For Indication only. No association, affiliation, recommendation, suitability, fitness for purpose should be assumed or is implied. Registered trademarks are owned by their respective registrants.