Owon ADS824A Battery Option Question

[TomC]

I recently bought a new ADS824A from a vendor at AliExpress. The item was on sale for an attractive price ($500), but the WiFi and Battery options were not available. Since I received it I've been able to add the WiFi/Bluetooth option, a wireless mouse, and a USB Drive via a compact USB hub (see attachment 1). However, although operating the scope from a battery is not hard to do, I haven't been able to hook up the Battery so that it behaves the way described in the manual, for example, display the battery status on the user interface as shown in attachment 2 (see item 2).

According to Owon, and vendors that offer the Battery option, this option has to be selected before the unit leaves the factory, and can not be added afterwards. But they say the same thing about the WiFi option, and all I had to do was to figure out which WiFi/Bluetooth dongle they were using and purchase one from Amazon for about $7.00. So I'm hoping the same thing is true for the battery. However, so far, I've been hitting a brick wall trying to figure this out. I opened the scope to see what clues I could gather by looking at components, connectors, etc. (see attachments 3 & 4). I believe the battery should be connected to CN16. I've done Voltage, Continuity, and Live tests involving this connector. The information I've gathered so far is compiled in attachment 5. However, so far, the only connection configuration that has allowed me to run the scope on battery power is connecting the battery in parallel with the Power Adapter. Unfortunately, this is a less-than-ideal configuration since the battery may always get a trickle charge depending on the PCM/BMS capabilities. In addition, the scope doesn't show the battery status because it can't tell whether the power is coming from a battery or the Power Adapter.

If anyone has an idea of how to properly connect this thing, or better yet, owns a scope with the battery option and can take some pictures of the battery and the connection, I would be very grateful if you post it here.

[runcyclexcski]

Have you figured it out? I've been wondering about the noise of this unit, as seen at the fines scale (when plugged BNC-BNC). Not looking to use all the 4 channels, or a battery, or, for that matter, its screen, as long as I can record and display the data in semi-real time on a laptop. Hoping for an internal noise RMS of 50 uV or less (sampling a signal at 10 kHz with 100 uV RMS).

[TomC]

I figured out that the battery is supposed to plug to CN4 after I had reversed engineered a small section (see attachment). But when I hooked up a 12V battery to test my theory, POOF! Something got smoked immediately, the scope wouldn't power on anymore. The fact that CN4 wasn't populated with a connector and that the heatsink was partially covering it should have been a red flag, but I was confident I've found the secret. Except, evidently, I didn't know all the details! After carefully looking with a digital microscope, I found that both U609 and U610 had been cooked (visible thermal holes).

U610 had no markings and the markings on U609 didn't lead anywhere but given its location in the circuit I strongly suspected a MOSFET. Anyway, after weeks of further reverse engineering I figured out a way to repair the scope. I'll post more details once I complete the job.

In conclusion, I should have paid more attention to Owon's warning regarding the battery option. Although I still think adding the option is doable, I didn't have all the details when I attempted it. However, I no longer think that the battery option offered by Owon is the best path to battery operation. Instead, I'm currently pursuing an alternative option using an external M12 battery.

[TomC]

I've got my scope fixed and have implemented an alternative battery system that allows practically unlimited battery operation by using easily swapped M12 batteries. This project took several weeks and involved three major efforts to achieve:

1. Reverse engineer the Power Circuit. I was hoping this would point the way to a fix as well as reveal the battery option secrets.
2. Design and implement a fix. Step 1 revealed that merely replacing the damaged components was probably unfeasible, so I used an alternative solution using different parts.
3. Decide on a system to use battery power and implement it. First I wanted the Owon battery option, but it's 2 hour limit made me seek a different path using M12 batteries.

While working on this project I took many photos and generated a number of documents. My plan is to share the details in three posts, basically following the three major efforts described above.

Reverse Engineering the Power Circuit
=====================================

Everything started when I thought I've discovered the Battery Option secrets and hooked up a battery to CN4. POOF, attachments 1 and 2 were the result of this experiment.

Prior to the mishap I had tried to positively identify U609 and U610. The markings on U609 didn't help, but the circuit topology I had diagramed at the time had me convinced it was a high-power P-Channel MOSFET used to Charge/Discharge the battery. That made sense for scopes with the battery option, but my scope didn't have that option, and U609 blew like a fuse.

I couldn't find any markings on U610, at the time I thought it was a battery charger IC. When it blew, I removed it and found markings on the bottom (see attachment 3). These didn't help. After reverse engineering the Power Circuit, I realized it functions as an ON/OFF Pushbutton Controller and possibly a battery status monitor. Now, based on the circuit topology and behavior of the scope before the mishap, I believe that U610 is a custom MCU or ASIC. An IC I probably wouldn't be able to source, especially since custom flashing would be required.

Finally, I realized that I probably would need a more detailed schematic, so I decided to trace the component in the vicinity of U609 and U610 and draw my own. Attachments 4 and 5 show the components I traced. This is a multilayer board, so a few components connected to a via that didn't go anywhere on the surface layers. I didn't try to trace the destination of those.

After I finished the Schematic, I had a much better idea of how everything worked and put together two block diagrams (see attachment 6). These represent my concept of the Power Circuit of a scope with the battery option and a scope without the battery option.

Attachments 7 & 8 are the actual reverse engineered schematic of my scope, and my description of the different ICs based on the topology of the circuit, the scope behavior observed prior to the mishap, and datasheets of identical or similar components I've looked at.

Attachment 9 is a PDF version of Attachments 6,7, and 8.

Edit: There were some items missing on attachments 7 - 9.

[TomC]

Repairing the Power Circuit
===========================

After considering several ON/OFF Pushbutton Controllers, I decided to use an LTC2954-1 (see attachment 9) to replace the functionality of U610. This IC features a wide voltage operating range 2.7V - 26.4V which is ideal, since it will be hooked up to the 12V rail. It also has adjustable on and off timers that will allow me to conserve the original Power Switch 1-2 second delay behavior. Finally, it also has a power kill input activated by an accurate 0.6V threshold comparator. The plan is to use this feature to implement a battery over-discharge detection circuit. This way, when an M12 battery is used as the power source, the scope will be automatically turned OFF if over-discharge is detected.

Next, I had to decide how to incorporate this new part into the existing PCB. The reverse engineered schematic revealed that the new IC could be easily added via a separate daughterboard with minimal alterations to the existing PCB. The daughterboard could use CN6 to connect to SW1, the Power Switch, CN13 to connect to the gate of U606, the Solid-State scope ON/OFF switch, and CN16 to connect to supply power. These are all the connections required to implement the circuit.

The LTC2954-1 Datasheet (attachment 9) was the main source of information used to design the daughterboard circuit (attachment 1). The only connections to the motherboard PCB are the two plugs (CN6 and CN13) and the connection to CN16 pin 4. R1 is optional but provides some additional protection to the IC. R6 was necessary to overcome parasitic currents on the motherboard PCB that made the internal PB' pullup insufficient. Without it SW1 failed to function once in a while. C1 and C2 set the ON and OFF timers to a 1 second delay. R2, R5, and R4 are a voltage divider that activates the KILL' input when VIN drops to around 9V. This is the battery over-discharge detection circuit intended to turn OFF the scope to prevent further discharge. R3 is a pullup for the open collector EN (enable) output and Q1 takes the place of Q72 in the original circuit.

The daughterboard circuit went through several iterations before settling on the version on attachment 1. When the two LTC2954-1 ICs I ordered arrived, I discovered that I didn't have the correct breakout PCB. The IC has a 0.65 MM pitch, and the closest I had was 0.5MM pitch breakout PCB. I ordered the correct breakout right away but couldn't wait to breadboard the circuit and try it out, so after some extreme gymnastics I soldered one of the ICs to the 0.5MM breakout (attachment 2), not pretty, but allowed me to test the concept (see the preliminary breadboard test, attachment 3).

Before hooking up the breadboard to the motherboard for final testing, I had to add headers to CN6 and CN13 and remove the damaged components from the motherboard to prevent interference with the new circuit. First I removed U610 (see attachment 4). This confirmed that pins 2 and 3 had no connection, something I suspected when reverse engineering the circuit, but couldn't confirm until now. U609 is in a very tight place, and since it's open anyway, I decided to leave it alone, so although it's not physically removed, it's electrically removed. The breadboard final test is illustrated on attachment 5, and everything worked as expected.

Attachment 6 shows the LTC2954-1 connected to the correct breakout PCB.

Attachment 7 shows the daughterboard assembled on a prototype PCB and connected to the scope via CN6, CN13, and CN16 as previously discussed. The extra 2 pins on the 3-pin connector at the top is something I added at the last minute. I plan to use them to monitor the battery voltage at the scope end when I implement battery power. The daughterboard and the connector are attached to the heatsink with industrial grade Velcro.

Attachment 8 is the complete schematic of the Power Circuit after the modifications. It shows the removed components, and the daughter board circuit on a container section at the bottom right.

[TomC]

Implementing Battery Operation
==============================

After integrating the daughterboard into my scope, it became fully operational and just as good as before the mishap. To implement battery operation, I purchased the following items from Amazon:

-----------------------------------------------------------------------------------------------------------------------------
1. A CaliHutt M12 Power Supply & Charger (see attachments 1 and 5). https://www.amazon.com/%2525E3%252580%252590Multi-Function%2525E3%252580%252591-CaliHutt-Replacement-49-24-2310-48-59-1201/dp/B08VS64H52

The CaliHutt M12 Power Supply & Charger listing claimed it could deliver up to 3A from its DC port. Most of the other devices I researched were listed as only capable of delivering 1.5A. However, in reality, the DC port shuts-down at about 1.5A. I don't think I can source a similar device with better than 1.5A capability. So I decided to add a second DC port connected directly to the battery terminals. On attachment 5 you can see the DC port I added; It's level with the top of the front label. If you could look inside, it's perfectly aligned with the battery connector, inside, there was just enough room to connect the port and battery connector terminals together.

Since I didn't alter any of the internal circuitry, all the advertised device features still work fine. You can use it to charge an M12 battery by connecting a 5V source to the Micro USB port (it internally boosts the 5V to about 12.3V to perform this function). However, it takes much longer than standard M12 chargers, about 8 hours for the batteries I bought, I'll probably won't be using this feature. There is also a USB-A port that can be used as a 5V power bank. And finally, you can use it to check the M12 battery capacity, press the "B" switch and it will display it for several seconds via the 4 LEDs, just like other M12 tools.

2. A set of four 1-foot Auropath DC 5521 M to USB-C Cables (see attachment 2 and 5). https://www.amazon.com/dp/B0FH2FGYLW/ref=cm_cr_arp_d_btm?ie=UTF8&th=1

These are generic dumb cables with just two 22AWG wires connecting the DC port to the USB-C connector. They work fine unaltered. However, I wanted to connect a Battery Capacity & Voltage Gauge (item 3 below) to the same DC port, and I also wanted a section of the sheath removed so I could connect a current probe. I ended up cutting the ends of one of the cables off so I could put together the cable illustrated on attachment 5. Most of it is 16AWG, for less loss, and I have all the features I wanted.

3. A pair of DFCROMI Battery Capacity & Voltage Gauges (see attachment 3). https://www.amazon.com/dp/B0D4YQ6L1Q?ref_=ppx_printOD_title_dt_b_fed_asin_title_0_0&th=1

These are programable and designed to support different voltages and battery chemistries (intended mostly for Golf carts). I consider these optional, since item 1 also has a battery gauge, but I wanted the extra capabilities. The battery capacity is always pictorially shown via a 7-segment battery icon, the right-hand side of the LCD can show the voltage or the battery capacity from 100% to 0%. There is a choice of one of two discharge curves to control the battery capacity display. One of them (S5 On) matches the batteries I bought quite well.

Edit: Added attachment 9, the manual for this device. This is a scan of my manual, as far as I know not available elsewhere in the web.

4. A pair of Nonbliep 9.0Ah M12 Battery Replacements (really 5.0Ah) (see attachment 4). https://www.amazon.com/dp/B0G352BC8D?ref_=ppx_printOD_title_dt_b_fed_asin_title_0_0

These are Milwaukee knockoffs, I have many of them in different sizes, mostly work fine for the price, but the battery capacity claimed by the manufacturers is seldom accurate. After testing this set of batteries, I find that their true battery capacity is right around 5Ah (see attachments 7 and 8 ).

-------------------------------------------------------------------------------------------------------------------------------

To test my new Battery Operation Setup, I performed a test with one of my new batteries where I tracked the Voltage, Amps, Watts, Capacity with S5=ON, Capacity with S5=OFF, Timer time (at 5-minute intervals), etc. (see attachments 7 and 8 for details). This test lasted 1 hour and 40 minutes, which is the length of time each of my new batteries can run the scope. I used this information to calculate the true Amp Hours of my batteries as shown in the attachments. Is not 9Ah as claimed, but it's a decent 5Ah. As can be seen from the table, the S5 = ON setting tracks the actual battery discharge (% Left) quite well. So am happy with the DFCROMI Battery Capacity & Voltage Gauges. One other thing that turned out to be a great feature, that I wasn't even expecting, is that the capacity reading also very closely reflects the number of minutes left before the battery reaches 9.2V and the scope gets turned OFF automatically.

Attachment 5 shows the tools I used for the test.

Attachment 6 is a screenshot of the scope showing the current probe reading 15 minutes into the test. It also shows the reading of the DVM feature available on the ADS scope series.

[TomC]

Have you figured it out? I've been wondering about the noise of this unit, as seen at the fines scale (when plugged BNC-BNC). Not looking to use all the 4 channels, or a battery, or, for that matter, its screen, as long as I can record and display the data in semi-real time on a laptop. Hoping for an internal noise RMS of 50 uV or less (sampling a signal at 10 kHz with 100 uV RMS).

Hi, I tried to reproduce the scenario you are looking for; however, the AWG I'm using only goes down to 1mVrms.

Attachment 1 shows the physical setup, BNC to BNC, there is a 50 Feed-through at the scope end.
Attachments 2 & 3 show the CH1 setup, notice the 20MHz bandpass limit, without that (All) the noise is as big as the signal.
Attachments 4 & 5 show the signal in Run and Single, with the scope powered by my M12 battery.
Attachments 6 & 7 show the signal in Run and Single, with the scope powered by the Power Adapter that shipped with it.

Evidently, the Power Adapter adds quite a bit of noise. Thanks for asking the question, I didn't realize powering the scope from the battery made this much difference.