OWON OWH9811 (1 phase) and OWH9830 (3 phase) AC Power meter

[flash2b]

There is a new AC Power meter available from OWON, see: https://www.owon.com.hk/products_owh9800_series_power_meter

Quoting from the product page:

• Two-channel design:
  CH1 (rear panel): AC input – Supports 3–600V / 0–20A RMS, with an accuracy of ±0.5%
  CH2 (front panel): AC/DC input – Supports 0–1000V / 0–20A, with an accuracy of ±0.5%
• Supports single-phase and three-phase measurements, with 16 common parameters available for measurement
• Harmonic measurement and analysis functionality, capable of analyzing up to 63rd harmonics
• Combination of input and output power information for both channels, with automatic calculation of conversion efficiency
• Trend waveform display, providing visual representation of dynamic parameter changes for easy debugging and analysis
• 3.95” TFT LCD screen with clear, graphical display of multiple parameters, offering intuitive operation
• Standard RS232 interfaces for remote monitoring and data logging, compatible with SCPI and Modbus protocols

Manuals are found here: https://www.owon.com.hk/download.asp?category=Digital%20Power%20Meter&&series=&model=&SortTag=

I have pulled the trigger and bought the Owon OWH9811 1 phase version which just arrived.
 
The built quality of the unit is great. The specified performance and accuracy is quite impressive however the bandwidth is limited.

The unit came equipped with accessories as a test lead with clips and but not the promised RS232 cable  . It does not support USB CDC so that is a pity but not a real problem.

Please note that on the picture below it is just the unit powered up out-of-the box.





I will be creating a device definition for this series to support automated testing/logging with TestController, since currently a driver does not exist.

Overall I am very happy with my new Owon OWH9811.

[flash2b]

First print sceeen, directly from the device...

(just noise from the inputs... no signal connected)

[Gertjan]

Interesting that this new instrument only offers the somewhat outdated  RS-232 for remote control. (No USB or LAN).
It makes me think that this might be an already existing instrument, that is now rebranded by Owon. Just like the Owon LCR20xx LCR meters are made by Anbai/Applent.

Nice screenshot.   Is it possible to retrieve screenshots over RS-232?

regards, Gertjan.

[flash2b]

I do not think is is a rebrand. Looking at the Owon XDM2041 it is the same form factor and positions on the back for the connections are the same.

Well the XDM3041 and XMD3051 do have USB + Lan, but consider that on the back of the OWH9830 there is not much space and also all connections carry high voltages / high currents.

There is no SCPI/Modbus command to retrieve the screenshot over RS232, not a big problem not many devices offer this functionality.

[kripton2035]

why is there an usb symbol on the screen, and no usb plug on the device only rs232

[smk]

why is there an usb symbol on the screen, and no usb plug on the device only rs232

It has a USB host interface on the front panel for saving screenshots or measurement data .

First print sceeen, directly from the device...

Does the instrument save the screenshots as JPG or did you convert it? I would prefer PNG or BMP.

Anyway, a very interesting instrument, just ordered a OWH9811 from Banggood.

[flash2b]

The output is BMP but that format cannot be uploaded to this forum, so I changed it into JPG.

[smk]

Thenks for the prompt response. JPG is a lossy format, good for photos but less ideal for detailed or text-heavy screen captures. The forum accepts PNG with lossless compression for screenshots.

 

[Martin72]

https://www.owontechnology.eu/product/21592622/owon-owh9811-1-phase-3-600v-0-20a-rms-power-meter

Only 247€... What's the catch?
A little further down, you can see the technical specifications, and that's why it's so cheap.
But it will almost certainly be better than the usual power meters that you can plug directly into the power outlet.

[Gertjan]

Only 247€... What's the catch?
A little further down, you can see the technical specifications, and that's why it's so cheap.

Hi Martin72,

As I am not very familiar with Power Meter specifications, could you expand on this? Where is the catch?

regards, Gertjan.

[smk]

Only 247€... What's the catch?
A little further down, you can see the technical specifications, and that's why it's so cheap.
But it will almost certainly be better than the usual power meters that you can plug directly into the power outlet.

247€ is the price without VAT at owontechnology.eu.
And it is not only one power meter, it is two power meters in one case, 230V AC primary (here) and AC/DC secondery. I am not aware of any other budget power meter that can do these measurements simultaneously to calculate transformer or DC power supply efficiency.

[flash2b]

I mentioned it before, the bandwidth is limited if you refer to the Operating frequency and Harmonic analysis in the spec sheet.

But it is still a very good deal imho.

[flash2b]

I see that the first batch of Owon OWH9811's is now sold out on Banggood, so I hope there will be some new users soon that exchange experiences with this device.

[Martin72]

Hi Martin72,

As I am not very familiar with Power Meter specifications, could you expand on this? Where is the catch?

regards, Gertjan.

Hi Gertjan,
First of all, I mean this in general terms and regardless of the price.
You can't expect more for the price, understandably.
The accuracy is somewhat poor at 0.4% of the measured value + 0.1% of the range plus 1 digit of the display.
The bandwidth is also low at 2.8 kHz, which is naturally reflected in the usable measuring range, which is specified as 40-70 Hz.
Finally, I don't really like the measurement repeat time of min. 100 ms.
Interestingly, this is highlighted as a key feature for dynamic measurements.
At a mains frequency of 50Hz, you already have 5 periods of blind time—a little long for dynamic load jumps or to correctly detect an inrush current.
The data interface is also somewhat outdated.
As I said, that's just a general statement; given the price, you can't expect miracles—and I'm afraid it won't stop me from buying a device to test, either.

[Gertjan]

Hi Martin,

Thank you for your detailed, and well-considered answer.
Indeed it is "you get what you pay for".

But still, when confining the use cases to DC and mains related, this meter seems to be very useful.
High accuracy is not necessary in most power related measurements. Missing short dips in the mains might be a problem. But probably that problem is only solved by much more expensive power meters.

All in all, I think one gets a lot of value for the money. And at this price point there is not much choice
I am looking forward to your experiences in actually using this meter.

regards, Gertjan.

[flash2b]

Well Martin72 I hope it will arrive soon. Mine took 2 weeks, but I ordered before there was stock (I used Banggood).

[flash2b]

I had some troubles connecting a RS232-USB adapter to the back of the OWH9811. It is located so close to the corner where the feet are elevated so the adapter did not fit. Luckily I had a Male-Female "Gender" changer. The RS232 do not need a null modem and will accept a male plug directly.



I played with the SCPI.  I could *IDN? the meter quickly from TestController since the response is conforming to the standard. I used the highest speed which is 115200 baud using a FT232RL USB Serial adapter. It also supports Modbus, but that specification is not available which is a pity.

One observation already, the device needs to be manually put in remote by pressing the button <local> to respond properly on remote commands over RS232. There is no SCPI command to do so and it does not automatically switch to remote after responding to SCPI command like most meters do.

[Martin72]

Hi,

It also supports Modbus, but that specification is not available which is a pity.

I am interested in Modbus, as I have had to deal with it more intensively lately.
That's why I wrote to Owon earlier and asked for a corresponding data interface protocol where you can see which registers, functions, and addresses are used.

[flash2b]

I believe also that retreiving all measurements through Modbus will be the quickest execution.

My last 5 definitions for TestController I have created for devices using the Modbus protocol and once you know which registers and double-registers you need to poll and what value divider/multiplier they need, it is straight forward. I can be become a bit unclear since the registers are numbered opposed to names with SCPI. (I did not us a define kind statements).

Hope they will respond to your query.

Also the Integration mode (counting energy, Wh) is not supported with SCPI. Only the Normal and Harmonic modes are.

[Martin72]

Hi,

Very quick response from Owon, I'm pleasantly surprised...

[smk]

Some time ago I received my Owon OWH9811. It took some weeks to arrive because I had not seen that the default shipping method to Germany was to via European Railway Direct Mail. I have only done very few measurements but I already took it apart to make some photos.

[1] After removing the front stand and the two screws hidden underneath and the four screws at the back the chassis with all parts can be sled out of the housing.

[2] On the left side is the power supply. At the back is the mains connector with voltage selector and a fuse holder and at the front a small PCB with the power switch. In between is the transformer with three separate sets of secondary windings.
The main PCB is divided in four parts isolated from each other and only connected through some digital isolators and photocouplers.
The main part contains the AC measurement channel(s) for the inputs at the back of the instrument. There is only one current transformer installed but all 3 voltage divider strings are populated.
In the front left, marked in cyan in my photo, are the voltage supply and some connection devices for the front panel.
In the front right, marked in magenta in my photo, is the AC+DC measurement circuit.
In the back right corner is the serial port which is connected to the front panel connection part but has some additional isolation of the signal lines and the power supply.

[3] The back side of the front panel.

[4] The DC measurement part with the current shunt, voltage divider and ADC in the bottom right and a part of the main circuit with the micro controller and the AC power measurement IC in the upper part of the image.

[5] The current transformer and the 3 voltage dividers and the range switching circuits.

[6] The micro controller on the main part of the PCB.
[7] The three phase AC power measurement IC.
[8] The ADC for the AC+DC voltage and current channel.

[9] The AC current binding post at the back are very cheaply made and are too short for 19mm long banana plugs.

[10] The front stand, even when not used, is higher than the bumpers at the corners of the case.

[Martin72]

The black current transformer isn't so bad after all:

https://xonstorage.z8.web.core.windows.net/pdf/hoptech_hct204kfh_apr22_xonlink.pdf

0.1% accuracy, 0.05% linearity, bandwidth up to 10 kHz...

Judging by the circuit board, it would be easy to convert to a 3-phase system, at least from a hardware standpoint.

[smk]

Judging by the circuit board, it would be easy to convert to a 3-phase system, at least from a hardware standpoint.

Theoretically that could work but without the proper firmware and calibration process it is impossible.
As both models are fairly priced it makes no sense to try it.

While I don't consider the OWH9811 to be a high precision instrument I think it is useful and quite well built in most parts.
It is built around two readily available power meter ICs,  a micro controller and a front panel with a display.

The AC measurement part utilizes an Analog Devices ADE7880ACPZ polyphase multifunction energy metering IC with harmonic monitoring and one or three current transformers.

The AC+DC channel utilizes a TI ADS131M02 two-channel, 24-bit, 64-kSPS, simultaneous-sampling, delta-sigma ADC and a current shunt.

No part markings are sanded off and all parts look genuine.

Some other parts on the PCB are:
GigaDevice GD32F303CGT6 ARM Cortex-M4 microcontroller
NXP 74HC4052D Dual 4-channel analog multiplexers/demultiplexers
TI ISO6762 Six-Channel Digital Isolators
EL817C Photocouplers

At the serial port are:
TI ISO6721B dual channel digital isolator
3PEAK TPT3232E-SO3R RS232 transceiver
Dexu Electronics F0303S-1WM DC-DC converter
Two unpopulated SO-8 footprints probably are related to an optional RS485 interface.

[Martin72]

Theoretically that could work but without the proper firmware and calibration process it is impossible.

That's why I wrote that, at least from a hardware standpoint.

[rs38]

would love to read any further experiences with this device

also I like to know if the 3-phase (4-wire) version of this energy meter can be repurposed to measure three independent single-phase loads separately. The goal is to log individual per-phase data rather than just the total 3-phase power sum. Is there any reason this wouldn't work as expected or hard to wire on? is there a flaw in my logic?