SIN90S, a multimeter with expanded functionality

While developing a device for testing and balancing multi-cell battery packs, I needed a multifunction meter. Since it was going to be used together with an electronic load, though, I needed a somewhat unusual kind of meter.
Inspection, description, tests, and a short cautionary tale below.


The battery-testing device project hasn't stalled; it just requires a lot of different modules, most of which I already have. I can now put together a rough list of what the device will consist of.
1. Adjustable DPS8005 DC-DC converter. It handles battery charging and plots the charging curve on a computer — link
2. Multichannel voltmeter module. It measures the voltage of each battery in the pack and sends the data to a computer — link.
3. 150 W electronic load. It discharges the battery under test at up to 2 A — link
4. Capacity measurement and electronic load control module, covered in this review.
5. Power for the device's power stage: two 48 V power supplies — link
6. Power for the control electronics and fans. 90 V is too much for them, so I'm still looking into options. It'll be either a separate power supply or a DC-DC converter.
7. Enclosure. I've already picked one out and will show it in this review.
8. 19 balancers rated for up to 2 A.
9. Heatsinks, fans, connectors, fuses, and other bits and pieces will most likely be bought locally.


I planned to measure the pack's capacity during both charging and discharging, which is the more accurate way to do it. That meant I needed something to control the electronic load. I didn't want to build everything from scratch, so I went with the approach shown in one of my reviews. It works pretty well, but the meter in that review was rated for up to 100 A, which is a bit... excessive for our 2 A. The accuracy would have been terrible.
Ideally, I'd use an ammeter-voltmeter like this one, but unfortunately these are often limited both in voltage (30 V) and in functionality. I needed it to control an external device—in this case, an electronic load.

Anyway, a friend and I searched online for different options and came across a meter called the SIN9020, which can measure up to 90 V and 20 A. But searching on AliExpress did turn up similar listings with titles like “4-byte red LED dual display.” Some even had the same photos from buyers, despite the main listing photo showing the meter we wanted. We didn't want to risk it.

1. The one described as a “4-byte red LED dual display” is, as you can imagine, a completely different meter.
2. An alternative was the VAC-1030A, rated for 120 V and 30 A, but we couldn't find one for sale. It seems to have been discontinued, even though it has a relay and is the lowest-current model in the VAC series.
3. While looking at other options, I also saw meters from the same series as the one in this review (judging by the display), including one with a non-contact Hall sensor. It costs only about $20.
4. Eventually, my search led me to Taobao, where I found a decent little shop that even offered free shipping to the forwarding agent. The same shop also had a meter rated for up to 1000 A.



No complaints about the packaging—it was excellent. The meter was wrapped in bubble wrap and placed in a bag, which was then wrapped in another large sheet of protective material. The only downside is that shipping costs more this way.


The seller offered two versions:
90 V, 20 A — double-sided — $12.25
90 V, 20 A — double belt protection — $14.29

Specifications for both versions. The supply voltage isn't listed here, but as far as I could tell, it can be powered from the measured circuit at 10–90 V, or from a separate input at 10–14 V.


Since I wasn't sure I understood what those names meant, I played it safe and ordered the more expensive one. This is what I received.


The package includes:
1. Meter
2. Board with shunt and terminals
3. Connecting cable.


The meter's enclosure will be familiar to many electronics hobbyists. The previous meter used the same one, as do various converters in the DPSDPH series. For example, on the right is a DPS8005 converter that will be used together with the meter under review.


Dimension drawing.


Removing the rear cover gives access to the internals. You can immediately see that the board is designed to have the shunts and power connector mounted directly on it.


There are actually two boards inside: a controller board with the display, and a measurement board with the power supply.


The controller board is nothing special: the display and three buttons are on one side, and an STM8S005 microcontroller with a simple supporting circuit is on the other.


The measurement board, on the other hand, has quite a few more components.


Power first goes to a pre-regulator, then to a PWM regulator based on the 34063.
The measurement circuit uses a dual op-amp MCP6002. I reviewed a converter quite recently that used exactly the same one.


The power board is a separate unit. I assume it isn't included with the cheaper version, and I don't actually need it anyway. But I wasn't sure I'd be able to find the control signal on the cheaper version (and it might not even have one), so I ordered the full kit.

On the left is the external power connector, with a two-position jumper behind it marked 3W and 2W. The first setting is for external power, the second for power from the circuit being measured. But there's a catch: in the measured-circuit-powered mode, the relay doesn't work; it can only be powered through the additional connector.
On the right is a four-pin connector that carries both power and the measurement and switching connections.

This is probably why the relay has a separate power supply: in one of the two modes, power can flow through its contacts, and the relay could also overload the power regulator. The allowable voltage is 10–14 V in external-power mode and 10–90 V when powered from the circuit being measured.


Internal wiring diagram for the main connections on the power board. In external-power mode, the Cb, B+ and B- contacts are used.


The manual describes how to connect the version without a relay. The wiring is essentially the same as on my version:
IN- = B-
IN+ = B+
Out- = Cb


There's nothing special on the underside. It uses a relay rated for up to 30 A on the NC contacts and up to 40 A on the NO contacts, which is more than enough for this device.
I measured the shunt resistance from the top and got 2.9 mΩ. At 20 A, that gives 58 mV, which doesn't quite match the standard 75 mV shunts. A standard shunt might still work, but it would need to be recalibrated.
The relay's NC contacts have a resistance of 2.36 mΩ. I measured it just to be on the safe side.




I've written several times about the difficulties of using ordinary relays to switch DC, and finally decided to give a real-world example of what can happen if....

It all started quite a while ago, almost 50 years ago. Three American astronauts decided that staying home wasn't very interesting, and that it would be nice to get together and spend a weekend on the Moon. Their names were James Lovell, John Swigert and Fred Haise.


As you do, they packed beer, sausages and a portable grill for their picnic. Since it was a long trip, they got themselves a big rocket and named it Apollo 13. It launched on April 11, 1970, a Saturday—presumably so they could make it to the Moon by Sunday and be back at work at NASA on Monday.


So the three buddies are flying along, and about 300,000 km from Earth, something suddenly goes bang. The first thought: Haise is messing around again, dropped a battery on someone's foot, the joker. But no, the reality was much simpler: one of the oxygen tanks was gone. And since this wasn't some favorite highway in Colorado lined with repair shops, it wasn't something they could just fix. Then came the radio message familiar to many:
Houston, we have a problem!

It also turned out that, besides the oxygen tank, they'd lost their GPS navigator, MP3 player radio, and something else had started leaking into space.


Houston replied, “Stay right where you are. We'll think of something.” Well, they came up with this idea: fly farther out, then turn around and slowly head back.
Along the way, the guys got rid of the container of sausages, beer, and Hayes’s beloved barbecue grill by dropping it (the grill, not Hayes, though they could have dropped him too) on the Moon. To recoup at least some of the cost, they booked it as a scientific experiment, even drew some nice graphs and said they were measuring the thickness of the lunar crust.


Along the way, they realized it was getting harder to breathe, and then came the second call:
Houston, we have problems again!!1

Houston replied, “Don’t worry. Just connect that thingy to that whatchamacallit that hose to that box over there, and everything will be OK.” It would have been OK, except for one problem: the hose was round, and the box was rectangular (no standardization whatsoever). In the end, a NASA specialist named Smiley came up with a solution. All they needed was the polyethylene cover from the cooling suit, the cardboard covers from the flight plan, a piece of Hayes’s favorite towel (presumably payback for the jokes), and a roll of blue duct tape. The photo on the right shows the whole contraption, made of junk and sticks; and since the photo is black and white, you can barely see the blue tape :(

But you have to give the guys credit: they’re still smiling in the photo (well, at least Lovell and Swigert), which is pretty optimistic considering they’re floating hundreds of thousands of kilometers from Earth in a half-working tin can, with no beer or cigars. On top of that, it was so cold that Hayes eventually got sick. Apparently, they accidentally jettisoned the liquid for warming up wiping things down, too.


The guys somehow made it home, climbed out, and decided to figure out who was who and who needed a punch in his smug ginger face. And while they were at it, they decided to give the photographer a hard time too, since he’d managed to tilt the horizon in the photo.


So the NASA director called in his chief engineer—let’s call him John Smith—and asked:
“Johnny, what the hell is this?”
“Oh, the relay glitched.”
“Johnny, where did you get the relay, you idiot?”
“Same as always, on Ali, from the well-known Store929593490. Rating 98.3, 5,435 sales. You told us to cut the budget.”
“Johnny, you’re an idiot.”
“Why?”
“Did you read the datasheet?”
“Who reads those? I’ve always done it this way, and nobody’s complained yet.”
“RTFM, Johnny, RTFM!”

It turned out to be simple and rather mundane: the clever Chinese American engineers used a thermal relay with 28 V contacts, but completely forgot that this wasn’t a truck with a 24 V battery—the system voltage was 65 V. Naturally, the contacts welded shut, and instead of staying at 27 degrees, the tank heated up to almost 540 (they’d even installed good thermal insulation, just in case). But that wasn’t all: the engineers had also made the instrument’s scale go only up to 27 degrees, so anything above that showed as 27 (so as not to upset anyone). A total mega-fail. Don’t believe it? Proof.

And the moral of the story is that contacts work very poorly with DC. For voltages above 28 V, other systems are usually used, because in this case the contacts will either weld together or an arc will form and burn them out. It all depends on the load type and operating conditions.
The photo shows an example from a NASA lab test, where the contacts welded together during the experiment. So, because of this mistake, they botched the Moon landing, wrecked the spacecraft, nearly sent three astronauts to the great hunting grounds in the sky, and NASA collectively shat a brick.



I sometimes criticize Chinese developers of cheap devices, only to find that good old Americans managed to screw up this badly on a space station worth several billion dollars. Be careful when designing a device, even if it’s not a nuclear reactor or a space station.



Quite by chance, I noticed that the two boards are not only the same size, but also have the mounting holes in the same places. What’s more, the connectors almost line up too, offset by just one pin. You could practically assemble everything into a single unit and get rid of the ribbon cable.


The instrument looks exactly like the VAC1100 and similar versions, with the same screen and the same three buttons on the right. Even the buttons feel just as stiff—really stiff. It’s downright ridiculous.


Before testing, I decided to measure the current draw and minimum supply voltage.
I found that the instrument starts up at 9.8 V, but only works properly at 11 V, drawing 40 mA. After a minute or two, the backlight switches off automatically and the current drops to 19–20 mA. The current is about the same at a supply voltage of around 60 V, but with the backlight on.
By the way, about the backlight: it’s nice that it can switch off, but it’s really bad that there’s no way to configure this feature.


Now, let’s look at what appears on the screen and how it all works. There’s a manual for this instrument online, but for the most part, everything is pretty clear without it.

1. Voltmeter
2. Ammeter
3. Battery charge indicator, calculated based on the capacity charged/discharged and the value set in the settings.
4. Capacity passed through the device in Ah and current power in watts.
5. Capacity passed through the device in Wh and timer
6. Control buttons
7. Operating mode selection.

The up/down buttons move the yellow cursor. A short press of OK turns a mode on or off; a long press (2–3 sec) opens the parameter settings.


1. When power is applied, the cursor is always in the middle position. In this mode, pressing OK turns off the relay (by default, the relay is always switched on automatically). The arrow shows the direction of current flow, from the load to the source or vice versa, although it’s all relative here. Also, if the arrow is red/green, the output is active; if it’s gray, it’s inactive.
2. OVP - maximum voltage limit setting
3. OPP - power limit, a kind of overload protection.
4. OCP - load current limit.
5, 6. The output is inactive, but the arrow showing the direction of current still works.
7. OFT - operating time limit.
8. OAH - battery capacity limit; the charge percentage shown on the indicator is calculated based on this value. Accordingly, current flowing in one direction increases the percentage, while current flowing in the other direction decreases it.
9. LOP - minimum voltage limit
10. A short press activates the required limits, which work according to a logical OR, i.e. the first limit reached triggers them.
11. If a parameter is already at its limit, the relay cannot be switched on. For example, here I could only switch it on with the last limit selected.
12. The battery is “discharged” in almost all the photos. For the example, I connected the device so that it counts in the “charging” direction: the arrow lights up red, and the Ah, Wh, and charge percentage counters increase, while the load power is displayed.

By default, all parameters are set to 00.00. As I mentioned above, a long press opens the parameter settings, then you adjust them with the up/down buttons. But if you haven’t changed anything, you can’t exit; you need to press up or down at least once.


Now for a few tests.
First, voltage measurement. I tested it with a power supply starting at 11 V and going up to almost the maximum possible, 90 V. Across the entire range, the accuracy is clearly within the stated 1% error, and even better.


To test current measurement accuracy, I had to use a fancier power supply and load. The device was connected so that its own current consumption wouldn’t affect the measurement result.



Current measurement is noticeably worse. To start with, even with no load, the device showed about 40–50 mA, but I decided to test it “as is.”
The result was that across almost the entire range, from 100 mA to 12 A, the device readings “lagged” the actual values by 40–80 mA. I measured 15 and 20 A with another device, since my main one only goes up to 10 A, and even the 12 A reading is far beyond its range.

Naturally, the question arises: how could it read 0.04 A with no load, but 0.01 A under load? It’s simple: the 0.04 A was flowing in the opposite direction, i.e. with a minus sign. In all the photos except the first, the arrow is green; in the first, it’s red.


The reason is that the device measures current in both polarities, so technically its range isn’t 0–20 A, but -20 to 20 A, or 40 A if you count the full span. This causes two problems:
1. The accuracy is lower than it would be if it measured from zero to 20 A
2. The zero point has to be set precisely.

You can even see in the product listing photo that with only the power connected, the display shows a value—in this case, 80 mA.


To calibrate it, go into the settings. To do this, hold down OK before applying power; when you release it, the calibration menu will appear. Use OK to switch between items and the up/down buttons to activate them.
1, 2. There’s a Mode setting that can be changed from zero to five, but after trying all the options, I didn’t notice any difference.
3. The 0.0 A mode is zero calibration. Turn on the device with no load and use the “down” or “up” button to get a zero reading. In my case, I had to press “down.”
4. Press OK until you get to “Save,” then press “up” or “down” to save the setting. The number on the screen changes, so to avoid messing anything up, I pressed “up” once and “down” once before disconnecting the power.

There are also calibration points for two voltage values, 10 and 40 V, and one current value each at -4 A and 4 A.


The device was already within the stated 2% accuracy, but calibration made it a little better.
But now I noticed another problem, shown in the last two photos:
1. Load current is 100 mA; immediately after a high-current test, the meter reads 140 mA
2. After 15–20 seconds, the readings return to normal. The shunt heating is presumably affecting them.


I’ve already bought an enclosure for the device. In this case, we didn’t try to save money and went with an almost-largest-size Bahar metal enclosure. It may seem too big, but considering how much needs to fit inside, it doesn’t seem so big after all. The size of the front panel also played a significant role here.


I’ve already talked about this enclosure in my review of the ZXY6020S converter.


Now for my final thoughts. I have to say I’m somewhat disappointed. Of course, I wasn’t expecting high accuracy, but I simply couldn’t find anything more suitable for my needs. The biggest problem is that the meter measures up to 20 A, while I actually need a range of up to 2–3 A. The options I see are either replacing the shunt with one of higher resistance, say 10 times higher, or increasing the reading as much as the calibration range allows.
With the first option, I’d have to recalculate the readings; with the second, I’d get a huge error at high currents. But the first option seems more practical.
The shunt’s temperature stability is also an issue. I can either live with it or replace it with a proper 20 A, 75 mV shunt.

And last but not least, the buttons are far too stiff. The photos of the internals show that they’re ordinary buttons, but they’re incredibly hard to press once installed in the enclosure...

Otherwise, it’s just a meter. Its accuracy is average, though it does meet the stated specifications. A full manual with details on the additional settings would have been useful. The table of contents does list a section on these settings, but the manual itself doesn’t describe them. For example, I’d turn off the power-saving mode.


That’s about it from me. I’d be glad to hear your questions and comments, and perhaps someone can suggest something more convenient for my needs.

The meter cost $15.31 and weighs 110 grams according to the forwarding agent. Shipping costs depend on various factors, such as the quantity and whether there are other items in the order.