Bench PSU always displays amperage as 0.00

[cjs]

So a while back this cheapish 30V 5A bench PSU stopped displaying the drawn amperage; it now always displays 0.00. Voltage display still works, and current limiting appears to work ok as well. (I can turn the current dials down and at what seems like an appropriate point the CV light goes out, the CC light goes on, and the voltage drops more as I turn the current limit down.)

The data connection to the displays is via CN1 on the display board to the lower-right connector on the mainboard (unplugged in this photo); the other connection appears to be just power. The device on right side of the upper heat sink at the left has a pin going to the shunt on the line labeled "5A" at the upper right, which then goes to the output. I'm fairly convinced this is the shunt for measuring amperage; it's about 2 ohms and looks just like the shunts I've seen in multimeters.

(In case it's confusing, the image of the back of the board has the part towards the front of the unit at the bottom and the back of the unit at the top. The "5A" shunt at the upper right in the top view, is at the lower left in the bottom view, with two solder-filled vias above it where the "10A" shunts are not installed.)

When I short the outputs on the front I can measure the voltage across the shunt at a few millivolts when the current limit is slow, and rising to tens of mV as I turn up the voltage, so I reckon that part of the current sensor system is working.

The large grey 1500 ohm 1% power resistor to its left is between the pre-shunt +voltage rail and ground. I've measured it in circuit and it takes quite some time to settle down (caps nearby?) but eventually settles on about 1460 ohms. I'm guessing this is probably part of the measurement system too, but I've no idea why it's there.

And that's about as far as I've investigated.

Any idea what's broken and if it's easily fixable? I don't want to spend too much time on this, but if it's something obvious that could get my amperage display back, that would be cool.

[CountChocula]

Can you trace the signals to the display board? It sounds like the actual current limiting circuit is working just fine, and there is just something wrong with the display, so I would rule out any broken wires/bad connection/broken traces to the display chips before looking into anything else.

PS: The shunt resistor should be much lower than 2Ω… at 5A, that would dissipate 20W of heat and probably negatively impact regulation. I think these things are usually a few mΩ at most, so it's possible that your DMM is not sensitive enough, or that somehow you have a broken shunt (in which case, though, current regulation wouldn't work, which doesn't seem to be the case).


—CC

[cjs]

I was thinking that it's not a display issue because, aside from power, there are only four wires going to the display (it's the connector on the lower right) and those are driving six digits of display at fairly fine resolution. So I'd assumed it must be digital. Also, when that cable is unplugged the displays don't come on at all: they remains blank rather than at 0.00 or some random number.

But now that I look more closely, it seems that maybe it is analogue, though that leaves me at a loss for why the displays don't come on even when disconnected. The evidence that points to this, it seems to me are a) there are a couple of trim pots, one under each display, on the display board itself, and b) the connector seems to be connected to a transformer.

Designating pin 1 as the one nearest the closer edge of the board (next to the screw), connector pins 2 and 4 have traces across the bottom of the board directly to transformer pins 1 and 3, and are also connected to the left side of D1 and D2. Connector pins 2 and 4 are connected to the right side of D1 and D2. The two diodes themselves show about .5 V drop with red probe on the left and black probe on the right, and over-limit with the probes reversed.

Does that look like any sort of circuit anybody's familiar with?

I guess I need to stick a scope on those pins and see what's going on? I guess whether they're analogue or digital, if it is indeed one pair for each side of the displays, I'll quickly find out whether the problem is on the display board or the main board.

[BrokenYugo]

Need better pics of the display board.

[CountChocula]

My guess is they are probably analogue—I've seen chips like those in countless voltmeter/ammeter combos from Aliexpress and eBay. Do try to make sure there is voltage going all the way to the ADC… if you can trace the current sense voltage all the way to it, chances are that it is toast.

[inse]

Looks like it‘s almost panel meters.
I see bridge rectifiers, voltage regulators, filter caps.
Check out what type of chips are on the meter board, look for datasheet and go debugging

[cjs]

Ok, so for the following tests I put a 10 Ω 50 W resistor across the output of the PSU, giving me (when not limiting) 100 mA draw at 1.0 V, 500 mA draw at 5.0 V, and so on, and I verified this with a multimeter monitoring the output voltage and a second multimeter in series with the resistor monitoring current.

I then had a look at what's going on on the pins on that connector. With 0V/0A output, pins 1 and 3 are at about 9.5 V and pins 2 and 4 are at 0 V, all having a weird very short ±7 V pulse at about 16 μs intervals, each followed by a ~1.5 V pulse about 2 μs later. These pulses never seem to change; perhaps they are just noise. You can see this in the first attachment below.

When I raise the voltage, pins 4 and 3 rise proportionally, to 10V and 19.5 V at 10 V output, and so on. The second attachment below shows the state at 8 V / 800 mA output. Pins 2 and 1 do not rise at all here. To get a more accurate voltage difference measurement I also checked the difference in voltage with a multimeter: it's not consistent, and varies between about 9.3 V and 9.6 V difference as I change the output voltage.

So this looks as if there's some sort of voltage control of the display going on here with, one would guess, pins 4 and 3 controlling the voltage display and pins 2 and 1 controlling the current display, and the current always measures 0 because the main board is not changing the voltage on pins 1 and 2. (I have no idea why they send what appears to be the same signal twice, except with one 9.3-9.6 V higher than the other.)

However, it gets more mysterious. I swapped pins 2/1 with pins 3/4 in the cable, expecting that I'd now see the voltage on the ammeter display, but I continued to see the voltage on the voltage display! So there's still something else going on here that I don't understand.

There is a notable difference between the volt and amp displays: the former always displays just volts, but the latter will display amps or milliamps, with the decimal point moving between two positions to indicate which it is. (This can be seen on the labeling of the meter front panel in my original post.) I don't know how this might affect things.

Thoughts? Could this still be an issue with the display board? Or is it probably, as it appears to me at the moment, an issue with the signal that the main board is sending to the display board?

[cjs]

I don't know why the second pic was not the one shown in the editor, but changed to a different pic after I posted. Let me try to attach the second 'scope pic again. (It's definitely showing the right one in the list of attachments below this post.)

[inse]

That’s just noise from the switch mode power supply, also depending on where you put your GND clip…
Can you identify the chips on the display board?

[cjs]

Here are pics of the display board as close as I can get with my current lens. Lighting issues unfortunately make most of markings unreadable, but below I've pasted my notes with almost all the chip designations and some tests of voltage levels. Markings that are missing are the surface mount caps and the two SOIC-16 chips: those are actually entirely blank on the top, with no signs even of markings being removed.


Vout = output voltage

Display board: MS-D3 / REV:170316 / Designed By Lys
- Mainboard side CN7:
  - 1=Vout; 2=GND (frame/output);  3=Vout - 10 V
- CN2: 1:`GND`, 2:`-V`, 3:`I+`
- CN1: pins 1,2,3,4; control inputs?
- LED1,LED2: SM320561D3B 3-digit 7-segment + decimal displays
- U1,U4 (SOIC-16): no label; 3-digit display controller?
- U2,U3 (3-pin + tab): AMS1117 3.3V XJS68; voltage regulators?
  - Pin 1 connected to CN2 `GND`
- C2,C4: ?,?
- E2,E3 (upper electrolytic caps): 100 μF 16 V
- E1,E4 (lower electrolytic caps): 100 μF 16 V
- RV1, RV2: 202 (2000 Ω) trim pots
- BD1,BD2 (4-pin): `+`/`-` on top, `MB6S` on bottom
- C1,R1,R2,R3: ?,472,102,102
- C3,R4,R5,R6: ?,333,333,472
- J2,J1 (back): both open
- Measurements relative to CN2 `GND`:
  - CN2
    - `GND` considered 0 V here, but -Vout on mainboard
    - `-V` is inverse of Vout (3.5 V Vout = -3.5V `-V`) (mainboard GND)
    - `I+` is -10 V w/Vout=0, see CN7 above
  - U2 1,2,3,tab: 0, 3.28, 7.82, 3.28 V
  - U3 1,2,3,tab: 0, 3,28, 7.91, 3.28 V

[DavidAlfa]

To me it looks like it's all around U7.

[inse]

Maybe this helps:
https://kupdf.net/download/ps305d-power-supply-schematic_5c7a7050e2b6f53a2a71e043_pdf

[cjs]

The schematic didn't make things a whole lot clearer for me, though I guess IC201 and IC211 have an A/D converter on pin 21 and translate that voltage level into a numeric display on the 7-segment LEDs? I guess I can check the pin connections on that schematic against the chips on my board and see if they're similar enough that I've identified the A/D input, I can see if both chips are getting what looks like an appropriate input.

The "AMS1117" look a lot like an SOT-223 LD1117 voltage regulator, which has pinout 1:ADJ/GND, 2:Vout, 3:Vin, tab:Vout.

I disconnected the board poked a bit more at BD1 and BD2. The four-wire inputs go to these and nowhere else as far as I can tell (I probed all the other pins on the board, though it's always possible I missed one somewhere). They do seem to be bridge rectifiers, as you'd expect from the markings, if I tested them correctly with my multimeter.

Code: [Select]

- BD1,BD2 (4-pin): "MB6S" diode bridge?
  - outputs on top: `-` left, `+` right; inputs on bottom unmarked
    - both inputs: positive → outputs `-`=OL , `+`=0.6
    - both inputs: negative → outputs `+`=0.6, `-`=OL
  - BD2 lower pins: CN1:2 CN1:1
  - BD1 lower pins: CN1:4 CN1:3
  - BD1 outputs: `-` → U1 pin 1; `+` → U2 pin 3 (regulator Vin)
  - BD2 outputs: `-` → U4 pin 1; `+` → U3 pin 3 (regulator Vin)

I'm having trouble wrapping my head around what they do. Perhaps the output on +/- is the difference between the two input voltages, which are always around 9.3 to 9.6 V apart in my testing?

At any rate, the chips are definitely not the same as the ones from the schematic, as the schematic versions have pin 1 not connected.

[inse]

Check if the pinning  14-pin IC on the main board matches a standard quad-OPAMP (LM324 e.g.) and verify the feasibility of the signal levels.
You could also swap both ICs on the meter board to narrow down the error path.
Usually those meter ICs have separate ground levels for supply and measurement, that means input- is different from supply GND.
Take that into consideration.

[Keri Szafir]

Yeah, I also cross-referenced the schematic with a ICL7107 datasheet (it's a pretty jellybean digital voltmeter IC), it's not the one.
There are discrepancies between the schematic and actual device - beware! It's not just same unit under a different brand. It may be similar but it's not identical.
It looks like both left and right part of the display board use a similar circuit, the front end is probably different. Closeup pic for reverse-engineering please, it may provide something.
Does this PSU display the set voltage/current when you adjust it, and then switch over to measured voltage/current?
If it displays the set values but not measured values, look for the problem in the measurement circuit on the main board, and if the current limit kicks in properly but the ammeter display doesn't work, that narrows the field pretty considerably to the driver circuitry for the ammeter.

[cjs]

It looks like both left and right part of the display board use a similar circuit, the front end is probably different. Closeup pic for reverse-engineering please, it may provide something.

Oh, I thought I'd posted the close-ups of it that I took, but apparently not. They're attached to this post.

The board's a bit tough to read, though, at least for me. I can pretty clearly see the traces that must be from CN1 pins 1 and 2 to the bottom pins of BD2, but not the ones from CN1 pins 3 and 4 to the bottom pins of BD1. (Perhaps they're going under the connector and chip.)

Does this PSU display the set voltage/current when you adjust it, and then switch over to measured voltage/current?

No, it appears to display only measured voltage/current. When I turn the knobs quickly, the voltage takes a bit of time to drop down to the new set point.

[cjs]

Check if the pinning  14-pin IC on the main board matches a standard quad-OPAMP (LM324 e.g.) and verify the feasibility of the signal levels.

Good thought! As it turns out, the upper IC on the main board (near the outputs) is marked "LM324M", so that provides a good place to do some more debugging. (And yes, I really do need to work out better lighting to capture the markings on the ICs. Dusting them off before photographing them would probably be a help, too.)

I'm not up to doing further probing tonight since it's now well past midnight here, but I did go through and find out what most of the other ICs are on the board, and dig up the datasheets for them:

Code: [Select]

- U2 (4 pin) [PC 817C][pc817] Optocoupler
- U3 (DIP-8): [SM7012] DIP-8 AC/DC PWM Power Switch
- U5 (SOIC-16): [TL494S][tl494] Pulse-Width-Modulation Control Circuits
  - Pinout: `1IN+ 1IN- FEEDBACK DTC CT RT GND C1`
          `‥ E1 E2 C2 Vcc OUTPUT_CTRL REF 2IN- 2IN+`
- U6 (SOIC-8): [OP07C] Low-Offset voltage (0.25mV) single op-amp
  - Pinout: `OFFSET_N1 IN- IN+ V- ‥ OFFSET_N2 V+ OUT NC`
- U7 (SOIC-14): [LM324M][lm324] Quad Op Amp
  - Pinout: `o1 i1- i1+ Vcc i2+ i2- o2 ‥ o3 i3- i3+ Vee/GND i4+ i4- o4`

[lm324]: [url]https://www.onsemi.com/pdf/datasheet/lm324-d.pdf[/url]
[tl494]: [url]https://www.ti.com/lit/gpn/tl494[/url]
[SM7012]: [url]https://datasheet.lcsc.com/lcsc/1810010808_Shenzhen-Sunmoon-Micro-SM7012_C130259.pdf[/url]
[OP07C]: [url]https://www.ti.com/lit/gpn/op07c[/url]
[pc817]: [url]https://pdf1.alldatasheet.com/datasheet-pdf/view/43376/SHARP/PC817C/+33_JJUO.vRPbSGwBDI+/datasheet.pdf[/url]

You could also swap both ICs on the meter board to narrow down the error path.

That's a bit tricky as I don't have SMT desoldering equipment myself, though I might be able to find a friend to do it for me. I'll keep it in mind if other avenues seem to be petering out.

Usually those meter ICs have separate ground levels for supply and measurement, that means input- is different from supply GND.

Yes, I'm realising that there's a good bit of "different ground levels" going on here, with things like Vout on the mainboard turning into GND on the display board, and the "-V" and "I+" pins. Still trying to wrap my head around it, but I guess that will come.

[MathWizard]

There's not that many parts on this thing. If it was me, I'd start at the output, and map back to the chips, making a schematic in LTspice.

The ASM11xx chips are regulators.

I've mapped out a few computer PSU's in full detail now. More often when mapping a section of circuit, I should find local Vcc ASAP, so I don't go off sideways into other circuit area's.