[SOLVED] UNI-T UTD2052CEL Oscilloscope SMPS repair

[wblock]

It might not be the Samwha capacitors, but... that would be the first thing I'd suspect.  An ESR meter can test some of those in-circuit.  If C8 is the startup cap, replacing it alone might get the thing going.  But it that does work, replace all of the eletrolytics from C8 to the right, because that would mean the rest are probably near failure, too.

[negativ3]

It might not be the Samwha capacitors, but... that would be the first thing I'd suspect.  An ESR meter can test some of those in-circuit.  If C8 is the startup cap, replacing it alone might get the thing going.  But it that does work, replace all of the eletrolytics from C8 to the right, because that would mean the rest are probably near failure, too.

Thanks all,

So... is replacing C8 the next logical step to take? I do not have an ESR meter

[tautech]

It might not be the Samwha capacitors, but... that would be the first thing I'd suspect.  An ESR meter can test some of those in-circuit.  If C8 is the startup cap, replacing it alone might get the thing going.  But it that does work, replace all of the eletrolytics from C8 to the right, because that would mean the rest are probably near failure, too.

Thanks all,

So... is replacing C8 the next logical step to take? I do not have an ESR meter

Yep, sometimes that's just all it takes.
But you do have a capacitance mode in your DMM don't you, if it measures more than 10% low, replace it for sure.

[retiredcaps]

So... is replacing C8 the next logical step to take? I do not have an ESR meter

If ESR is high, you can try warming up just C8 with a hair dryer or adding some fresh solder to the bottom of the legs with your soldering iron.

Heating C8 up slightly will lower its ESR maybe enough to start the power supply?

[negativ3]

Yep, sometimes that's just all it takes.
But you do have a capacitance mode in your DMM don't you, if it measures more than 10% low, replace it for sure.

If ESR is high, you can try warming up just C8 with a hair dry or adding some fresh solder to the bottom of the legs with your soldering iron.

Heating C8 up slightly will lower its ESR maybe enough to start the power supply?

I tried heating the pins on C8, no difference. My DMM does not have cap mode  .

I removed C8 and measured across it in ohms mode and it shows charging and discharging (+3.25 meg connected correctly and -3.25 meg connected reverse).

C8 is a 47uF 50V. Closest I have is 22uF or 100uF 25V which are waaay off.

[retiredcaps]

C8 is a 47uF 50V. Closest I have is 22uF or 100uF 25V which are waaay off.

Either 22uF or 100uF should be "good enough" for testing purposes.   I have done the same replacement many times and it works.  That is, if C8 is at fault.

The voltage across that cap should only around 5.7VDC if I'm reading the datasheet properly.

[negativ3]

C8 is a 47uF 50V. Closest I have is 22uF or 100uF 25V which are waaay off.

Either 22uF or 100uF should be "good enough" for testing purposes.   I have done the same replacement many times and it works.  That is, if C8 is at fault.

The voltage across that cap should only around 5.7VDC if I'm reading the datasheet properly.

Thanks all for the incredible assistance with this. 

Ok, I have replaced the cap with 100uF type and it is the same, all outputs zero V. 0V across C8 with reference to case/earth point on the PCB.

[StillTrying]

The voltage across that cap should only around 5.7VDC if I'm reading the datasheet properly.

I was just reading that! The control pin charges the cap to 5.7 - 7V, so 100uF 25V is close enough to check.

It's looking like the TOP225YN then.

[retiredcaps]

Ok, I have replaced the cap with 100uF type and it is the same, all outputs zero V. 0V across C8 with reference to case/earth point on the PCB.

C8 is on the hot side.  You should be using the -ve leg of the main filter capacitor as ground or you can just measure the voltage of C8 across the legs of the capacitor.

[retiredcaps]

@OP, does R21 measure as 9.8 ohms?  In a post, you said all passives check out okay, but I just want to verify.

There is one more simple check you can do . See if TOP225YN is shorted.  With power off, see if S-D, S-C and C-D are shorted.  You can make this measurement in circuit.  Any reading less than 100 ohms is suspicious.

TOP225YN is available at mouser.com for $4.03 USD.    If you are not in the USA, TOP225YN can be found on ebay and aliexpress.com

[negativ3]

R21 = 9.8 ohms.
S-C = 39.8 ohms.
S-D = 3 Meg ohms and rising.
C-D = OL.

From this the TOP225YN looks favourite?


Sent from my iPhone using Tapatalk

[retiredcaps]

S-C = 39.8 ohms.

39.8 ohms doesn't look right.  You might want to desolder the IC and see if you get the same readings out of circuit.

I'm not 100% familiar with this off-line PWM switch and a 20 page datasheet at midnight isn't my idea of bedtime reading.  In 30 minutes, I'm off to bed.

Having said the above, if it still reads 39.8 ohms out of circuit, I would say it is bad.  As you might have noticed, others have more experience and knowledge with this type of power supply.

[retiredcaps]

One other tip before bed.

If/when you order a new IC, measure it before installing it.  All three measurements should be open circuit.

[negativ3]

TOP225YN removed and C-S reading 40 ohms. I will read the datasheet again. Thanks retiredcaps.


Sent from my iPhone using Tapatalk

[negativ3]

Are these sleeves used for isolation or something else? Tempted to remove them and test again...

[Shock]

Are these sleeves used for isolation or something else? Tempted to remove them and test again...

They are ferrite beads.

[tautech]

Are these sleeves used for isolation or something else? Tempted to remove them and test again...

Leave them be for testing, they're there to manage EMI.

[Shock]

Just something I observed from your posts, when testing the low voltage side did you confirm if there is good continuity between the test point you are using as ground and the pin out grounds mentioned by JFJ? Because those are good ground reference points (for the low voltage side) you can clearly see they are common on the PCB.

Next have you tested for voltages both with the output cables connected and disconnected? I'm not sure if I saw this clearly mentioned anywhere. Often when diagnosing power issues people tend to jump pretty quickly to the power supply itself but I don't recall reading you have confirmed the output voltages one by one, or tested across the secondary caps for voltage, or when the smps is removed from the mains and discharged checking the continuity across those outputs to see if any short is present.

[negativ3]

Just something I observed from your posts, when testing the low voltage side did you confirm if there is good continuity between the test point you are using as ground and the pin out grounds mentioned by JFJ? Because those are good ground reference points (for the low voltage side) you can clearly see they are common on the PCB.

Next have you tested for voltages both with the output cables connected and disconnected? I'm not sure if I saw this clearly mentioned anywhere. Often when diagnosing power issues people tend to jump pretty quickly to the power supply itself but I don't recall reading you have confirmed the output voltages one by one, or tested across the secondary caps for voltage, or when the smps is removed from the mains and discharged checking the continuity across those outputs to see if any short is present.

Thanks Shock,

Ok, I have verified the HV cap -ve and chassis earth are not connected to secondary GND.

I have measured all outputs (cable connected and disconnected) WRT DC GND and they all show 0.000V DC output.

With the PSU discharged, between DC GND and outputs, I measure the following resistance(s):

3.3V = 0.732K
5V = >1M
Line Trigger = 0.97K
20V = 3.81K
-11.5V = 1.62K
~8V = >1M

[negativ3]

The transformer has been measured with the following:

Primary:
1-3 = 1.2 ohms
4-5 = 0.4 ohms
3-4 = Open cct

Secondary:
6-7 = 0.2 ohms
6-8 = 0.3 ohms
6-9 = 0.3 ohms
6-10 = 0.5 ohms

Open cct between primary and secondary.

[JFJ]

.. the HV cap -ve and chassis earth are not connected to secondary GND.

The secondary GND should be connected to chassis earth via the main PCB (through the 12-way ribbon cable).

Incidentally, I notice that you only give single values for each of your resistance measurements. When measuring resistance in-circuit, it's best to do it in both directions (by reversing the meter leads). This can help to reduce some of the misleading resistance readings that can occur in the presence of semiconductor junctions.

A long time ago, while testing faulty PCB, I checked the resistance of an 18K resistor that I suspected might have failed open circuit. The reading on my analog multi-meter was exactly 18kohm. I then wasted several hours performing other tests, before I noticed that there was a small diode (obscured by larger components) connected in parallel with the resistor. The forward biased diode just happened to read 18kohm on my meter (reverse biased it read infinity).

[negativ3]

The secondary GND should be connected to chassis earth via the main PCB (through the 12-way ribbon cable).

Incidentally, I notice that you only give single values for each of your resistance measurements. When measuring resistance in-circuit, it's best to do it in both directions (by reversing the meter leads). This can help to reduce some of the misleading resistance readings that can occur in the presence of semiconductor junctions.

A long time ago, while testing faulty PCB, I checked the resistance of an 18K resistor that I suspected might have failed open circuit. The reading on my analog multi-meter was exactly 18kohm. I then wasted several hours performing other tests, before I noticed that there was a small diode (obscured by larger components) connected in parallel with the resistor. The forward biased diode just happened to read 18kohm on my meter (reverse biased it read infinity).

Hi JFJ,

I meant to state the resistances quoted were the lowest readings with probes in both directions during test.

You are of course correct, the DC ground goes to chassis via the DSO main board.

[retiredcaps]

I meant to state the resistances quoted were the lowest readings with probes in both directions during test.

1. If IC is shorted both ways (40 ohms) out of circuit, then I think it is safe to say it is bad.

2. JFJ brings up a good point about a multimeter's voltage possibly turning on junctions.  There was a thread about the this at

https://www.eevblog.com/forum/testgear/multimeter-with-low-continuity-test-voltage/?all

jjoonathan asks about this in

https://www.eevblog.com/forum/testgear/need-suggestion-about-dmm-kyoritsu-1012-vs-hioki-4252-vs-fluke-17b/?all

3. If you are going to order the IC from mouser, you might want to think about changing out the Samwha caps to something like Panasonic, Rubycon, Nichicon or United Chemi-con.

4. Measure the resistance of the empty IC pads on the pcb to make sure they are open circuit.

5.  Please report back if a new IC fixed your problem or not.

[negativ3]

New IC and caps ordered. Will take about a week for delivery. I will for sure update when I get to the next step.

Thanks all for your fantastic support.

[negativ3]

New caps and IC delivered. Old caps removed and replaced with new.

Now working 100% and on an extended test.

Thanks again all who helped me out with this.