Manson SPS-9400 Switching PS repair

[Itsu]

So it looks like the TOP222y is doing what it suppose to be doing.

I will re-insert it and try to solder an insulated wire between the lifted Drain terminal and the PCB (as short as possible) and put a current probe on it.


Itsu

[Verpies]

No,  but what do you mean by test gear,  the gear under test (defective PS) or gear used for testing (scope etc.)?

I am so paranoid about this that I use isolation transformers on all my test gear power, power supplies and DUTs.
If I need to ground any of them I use separate bannana plugs for that.

[Itsu]

Ok,  i will try to install at least an isolation transfomer for the DUT.

[Itsu]

I re-installed the TOP222y with a loop wire from its drain to the D3 / TR2 prim. junction.

Current probe measurements during the short 15s window shown below.

The last screenshot was ignoring the negative going spikes, and should be similar ( i think) to the IOUT signal in the signal picture (Fig. 6 TOP222y data sheet) at the bottom of this post.

[Itsu]

Guess the current probe should be reversed (or the channel), so here the above last screenshot with the current probe reversed (normal)

Looks to me some sort of flashover occurs.

[Verpies]

Yup.
The current through an inductor should not reverse like that. Even in the switching spike.  If this was a scopeshot of voltage then it would not be anything unusual.
Also, outside of the spike the current is decreasing with time, so I am assuming that during that period the TOP222 is turned off and is allowing the current built up during the spike to decay.

[Itsu]

I will remove TR2 and see / measure it.

The only to me known supplier of this part (Andrews Comms) emailed me that this part is currently not obtainable "due to coronvirus issues".

Itsu

[Verpies]

I will remove TR2 and see / measure it.
The only to me known supplier of this part (Andrews Comms) emailed me that this part is currently not obtainable "due to coronvirus issues".

If you convince yourself that it is bad internal isolation, then you can repair it by rewinding it...or:
...use a low viscosity epoxy to impregnate the windings inside.  To do that correctly you should use a vacuum chamber, so the epoxy is degassed and pushed inside the windings by the atmospheric pressure after removing the submerged transformer from the chamber (the epoxy still should be liquid at this point).
Heating during this procedure also improves results.

If you don't have a vacuum chamber and don't want to buy one, then it is possible to rent the use of one.

[Itsu]

thanks for the good advice.

After removing TR2, and measuring the primary (which is the most internal    ) i found it to be still around 2.5mH, but when squeezing the coil, there is a spot that it decreases to about 1.5mH.

So i convinced myself that its the TR2 primary internal isolation that is the problem.

Looking for a replacement first.

TR2 (transformer) listed in the service manual as:

1 8710-9400-3199 SPS-9400 AUX TRANSFORMER 1.00000 PCS IN SUB-ASSEMBLY CE VERSION Location : TR2

[partzman]

thanks for the good advice.

After removing TR2, and measuring the primary (which is the most internal    ) i found it to be still around 2.5mH, but when squeezing the coil, there is a spot that it decreases to about 1.5mH.

So i convinced myself that its the TR2 primary internal isolation that is the problem.

Looking for a replacement first.

TR2 (transformer) listed in the service manual as:

1 8710-9400-3199 SPS-9400 AUX TRANSFORMER 1.00000 PCS IN SUB-ASSEMBLY CE VERSION Location : TR2

Itsu,

I found this bit of info on that transformer quoted from a user-

I have the schematic for the power supply and I've traced the fault to a blown primary on the aux power supply transformer (TR2). It is an E core in a 10 pin former.
It has a Manson part number (A9400000) but does not seem obtainable from Manson or from the company that sold the psu. Is it possible to buy one or to find the specification. It is driven by a TOP222Y. It has a 130 turn primary and three secondaries, but I have no information on the secondary turns. The schematic does not give me a value for Vcc either. Is it likely that the TOP222Y has failed and blown the primary, there is some suspicious black deposit near its D pin.

If the primary is wound over the secondaries, you might be able to rewind the primary if you can successfully get the core apart.  If you do go to all the trouble, I would attempt to find wire that is double insulated or coated polyurethane as it will have a greater resistance to breaking down.

Regards,
Pm
Pm

[Itsu]

Thanks PM,

the E-core's are cracked, each one has a leg broken off, probably done by me, i will try to glue it.

The primary looks wound on the inside, so i am now unwinding the secondaries one by one taking note on the number of turns etc.

Hopefully it can be restored as indeed the part is very hard to find.

Itsu

[Itsu]

Hmmm,  i removed the 3 secondaries ontop of each other and the primary, no obvious damaged seen.

Then i removed the (2 layers with iso tape inbetween) primary, but other then some disturbed turns i the middle no obvious damage seen.

I will try to reassemble it.

[Itsu]

After reassembling, the inductance of the coils is roughly half of what it was, so the PS now does not "light up".

Seems the ferrite E-cores are busted to much still after glueing.

Looking for E20/14/5 E-cores which are hard to find, i have a quote outstanding in the UK,  but no response till now.
https://www.acalbfi.com/nl/Magnetic-components/Cores/Ferrite/p/E-Cores---Core-halves-of-high-permeability-grades/00000000H2


Else i have to build my own transformer, using the measured inductances.

itsu

 

[Verpies]

Hmmm,  i removed the 3 secondaries ontop of each other and the primary, no obvious damaged seen.
Then i removed the (2 layers with iso tape inbetween) primary, but other then some disturbed turns i the middle no obvious damage seen.

The damage to the insulation can be invisible and a cracked core cannot be repaired.
Did you count the turns as you have unwound them ?

[Verpies]

Seems the ferrite E-cores are busted to much still after glueing.

I have never, never been successful with gluing a core. The reluctance increased a lot because of the small glue gap and the magnetostriction breaked any glue I've tried long-term.
Not to mention the temperature...

The only thing that kind of worked was ultrasonic soldering, but the eddy currents melted it eventually when the power was increased.

Looking for E20/14/5 E-cores which are hard to find, i have a quote outstanding in the UK,  but no response till now.
https://www.acalbfi.com/nl/Magnetic-components/Cores/Ferrite/p/E-Cores---Core-halves-of-high-permeability-grades/00000000H2

Does the core look exactly like this ?

[Itsu]

Did you count the turns as you have unwound them ?

Yes,  i have noted down the specifics of each.

I was afraid that glueing would not work, i have done that before without decent results.

Does the core look exactly like this ?

Yes, its the same,  only half thinner and the "legs" longer, very fragile.

[Verpies]

Yes,  i have noted down the specifics of each.

So there is still hope of rewinding it.

Don't throw away the broken core. A core fragment can be used to determine properties of the material it is made of. i.e. the A_L value, permeability, resistivity.

[Itsu]

You mean like in the website under "Measuring initial permeability mi"?

http://home.scarlet.be/on9cvd/E-Zelf%20meten%20aan%20spoelkern%20materialen.htm

Itsu

[Verpies]

Yes

[Itsu]

No luck finding the A94000000 transformer or E20/14/5 E-cores.

I have removed TR2 and the TOP222 and want to install a 220V / 2 x 12V transformer instead (floating above the board).

220V will be tapped from C2, 1 x 12V ac goes to the D5 circuit for supplying VCC, the other 12V ac goes to a new board with rectifier, 12V regulator for supplying 12V dc to DM VDD.

Any objections spotted?

Itsu

[Itsu]

I made the above mentioned modifications which seems to work.

The PS is back to life and shows correct voltage and current.

There is however a heap of noise visible (S-meter) and audible when connected to the receiver, so some further investigation and decoupling is still needed.

Could be some of the old now unused circuit of the TOP222 is oscillating.

Another point i am puzzled about is the DM VDD circuit.
It comes from a secondary of TR2, so must be AC, but the connection shows + and -  DM VDD.
I am feeding it now with 12V DC, but maybe it needs AC.


Itsu