I found my Ham radio 14V /40A Switching mode PS (Manson SPS-9400) failed to poweron (after 15s switching off).
Looks like a protection circuit is doing that, its TOP222Y is getting very hot fast, could be defective or the transformer it drives.
https://www.austech.info/showthread.php/55513-MP3090-SPS9400-power-supply-repair-hints
So will be troubleshooting that.
Itsu
Quote from: Itsu on 2020.10.06, 10:40:00
Also found my Ham radio 14V /40A Switching mode PS (Manson SPS4900) failed to poweron (after 15s switching off).
Oh shit!
Does this schematic match that power supply ?
Yes, thats the one i also found and matches the layout i have up till now.
The zener ZD1 and diode D3 around the TOP222Y measure OK, also the TR2 primary seems OK as i measure (in circuit) 2.5mH.
Quote from: Itsu on 2020.10.06, 11:15:14
the TR2 primary seems OK as i measure (in circuit) 2.5mH.
That is very good news.
If I were you, I'd remove the TOP222Y and the ZD1 (or D3) and remeasure the TR2 primary.
I would not be quick to assume that the TOP222Y is bad. I would test it according to its basic application circuit in its datasheet.
Failed ZD2 or D6 on the secondary side could cause it to overheat, too.
Right, D6 and ZD2 and D7 also measure fine.
The TOP222Y measure (DMM diode setting) D / S as a diode, so looks OK too.
Yes, i will follow the TOP222Y datasheet i just downloaded to check on it.
I measured my new TOP224Y, which is a 3x higher power version of the TOP222Y, and I discovered that:
- Between the Control terminal and the Source terminal the Diode measurement by my multimeter yields 0.5V when the negative probe is on the Control terminal (open circuit in the other direction).
- Between the Control terminal and the Drain terminal the Diode measurement by my multimeter yields 2.2V when the negative probe is on the Control terminal (open circuit in the other direction).
- Between the Source terminal and Drain terminal it measures as a Diode with 0.5V voltage drop.
When connecting it as shown in Fig.14 of the datasheet and with +40V DC applied to the drain through a 500Ω 5W resistor, it generates 300kHz pulses on the Drain terminal for 3.85ms and then nothing for 67ms, when the Control terminal is left floating but bypassed with a 4.7µF cap to the Source terminal. When that capacitor is increased to 47µF then these 300kHz pulses stay ON for 38ms and OFF for 666ms.
Ok, thanks, good to know.
Need to remove the TOP222y from the pcb still.
First 15 seconds after poweron its ok, temperature gradually increases on the TOP222 from 20°C to 40°C then it shuts down the PS, but with the poweron button still on ON the temp still continue to rise
untill i switch the switch to OFF.
In those 15 seconds, all looks and works OK as far as i can see / measure.
Itsu
Quote from: Itsu on 2020.10.06, 13:46:53
First 15 seconds after poweron its ok, temperature gradually increases on the TOP222 from 20°C to 40°C then it shuts down the PS, but with the poweron button still on ON the temp still continue to rise
until i switch the switch to OFF.
Maybe something is shorted on the secondary side and is overloading it.
You can disconnect the secondaries by removing D5, C71 and the wire to the DM VDD (in that order).
The D6 is a part of the feedback voltage circuit so leave it for last, because removing the feedback voltage will increase the duty cycle of the PWM inside the TOP222Y and cause it to overheat even more (maybe that is happening already).
The voltage between the Source and Control pin should be around 5.7V if the feedback circuit is working properly.
If it stops shutting down with the secondaries separated then you will have a confirmation of secondary overload.
You can check that the TOP222Y is operating by putting an ISOLATED! current probe on its Drain terminal. If you see current pulses on the scope then it is operating.
Using differential probing with my old Owon scope i see that there are pulses between drain and source of the TOP222.
These pulses continue after it shutsdown the PS, so it looks like the TOP222 is working OK.
Will try to disconnect the secondaries.
Source and Control pin is 5.8V, so looks good. I find D3 also getting hot, but ZD1 not, also TR2 not.
VCC measured 13.8V then dropping to 12,5V then shutdown.
I cut D5, still heating up but no shutdown.
C71 not reachable from the top, so have to remove the PCB first.
Quote from: Itsu on 2020.10.06, 14:41:46
Source and Control pin is 5.8V, so looks good. I find D3 also getting hot, but ZD1 not, also TR2 not.
VCC measured 13.8V then dropping to 12,5V then shutdown.
I cut D5, still heating up but no shutdown.
C71 not reachable from the top, so have to remove the PCB first.
If D3 is getting hot but not ZD1, ZD1 is most likely shorted.
Pm
Hi PM, thanks, but ZD1 measures OK with the DMM in diode setting.
I mentioned TR2 not heating up, but when measuring on the other side it does, so we have TOP222, D3 and TR2 heating up then shutdown after 15 seconds.
Quote from: Itsu on 2020.10.06, 15:12:36
Hi PM, thanks, but ZD1 measures OK with the DMM in diode setting.
I mentioned TR2 not heating up, but when measuring on the other side it does, so we have TOP222, D3 and TR2 heating up then shutdown after 15 seconds.
OK, but the diode test would not test for the zener breakdown of ZD1. For example, if ZD1 breaks down at some lower voltage than 200v, the current to charge the primary of TR2 would possibly go continuous thus raising the dissipation in the TOP222 and D3 along with the primary of TR2. The value of the zener at 200v is set to allow the primary flyback current to fall to near zero under the the maximum duty cycle to keep the current in a dis-continuous mode.
I would not recommend that you remove ZD1 and try the circuit because the TOP222 will more than likely avalanche and could destroy itself in a short period of time.
Regards,
Pm
I removed ZD1 and tested it with a 250V supply and it works OK as it was able to regulate it back to 200V.
I will tomorrow remove the TOP222 and test it like verpies showed.
Also then the TR2 can be removed and measured for correct workings (no shorts / opens).
Thanks for now, Itsu
Itsu,
Sorry for the wild goose chase! :-[
Pm
Quote from: Itsu on 2020.10.06, 20:35:29
I will tomorrow remove the TOP222 and test it like verpies showed.
Also then the TR2 can be removed and measured for correct workings (no shorts / opens).
A scopeshot of a signal from the current probe placed on the insulated red wire (see schematic) would tell us a lot about the health of TR2 in circuit.
We could also see whether the current through TR2 is discontinuous or continuous and whether the magnitude of this current exceeds TOP222Y specifications.
PM,
QuoteSorry for the wild goose chase!
don't be, i appreciate your input.
verpies,
see picture of backside of the pcb, your red line is only like a 5mm trace.
Removing DM VDD (connector) still heating up the TOP222
removing D5 still heating up the TOP222
No heating of D6 or ZD2 itself.
So secondary loads are probably OK, could be the TR2 secondaries inside (or primary).
I measure the heating up of primary the TOP222, then secondary C70 (very close to the TOP222), D3 and TR2.
There is 414V DC (DMM) across C70.
I will remove the top222 and TR2 for measurements.
Itsu
TOP222 removed, and measuring with the DMM in diode setting, comparing with below verpies results:
QuoteI measured my new TOP224Y, which is a 3x higher power version of the TOP222Y, and I discovered that:
- Between the Control terminal and the Source terminal the Diode measurement by my multimeter yields 0.5V when the negative probe is on the Control terminal (open circuit in the other direction).
- Between the Control terminal and the Drain terminal the Diode measurement by my multimeter yields 2.2V when the negative probe is on the Control terminal (open circuit in the other direction).
- Between the Source terminal and Drain terminal it measures as a Diode with 0.5V voltage drop.
Mine shows:
- Cntr - Source = 0.5V (black on Ctrl) open other direction
- Cntr - Drain = open (black on Ctrl) open other direction
- Source - Drain = 0.5V (black on Drain) open other direction.
So it seems there is a difference between Ctrl and Drain.
Some Fig. 14 setup screenshots with 47uF on Ctrl-source and 40V on drain via a 470 Ohm resistor.
Itsu
According to the TOP222y data sheet page 12, its "Thermal Shutdown Temperature" is 135°C
So why does it "look like" its shutting down around 40°C? Is it the thermal delay between inside of the chip and the outside casing?
Quote from: Itsu on 2020.10.07, 14:47:05
- Cntr - Drain = open (black on Ctrl) open other direction
...
So it seems there is a difference between Ctrl and Drain.
Could be, but maybe my multimeter applies a higher test voltage or the lower version of the TOP controller has a higher voltage drop that is above the range of your Diode test.
Quote from: Itsu on 2020.10.07, 15:26:15
Some Fig. 14 setup screenshots with 47uF on Ctrl-source and 40V on drain via a 470 Ohm resistor.
Looks like it is working, although that test does not excercise the Ctrl pin exhaustively.
Quote from: Itsu on 2020.10.07, 16:33:55
According to the TOP222y data sheet page 12, its "Thermal Shutdown Temperature" is 135°C
So why does it "look like" its shutting down around 40°C? Is it the thermal delay between inside of the chip and the outside casing?
I am sure that there is a thermal delay between the chip and the package but 15sec seems too long to explain the shutdown this way. If a freeze spray extends the shutdown time then it would be a confirmation of this explanation.
This PWM controller chip also has several other shut down modes, such as overcurrent, overvoltage and undervoltage. It is worth investigating them.
The first one can only be inspected by soldering an insulated wire between the lifted Drain terminal and the PCB (as short as possible) and putting a current probe on it.
Quote from: Itsu on 2020.10.07, 09:02:55
There is 414V DC (DMM) across C70.
Yeah, if the caps or resistors in the probes break down then this can destroy the frond end of the scope. Even when probing differentially.
Are you running your test gear on galvanically isolated separation transformers ?
QuoteEven when probing differentially.
Good to know.
QuoteAre you running your test gear on galvanically isolated separation transformers ?
No, but what do you mean by test gear, the gear under test (defective PS) or gear used for testing (scope etc.)?
I understand that the best thing to do is to have the DUT isolated, so in this case the defective PS.
By the way, i will try to move this SPS4900 repair part to a new thread, so don't be surprised if it is moved later tonight to this thread:
"Manson SPS4900 Switching PS repair"
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
Quote from: Itsu on 2020.10.07, 19:56:37
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.
Ok, i will try to install at least an isolation transfomer for the DUT.
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.
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.
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.
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
Quote from: Itsu on 2020.10.09, 11:29:38
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.
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
Quote from: Itsu on 2020.10.09, 14:57:30
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
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
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.
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
Quote from: Itsu on 2020.10.10, 20:20:43
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 ?
Quote from: Itsu on 2020.10.13, 19:23:48
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 (https://youtu.be/WuYdsStS1MQ), but the eddy currents melted it eventually when the power was increased.
Quote from: Itsu on 2020.10.13, 19:23:48
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 ?
(https://en.aet.com.pl/Portals/en/ZelNet_HermesNet/ProductImages/300x240/E-020-006-0020%2000531-RDZEN-E20-10-6-F-887-S-0-20-ROHS.JPG)
QuoteDid 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.
QuoteDoes the core look exactly like this ?
Yes, its the same, only half thinner and the "legs" longer, very fragile.
Quote from: Itsu on 2020.10.14, 07:48:35
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.
You mean like in the website under "Measuring initial permeability mi"?
http://home.scarlet.be/on9cvd/E-Zelf%20meten%20aan%20spoelkern%20materialen.htm
Itsu
Yes
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
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