EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: OH2LIY on November 05, 2018, 07:35:49 pm
-
Hello, once again my R&S FSIG SA refused to work. This time symptons was just blinking stby led. This was 3. or 4. time when this SA was broken. Every time I have managed to find and fix broken components.
Once again I grabbed manuals and find that there is only very short description of power supply functions, not even block diagram, only voltages&pins was described. Just to remove PS from SA takes closer to one hour. After that, fun begins. Opening power takes other hour... After long and hard thinking and testing, I did managed to find short circuited ceramic cap. Capasitor was between 4/8 pin on CA258 (LM358) and it short circuited +12V and -12V lines.
After replacing cap, stby power started nicely and also main power started after pressing front panel main switch!
-
Well done!
Sent from my WAS-LX1A using Tapatalk
-
Long time not seen our PSU for R&S 8)
Must be almost 15 years old :-+
Can you please take a picture of the label at the side of the PSU
-
My mistake, I didn't take photo from label (actually, I did this repair over month ago and last weekend was time to fix HDD, and again I did have to remove power and so on). I'll take photo next time (hopefully not too soon)...
Ramppa
-
Huluvu, forget to ask, but do you have schematics for PS?
Ramppa
-
Little bit long time, but PS is:
Ascom Frako
77-953-0300
R&S 1043.9941.02
Ramppa
-
I have R&S ZVK which wouldn't power up. In fact, there is zero reaction after trying to power on. The power supply is the same model Ascom 77-953-0300. The fuse is good. How do you try to power it on to see if the power supply is dead or not? I see there are two bundles of power cables and one ribbon cable. Do I need to fake a connection to emulate the motherboard, and what simulates the power on button? Any tips would be appreciated on debugging it.
Thanks in advance for your help.
-
Zcab, the service manual for the FSEA (using the same power supply) shows the test procedure (p. 339 ff). The manual is on the KO4BB site.
Operating the power supply on the bench requires the following connections:
Ribbon cable connector:
- pin 4 ( ON/OFF control ) to GND (power supply chassis)
- jumper wire from pin 5 to pin 6 (overtemp. sensor input)
The power supply requires a minimum load on the 5.2 V (digital) and 15 V (analog supply) outputs in order to operate properly. The required minimum load currents are 2 A and 0.1 A. Connect a 2.7 Ohm load resistor to the 5.2 V output ( square block connector, pins 1 and 4, see below ), and a 150 Ohm load resistor to the 15 V output ( single-in-line connector, pins 3 and 5 ).
I used a 12V car headlight lamp as a load for the 5.2 V and a 75 Ohm resistor for the 15 V output.
Here are some personal repair experiences ( power supply had been running continually for many years in a FSEA spectrum analyzer ):
- 15V output was unstable ( AC component ca. 5 V pp with a few hundred Hz ). Output stabilizes after a few minutes operation. Caused by defective 220uF electrolytics in the 15V regulator.
- -15 V output shows very high AC component ( several volts, a few ten kHz ). Defective 220uF electrolytics.
- 12V standby output collapses after switching on. Defective 1000uF electrolytic in the auxiliary power supply.
All of the above problems werde caused by defective capacitors of a particular brand (dark blue coating).
Good luck !
Martin
-------------------------
Here is the pin-out ( but please check again with the service manual):
Cable W20 (ribbon cable, 16-pin connector):
1 sysreset pulse output
2 Power Good (Lo) , equivalent to the green LED
3 Line trig out (50/60 Hz)
4 activation input ( GND: ON ), goes to front panel power switch
5 temp sensor ref inp.
6 temp. Sensor LM335 inp. > 3,35V: Shutdown
7 fan -
8 fan +
9 fan -
10 fan +
11 nc
12 GND deactivates protection functions (for service purposes only)
13...16 n.c.
--------------------------------
Cable W21, single-in-line connector
analog supply
X21/1 5,5 V out
X21/2 5,5 V out
X21/3 Analog GND
X21/4 Analog GND
X21/5 15 V out
X21/6 15 V out
X21/7 Analog GND
X21/8 -15 V out
X21/9 Analog GND
X21/10 +28 V out
--------------------------------
Cable W23, square block connector
digital supply
X58/1...3 5,2 V out
X58/4 Digital GND
X58/5 12V Standby out
X58/6 12V out
X58/7...9 Digital GND
-
Hello OH2LIY,
Thank you for this exelent post.
My FSEA had exact same capacitance broken.
It seems to be a common issue for this age?
Replaced with 100nF/100V and analyzer returned to be alive.
Br.
Aheld
-
I'm in the middle of a PSU-repair right now. A schematic would be great...
When you check these, look at pin 8 of the UC3845. There are at least 4 of these ICs in the PSU and I had one broken. It should be 5VDC on pin 8.
I'm now changing all the electrolytics. Have 12V-standby to operate and power to the fans, but still shut down after a couple of seconds...
Regards Jens
-
Hi Jens,
The PSU of my FSEA is also broken a few days ago.
I lost the +5.5V on X21.1/2. The PSU switches off after a few seconds.
The capacitors seems to be ok, but I will change all the electrolytics.
At the moment I'm wondering because I can't follow the 5.5V output trace on the pcb backwards after a larger inductance.
Maybe we can exchange some of our experiences during the repair. That would be helpful.
Regards Markus
-
Hello Markus,
I'll be happy to share my information but I don't think we have exactly the same fault, or mine is more faulty than yours.
Just be careful when you change any components. I think it's a 4-layer PCB and it's easy to break the lines inside the PCB...
I have changed all electrolytics on my board and a check with RLC-meter didn't show any that was really bad. Seen to the risk to damage the board when replacing, it might not be the first thing to do - if there isn't any obvious faulty/thick caps.
The unit is a real pain to get in and out from the chassis...
I get the standby-power of 12V at X23.5 and this is the only voltage that really works. I have 6V to the fans and they spinn when I start.
My 5.2V at X23.1 is only up to 4.1V and then the PSU shots off, (and fans don't spinn). If I load with 2.5ohm I only get 0.5V. (Load with 100ohm at the same time on 15V at X21.5.)
Hard to follow the wires... check the voltage on UC3845 at pin 8. There are 3 of them and pin 8 should be 5V. (a voltage-reference and one of my ICs was bad)
I have now changed a double diode, (with the PCB on the table and fans pointing from you, it's the one at the right hand side and the outerside of the small cooler that the right fan is blowing through.). I thought I've checked it earlier, but perhaps not.
I now have one more that might be faulty. Could you check the doubble diode, (on my attached picture it's the second from the main transformer on the left hand side of the board (towards the cooling). It's a BYV72EW-200 and with a ohm meter I get about 7.95ohm between the pins and with diode-test more or less 0V. Could be a transfomer but could also be another faulty shottkydiode...?
Please note that I don't have any schematic so the marked things in the attached picture is what I assume/think is right.
Regards Jens
-
Hello Jens,
My power supply is working again!
It wasn't a big thing. I changed the LM358 on the little plug-in pcb as shown in the picture from OH2LIY in the first post here.
Now the psu starts and does not switch off after a few seconds. All voltages are back.
However, I still have strong ripple on the +/- 15V lines. I will change the 220uF as Martin reported above. Thank you for your advice, I will only swap the necessary capacitors.
I checked the double diode BYV72: With diode-test I measured about 0.32V in forward direction and "OL" in reverse direction. The outer pins are connected together. Probably the DC output.
You should replace the diode first!
Thank for you picture with the labels! I know it is reverse engineering but it's very useful for further repairs.
While I'm waiting for the capacitors, I can do some more measurements for you. Let me know if you need anything.
Best regards
Markus
-
Congratulations Markus!
I've been out traveling and will not have any time for this repair until mid February. I'll order some BYV72 and test to exchange it. Many thanks for your information.
What voltage do you have to the fans when you start the PSU? You can easily measure this on the upside-down PCB at the connector.
Do you load it in some way when you start up the PSU, as adviced earlier in this thread?
73 Jens
-
When suddenly the fan noise in the lab goes down, you know something is wrong. My FSEB just shut down and at first I thought it overheated since it sits in a corner.... so I blew a large fan at it for a while and tried to restart, but nada. It turns on, starts to boot and then shuts down 5 seconds later. Top and bottom covers were off anyway since I'm hunting another error, so I just flipped in on the side and started measuring the voltage rails on the bottom main board. All were good except the 5.5V analog which starts at 5.5V and then rapidly drops to some ~3V when the supervisor circuit has enough and shuts down.
So, out comes the PSU which by now I can do in the dark with my hands tied behind my back :)
Taking off the top cover I do smell something "electrical", but nothing is obvious. Thus I hooked up my PSU test jig, preloaded the rails as per the service manual and tried again - same thing, the analog 5.5V drops and it shuts down.
Since this was before I found this thread I had to do my own staring at the PCB and following the traces but eventually I found out that the little add-on PCB is the DC/DC for the 5.5V analog. And it did smell a bit "burned", mainly the inductor which showed some signs on overheating. At first I thought this was a transformer, but it has only 2 connections and measures as ~30uH, so I am hoping it is still good. Next I desoldered the MOSFET and the dual diode, but these also seem OK. The inductor that got too hot is right after the diode rectifier, so if it got hot something must be drawing too much current to ground. So out came the three 1000uF/10V caps and lo and behold what is there on the PCB?
After some scraping and cleaning, the resistance from 5.5V to GND went back from the Ohm range to MOhms. The PCB is pretty much eaten away around the via for the positive cap lead. Luckily ground is on one side and the 5.5V is on the other, so a repair should be fairly easy. Ordered new caps and UV curable solder mask, I'll be back with hopefully good new when these get here...
-
And now for the conclusion:
- repaired the PCB with UV cure solder mask, the UV lamp did not arrive but the sun was shinning 8)
- soldered the replacement caps
- put it all together for a dry run, 5.499V analog on the meter :-+
- reassembled the PSU and back in the case, the FSEB is one happy camper again
-
Hi. Does anyone have the part numbers for the 9 way mating connector? Please.
I need to test my PSU.
Thanks
MPC
-
I bought this one: https://www.mouser.de/ProductDetail/TE-Connectivity/350712-1?qs=OHKvVYVKKrfVmRRZ4oQI7w%3D%3D (https://www.mouser.de/ProductDetail/TE-Connectivity/350712-1?qs=OHKvVYVKKrfVmRRZ4oQI7w%3D%3D)
-
For R&S boat anchors that don't boot up see this thread:
https://www.eevblog.com/forum/testgear/rohde-schwarz-fsae-30/msg3238578/#msg3238578 (https://www.eevblog.com/forum/testgear/rohde-schwarz-fsae-30/msg3238578/#msg3238578)
The problem was a dead resonator on a PIC chip on the CPU board. Symptoms are a black screen, no boot, no sound, almost seems like a dead power supply
-
Hi. Do not use the Molex version of this connector the crimp pins are too large. The TE Mate-N-Lok works okay.
https://uk.rs-online.com/web/p/pcb-headers/0849041
I am having a problem with the PSU shutting down when switching from standby to ON. The fans give a short rotation, then shutdown. I have noticed if I leave the analyser on for several minutes in standby then the unit will power up.
I have setup a test jig for the PSU and found the 15V is going over voltage. Please see oscillogram.
I noticed in this thread there is mention of the UC3845 reference going faulty. Any idea which is the UC3845 controlling the 15V.
Thanks. MPC
-
looks to me like +15V and +5.2V appear briefly, then go out?
-
Yes but you can see the 15V is heading towards to 20V which is tripping the PSU.
I have remotely power each UC3845 device and checked the reference voltage and they all show 5V.
Any thoughts greatly appreciated.
MPC.
-
My findings so far....
Using the picture attached. The RED circled devices all run from the 12V standby. The yellow circled device runs from a special 25V supply which is developed using the 15V supply with a 10V shunt regulator on top to make 25V. The 2K2 provides the shunt resistance from the main 45Vdc.
The yellow circled device is powered from this 25V supply and also appears to be used to check for under/over voltage on the 15V supply.
For this PSU unit the TL431 has gone open circuit. Not sure that this is causing the 15V over voltage error but I hope to replace the TL431 and test.
Does anyone know if you can run the PSU without the auxiliary boards?
MPC.
-
[attach=1]Finally working .....
In the end the original TL431 was working ok.
The PSU unit is now working. It must have been a dry joint because soldering wire probes onto the PCB to test the voltage/current error amplifier circuit appears to have fixed the issue. The unit started up okay. This would make sense as the PSU would start if left on for several minutes.
For future reference I have update the photo with the circuit blocks identified. Also attached is a partial circuit diagram of the 15V DC/DC converter and the oscillogram of the units start-up.
Green is the 15V Supply
Yellow is the 15V into pin 3 LM358
Purple is pin 7 of the LM358 error amp
Blue is pin 1 of the LM358 error amp
MPC
-
AND finally.....
The unit failed start-up the next morning.
The first oscillogram shows the 15V is overshooting causing the PSU to shutdown.
Green is the 15V Supply
Yellow is the 15V into pin 3 LM358
Pink is pin 7 of the LM358 error amp
Blue is pin 1 of the LM358 error amp
The TL431 10V fail oscillogram shows the green trace 45/50V supply working which is connected to the top end of the 2K2 shunt resistor
The Blue trace is the cathode of the TL431 which should regulate to 10V above the 15 supply. Here you can this is overshooting the 25V and reaching 30V
The Purple/yellow trace is the 15V. The difference between the Blue and Pink should be 10V but this overshoots. This suggests the TL431 shunt is not regulating.
Removing the TL431 and some parts around the TL431 to isolate the reference input track into the TL431 showed a resistance. On the top of the PCB around the TL431 there was a glossy layer of something which looked like lacquer. Also a copper track running close to the TL431 was black.
The PCB photo shows a conductive coating over the PCB around the TL431. The corrosive liquid looks to have leaked underneath the solder resist and caused the copper track to corrode to black. This conductive layer had leaked onto the TL431 refence input and caused the TL431 to fail.
The final photo shows the leak and black copper corroded track.
Cleaning up the top surface and replacing the three caps close to the TL431 has now fixed the issue. The problem could have been solved just trough observation!
TL431 10V fail photo.
Green is the 45/50V Supply
Yellow/Pink 15V supply
Blue is the 25V supply reaching 30V because the TL431 10V shunt has failed.
Interesting link https://electronics.stackexchange.com/questions/102048/what-effects-might-an-electrolyte-leak-have-on-nearby-electronic-components
MPC.
-
Hello all.
I repaired this type of PS of my ZVC 5 years ago.
Back then, I replaced two leaky electrolytic capacitors and cleaned the pcb.
Unfortunately I didn't replace all of them.
Now it starts to fail again sometimes and I want to replace all electrolytic capacitors.
As so often it seems that they are a also weak point in this PS.
Does anybody have a list of electrolytic capacitors with capacity and voltage values, so that I can buy them before take my ZVC out of service and disassembling the PS?
Regards Matthias
-
Hello. - Hello.
Attached is a picture of the electrolytic condenser information mounted on the power supply board.
There are two types of power supply boards. One with an electrolytic condenser of 10uF/50V and the other with 47uF/50V. (The part described on line 2)
We recommend replacing all electrolytic condensers on this board except the 470uF/630V. The total number of capacitors is 40.
The defective LM358 is caused by a leaking electrolytic capacitor (220uF/25V) attached to the sub-board above it. Therefore, if the LM358 is defective, the surrounding components should also be checked and replaced.
There was a mistake in the text, and the following is the corrected content.
condenser => Capacitor
47uF/50V=>47uF/63V
-
Hello,
I try to repair this PSU as well. I got a signal analyser with completly dead PSU without any blown mains fuse at the socket.
After disassembly there was only one visual dead capacitor on the PFC side. After removing it the Short disapeared.
Without the SMD capacitor the big mains capacitor mounted sideways only charges up to about 19V at regular line voltage -> nothing happens, -> only very small mains current consumption
When I charge the capacitor with an external DC PSU up to 70V the 12V standby voltage comes alive.
It seems to me that the PFC circuit responsible for charging the main filter cap is dead.
What do you guys suggest to continue measuring. I am a bit lost without a schematic and layout.
Shouldn't the main filter cap always charge to line voltage even the PFC is off? The PFC charges the capacitor only higher?
I realy appreciate any help, thank you in advance
-
Hi Andi555!
Next week I will find broken PS and have a look for PFC.
ML4821 is responsible for PFC work.
Please find in attachment the connection diagram for PS standalone test.
73
Harry
-
Hello,
Without the SMD capacitor the big mains capacitor mounted sideways only charges up to about 19V at regular line voltage -> nothing happens, -> only very small mains current consumption
Hi!
Even with inactive PFC, the voltage across the output capacitor should be up to peak rectified voltage. But you have significantly less.
Let's look at the example diagram from datasheet to see what could cause this.
ML4821 operates with the current sence resistor. Most likely it overheated and burned out as a result of SMD capacitor short.
Could you find it on PCB and check.
-
I found PS with a similar malfunction. SMD capacitor also burned out.
Unfortunately the customer tried to fix it himself but did it carelessly.
The circuit is a little more complicated (please see the picture). There is a soft start on the IGBT transistor. Check it and both MOSFETs.
-
Thank you very much,
I will test the boards with you info the upcoming days.
There is a IGBT / Mosfet ? beside the rectifier which switches the recified DC (Minus) filter cap to Ground (Rectifier) -> Maybe this is some kind of overvoltage protection ?
-
Hi Andi555!
I added soft start resistor to the diagram above. In my PS it burned out and was replaced with 20 Ohms. Through this resistor a smooth charge of the 470uF capacitor is carried out. Then the transistor opens shunting the resistor.
Most likely this resistor burned out.
Check all the parts circled in red.
-
Thank you Harry for your detailed description!
One hint: The diode symbol right beside the 6,3A fuse is drawn in the oposite orientation. I checked all the marked parts and indeed the 20Ohm start resistor is burnt to "open". The soldering joint also looks like it heated up for a long time.
For the test setup I shorted the resistor and turned up my variac slowly .... -> The PSU now works! :clap:
The burned capacitor is still not in place, maybe it is not realy necessary. Do you know the value? Otherwise I will leave it open.
-
Hello
I soldered one like this.
https://pl.farnell.com/murata/grj55dr72j224kwj1l/cap-0-22-f-630v-10-x7r-2220/dp/2218826
-
Hi Andi555!
I corrected the drawing.
The similar SMD capacitor on the bottom left has 0.15uF.
Use the one suggested by Jarek.
Could you make a photo of unsoldered soft start resistor to find a replacement.
-
Hi Harry,
thats the charge resistor on the primary side of PFC. No visual burn marks, but "open".
-
This is Vitrohm fusable resistor.
https://www.vitrohm.com/content/files/vitrohm-ktf-wirewound-fusible-resistor-datasheet.pdf (https://www.vitrohm.com/content/files/vitrohm-ktf-wirewound-fusible-resistor-datasheet.pdf)
You can use the following A5MC
https://www.uchihashi.co.jp/en/fuse/a5mc/ (https://www.uchihashi.co.jp/en/fuse/a5mc/)
-
These special resistors seem to be very hard to buy from a "normal" supplier for low quantities.
Its a bit risky to replace it with a regular 5W ceramic power resistor and rely on the mains fuse...
-
Don't be afraid to install a 5W resistor without a fuse.
Your resistor was burned out abnormally. Its jumper didn't unsolder during long heating as it was intended, but the coil instantly burned out due to a short circuit.
This is because it has thermal protection against breakage of the IGBT transistor. In this case, it turns out to be connected in series with the PS and a lot of power begins to be released on it.
Below you can see the diagram of this unit from another R&S serial power supply. There a resistor without protection is used. It burns out when the PFC key is broken.
For fun I disassembled both resistors.
You can find fused resistor on Aliexpress. It is original.
-
Hi,
amazing news -> Analyzer is working !
I changed the resistor to a diy fused one (-; With low melt solder it should open if the resistor gets red hot. I hope not.
There are two new 100nF 630V capacitors as a replacement soldered on.
At the first start the analyzer found no HDD and the BIOS stopped. At the second try and several times later it was no problem, the analyzer was booting well. Date and time was all right. BIOS Battery seems to be still ok. Next step is to clone the HDD for backup and clean / lubricate all the fans. Some sound a bit rough / noisy.
Thank you for the support!
-
Hi Andi555!
Well done.
Congratulations!