Anritsu / Wiltron S331A - Service Manual / Schematic

[rg58]

Hi there,

I was lucky to get hold of one of these Anritsu S331A analyzers, but this one has at least 2 faults.

1) Unresponsive to key press, sometimes beeps (only the ON button always works to turn it ON, but not OFF)
2) Integrator Error during self-test (it shows SWR of 1.0 even nothing is connected to the antenna port)

Funny thing is, trying to turn it off (ON/OFF button), it produces vertical lines in the graph while scanning AND seems to move the "cursor" one step forward during scanning. All this is pretty weird. I already cleaned the keypad rubber which measures good (<200 Ohm), also the keys get about 5V on the PCB.

Does anyone have a Service Manual maybe? Or could share some advice repairing this?

[metrologist]

Many times the integrator failure is due to burned input circuit.

The other issues sound like connectors.

https://dl.cdn-anritsu.com/en-us/test-measurement/files/Manuals/Maintenance-Manual/10580-00008.pdf

[rg58]

Thanks, I will have a look at the input circuit (what I already suspected anyway, but was not eager to touch it before knowing more).

Regarding the keys, I think you need to press them really hard to make them work. What would be a normal resistance of the black conductive rubber pads that press on the copper traces on the PCB? I measured around 150-200 Ohms for all of them. Isn't it possible the property of the conductive rubber is changing over time, and the resistance should be much lower?

[rg58]

Anyone knowing what this part is? Could that be a package of 4 RF diodes for the SWR bridge? It is connected to 51 ohm resistors as well as 1kOhm resistors.

During frequency sweep, SWR shows always 1 (with real antenna connected, or with open end anyways), so there is no (reflected) signal being measured at the bridge or input section. There *is* an RF signal coming out, however, it also looks quite small, well below 0dBm (<1mW).

Error message shown was "Integrator Error".

Would it be safe to measure this sensitive part with a multimeter on diode test?

[metrologist]

What's the board #? It's likely a 3-xxxxx number.

You would use an RF probe and SPA to trace the signal.

The burned boards I've seen have the PCB charred where you see that bodge. That's because the telcom gorillas forget to turn off their PAs before connecting.

[rg58]

Yeah, that probing with the spectrum analyzer was scheduled for today. Last night I removed the soldered brass shield which was a real pain. I was afraid it might not even be necessary, but just out of curiosity I removed it anyway.

The RF output from 25 to 1000 MHz is mostly -13.4dBm with the exceptions -12.2dBm ~300MHz and -18.6dBm ~750MHz (non-linear) which could be related to the fault or the 2 metal covers that I removed which might have de-tuned the circuit underneath. From the output level, do you think that's enough? Looks a little low, not even 1mW. I would imagine such devices to output in the region of 10mW (+10dBm), but this is just a guess and from what I remember from other analyzers I had on my table (I might be wrong about that).

The board number is S400A-D-49222 Rev. E (49222-3B as the sticker says) from 1998. I could not find any number like 3-xxxxx. It might be a little older then.

Attached you the photos of the input section. There are MMICs and something that looks like an ATF... HEMT transistor, which might be the one which is gone.

U35 (which looks like a TO92 transistor) is also pretty close to the input. Because the signal passes over the artfully crafted "SMD bridge" to the right, leading to this mystery component. From there 2 traces continue through vias under the shield to the other side, then eventually to that U35 which is the first active component as it seems. I assumed that the path to the right is for REFLECTED POWER to be measured (missing), the one going straight into the brass-shielded compartment might be the VCO, mixer and power amp stages. That's why I asked first for the mysterious component to the right of the input socket.

[metrologist]

TP9 12V
TP10/12 5V
TP11 2.5V
TP29 and the sampler above is the incident signal
TP30 and the sampler above is the reflected signal
Synthesizer output is TP18, tracking synth is TP19
And look at the RF path around U28

[khach]

Anyone knowing what this part is?

Phase sampling detector like MACOM MSPD https://cdn.macom.com/datasheets/MSPDxxxx-x%20Series.pdf

[rg58]

Ah OK, thanks @khach! Never came across any of those ones yet. The datasheet is very interesting to read.

@metrologist
Thank you very much! This is very helpful indeed.

Here are my results (@14.7V):

TP9 12V   => 11.5V (little low)
TP10 5V   => OK
TP11 2.5V => OK (2.48V)
TP12 5V   => OK

Few more that I tested:
TP3  =>  5.2V
TP4  => 13.4V
TP5  => -7.4V
TP6  => -7.5V
TP13 => 0.0V
TP19 => 9.0V

With frequency set to 50 MHz (unterminated input):

TP29 incident signal => -45dBm @ 49.8MHz
TP30 reflected signal => -46dBm @ 49.8MHz (unchanged with termination 50 Ohm!)
TP18 synthesizer output => -49dBm @ 80 MHz (highest peak), lots of other signals/mixing products
TP19 tracking synth => -41dBm @ 49.8 MHz (-200kHz below signal)

I see a frequency of -200kHz below the set frequency in many places that I probed. All is a bit confusing if you don't have a schematic or block diagram at least.

Signal at the INPUT connector was now -22 dBm at 50.0 MHz (did I just damage my SA input? before it was -13dBm)

RF path around U28 => SMD code "E3" should be an ERA-3 MMIC with input at the dot, output on opposite side

input: -17.0 dBm @ 50.0MHz
output -18.6 dBm @ 50.0MHz

This can't be right! 1.7V DC bias voltage (on output) is too low as well, recommended are 3.5V. I am getting 8V before the bias resistor, so this thing draws too much current! On its *input* I can measure 1.2Vdc, not sure if this is normal. Voltage drop of 6.3V over 196 Ohms equals 32mA which is close to the 35mA as stated in the datasheet. How can that be? The resistor measures correctly.

Continuing through to the next component, the white pill with SMD code "Bd" has the -18.6dBm signal on its input(?) and -40.6 dB on its output(?). So either I am looking at it the wrong way and 2 signals are being mixed together here *OR* this component also attenuates the signal by 22dB and might be blown as well.

The whole unit also makes a rather loud high-pitched beeping sound from some DC/DC converter coils. Additionally there is a rhythmic buzzing/scratching sound when the analyzer is doing something, probably also coming from those coils. Is this normal?

Will continue tomorrow.

[metrologist]

The "white pill" looks to be an NE021.

The device below the U28 silkscreen on a couple of your images looks like it has a piece blown off. Might just be the image?

[rg58]

Yeah, I though this as well right from the beginning and looked at it closely.

This was just some flux residue. Wiped it off, just to be sure.

[rg58]

Thanks for the type of the "white pill"! You are very knowledgeable indeed, sir! Found the datasheet for it, it's an NPN transistor. Nowhere to find this SMD marking code in all my lists that correctly represent this part in this package.

If "E3" is in fact an ERA-3 with something that appears to be a dot (although it looks more like a lasered c in a circle), then the dot marks the input, opposite site is output. The "white pill" besides it marked "Bd" seems to have its base connected to the output of this MMIC. It measured good with a diode tester (2x ~0.77V), but it's still a little strange as it attenuates the signal by 22dB.

In the correct orientation of the letters "Bd" the base would be on the RIGHT, being connected to the output of the MMIC. Find this a little strange too. Could be a DC problem here, so I will do some more AC/DC probing first. Let's see if those guys are clean

[rg58]

The VCO output looks aweful too. I tried with 2 different ground points and 2 different probes. First peak on the left are 50MHz, the marker sitting on 80MHz. Looks like a comb generator, not a VCO 

Supply voltage 11.5V clean
Tuning voltage 14.8V clean (50 MHz)

As for the rest, the transistor and the MMIC, I can't make sense of all that.

*If* I follow the signal from the other direction, everything makes sense and the signal increases, but only if I omit the pinouts of both components. Are the types correct? ERA-3 and NE021? Why would we connect a noisy MMIC infront of a low-noise transistor? All this makes no sense to me. The pinouts must be completely wrong. Although from B to C and E I can meassure 2x 0.7V.

Signal path (backwards?) looking like this 

Transistor C -> 0.75Vdc / -56.6 dBm
Transistor E -> 0.62Vdc
Transistor B -> 0.94Vdc / -32.6 dBm (Vbe only 0.32V?, diode test was ~2x 0.7V)
MMIC output -> 1.72V / -32.6 dBm
MMIC input -> 1.20V / -16.6 dBm

I have the feeling this thing is a complete mess... 

[metrologist]

This instrument uses harmonics for frequencies outside of baseband, so that transistor is used to suppress or enhance those by switching in reactance via bias circuit on its base. Those are the correct part numbers. Set start/stop freq to the the same value in baseband so it's not sweeping. That's an 800-1600M VCO.

Also, look at the traces just on the MMIC output and the two diodes connected to the 51R1 resistors to the right, I think the lower diode is baseband path, and the other path is to the left and through the inductor above the top diode, and this is also controlled by the bias circuit connected to the top side of the inductor.

I believe that the 2N2907 transistor just above the plastic mounting peg to the right of the VCO is the voltage switch to the MMIC, which goes through the 1100 and 1960 resistors.

[rg58]

Thanks a lot! You were right about the VCO.

@50MHz: 1600 MHz / -28dBm - funny, this is the highest frequency that you mentioned (didn't expect THIS high, wow! needed to change SA)
@500MHz: 1000 MHz / -19dBm

Not sure how they do the mixing as these are 600MHz, but set frequency is only +450 MHz. But as you said, they probably use harmonics somewhere in the circuit and do some switching.

As for the rest, will see if I can take more measurements tomorrow  and report back to you. Seems some DC measurements from above were incorrect as the RF is effecting my multimeter.  Having a few busy days here...

[rg58]

So that PNP transistor to the right of the peg seems to be OK, although in-circuit measurements were a bit inconclusive on C-E. At least not shorting, assuming that part is good.

However, the ERA-3 is officially dead. Although from the DC current draw of 32mA it looks OK (35mA according to datasheet), it measures a DC resistance of 73 ohms on it's output which can't be right, but adding up in my calculation now: 7.9V/(196+73ohms)=29.4mA + current drawn on the input. Because the datasheet of the ERA-3+ tells us the (simplified) internal circuitry. There are 2 transistors in darlington configuration inside, plus 4 biasing resistors. The 2nd transistor seems to be shorted, so we are left with 1 resistor to GND probably in the region of... 73 ohms. This makes sense now and explains the signal loss of -30dBm - -16dBm = -14dB instead of a +23dB gain. The datasheet also tells us the supply voltage to be 3.5V at 35mA, but it breaks down to 1.62V at 32mA.

Checking transistor NE021 in circuit again. Diode test: Vbe=0.78V, Vbc=0.78V, Vce=nothing, reverse polarity all 3 nothing. Looks a bit HIGH for silicium, otherwise that part seems to be OK. But why the loss in signal? At 50MHz: Base: -30dBm, Collector: -55dBm, another -25dB loss! DC voltages in operation: Vb = 0.94V, Ve=0.51V => Vbe=0.43V (unusual LOW this time!) and Vc=0.62V. I can't find any RF short on it's output, it only continues to a SMD capacitor of 10pF (as my LCR meter says). Behind that... oh, I see it now. These are 2 switching diodes, behind the upper diode some low impedance circuitry. Makes sense now, this could be about right. The 2nd transistor acts as a power amplifier it seems, so the current is not enough. Hence, my 50 MHz signal nearly disappears in the noise while a weak 49.8 MHz signal appears here now.

Will replace ERA-3 when I received it and let you know the outcome. Hope this solves the problem.

[metrologist]

I don't believe that transistor is biased in baseband. The bias line is from the feedthrough to the left on that small blue resistor.

Consider why that mar is fried and the low impedance circuitry on the other side. There is little to protect the samplers. Also, follow the trace from TP29 to the J310S fet, which is next in line. Then 2N5087S and a few 2N7002S around there. Similar circuit on the reflection side (other side of bridge).

Assuming a mild overpower input, that should cover it. HV destruction then toss it but I've seen these with a melted bridge and black charred PCB so it might be salvageable.

[rg58]

Yesterday I received the ERA-3 and just soldered it in.

Here is the result:

ERA-3 input: -19dBm
ERA-3 output: -5dBm
= Gain: +14dB (good!)
Vcc: 3.05V (good!)
Level at N-socket: -9.8dBm (good!)

All looking good now! Well, except the device still doesn't show valid measurements   The SWR is still stuck down at 1.

The via (feed-through) behind the blue resistor at the base of NE021 ends up unconnected in the middle of a ground plane. Is this a 4 layer board? I can't tell from where the DC is coming. The cap at the other end of the trace is connected to ground, then there is a coil (choke) in a black package connected to it. All this makes sense, but where is the DC coming from?

When I approach such repairs I always think what might have been fried first, so my main interest was in those samplers right from the start. In the meantime I did some diode testing (in circuit) and the 3 diodes in both samplers seemed to be OK, although I am not 100% confident with the results. One could just be damaged, not completely fried. Would need to investigate them more thoroughly, signal-wise.

Next is the transistor connected to TP29 which you identified as J310S. I can't find a datasheet for this. Marking code "6TU" is not in my lists either. I did find a datasheet from Fairchild for MMBFJ310 (6T) and MMBFJ309 (6U), so telling from the marking code its a combination of the two   Assuming the pinout is correct, it has a D-S short. You were spot on once again! Thanks!   As for the remaining transistors around there, none of them is shorted as it seems.

Will try to source the J310 with an "S" at the end or look for the MMBFJ310 which *should* be the same. Once received I will let you know again. Btw, my time for electronics is very limited these days.

[rg58]

Received the MMBFJ310 today and put it in. To my surprise it seemed to be blown again.

That's when I realized it's a J-FET!! Silly me 

Of course I knew it, but somehow overlooked the fact. The J310 is quite common and the name already suggests it. Oh well...

Back to square two...

[rg58]

While the device was switched on and sitting quietly on my table, the display suddenly disappeared for no reason. Nothing is shown on it anymore. The backlight is still working. When I turn the device off and on again, about 3-4 horizontal lines appear for a fraction of a second, same as it did normally when it started up. It still reacts to keypress with beeps and I can turn on and off the backlight.

I have already checked the voltages from the test points in my earlier post and on the voltage regulators themselves, but everything looks fine as far as I can tell. I have also taken out the display again, removed the ribbon cable, wiggled it while it was on, but still no display. 

Has something suddenly died ?

What happened

[z1zmail]

Ciao, hai controllato la tensione al pin 4 del display ( V0)? È la regolazione del contrasto; se  è inferiore ai 10V, sembra spento.

[vishaldotgupta]

 i am not sure if you still need any help.  but

1.  you need to make sure if the system is still booting. to do so, press on button and wait.  do you hear any beep sound?   does backlight switch works.  if yes, press contrast button and try to increase contrast. 

2.  if increasing contrast does not help,  make sure contrast voltage is present.  check LCD data sheet and check  pin no for for contrast voltage value.   if not in limit, the contrast driver could be faulty.

3.  if still not working, check if there is activity on data and clock lines of LCD.  if  it is there, then probably you LCD has gone bad.  if no activity, then system is no booting.

4.  remove PLCC EEPROM, clean contact with plug it back.  Try switching on again

hope it helps

[vishaldotgupta]

the way S331 works is as follows.

The RF injected into the coupler bridge, two measurements are made.  one is called Reference and the other is Reflected.

When for some reason  reference is missing, sweep will be very slow.  and when Reflected cannot be detected, the VSWR will be 1.  and a flat line.

so in your case you need to probe the reflection measurement circuity right upto the A/D converter

[vishaldotgupta]

this most likely is a RF diode.

just try touching its pins using a tweezer.

when you touch does the sweep speed increases and you get VSWR graph?  if yes, most likely it is faulty.  u need to also check the capacitor and resistance around it to be sure

[rg58]

Thanks for all that effort, vishaldotgupta! All correct what you said.

However, unfortunately your reply was 11 months after my last post, so the unit is already long gone. Still not exactly sure what cause the sudden death while the unit was sitting quietly on my table. At the end I probed around and (if I remember correctly) I could trace it back to a faulty negative voltage supply for the display itself, but could not figure out the reason at that time. Something seemed wrong with one of the DC/DC converters.

As the unit was in an overall rather messy condition, I decided not to waste more precious hours on that one.