Monday, September 23, 2019

Marconi 2024 RF Signal Generator Error 508: Amplitude modulator unlevelled Repair

I have recently bought a used Marconi 2024. But instrument was partially faulty having issue a minor issue with LCD screen and a comparatively major issue with Error 508 : Amplitude modulator unleveled. So it needed a repair First i will try to fix issue Error 508 because replacement compatible LCD will arrive later.

This post is going through will be mostly what to look for when you are attempting to fix same issue in your marconi/IRF or now aeroflex 2024 or similar marconi 2023.

This article will quote  service manual for this instrument multiple time, service manual can be easily found with just simple internet search or on ko4bb

As i have already fixed my issue now i know with my instrument issue was with carrier RF signal level being too low. But issue with other instrument many be different so  i will still describe here how i traced to the faulty component.

I assume all power supply voltage are fine with voltage and ripple.

There are two version of service manual available. old service manual have There are a quite a few error in newer service manual's schematic.

As per Maintenance manual  Error 508: Amplitude modulator unlevelled appear when pin diode based modulator loop is running into issue with either AM level or the level of carrier RF signal or detection circuit of AM modulated RF signal.

Service manual says  "If the error only appears at some frequencies and levels and not at others, the most likely cause is insufficient RF level present at the input of the modulator. When this  occurs, the AM distortion and AM accuracy will get worse with the loop running out of the leveling range. One of the main reasons for insufficient RF input level is most likely to be a failure of  band-pass filters which divide the frequency from 1.2 to 2.51 GHz all the way down to 10 MHz. The following table lists the changeover frequencies for each filter:"

No. Carrier frequency range Filter type selected
1 10 kHz to 10 MHz BFO band
2 >10 MHz to 18.75 MHz 18.75 MHz LPF
3 >18.75 MHz to 37.5 MHz 37.5 MHz LPF
4 >37.5 MHz to 75 MHz 75 MHz LPF
5 >75 MHz to 150 MHz 150 MHz LPF
6 >150 MHz to 300 MHz 300 MHz LPF
7 >300 MHz to 420 MHz 420 MHz LPF
8 >420 MHz to 600 MHz 600 MHz LPF
9 >600 MHz to 1200 MHz 1200 MHz LPF

if you have error only any one of these ranges the you problem is relatively easy to fix. Filter section is on sheet 7 of 10 (issue8) AA1 board in the service manual and each filter band has it's own network of passives and diodes. These filter are individually switched using LM324 opamps IC604 IC605 and IC606 on same page. Opamps are driven as comparator using 3 to 8 demultiplexers IC603 and IC604  when a particular filter is active with SWn pin will go low to around -11V. 

Here is the table how these SWn pins are activated by digital circuit.  you may want to verify if you filter band related issue.

B3 B2 B1 B0 filter band Active demultiplexer Active SW pin
1 0 1 0 10 - 18.75 MHz IC603 Y2 low SW10 SW10 active
1 0 0 0 18.75 - 37.5 MHz C603 Y0 low SW9 active
0 1 1 0 37.5 - 75 MHz IC608 y6 low SW8 active
0 1 0 0 75 - 150 MHz IC608 y4 low SW7 active
0 0 1 0 150 - 300 MHz IC608 y2 low SW6 active
0 0 0 1 300 - 420 MHz IC608 y1 low SW5 active
0 0 0 0 420 - 600 MHz IC608 y0 low SW4 active
1 1 1 1 600 - 840 MHz IC603 Y7 low SW3 active
1 1 1 0 0.84 -1.2 GHz IC603 y6 low SW2 active
1 1 0 0 1.2 - 2.51 GHz IC603 y4 low SW1 active

I was not having filter band specific issue. With my instrument error 508: appears on these frequency

370 - 447 Mhz
583 - 600 Mhz
705 - 720 Mhz
843 - 849 Mhz
1125 - 2500 Mhz

As you have follow with table above this does not fall into any specific band. so i have to look aways from filter issue to other circuit driving AM.

AM modulation RF circuit can found in the manual on Sheet 6 of 10 (issue 5) of AA1 board. Detector and compartor is also on same page with IC802(a) and IC501(b).  ( Caution: There a error in newer version service manual on this page. compartor IC802(a)'s negative and positive input are swapped )
here is the view of schematic around IC802(a) form original service manual.

If error appear on all frequency bands then diode D510 or D511 may be faulty. in my case i have replaced both the diodes, voltage level and error 508 was still there.

AM modulation digital circuit is partially located on Sheet 4 of 8 (issue 5) of AB1 board. PIN 11 of PLD  connector (connector connecting RF and digital board) connects RF sections's IC802's input to pin 14 of IC33 through a resistor. IC33 is driver of AM signal (with offset)  into RF board.
When As AM signal generated internally it generated by DSP with DAC on Sheet 2 of 8 (issue 5) of AB1 board.

DSP drives DAC IC13 for AM signal and DAC drive opamp filter IC207 and IC15 This filtered signal from IC15(pin7) get fed into Analog MUX IC31  which drives reference of another DAC IC35(a). 

From Sheet 4 of 8 AB1 AM offset 

From Sheet 2 of 8 AM DAC and AM Filter 
I have verified that Both the DACs Mux and Opamps were working as intended in both of the cases when IC13 was generating AM signal or IC13 was idle.

My though while looking and Digital AM section was the offset of around 900mV on pin 14 of IC33 input to IC802 is incorrect being only at ~900mV  while service manual documents "possibly incorrectly" says it should be around 1.4V. I have found no evidence at all that offset should be at 1.4V, i have found many  mistake in the document this many also just be one.

I have done LT spice simulation of both DAC filter and offset generator both circuits were same as simulation.

This offset is actually a calibration parameter of the instrument and accessible to user for software calibration. under calibration menu with parameter number "Util 112" Cal AM adjust DAC A: xxx for 30% AM is directly setting AM offset. just re-calibrating for this "may" fix issue.

Or you if just want to hide error message setting it to extreme end may make error "message" disappear  while error may be still there and you will see effects of it when you try to do AM Modulation. i verified setting it to 0 make my errors "message" disappear for all then ranges expect for 1.2Ghz -2.4Ghz band. i verified AM was not usable.  Because ultimately AM level is incrrect after pin modulator.

I verified all three modulation related DAC for correct operation. and they all were setting  correct voltage. at first i had hooked my scope to digital bus of DAC and mux later in process i find out there was no need to do any protocol decoding on these DACs as most of them are one or another instrument parameter/settings and most of these instrument parameter/settings are accessible in instrument's calibration menu.

That was all what can be verified in the Digital section.

Now when i look at RF section i probed bias voltage on every RF transistor and mmic. Most voltage are mentioned in the service manual or can be found using part's data sheet.  A simple multimeter can be used to verify this.

By this time was  sure all RF transistor and MMIC in moudulator and output section were functional

Then i looked at filter for correct filter to be activated. With table above mention in this post it is little easy to determine if correct filter is activated.
Later i was probing filter input and output section together with output on D510. my siggen's first error point was with 370Mhz so i was probing with scope diode D510 output (even a multimeter is enough to measure at this point)  and at the same time going from 360 Mhz towards 370Mhz error point.
As i was at 360Mhz input of compactor was was good +< - but as i went near to error point frequency voltage on D510 output started to drop form 30mv to slowing getting to point when - > + make error messare appear

Same behavior of carrier signal dropping in amplitude was also there on the filter input  and both of the divider output IC601 and IC602 (sheet 7). At this filter stage i was using my 500Mhz scope while instrument was outputting only 370Mhz.
Spectrum can also be used to verify carrier signal at filter stage. or you have to make a little rectifier circuit (like with D510) and solder it were you want to measure with multimeter.

After all this i checked that amplitude into divider/filter section was actually going quite a bit up and down as i adjust frequency. so i traced back the signal and found out in the carrier generation section  Sheet 1 of 10 AA1 board IC106 was to be faulty, i replaced it then it worked with a catch.  Before changing the IC106 error 508 was present on almost 60-80 % range of 2.5Ghz total but after replacing IC106 , one or two really small  frequency ranges  (<1% of total 2.5Ghz range) was still which do have correct output amplitude but still show error 508, and last very very time consuming catch was, as i was testing the whole thing, metal lid was open and after i fedup i closed the lid and whole range was error free.

Output Bias IC106

Input bias IC106
Sheet 1 of 10 AA1 Board Carrier Gen Section

If any of mmic in carrier gen section is faulty you may not need scope because service manual mention bias voltage for IC104-IC106 to be ~4.5 at output and 1.5V at input. you can measure with multimeter. if you need to check gain at each stage you will need a fet probe for your SA. i was always using a cheap 10$ fet probe on other projects as well without any issue.

Document is giving us part number of IC106 to be MAR2-3 This part seems like not very famous so there is no old stock. And it is a 2Ghz MMIC, while signal will reach 2.5Ghz on this signal path. Documentation is wrong or service manual was is for other 1.3 Ghz model.

Which my very little experience marking "A02" on these tiny mmic packages is quite unique .
Equivalent part number will be MAR-2   or  MAR-2SM  or   MSA-0235  or  MSA0285  or  MSA-0286 or  MSA-0270 ( while MAR-2 and MAR-2SM are 2GHz parts)

But if you look at the board it self you can a clear indication that this part is MSA-0286, because  IC104-IC106 looks exactly same as IC101 and as per schematic IC101 is MSA-0286. And also IC107 is MSA-0486 with marking also on chip A04, i see no reason why would they go for MAR2-3 in the middle. I have verified on my board itself IC101 , IC104-IC106 are  MSA-0286 and IC107 MSA-0486 may be at some point later in production they have changed IC104-IC106 from MSA-0286 .

Adding to possible hardware version story some where on internet i have seen digital PCB newer model under brand name IRF.  This new digital board had a cpld to all logic stuff and flash based mcu instead of EPROM.

I did not have any of these in stock so replaced with HMC476MP86  it is 20 dB gain block. Part i was replacing is around 12dB Gain but as P1DB limits max output to so i feel (with almost non RF experience though P1db of MSA-0286 is only 4.5db and HMC was 3 time around 12dB P1db) it will work for testing and i verified with a diy fet prob gain HMC476MP86 was having in circuit was from 6 dB to 12 dB depending on frequency selected.

I have bought MSA-270 and MSA-0286 both. Some time later when new LCD display arrive from china i will also replace HMC476MP86 part.

If still have any question please feel free to ask.

No comments:

Post a Comment