EEVblog Electronics Community Forum

Electronics => Repair => Topic started by: RedShoeRider on February 24, 2018, 07:35:23 pm

Title: Anritsu MG3632 Signal Generator Repair - Not quite sine wave
Post by: RedShoeRider on February 24, 2018, 07:35:23 pm
Good Afternoon, All:

Once again, my scalp has finger marks in the from the head scratching I'm doing. Before I make myself more bald....

The Patient: Anritsu MG3632A Synthesized Signal Generator. 100kHz - 2ghz
The Problem: Bought it broken; it wouldn't power up. After fixing a bad power supply (blown rectifiers), I started taking a look around the unit. Here's the waveform at 500mhz:

( (

Not bad, right?
But, here's the waveform at 50mhz:

( (

From about 300mhz to the end of what my scope can trigger on at about 1.2ghz, it looks like a proper sine wave. Below 300mhz, looks like that crap wave. At 10mhz, it looks like a rounded square wave.

Thoughts? Suggestions?
The user manual is just about useless for troubleshooting this, and the service manual isn't findable for this unit. 

Thank you, all!
Title: Re: Anritsu MG3632 Signal Generator Repair - Not quite sine wave
Post by: Chris56000 on February 24, 2018, 09:03:27 pm

I'd begin with looking at as many of the power supply pins to the various analogue chips on the main/signal PCB as you possibly can!

To do this, write down all the device numbers on a piece of paper, then download the data–sheet for each, then print out the page showing the Vcc and GND pins of each, usually the first page.

Remember that in equipment of this quality and complexity, the power supply rails that you can identify on the PSU may not feed all chips direct, there could be additional regulators on the main/signal PCBs, which could be either linear or switch–mode type!

What you are looking for is:–

a) Adequate supplies to each device, both for magnitude and freedom from extraneous noise or ripple;

b) Split (positive and negative) DC power supplies correctly balanced with respect to zero.

There's something distinctly wrong with these waveforms you've shown us that points to an output driver/amplifier issue, as the amplitudes are negligible and the distortion/noise you're seeing is due to the oscilloscope acquisition–stage I/P amplifier stages going flat out to read what little there is!

The 50MHz waveforms I estimate to be 5–7 mV RMS which'll simply get lost in general hash, noise pickup on connecting–leads, etc., etc!

I would have thought such a generator to be capable of 100mV RMS at least over it's full rated frequency range!

Give my suggestions a try and report back to us what you find!

Chris Williams


I don't know how complete it is but I believe Artek can supply the S.M. for the Anritsu MG3631/3632 – there's certainly a pic of one of their CDs they make for that model – I think Artek supply it both as download or cd, and the cost, about £17–£20 (same in USD!) is well worth it – Artek do not supply Chinese–scanned or photocopied rubbish!


I've just checked Artek's site –an S.M. with schematics is available for the MG3631a/MG3632a for $12.50 – buy yourself that and let us know what you find, please note – you'll need to find a means of extracting sections of schematics (print/scan or use a free pdf extract tool) to post in order for us to advise please!
Title: Re: Anritsu MG3632 Signal Generator Repair - Not quite sine wave
Post by: RedShoeRider on February 25, 2018, 02:10:55 am
Hey Chris:

THANK YOU! I didn't know about Artek's site; It didn't come up with a Google or DuckDuckGo search. I ordered up the schematics for the unit a little while ago; I'm excited to see how everything bolts together in there.

And an equally large thank you to you for the pointers in troubleshooting it. I usually do things like stereo's and amplifiers, not my test equipment itself. I'll hopefully have the schematics tomorrow and can start getting in to it.

Title: Re: Anritsu MG3632 Signal Generator Repair - Not quite sine wave
Post by: Chris56000 on February 25, 2018, 12:32:36 pm

My pleasure – please keep us all posted how you get on – we're all here to help!

Chris Williams
Title: Re: Anritsu MG3632 Signal Generator Repair - Not quite sine wave
Post by: RedShoeRider on March 26, 2018, 01:42:54 am
Good Evening, All:

To close the loop, as they say: problem solved!

After going though all of the power supply rails and finding only one odd thing, which was the 5V rail on the Main Loop board was reading about 3.2v, my search narrowed to the signal path itself. Tracing the signal though was fairly easy, as Anritsu laid the boards out in a very logical pattern and everything is labeled well. Where it all went wrong was Q216; the signal went from a clean sine wave to a distorted mess. Q216 is a Mitsubishi MGF0904A, a GaAs FET amplifier that acts as a preamp feeding the main amplifier. Turns out that the FET had partly failed, so it was distorting the signal at the same time it was amplifying it, and was conducting to ground, dragging down the 5V line that fed it. R289, which is a 2W 22 ohm resistor that current limits the FET, was getting hot enough to cook hot dogs on.

Got a replacement FET, put it in.....and we're in business. A sine wave the way Anritsu indented.

My thanks to you, Chris. Couldn't have done it otherwise!