Advantest U3661 Err 900: NO LOCK DET

[Cyclotron]

I picked up this Advantest U3661 with a failure to start-up issue.  I believe I solved that, but now that it boots up, I get this error "ERR 900: NO LOCK DET".

I haven't found any information on the error or possible solutions.  Anyone with knowledge or a hint on a path forward would be great!

[m k]

It's a big number.
U3700 series has different numbering system but sort of there has warnings.
R3267 numbers end to 6xx file operations and extra 800 is information that process is in progress.

Maybe it's a message that no signal can be detected.

[Cyclotron]

It's a big number.
U3700 series has different numbering system but sort of there has warnings.
R3267 numbers end to 6xx file operations and extra 800 is information that process is in progress.

Maybe it's a message that no signal can be detected.

Based on what I have found, this error is related to the inability to lock with one of the PLLs.  I understand it impacts the lower frequencies.  I have had several other projects on my desk and haven't gotten back to this one, but will tomorrow or the next day.  There isn't much information on the cards and/or schematics.  I will try to isolate which I/O card has the issue, then try to find the PLL circuits on it. 

The other issue is that there is no easy way to operate the unit while having access to probe the internals.  I may explore disassembly of the chassis completely to see if I can get the bits to run on a make-shift jig.

I'm also curious what a shop might charge to repair the unit, as there appear to be some out there that still work on it.

[Cyclotron]

Adding some pictures of the boards.

[Cyclotron]

More pics

[Cyclotron]

More pics

[Cyclotron]

more... uh, pics

[Cyclotron]

last

[m k]

Dedicated jig for a single machine is not very practical, if not easy to do.
Maybe you can do something that allows a star like connection of modules.
I've used extra wires, somewhere is always a route out.
If you've not done so, be noted that it starts annoying pretty fast if you need more wires too often.
In your case even more, if low noise is important.

U3461 ERR 900 is Center frequency setting problem.

[fabovero]

Have you solved the no lock problem ?? I found this problem on 5 u3661 , it's a congenital disease on these devices, it's due to a degrade of microstrip print circuit. If you like i can send  pictures and suggestions about solving no lock and other trouble

[Cyclotron]

Have you solved the no lock problem ?? I found this problem on 5 u3661 , it's a congenital disease on these devices, it's due to a degrade of microstrip print circuit. If you like i can send  pictures and suggestions about solving no lock and other trouble

Feel free to provide anything you might have.

[fabovero]

OK, before to send you the pics (in this moment are not here) i found 2 problems in the3661:
- lock error is normally due to 3840 mhz 2nd if oscillator, you open the visible lateral cover (a lot of M2.5 screw) and measure the pll voltage
if it is not locked you will find a voltage around 9 Volt, in my devices i have solved removing the trimming plate, in this way the pll voltage decrease 6-7 volt and the pll is locked.
- after solved pll error you can verify if the input level is ok after calibration, if yes   you can close the box, if not the istory become very complicated. If you find a loss of signal in band 0-3200 MHz, the gulty is the filter of 2nd if (4061mhz); if you find problems in band 3000-26000 you must go in a church and put a candle to yor saint !

as soon as possible i send some picture to clear th mind

[fabovero]

Here is the ol 3840, the test point to verify pll voltage is near the operational amplifier driving th 2 varicap diode. In some case you dont eliminate the plastic plate, but only adjust to reach the pll locking. I made the hole on the cover to test the voltage at colsed device.
For execute all the works is better to extract th entire module, but it require 5 supply voltage : +5 , +7 , -5 ,+10 , -10 max current 200 mA on +10 V . Since the module dont receive the freq reference from mainframe , you must connect to a 200 MHz generator 0 dBm.
Let me know is clear, sorry for my bad english !!
Next time i send you other pictures with supply connections

[Cyclotron]

Here is the ol 3840, the test point to verify pll voltage is near the operational amplifier driving th 2 varicap diode. In some case you dont eliminate the plastic plate, but only adjust to reach the pll locking. I made the hole on the cover to test the voltage at colsed device.
For execute all the works is better to extract th entire module, but it require 5 supply voltage : +5 , +7 , -5 ,+10 , -10 max current 200 mA on +10 V . Since the module dont receive the freq reference from mainframe , you must connect to a 200 MHz generator 0 dBm.
Let me know is clear, sorry for my bad english !!
Next time i send you other pictures with supply connections

This is fantastic! I plan to get back on that project in the next month and I really want to get it working. I plan to donate it to a charity run to teach kids technology and engineering.

[tszaboo]

Here is the ol 3840, the test point to verify pll voltage is near the operational amplifier driving th 2 varicap diode. In some case you dont eliminate the plastic plate, but only adjust to reach the pll locking. I made the hole on the cover to test the voltage at colsed device.
For execute all the works is better to extract th entire module, but it require 5 supply voltage : +5 , +7 , -5 ,+10 , -10 max current 200 mA on +10 V . Since the module dont receive the freq reference from mainframe , you must connect to a 200 MHz generator 0 dBm.
Let me know is clear, sorry for my bad english !!
Next time i send you other pictures with supply connections

This PLL section looks very similar to the Advantest R3132.
I was able to tune the 2nd LO in that, there is entire user guides about it on this forum. You only need one power supply and some copper tape for that.

I still have problems with the sensitivity of my R3132, it's down and it is compressed. I was suspecting a mixer issue, but actually I would look at the filter you mentioned, maybe the issue is similar.

[Cyclotron]

I wish I had a schematic, but the hints here are great. I will need to go get mine out of storage, but it's probably going to be a couple of weeks.

[Cyclotron]

Ok, I got my U3661 out of storage. I have two projects on the healing bench and then I'm going to take a look.
I still don't have schematics but this reference might be useful.
https://manuals.repeater-builder.com/te-files/ADVANTEST/ADV%20U3661%20Maint.pdf

[M0PUB]

Hi FaboVero, thank you for the useful information. I have an Advantest U3661 which shows low input level in the low band, but works well otherwise. I am very interested in your comment about the 2nd IF filter. Do you have any more information, please? I have not even found the 2nd IF filter yet!
Best wishes, Alan, M0PUB

[Manolo Mos]

Hi. I've bought an Advantest U4941 that its hardware is very similar to your Advantest.
Here is a good thread about the Advantest U4941 that however, its cards are a bit different to mine Advantest, but it don't mind.
https://www.eevblog.com/forum/repair/advantest-u4941-portable-spectrum-analyser-repair-by-fraser/
Well, mine Advantest had similar fail than yours. I found that tiny SMD capacitors were dry, with big ESR, especially 6,8uf 25V. Fortunately it's caps did not spilled electrolyte to the pcb.
The other capacitors had a bit of ESR, but not fatal, more the big capacitor is more good the ESR is.
However, I replaced ALL surface electrolite capacitors and, it worked.
I passed the calibrattion and it worked.
I found another fail, the signal amplitude changed between high and low RBW, but it solved itself changing RBW, I suspect a relay located in the RF module.
I replaced 6,8uf 25v caps by 10uf 50v value. I replaced surface capacitors by normal radial capacitors, placed carefully.
I hope this information can be useless and your Analyzer get repaired. Best regards