Agilent N3300A DC Electronic Load Mainframe

[Mad Professor]

Hi All.

I have picked up a faulty Agilent N3300A DC Electronic Load Mainframe.
It is fitted with the following modules.
4x N3302A 60V/30A 150W.
1x N3303A 240V/10A 250W.
1x N3304A 60V/60A 300W.

This unit shows very little in the way of life.
When powered on nothing is displayed, and the fans do not come on.
However if you press the blue shift button then shift is shown on the display, but as soon as you push any other button it goes off.
So far I have only done a little bit of research and testing.
I have found this YouTube video: .
This shows that the unit's fan's and display should work even with all modules removed, however mine does not.
I have also been able to find the schematic here: Agilent_N3300A_N3301_mainframe_eloads_schematic_5989-6688EN.pdf
I have had a very good look around the logic board and I can not see any damaged, blown, burnt components.

So onto some basic voltage checks, and I can confirm the following.
J12 (mains supply voltage) in my case 250vac.
J8 (24volts).
U21 output (5volts).
U23 output (3volts).
U20 output (2.8volts) unless Q3's gate is pulled to ground with a resistor then I get (12volts), this rail looks to be only for the cooling fans.
U19 Pin1, my DMM shows just shy of 50Hz.
U1 Pin3, my DMM shows 40MHz.

Any help or advise would be greatly appreciated.

Thanks for your time.

Best Regards.

[Samogon]

Looks like CPU is not running i would check U15 x_rst and pwrdwn pins.

[bitseeker]

I concur, Samogon. Just what I was about to type. 

Mad Professor, although you indicate U23 has the correct output voltage, also double check that U15 is getting that power.

Regarding the shift button, its state is temporary, not latched. It's normal that once you hit another button, it turns off.

[Mad Professor]

Thank you for your replies.

I have checked U15 (X_RST) and (PWRDWN), Both are at 3.3v.

I have also checked voltages at a number of bypass caps.
U2 bypass caps (C2-C9) 5.1v
U15 bypass caps (C28-C43) 3.3v

[edpalmer42]

I'm not familiar with this unit, but since the display reacts to keystrokes, doesn't that mean that the CPU *is* running?  Does the front panel use a separate CPU from the rest of the unit?

[Mad Professor]

edpalmer42: Looking at the pdf from my 1st post I would say that the front control panel does have it's own processor, and that data is streamed between the boards.

I can not see any data / activity on the RX & TX lines of the main board.
At power on the RX line go from 0v to 3.2v for 90ms then comes down to 2.1v and stays at this voltage.
The TX line go from 0v to 3.2v and stays at this voltage.

[bitseeker]

I haven't worked on one of these loads before, either. However, it's not unusual for HP gear to have a separate processor for the front panel and serial comms between it and the main processor. Looks like that is the case here, too.

It's good that those U15 pins seem to be correct. Since the processor is on the GPIB board, are you able to interact with the load via GPIB at all?

[Mad Professor]

bitseeker: I have no way of interfacing with the GPIB port, I have never used GPIB before.

What waveform type (sine/square) and voltage levels would you expect from the SG-615PH crystal oscillator?
As stated before my DMM reads 40MHz, but I only have a 2MHz USB Oscilloscope. This is showing as a sine-wave floating above ground.

[bitseeker]

The datasheet has the specs. V_out should be at least V_cc-0.4V. The H version has a 5V V_cc. The speed should be printed on the IC.

The output waveform will look like a sine due to the low bandwidth of your scope (and the amplitude will appear low). If the speed on the IC is 40 MHz, then it seems like it's probably OK.

Regarding GPIB, if you have an Arduino UNO, it can be made into a rudimentary USB-to-GPIB interface. A search online will surface several projects with free firmware.

The serial lines between the GPIB board and front panel seems odd. I wonder if the whole thing might be held up because they can't communicate with each other. On other equipment, this would display an error, but such is not possible if the front panel isn't working properly.

[Mad Professor]

I do indeed have an Arduino UNO, I did not even think that someone might have already made custom firmware to turn it into a USB-to-GPIB interface. 
I will look into this, give it a try, and report back.

Thanks again.

[bitseeker]

Sure thing. I'm very curious what you find because I see these mainframes available "for parts" on a somewhat regular basis and wonder if yours may be representative of a common failure. It'll be interesting.

[coromonadalix]

http://egirland.blogspot.com/2014/03/arduino-uno-as-usb-to-gpib-controller.html

[nctnico]

I'd pull out an oscilloscope and probe signals around the processor. Things to look at are power supply voltages, crystal oscillator and reset line. If there are external data / address lines then check these as well for activity. If there are any components socketed then pull these out a bit and then push them back (reseating).

[Mad Professor]

I've been busy working on other projects and repairs, but it's now time to blow off the dust and start looking at this unit again.
Since I last posted, I have not done much in the way of testing the board, and I have yet to make a Arduino UNO USB-to-GPIB interface.

I have picked up some used test gear, psu's, meters, etc. but my new main toy is a GW Instek GDS-3254 Digital Storage Oscilloscope.
I have brought some PicoTech TA131 250MHz Probes off eBay dead cheap, unsure on the quality but for the price I can not complain.
I have had the Oscilloscope so still trying to learn my way around it.

[Mad Professor]

I have now been able to borrow a Saleae Logic Pro 8 to help with my fault finding, but I only have this for a few days.
The first thing I tried to log was the Xilinx Boot ROM, at power on you can clearly see a burst of data and clock.
As people have said it looks like the processor U15 is not running, this is the area I am looking at the most.
I can not see any data being sent out the diag pins, and nothing on the TX & RX lines between this board and the front panel.

[Mad Professor]

I am now trying the Arduino UNO USB-to-GPIB interface as stated above, but I have never dealt with GPIB so I am very very unfamiliar with the required command strings, or how to get started.
I have programmed the Arduino UNO with GPIB6.1.ino.
I have found the following document: Agilent Technologies DC Electronic Loads Programming Guide.
Would I not need to know the port address, etc of the unit, something that I don't know as nothing on the display.
Or is there a way to scan for devices on the GPIB interface?

[bitseeker]

Try sending a *IDN? query to each GPIB address from 0 through 30 and then read back from it to see which address responds. Of course, don't send to the same address as the UNO. There is a chance that the load is on the same address as the UNO, so if it doesn't respond to any of the other addresses, try changing the UNO's address and then query that remaining address.

Note that if the load is set to use RS-232, it won't respond to GPIB. So, if you get nothing from all the GPIB addresses, try the serial port (or maybe try the serial first).