Rigol DG1021 Repair

[Nilres]

Hey guys,

I was able to pick up an used and not working Rigol Function Generator nearly for free. It came without any description only that's not working. So I plugged it and powered it on. Everything seems to work. I was able to reset it to default values and also the selftest completed without any errors. But when I probed the output with my scope it showed up nothing (after that I also tested the built in counter which seems to work). After that I opened the case and started with checking the voltages. There are two connections from the Power Supply. The first one contains 4 cords: black, red, yellow and blue. I meassured 5V from black to red, 3.8V from black to yellow and again 5V from black to blue. I'm not sure about the 3.8V but maybe there are for some local 3.3V regulation or so. The other power connection contains 6 cords: two black, two red and two blue. Red: +20V and +8V and the yellow ones -20V and -8V. For me that all fine and plausible so I moved on.

I started to look where the signal is generated and then tried to follow the path. In the attachment you can see a photo of the PCB. In the top right is the output and in the center left the signal generation. So there is a IC where somebody scraped the part number (I believe it's an analog device). The signal path is different when the output voltage (VPP) is 2V or below. I was only able to figure out which way the signal is going when the output voltage is set up to at least 2.001V. It goes to the two relays in the top left (scoped at TP325). And then trough the third relay and then trough a resistor into another device where the part number is scraped but it never comes out of that device. And I have no idea what's going on there. Maybe you can guess what the device there is? Am I on the right track? The other thing I tried is to find the signal path from the output terminal back but because I haven't any signal there it's a little bit harder (only continuity testing on the DMM).

In the picture you can barley see the sma connection under the glue. The big trace on the bottom and top is the outside of the BNC  (I marked it black). The inside of the BNC is connected to the small traces in the middle of the output stage (marked yellow). There are two SMD commponents which I can't identify. The right one looks like it is some sort of inductor?! But I have no idea about the second one (in the blue case). But there is no continuity through the device. Then there is the big SMD part with the four terminals which seems to connect the output stage to the rest of the circuit. But it also seems to be scraped. So no idea what it is. The connection in the top right is connected to the BNC outside and is high resistance to all the other pads. The connection in the bottom right is connected to the blue device and is zero ohms to the pad in the bottom left.

Do you have any ideas about the parts I couldn't identify? Or any other ideas? I wasn't able to find a service manual or schematic for it what makes the repair a little bit difficult.

[SeanB]

Blue device is an inductor, like the one next to it. If high resistance it will be open circuit, as it looks like it has cooked in the image. The black item to the left is a SMD common mode choke, it should have conductivity on the top 2 pins, and the bottom 2 pins, with high resistance otherwise. The SMD device is an opamp, and this has likely been blown, along with the inductors and the common mode choke, by somebody applying either RF at high power or DC to the output. The replacement has been discussed here before, it is a common very high bandwidth LT part. I would also chech the sSMD device between the relays, near TP329, as it looks like it is a voltage clamp that might be faulty.

[Nilres]

Fantastic thanks.

That sounds not too bad. I searched the last 30 minutes where I can find the information about what opamp it is but I couldn't find it. Can somebody give me a hint where I can find these information? And also where I can find the specs for the Common Mode choke and the inductor?

And thanks for the tip with the diode between the relays you were right it seems faulty.

[SeanB]

Won't say I never smoked a signal generator before with RF, but I do now have a few RF fuse assemblies and a lot of fuses that fit in them. They also make very nice gender and style changers. Originally designed for spectrum analysers.

[Nilres]

I found this thread: https://www.eevblog.com/forum/testgear/rigol-dg1022-problem/
But they never really figured out what op amp there is populated. Maybe I can also ask Rigol about the parts but I didn't expect much from it.
So if anybody has a idea with what I can replace the faulty parts please let me know. Would be a shame if I have to throw the device to the trash only because of two blown inductors and an opamp...

[SeanB]

Easy enough to try a regular opamp in dip8 or SO there as a temporary test to see if it gives low frequency back, and then order the right one. In a pinch you could use the OPA551 in DIP8 and simply not connect pins 1, 5 and 8. you might lose the output enable ability, but it will at least then work.

edit: try this LM7321 will likely be it or at least a drop in replacement.

http://www.ti.com/lit/ds/symlink/lm7321.pdf

or LM7121

http://www.ti.com/lit/ds/symlink/lm7121.pdf

These will fit the pinout, except for pin8 which in both is NC but on the Rigol seems to be used for shutdown for the output.

[Nilres]

I will order one or two of those. I measured the CMC (the working side) at it's around 1,9µH (measured with a agilent U1733C). The closest I could find by the local distributors is a 2,2µH one but also wrong package size but I thing it's good enough. So the only value I'm missing is the inductance of the blue inductor at the output. Any idea what I can put in there?

[SeanB]

Replace with a wire link, or a 0.1nH SM inductor. That will work there.

[Nilres]

Ok thanks I think I will wait with my order a until the end of next week because I may have to add some other parts for other projects so the repair could take a few weeks but I will come back and report!

[Nilres]

I was able to pick up some parts to replace the broken inductor, the diode and the op amp (I used the lm7121). I wasn't able to find a replacement for the common mode choke so I just wired the broken side through (just for testing). Now I was able to trace the signal through the op amp to TP327 (see picture for details). There are four 50Ohm resistors (two time to in parallel in series). I'm not sure what the purpose of this arrangement is. Can some one help me. But I was able to trace out what probably should happen. I draw have drawn the estimated signal path into the picture in red. Now here is the problem: I can see the signal after the first pair of resistors but it disappears after the second pair of resistors. What happens there?

[mij59]

I was able to pick up some parts to replace the broken inductor, the diode and the op amp (I used the lm7121). I wasn't able to find a replacement for the common mode choke so I just wired the broken side through (just for testing). Now I was able to trace the signal through the op amp to TP327 (see picture for details). There are four 50Ohm resistors (two time to in parallel in series). I'm not sure what the purpose of this arrangement is. Can some one help me. But I was able to trace out what probably should happen. I draw have drawn the estimated signal path into the picture in red. Now here is the problem: I can see the signal after the first pair of resistors but it disappears after the second pair of resistors. What happens there?

The four 50 Ohm resistors act as one 50 Ohm resistor with 4 times the power handling capability of one individual resistor.
Maybe the second pair of resistors is open, or the output is shorted to ground.
 

[Nilres]

I checked it again today at it seems I was wrong yesterday: the signal didn't get lost through the resistors the amplitude simply drops enormous. I don't understand that. Can anybody explain this to me?

I also started to draw a simple schematic for the parts of the circuit I already discovered. It's not complete or anything but it may help.

[mij59]

I checked it again today at it seems I was wrong yesterday: the signal didn't get lost through the resistors the amplitude simply drops enormous. I don't understand that. Can anybody explain this to me?

I also started to draw a simple schematic for the parts of the circuit I already discovered. It's not complete or anything but it may help.

What is the dc voltage on the anode of D1 ?
Check  if there is a short between the center  pin and ground of the connector X1

[Nilres]

Thanks a lot for that advice. I screwed the replacement of the diode. After placing in the right diode everything works fine. The only problem that now can occur is that I can't really cool down the new opamp but yeah the future will show I it will work without extra investigation in the cooling. I currently don't have the tools to do necessary measurements. I tried to read the temperature with a thermocouple but it was hard to get real contact.

Thanks to everybody that helps repairing this gear.

[slowolitus]

For everyone who faced the same problem:
The original one OPamp is THS3095, do not use  lm7121, it will be overheated very fast. The same IC (ths3095) used in SIGLENT waveform generators too. See here https://archive.org/details/EEVblog-50_20151007