EEVblog Electronics Community Forum

Electronics => Repair => Topic started by: giosif on May 28, 2016, 11:48:05 pm

Title: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 28, 2016, 11:48:05 pm
Hi all,

I've recently got a Rigol DG1022 ARB function generator off eBay, I'm trying to repair it and I am hoping that, with your help, I'll manage to restore it.
I hope you'll have the patience to go through the whole ride.

The problem reported when I bought it was no signal on channel 1 after putting some voltage into it (see original report of the issue here - https://www.eevblog.com/forum/testgear/rigol-dg1022-problem/msg451006/ (https://www.eevblog.com/forum/testgear/rigol-dg1022-problem/msg451006/)).
On investigation, I found that there actually is an output on channel 1, but much lower amplitude than what the device is reporting on the display.
More specifically, for an amplitude of under 2V p-p, the output is correct.
For an amplitude of 2V p-p or above (up to 20V), the signal is 10 times weaker than it should be.
Thankfully, channel 2 is working, so I can use that for reference to some degree.

Now, for the things I managed to work out (please use the attached picture and annotations on it for reference):

Anyone still reading this at this stage?  ;D
If yes, I thank you for your patience and would appreciate any answers/pointers/sugestions/wishes/etc or, indeed, even other quesions that would allow me to get this device fixed.

Thank you!

Regards,
George
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 29, 2016, 09:52:25 pm
Just an update: today, I mustered enough courage to try de-solder the two shift registers SR1 and SR2 and replace them with new components.
I've decided this after seeing a video on xDevs, watching Tin de-solder some components, including SMD, with just a regular soldering iron.  :-+
I'm happy to report that, after some sweat, emotions, etc. I've managed to replace the shift registers with new ones and that did the trick: relay 2 on channel 1 and the pad associated with pin 8 of the (still missing) op-amp are now being turned on when the signal amplitude is set to 2V p-p or above.  :-+
So, questions 2 and 3 were answered.
Question 1 still stands, but I'm not too bothered about it right now, as the behaviour is the same for both channels 1 and 2 (and channel 2 is working properly).
The main outstanding item is the op-amp.
I am yet to make any significant progress on that.  :box:

Regards,
George
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 30, 2016, 10:14:02 am
Another update: I've just got the response from Rigol (I asked them for schematics of the main board and/or info on the op-amp) and they kindly told me to bugger off.
Actually, they suggested I buy a replacement main board as the only way of fixing the function generator.  :clap:
Oh well, back to searching for an unofficial replacement op-amp...
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: 1kostik1976 on May 30, 2016, 03:07:34 pm
Hi giosif
how you determined that the amplifier is faulty? very warm?
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: TurboTom on May 30, 2016, 03:10:26 pm
You can probably try a current feedback operational amplifier like the Linear LT1227CS8. Since it's got offset null inputs on terminals 1 and 5 you'll have to cut terminal 1 since it must not be connected to GND. Otherwise it should fit on the foorprint.

Good luck,

Thomas
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: 1kostik1976 on May 30, 2016, 04:05:22 pm
THS3095
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 30, 2016, 11:16:18 pm
Hi giosif
how you determined that the amplifier is faulty? very warm?

Yeah, I dreamt that dream too (i.e. hoping the op-amp was ok), but it wasn't to be. :-)
I reached the conclusion that the op-amp was dead because one of its input pins was shorted with pin 8 (enable/shutdown pin).
Also that pin 8 and, by consequence, the positive terminal of relay 2 were being driven to -4.5V, which was definitely wrong.
And that wasn't because of the relay, since I desoldered it and the -4.5V was still there.
At this stage, I desoldered the op-amp and the negative voltage went away.
It's most likely that this negative voltage messed up the shift registers as well.

Regards,
George
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 30, 2016, 11:24:37 pm
You can probably try a current feedback operational amplifier like the Linear LT1227CS8. Since it's got offset null inputs on terminals 1 and 5 you'll have to cut terminal 1 since it must not be connected to GND. Otherwise it should fit on the foorprint.

Good luck,

Thomas

Thanks, Thomas!
I'll give that a look.

Regards,
George
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 30, 2016, 11:36:23 pm
THS3095

Nice. That looks like the most promising candidate.
Or do you actually know this is what Rigol are using for the DG1022?

Thanks,
George
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: 1kostik1976 on May 31, 2016, 03:30:40 am
HI
I do not know. all the answers are in the photo that you posted. + 30 minutes of searching on digikey ......
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 31, 2016, 06:58:46 am
HI
I do not know. all the answers are in the photo that you posted. + 30 minutes of searching on digikey ......

Ok. Thank you for your help and time spent on that!
I ordered the part and will let you know once I have the component in place.
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on June 01, 2016, 11:53:26 pm
IT'S A...LIVE!!!  8)
Thanks to 1kostik1976's suggestion for the op-amp (i.e. THS3095), I now have a fully functional ARB function generator.  :-DMM
Title: Re: Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on June 01, 2016, 11:56:13 pm
I should also mention that I eventually figured out the answer to question 1 from my original post: relays 3 and C kick in when the signal needs to be atenuated into the tens of mV p-p area.
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: 1kostik1976 on June 02, 2016, 01:50:01 am
HI
Congratulations.
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on June 02, 2016, 06:10:40 am
HI
Congratulations.
Thank you, sir!
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: Giovanni28 on April 16, 2021, 08:03:42 pm
Is it a THS3095DDA ?
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on April 17, 2021, 10:44:54 am
Is it a THS3095DDA ?
Yes, I believe so.
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: Giovanni62 on May 08, 2021, 09:54:10 am
Thanks. I replaced the op amp, but apparently nothing has changed.  Is it possible to test the circuit without installing the heatsink?
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 09, 2021, 02:47:57 pm
That's what I did back then.
Provided you keep the output off most of the times and turn it on only briefly, when checking things, I believe you should be fine.
That said, it's been a few years since I did this...
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: Giovanni62 on May 10, 2021, 10:29:40 pm
Hi! I made some measurements, and I saw that the sinusoidal signal is present at the output of the OP AMP THS3095, but then it does not arrive at the output of the channel, what do you recommend?
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 11, 2021, 09:12:22 am
I would say:
1. check the + and - power rails for the op-amp
2. check the output of the op-amp is not shorted to GND

If all of the above is ok, then I'd suspect THS3095 is faulty.
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: Giovanni62 on May 11, 2021, 09:53:55 am
I checked, the THS3095 is ok.
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: giosif on May 11, 2021, 07:43:02 pm
Sorry, I misread your earlier post!
In that case, there is a problem somewhere along the path from the output of the op-amp to the BNC connector for CH1.
There are a few things on that path, so you will need to check for the signal at different points: resistors, relay (it is relay marked with 1 in my drawing above), inductors, etc.
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: Giovanni62 on May 12, 2021, 06:53:19 pm
On pin 6 of the THS3095 the signal is present, but it does not reach the output of relay 2 and the input of relay 1. Could it be the TL072?  What function does this OP-AMP perform?
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: HB9EVI on May 19, 2021, 11:55:54 pm
hi there

when working on my DI-box, I obviously fried the opamp on my DG1062 with the 48V phantom; CH1 cannot be turned on anymore with warning of overload;
checking with the scope I briefly see the signal entering the opamp when turning on the channel, but no amplified signal on the output
so I'll order a THS3095DDA and report again, but the case seems to be quite evident

anyway, interesting is, that there's a proper signal when activating square wave, but the overload warning persists and disables the channel again; so far I cannot figure out the overload detection mechanism - would be good to know though
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: HB9EVI on May 20, 2021, 10:10:44 am
a small follow up so far

I removed the supposed broken THS3095DDA; now the channel can be activated again and outputs the square wave signal again.
so no doubt that the opamp is through and also responsible for the overload protection.
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: HB9EVI on May 26, 2021, 06:26:18 pm
problem solved!

a new THS3095DDA arrived today from Mouser; so the device is up and running again
Title: Re: [SOLVED] Rigol DG1022 attempted repair (and reverse engineering?)
Post by: mojorizing on June 02, 2021, 08:03:49 am
This post motivated me to get out my broken DG1022. CH1 failed for some unknown reason, but it had been working when I put it away the last time I used it.  A square wave output would have a short pulse (looked like a tooth) on top of the square wave, and there was a -2vdc offset when it should have been zero.  As I trouble shot it, the CH1 DAC basically failed altogether with no output on pin’s 21 and 22. I measured the voltage outputs from CH2 pin’s 21 and 22 setup for a 1Khz, 2vP-P, no offset square wave with an O’scope .  I removed the Ch1 DAC and applied the waveforms from a Rigol DG1032 with the wavforms I measured from the working CH2.  Now CH1 had the 1Khz square wave on the front panel output but with an offset.   I installed the CH2 DAC to CH1 DAC location and still had the offset. The offset was coming from the first opamp HKJ downstream of the DAC.  I replace the DAC with a DAC904E  (https://www.mouser.com/ProductDetail/595-DAC904E (https://www.mouser.com/ProductDetail/595-DAC904E)) and the HKJ with an AD8009 (https://www.mouser.com/ProductDetail/584-AD8009JRTZ-R7 (https://www.mouser.com/ProductDetail/584-AD8009JRTZ-R7)). CH1 now works.  Funny thing, while I had both DAC’s removed I performed a self-test and the unit passed…..
Pictures are with the DAC’s removed and bad opamp circled. The shifted cap to the right was pushed aside when I hot-aired the DAC’s off the board.