-
Druck DPI-510 Repair
Posted by
Nunes
on 20 Mar, 2023 20:35
-
Hi, I'm new over here, thanks for letting me in.
I'm trying to repair an old Druck DPI-510 pressure controller, but up to now the information I could find about this unit is a user manual and a RS232 protocol manual, nothing else.
My unit came without a power supply, but I could figure out the voltages and pinout and this problem is solved, but now I'm facing a bunch of other problems, first of all, I cannot configure the sensors, it came with 2 absolute pressure sensors, but the display shows it as gauge sensors, besides when I try to calibrate it, I have no success.. after the calibration the pressure display is always showing a random value... even if I calibrate as an absolute sensor or as a gauge sensor (using the atmospheric pressure or not)
Up to now, I've tested everything I could, but I'm really missing a service manual or at least the schematics for this unit.
Is there anyone around here that could have such a manual or schematic? I've tried all companies I could, but no answer, and Druck doesn´t exist anymore, now it's Baker Hughes, which didn't answer me either.
-
#1 Reply
Posted by
daisizhou
on 21 Mar, 2023 08:25
-
You send some pictures up.Have you ever replaced the pressure sensor individually? As far as I know, pressure sensors are all one-to-one correspondence
-
#2 Reply
Posted by
Nunes
on 21 Mar, 2023 19:10
-
Yes, I have inverted the connectors of the sensors, besides, I tested them outside the 510, and they are working. Some newer Druck sensors, like the ones used on the 610 or 615 families have integrated flash memory with its calibration data as well as the range, but it´s not the case with the ones used on the 510, they are the simple bridge topology.
I'm looking for how to say to the 510 that my sensors are absolute, not gauge (I'm sure it's a software feature), and how to set the correct range.... I believe it also has some electronic issues, that's why Im looking for a schematic, or service manual, if available.
-
#3 Reply
Posted by
daisizhou
on 22 Mar, 2023 00:18
-
If the sensor on the 510 is just a simple bridge topology, as you said,So the irregular numbers you see are the same every time?It seems that the physical hardware of your 510 pressure sensor is normal, and may require calibration internally.--"I'm just referring to the possibility. If you want to confirm, you'd better take a video to confirm."
(My guess is that it exceeds the range and needs to be recalibrated inside the firmware)
-
#4 Reply
Posted by
Nunes
on 22 Mar, 2023 20:36
-
The measurement results are not the same every time I try it. I already did the calibration via the usual menu, I've used a DPI-615 to do it. After the calibration, the measurements are inconsistent and sometimes it's simply zero or stuck at some random value, that's why I believe there is a hardware problem. I also checked the sensors, and they are ok, according to linearity and repeatability.
That's the reason I'm looking for a service manual (which I believe doesn't exist) or the schematics.... besides, maybe It will need some work in the flash memory (I already used the factory recovery option from the DPI-510, no luck).
-
#5 Reply
Posted by
daisizhou
on 22 Mar, 2023 22:09
-
According to your opinion,This pressure sensor is composed of a bridge resistor,Then an operational amplifier must be connected to the back of it,Have you ever tried to measure the output pin voltage of this operational amplifier?It is an analog voltage。You can check whether the voltage is linear under different pressures
PS--I suspect that there may be a fault with this operational amplifier or the ADC behind it. You can try measuring and then make a judgment
-
#6 Reply
Posted by
Nunes
on 23 Mar, 2023 18:38
-
I have already done that, it's linear according to pressure. The conversion is not a trivial ADC, it uses the DAC and voltage comparator topology to digitize the signal, something looks wrong on this part of the circuit, maybe it's just a broken trace, so I'm looking for a schematic, otherwise, I'll be chasing gooses. If I cannot find an original schematic I will reverse engineer it. Still, the point is, it's a tool that I need to work on other items, right now I don't have the necessary time to reverse engineer the hardware and software to figure out what happens, if a repair manual or schematic were in hand it would be much faster.
-
#7 Reply
Posted by
fzabkar
on 23 Mar, 2023 21:41
-
ISTM that you have two separate channels, each with its own quad op amp and trimmers.
https://www.analog.com/media/en/technical-documentation/data-sheets/op400.pdfAm I right? If so, then it's unlikely that both channels would be faulty, in which case the culprit must be the ADC circuitry.
I presume that one op amp provides a constant bias current for the Wheatstone bridge, and the others constitute a differential amp for the pressure signal.
Is the ADC the socketed gold IC? To me it looks tarnished. In any case my first thought would be to reseat each socketed IC.
-
#8 Reply
Posted by
Nunes
on 24 Mar, 2023 17:31
-
You are correct about the analog part of the circuit, the opamps you mentioned work as the signal conditioner, and I already checked this stage, there are some test points after the opamps, and the signal is there, linear and constant.
About the ADC, that's where I believe my problem is, but it's not an ADC it's a DAC, the processor uses it to generate a ramp and a voltage comparator sends a signal to the processor when the analog value matches, and so the processor knows the digital value, quite a common topology back in the day, at least using discrete components.
But there is a catch, the part number of this DAC is partially erased, and I already checked all Analog Devices catalogs I could find but no luck finding such DAC (It's an AD part, but maybe not a commercial one)... the fault could be a simply broken trace, or bad solder.... I already reseat all ICs that I could... or even some tantalum capacitor going bad, wouldn't be my first time on it, but the main point is my lack of time to reverse engineer this equipment right now, or at least draw a block diagram of it, it's a tool that I would need to finish some other jobs, with a schematic in hand it would be much faster, but as I can see nobody ever saw a schematic from Druck, I believe I'll have to put it on the shelve and wait till I have time to draw my own schematic of time, right now, unfortunately, I cannot spend more time fixing the tool than doing the work I need it for.
-
#9 Reply
Posted by
fzabkar
on 24 Mar, 2023 21:50
-
-
#10 Reply
Posted by
daisizhou
on 25 Mar, 2023 01:24
-
Are you sure it's a DAC chip?Analog switching chips are typically used.
As your picture is not very clear, I can only find the chip as shown in the following figure. I don't know if it is accurate
It seems that the PCB circuit board is of good quality and there is no sign of breakage
-
#11 Reply
Posted by
fzabkar
on 25 Mar, 2023 01:33
-
-
#12 Reply
Posted by
Nunes
on 25 Mar, 2023 18:32
-
Thanks for the datasheet link Fzabkar !! Didn't know this one.
About the Wikipedia ADC explanation, it's exactly what I believe I have here.
-
#13 Reply
Posted by
Nunes
on 25 Mar, 2023 18:50
-
Ok, the DAC vs ADC dilemma... Let's go a little further.
In the DPI510 user manual, in the calibration section (very poorly explained, just for day-to-day calibration), they mention an automatic DAC calibration routine, during the actual calibration (which I did), the display shows something like "Calibrating DAC" and the + and - values of the voltage references (~2.5V, as expected for a bipolar conversion and a 5V processor).
As for the IC itself, the part number is really faint, unfortunately, I could not take a good picture of it, I use an analog Zeiss stereoscope, no camera on it, and the cell phone trick didn't work in this case, the image was cropped... but what I can see clearly is the AD logo, and some parts of the code: AD42xx7, the xx looks like a 2 and 9.... in this case, we would have AD42297.
I did an extensive search on this part, I found it for sale, and I saw pictures of it but not a single mention in the databook (I already checked the ones you sent me Fzabkar).
Right now I did another search, with no luck with the datasheet again, but I found an AD document (Product/Process Change Notice - PCN 07_0024 Rev. E) that linked the AD569 to AD42297. Checking a few pins of the AD569 to the circuit of the DPI510, I believe it's the actual part number, and it's a DAC.
About my belief in a mechanical failure: I bought this unit second hand, it was totally disassembled, lacking the power supply, and was stored in a beach region (salt water)... and the packing was far from good, so, summing it to the apparent aleatory faults, probably it's a trace broken, or a bad solder joint, a corroded trace under some parts, tantalum capacitor dying...... the list is extensive, with a schematic in hand, which I'm now sure I'll not have unless I do it, the fix would be faster... lack of time is my main problem, I need this tool to work on some actual works.
Besides, maybe there are some configurations (like the range and type of sensors) that were stored in the factory, in the eeprom, which was lost, in this case, a service manual would be great.
-
#14 Reply
Posted by
fzabkar
on 25 Mar, 2023 20:46
-
-
#15 Reply
Posted by
Nunes
on 25 Mar, 2023 21:52
-
Once more, thank you fzabkar !!
One of these documents was the one I used to correlate the AD42297 to the AD569, and the signals (+/-Vref, Vcc, Vdd, Gnd) were correct.
After your hint, I checked the AD7846 pins, and unfortunately, none were compatible with the circuit of the 510, I believe the correlation will be the AD569, but I will keep checking.
But there is good (maybe bad) news, I found one of the root causes for the nonconstant error, one of the pressure sensors has some kind of intermittent problem with the voltage supply wires, the resistance should be around 3K, but Im getting 100K, I remember checking it in the begin of the troubleshooting, it was correct, so I did a gentle tap on it and the value changed to 10k, I hope it's fixable.
But now Im facing another problem, the display that should give the actual output pressure is stuck at .1 Bar, if I remove the sensor it goes to 1bar, so, the circuit is somehow working, and if I overpressure above the max calibrated pressure, the alarm goes on, which means the conversion is working, but something in the digital part stopped.... I still believe most of my problems were caused by the poor handling of the unit and bad packing
EDIT: The sensor is fixed !! I removed the protective silicon and found a bad solder joint, it's now stable, 3.4K. But I'm still lacking the panel reading. Probably another bad handling issue (I hope), as it sometimes worked somehow.
-
#16 Reply
Posted by
daisizhou
on 26 Mar, 2023 00:44
-
If the fault is caused by a chip, this will be easy because the chip is purchased with a replacement.
According to your description, I think you should thoroughly clean the PCB and the PCB inside the sensor
Use 95% ethanol to clean or special PCB cleaning water (note that plastic and rubber will melt)
-
#17 Reply
Posted by
Nunes
on 27 Mar, 2023 11:57
-
Hi daisizhou,
Thanks for the reply !! The PCB was already washed with IPA and the sensor was fixed right after I found it was faulty, as I described previously, as for cleaning it, I did, but not because of contamination from the salty environment, but to help remove the protective silicon resin that encapsulated it. Up to now I didn't re-encapsulated it, I'll do it after it's completely fixed (if I manage to do it), I live really far from any salty environment and my equipment is stored in a controlled lab.
-
#18 Reply
Posted by
daisizhou
on 27 Mar, 2023 12:34
-
Are you still repairing it? What's the latest progress?
-
#19 Reply
Posted by
Nunes
on 28 Mar, 2023 21:20
-
Yes, I'm, but slowly, the time is still short. after repairing the pressure sensor, I could make correct pressure measurements with the 510, they are stable now, but the controller is not working correctly, if I set the output pressure to 5.0psi it adjusts the output to 1.1psi, and so on, I've tried some values and none are correct.
I believe, after messing with the configuration eeprom content (I saved the original before), that now I'm having a firmware problem, no more electrical or electronic problems... looks like the original content was corrupted (some parameters have shown garbage on the display, which is a big clue of corrupted content) and I cannot reconfigure all the parameters via the keyboard, probably some were factory set.... if it's the case, I'll have to reverse engineer the parameters or have luck finding a DPI510 with exactly same sensors configuration and someone willing to dump the firmware.
I'll keep posting the advances here, but I believe I'll stop for now and find another pressure controller to use for a while.
-
#20 Reply
Posted by
daisizhou
on 29 Mar, 2023 01:35
-
It seems that you need to calibrate this state to be correct,First of all, I think you need to enter a mode where the pressure value of the main engine corresponds to the actual pressure value, and then test whether the pressure is correct after the main engine is automatically adjusted.
Generally, the data in the flash memory will not be lost unless you enter calibration mode to make incorrect adjustments
-
#21 Reply
Posted by
Nunes
on 29 Mar, 2023 20:00
-
I did a complete pressure calibration using my DPI615 as a pressure source, so, it's calibrated.
As for the eeprom, as I mentioned in the first post, the unit came without a power supply and was disassembled, who knows what happened in the past, but I'm sure the data is corrupted, maybe some power surge.
But for now, I'll stop working on it, I spent more time that I could on it, it's a tool, and I have to focus on the real work now, probably I'll end up buying another calibrator, do the work I need to do and, in the future, I'll come back to the repair, and if get something new, I will post here on the forum.
Until then, if someone happens to have a schematic or service manual that could share, or have a lead to where to find it, it would be wonderful