-
Too high a output voltage could be from something like broken (fused) transistors in a linear regulator stage. One can usually check the transistors in circuit (look for a short).
-
Apologies, I haven't had much time this month to work on this problem.
My friend lent me his old FLIR thermal camera (emissivity was set to 0.95). Attached are some comparison photos of my Advantest R8340 after a warmup of 1 hour with an open case. I hope this is useful for some people.
Cheers.
EDIT: FYI, this Advantest has issues with the linear voltage regulators. I will consider performing another IR analysis to compare with this one.
-
Check three consecutive resistors. Burn often.
Hi serg-el,
I am making a list of replacement components for the R8340A and would like to know what you used (or anyone else for that matter) as replacements for these resistors. Please advise.
Thank you. -
There are the most common resistors. Since they work in heavy duty, the soldering needs to be updated. Desolder, re-tin and solder back. If you change to new ones - solder with a gap from the board for better cooling.
-
There are the most common resistors.
Any idea what the current/voltage/power would be? I am too afraid to probe the active board as it has already shown signs of overvoltage.
The 4-band colours are not clear, so I think I will have to desolder each one to confirm their tolerances before ordering...
I was thinking of replacing the generic resistors with these Vishay types:
https://www.mouser.ca/datasheet/2/427/vr25vr37vr68-1762017.pdf -
Too high a output voltage could be from something like broken (fused) transistors in a linear regulator stage. One can usually check the transistors in circuit (look for a short).
You are correct. I found a faulty linear regulator at U50 with markings NEC 7915 and replaced it with an LM7915CT/NOPB.
The following linear regulators were identified while examining the board and have all been replaced.
EDIT: Sorry for the upside-down picture... Windows displays it upright but when they are uploaded they become reversed.
-
Thermaly related issues are often caused by faulty El-caps.
All electrolytic capacitors have been replaced with the exception of C113 and C114 because I cannot find suitable replacements. The removed components that I was able to test seemed to be operating within tolerance.
-
The 4-band colours are not clear, so I think I will have to desolder each one to confirm their tolerances before ordering...
I have itemized the following resistors after desoldering them from the board and measuring them. They will be replaced with higher quality equivalents of greater performance.R188 to R192R156 to R159 were operating at ~ 180 °C which damaged the mainboard and caused the silicone coating on the bottom-side to vaporize over time.
-
Does anyone have recommendations for the following components (pictured and itemized)? I have not been able to properly source these components yet.
I plan on replacing them all just to be certain. There are some signs of thermal damage on the backside of the board that I am concerned about.
Thanks.
EDIT: Sorry for the upside-down picture... Windows displays it upright but when they are uploaded they become reversed.
-
I used acetone to remove portions of the protective layer on the mainboard so that I could replace critical components such as resistors.
A list of resistor candidates is attached. Yellow indicates the component selected and installed onto the mainboard.
The coating might be silicone. Does anyone suggest recoating the board before testing? If so, what product would you recommend?
I also replaced the linear voltage regulators pictured below. Just now I realized that the aluminum support bracket may have been electrically insulated from the regulators with a plastic film and an insulated washer on the screw. At the time, I thought it would be better for heat dissipation to discard the plastic film, but now I am debating whether or not to include it again. Does anyone know why the design has implemented this feature?
Thanks.
-
You cannot mount a 79xx regulator and a 78xx regulator onto the same heatsink without insulating one of them. This is because the thermal pad of the positive regulator is Gnd and for the negative regulator it is Input. In your case, if you insulate the 7915 once more, it should work. The heatsink will then be grounded by the two positive regulators. If this isn't desirable, you also need to insulate the positive regulators once more. Then you can connect the heatsink as guard, for example.
Regards, Dieter -
Since the radiator is screwed onto the housing, the regulators must be insulated. Otherwise, an unnecessary ground loop may occur. It is possible to use insulated flange regulators, but they have higher thermal resistance, so this is not the best option. Resistors R156 ... 159 are best mounted horizontally. It is possible at different heights from the board. Cool down better. The insulating tubes on them must be pulled to the very body of the resistor.
-
... then applied a urethane-based conformal coating on the top after cleaning it, which seemed to play fine with the remaining acrylic coating.
Can you let me know what urethane coating product that you used? I am at the stage where I need to reapply a coating.
Thanks! -
You cannot mount a 79xx regulator and a 78xx regulator onto the same heatsink without insulating one of them. This is because the thermal pad of the positive regulator is Gnd and for the negative regulator it is Input. In your case, if you insulate the 7915 once more, it should work. The heatsink will then be grounded by the two positive regulators. If this isn't desirable, you also need to insulate the positive regulators once more. Then you can connect the heatsink as guard, for example.
Regards, DieterSince the radiator is screwed onto the housing, the regulators must be insulated. Otherwise, an unnecessary ground loop may occur. It is possible to use insulated flange regulators, but they have higher thermal resistance, so this is not the best option.
These are two great reasons. I originally decided that the plastic insulators were to prevent ground looping, as sergi-el pointed out. At the back of my mind, when performing part sourcing, I did notice that the 78XX and 79XX series regulators were oppositely polarised. I wonder why Advantest decided to go with this design.
Attached are some pictures of the refurbishing process. Thermal paste was used to ensure good conduction through the materials. The excess paste was removed to prevent shorting. The high-voltage assembly was not provided thermal paste directly on the backside of the regulators as not to compromise the yellow insulation pad.
-
Resistors R156 ... 159 are best mounted horizontally. It is possible at different heights from the board. Cooldown better. The insulating tubes on them must be pulled to the very body of the resistor.
The instrument has been running for more than 2 hours without experiencing the problems that I mentioned previously. Before installing the shield and cover of the enclosure, I measured R156 to R159 (OHMITE MOX5002703GSE, 4 x 270 kΩ, 2 W/resistor, 100 ppm/°C, 155 °C max operating temperature) with an IR camera. The temperatures were measured at the base of each resistor. 75 °C was measured at R159, 85 °C was an average value and the maximum temperature was found to be 105 °C at R156. This is almost half the temperature that was observed previously when inspecting the original generic resistors. If one can source a high-voltage resistor of equal or greater power, then this setup should work fine. Your suggestion of positioning the resistors horizontally would improve thermal dissipation, but I wanted to avoid crowding the resistors and inhibiting airflow which also must be considered. Overcrowding is a problem with this Advantest design. Note that I applied your horizontal suggestion for most of the other power resistors.
Another alternative is to further divide the heat load amongst additional resistors, but this might get complicated. Otherwise, if the OHMITE MOX (2 Watt) or Vishay VR68 (1 Watt) resistors are unsuitable, then one can try and contact Japan Fine Chemicals https://www.jfine.co.jp/eng/e_c/resistors/lead/. These resistors were already incorporated in the original Advantest R8340 design and are rated for the kV range.
---
Surprisingly, all the taps that I measured previously have similar voltages to what was measured presently. If anyone has measured the voltages at these positions, I would appreciate it if you would share them for comparison.
At this point, I am assuming that the instrument is working properly and the overheating issues that were experienced previously have been dealt with. My guess is that the problem originated from the R156 tp R159 resistors operating at 185 °C and a linear regulator located at U50 that was showing signs of failure.
Eventually, I am going to have to recoat the board with some protective layer. As I have never seen coatings like this on circuit boards before, I would also appreciate it if someone with experience could provide a recommendation on a product that they have used.
Thank you all for your help and comments.
-
Apparently your transformer is switched to 220V. The modern standard for mains voltage is 240V. Switch the transformer to 240V. This will reduce the voltage at the unregulated outputs, and lower the temperature of the stabilizer microcircuits.
-
Apparently, your transformer is switched to 220V. The modern standard for mains voltage is 240V. Switch the transformer to 240V. This will reduce the voltage at the unregulated outputs, and lower the temperature of the stabilizer microcircuits.
The instrument labels claim to be configured for 90 - 110 VAC. My home operates at 117.2 VAC. When you suggest that I 'switch the transformer to 240 V', will I find a switch inside the transformer shielded box? Keep in mind that I am a beginner.
-
Clear. So your transformer is configured for 100V. Switch it to 120V.
-
Clear. So your transformer is configured for 100V. Switch it to 120V.
Big thank you, serg-el.
These are the kind of problems I would not have thought about myself because I only began learning about electrical engineering last summer. I am so fortunate to be able to learn from you and other members of this forum.
The average voltage of the unregulated taps has dropped 26 %. This will definitely prolong the life of the instrument!
The maximum temperature of the R156 to R159 is now 85 °C from 105 °C, a reduction of 19 %.
-
What a nightmare... Not only did this instrument destroy my Keithley DMM6500's OHM and DCI modes for no apparent reason when I was operating in DCV mode, but it also continues to show the same error message as before.
I left the instrument on standby to determine whether the error would return for a few days after the repair. Interestingly, the Advantest R8340 seems to be in the same calibration from pre-repair. I guess the power resistors and capacitors have little to do with the precision.
Yesterday, I received the same message "Overvoltage Detection Message" from Section 3-6 (page 36) when I set the instrument to source 100 DCV. I wonder why the instrument is reporting that an external voltage is interfering with the instrument. Could the instrument be shorting somewhere? When the error occurs, the instrument squeaks and then returns to standby. I can set the instrument to any value, but 100 DCV. Nothing is connected. The GUARD, GROUND, and LO are connected by the short bars. If I leave the instrument OFF for a while and come back to it, the error goes away again.
Does anyone notice that the high-voltage light flashes when the instrument is turned ON, and when switching from DISCHARGE, CHARGE, and MEASURE?
Figure 10 - 1 (page 267) indicates that there is a dedicated amplifier for the 100 DCV range. Maybe I have to start there.
I am so lost and frustrated.
-
...
Inside there is still another metal cover for the input I-V amplifier:
...
An LMC6081, Ibias of 10fA, can be seen in the beautifully constructed IV converter. It has an ink white dot over it, so maybe it is a selected part. This is the topology according to the manual:
...
There are some differences between the R8340 (bottom) and R8340A (top) in the shielded interface section of the board:
U1 - NEC C252A
R15 - COPAL 20kΩ 50E K-9W
R16
C7
C?
D2
D3
-
If it is not due to overheating it could be a cracked solder joint.
Apart from the disgusting spray-on coating that the manufactures included on the front of the board, I do not see anything that appears flawed. Maybe I am not looking in the right place. Maybe I will have to completely remove the board from the chassis, and have a look at the underside.
I will keep my eye open for anything suspicious-looking solder connections. Thanks.
This seems to be a bad spot. I am going to add some solder here.
-
Attached are two separate characterizations of the R8340 high-voltage source at 1000 DCV at 0.0005 NPLC and 10 NPLC. The instrument does not overshoot, is surprisingly stable and repeatable.
-
My Advantest R8340 seems to overshoot consistently on the maximum output of the 10.003 to 100.00 DCV range. The instrument might be experiencing 'overvoltage' due to this characteristic. When the instrument becomes too hot, the temperature coefficient of a component changes just enough to through the overshoot value above the 'overvoltage detection' limit. The overvoltage is well beyond the 'generation accuracy', outlined in the datasheet.
-
I think I figured it out. The last owner had replaced the fan and must have reversed the fan direction in the process.
Can anyone confirm that the fan should blow outwards? If pointed inwards, the stagnant air eventually warms up enough to cause some components to misbehave. The instrument has been running smoothly for several days now and I have not been able to repeat the problem.