EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: Hendrik on March 25, 2023, 07:12:25 pm
-
Hello guys,
I bought an advertised working fluke 335d, turns out it isn't working. After price renegotiation I decided to keep the unit and see what I can do.
The unit does not go into operate-mode, the current limit indicator is constantly lit.
The Master Reference Testpoint reads 14.9990V constant.
The current limit, trip and vernier are at the correct settings.
I tried following the Power-Up Fault Isolation instructions. The A8A2 Pre-Regulator is not humming loudy and the R8 resistor looks fine, although resistance measurement in circuit shows 82.6Ohm instead of 430Ohm (Would that only be measurable out of the circuit?). Which lead me to Step 7 - Troubleshoot and repair problem pcb board or boards.
[attach=4]
I already found a pretty dead cap (C13) on the Chopper Amp board. I cut that one off.
From my limited electronics understanding, but seeing the schematics I would assume that not having this capacitor would already make the device unfunctional?
[attach=1] [attach=5]
One other test I did was measuring the voltage drop across the shunt resistor R22 on the Series Pass Driver board.
The range is set to 10V. This should give me 0.01mA according to https://xdevs.com/doc/Fluke/FLUKE%20335D%20Instruction.pdf 3-11 (p.20/58) but I observed a voltage drop of 0.01V which would be equal to 1mA.
For 100V and 1000V it gave me 12mV=1.2mA and 70mV=7mA when it should've been 0.1mA and 1mA (but I don't know if this is the correct spot to measure).
From what I can tell for now is that when the device is powered up from cold the voltage drop looks like this: [attach=2]
otherwise like this:[attach=3]
Changes at different voltage dials on the front lead to voltage drop spikes ranging from 10-30mV at the shunt. It sounds like a relay engaging at about 22mV.
https://streamable.com/rbbzs9
Here cold boot at around 22mV voltage drop the device produces an very audible sound (the same as above?) but longer and higher frequency in comparison to the video above since the voltage is slowly rising (it also does the same sound on the falling side). I would think a very fast switching relay?
https://streamable.com/9zj283
The mains power consumptions stays at all times at about 35W. Nothing else changes value during the audibile sound, no extra power consumption, voltage ref stays constant.
Before I order a whole set of new capacitors for the pcb's I would like some input on the matter. Does this ship already appear sunken or is it salvagable?
Are there any quick checks I can do? Is this even testable without the C13 on the chopper amp pcb?
Thank you for reading this convoluted mess!
Datasheet: https://assets.fluke.com/manuals/335d____imeng0100.pdf
Kind regards,
Hendrik
-
I'm sure you can sell it here if you get bored.
-
C13 couples the output into the differential amp, so testing up to that point may be possible but not after.
I would not jump to a full recap (yet), but a good starting point would be to go around and check any electrolytics and tantalums for leakage or signs of failure. If you had one fail there will probably be others.
-
Bottom side of manual page 8-10, figure 8-5, has a current limiter.
It's +/-35V operated, start checking those voltages.
-
Hello,
Thank you guys for the input and sorry for the long delay.
I have checked most of the electrolytic and tantalum capacitors on the boards and have replaced bad ones with mostly identical Vishay parts. I have found a couple of resistors which seem out of tolerance, I will be replacing those as well but as soldering through the conformal coating is not so pleasant I am currently waiting for a professional solder fume extractor which should arrive after easter.
In the meantime I have checked all the power rails for their values stated in the Power Distribution Figure 8-8 on page 8-16 (last) of the manual.
The +10V rail provided by the A8A2 Preregulator seems too high with over 30%. All the other rails stay roughly within a 5% tolerance. Any thoughts?
[attach=2]
K2 relay on the A8A2 Preregulator is open. The voltage sitting at that relay between Pin 6 & 7 is 98.2 VAC.
There is no voltage on the interlock circuit but it is closed. The trip circuit is engaging as soon as I turn on the device, effectively grounding the +35V rail which would power the K2 relay.
I am currently gathering all the information in a spreadsheet and reading the manual over and over as I am super new to analog circuitry and given the complexity of the device this gonna take some time.
https://docs.google.com/spreadsheets/d/1iPhhbaXig34eY5xCBQ2r04t6HceobkJnf08qXb3mNkk
As suggested I also checked the current limiter and can confirm that all the -35 and +35 voltages are present.
I tested the chopper amplifier with my oscilloscope to see if the traces match with the manual. The chopper drive adjustment is too high and the output from the opamp is dropping way too low (using alligator clip ground lead). Anyone know how to interpret this? I have changed every electrolytic capacitor on that board but missed tiny blob tantulum C22. Will test that one later.
[attach=1]
The Master Voltage Reference is giving me a rock solid 15V leading me to believe that this board is OK.
I will continue my learning and investigation.
Until then happy holidays,
Hendrik
-
Lower picture has the power part of the current limiter, there +35V from pin 3 goes to A8A2 pin 1.
So K2 should be closed, and K1 can't close without it.
Maybe you mixed those +V and +10V.
-
Hello,
from my current understanding and sorry if I am horribly wrong.
Depending on how high the current is the voltage drop across R22 and R24 becomes high enough that the voltage at Q4 base is low enough for opening.
Q1 has just the output voltage sitting at the emitter and with the voltage venier and switch we change the voltage at the base. When either Q1 or Q4 turns on Q2 turns on resulting in the trip circuit activating and grounding the relay power for K1 and effectivly K2 on A8A2.
I tried to capture this here with this image:
[attach=2]
And regarding the V+ and +10V, I don't think that I mixed them up but who knows :-//:
[attach=1]
edit: Oh, I just realized the excerpt said Relay A8K1 not A8A2K1... will have to look into that tomorrow.
-
I'm reading
335D
Voltage Standard
Differential
Voltmeter
Instruction Manual
P/N 353300
May 1972
Rev. 1 6/78
There page 8-16 primary side of mains transformer and under mains switch is A8A2 AC input, pins 6, 7, 8.
This continues to lower left of page 8-7 and there K2 switch connects to pin 6 and pins 1 and 2 connect to coil of K2.
Pins 1 and 2 are also guided to A5A6 PIN3 (+35V) and (COMMON).
Back to page 8-16 and level with primary winding is A5A6 current limiter power part and its +35V to pin 3.
So if +35V is present K2 coil is constantly energized.
Normally open (NO) switch of K2 is pictured right from coil, not optimal but things happen.
Then page 8-8 and its through and through ground connection chain.
I can't find A5A6 pin 4 from anywhere else, its triangle ground is also A8A2 preregulator bridge only.
K1 diodes on top of the chain are also a bit strange.
Still a bit above and right is a guide to A5A6 pin 5, from A8A2 pin 1, the same pin pointed to A5A6 pin 3 elsewhere.
That through and through ground connection chain is more like a startup protection so A5A6 pin 5 doesn't feel right, more right would be preregulator bridge a.
But even more so is that topmost stump connection and triangle ground are both wrong and connection is normal common and +35V of current limiter.
Page 8-7, bit above and right of high voltage bridge is another pair of K1 and K2 relays.
Parts list has K1A and K1B.
+10V,
measure again but this time over C4 and CR5, and R1 if you can.
Being less over diode and capacitor is not very rational.
-
+10V,
measure again but this time over C4 and CR5, and R1 if you can.
Being less over diode and capacitor is not very rational.
Thank you for your lengthy answer, I will look into that.
For now with the +10V I have attached another image.
[attach=1]
Doesn't matter from where I take the Ground: Pin 13, R1, C4. After the diode the voltage reads ~13.5V.
Voltage across the resistor R1 reads 9.16V aswell.
The schematics and my board also match mostly. Just an additional Resistor (R20) and plastic Capacitor(0.5uF 2.4Ohm ESR) visible in the pictures.
Greetings and happy easter,
Hendrik
-
What there's over CR5 alone?
-
Over CR5 alone I have a voltage drop of -4.283V. See Image. With reversed leads it's 4.315V.
-
Parts list states that it's 1N4817, normal silicon rectifier diode.
So voltages are real, at least sort of.
Page 8-3 top left has sane interlock circuit, but A8 is missing, it's on bottom right of the next page.
So page 8-7 A8A2 preregulator is just that and +V is for R18 only.
Q6 overloading?
-
I'm sorry what do you mean by overloading? I mean how would I test for that?
I tested the transistor in circuit and got diodes from base to collector/emitter (both 0.55v drop) (so nothing wrong there I guess?)
I meassured the voltages with power if that is of any help.
-
How much over R18?
I have hard time understanding how this triangle ground is connecting.
But anyway,
Q4, Q8 and Q9 can operate when +10V and -V are present.
Q5 can operate when +V is higher than +10V.
So all for safety, and so when +V is below +10V nothing is pulling Q3 base up.
But +V is not DC.
What happens when one positive side rectifier diode is cut?
C4 can compensate if load is low but DMM is measuring lower value for +V.
-
There is 8.4V over R18.
I removed the CR5 (1N4820 from W43 1980) diode I think you were talking about. The +10V rail is gone to noones surprise. The V+ rail still measures ~+9.2V.
-
So a bit over 100mW for R18, not very much.
How are CR1 and CR3?
-
I tested CR1-CR4 out of circut. All are between 0.48V-0.5V VD and OL other way around.
I also checked if the problem would shift if I swap CR 2 & 4 with CR 1 & 3, but the problem stays the same with the +10V at +13.5V and +V at +9.3V.
-
For now +10V being high is fine, nothing is loading it so no drop.
It also means that some sort of 13.5V is available from the bridge.
Is it still the same if you remove R18?
If it is check CR1 and CR3 traces.
Check R18 resistance also, maybe assumed 100mW is something else.
If your meter has a peak function, try it also.
-
R18 measures 584.6 Ohm.
With R18 removed the +10V rail drops by 1V to +12.6V. +V stays the same.
This looks at device startup like this:
Without R18 the +10V oscillation never gets smooth.
I found some more resistors out of tolerance. R5 for example with 24.5kOhm which should be 22k +-5% (23.1k). I will make an order ready for a set of new resistors.
-
Dang, a wild goose chase then and everything is actually fine, drop over CR5 is also visible.
And picture was many words, again.
My bad, sorry about the inconvenience.
But why no comments from anybody.
Over 1k views and I'm quite certain I've not clicked even a tenth of them.
Would be pretty sad if all others are just bots.
There are 3 voltages but only 2 are filtered, -V and +10V.
+V filtering is blocked by CR5 and DMM tried to inform it.
I even wrote that +V is not DC but somehow it didn't connect with the meter.
Manual reading is not hurting much either, before hand, 3-80 is clear enough.
I don't think those resistance values are a problem.
Triangle ground is not isolated so use differential probe when scoping.
Q1 is pulsing when everything is ok but it may need a guidance from A8A1 side of A8T1.
I've not noticed a bridge use before as is that pregulator bridge.
It and Q1 together are a switch and current flows when Q1 allows it.
Quite innovative.
So back to the K2 of A8A2, I'm not sure how it is connected.
A8A2 pin 2 is shorted to A5A6 common, that is clear, through all those other module loops or not.
But that +35V side is uncertain, A5A6 pins 3 and 5 are mentioned.
3-27 describes the trip circuit and 3-29 indicates that interlock circuit is different.
-
No worries! Even if I am probing something completely working it is still a learning experience and I find measuring stuff fun anyways!
And I get why noone is replying, the complexity of the device combined with my lacking information due to experience is quite cumbersome for anyone to follow + streched over so many posts. I often read the repair posts just to maybe learn something.
Just to be sure I measured Q1 and Input/Output of A8 T1:[attach=2]
Q1 Q-E looks a bit weird but I can't say for sure.
I also took some thermal images of the boards after about 5 minutes of on time:[attach=1]
Triangle ground is not isolated so use differential probe when scoping.
I don't have any differential probes. Is it sufficient to have the DUT on an isolation transformer?
I will shift my focus back to K2 and reply if I have any news.
-
My CR5 take was a school example of confirmation bias, onwards from the first electrical comment, that +V is too low.
Somehow 13.5V/sqrt(2) level never materialized.
Anyway, those heat maps are interesting, Q1 C-E also.
A8A2 R1 and R11 and A8A1 R1, R2 and R3 are sort of acceptable.
A5A6 hot spots are 2.4k and for LEDs, one would expect that less is enough.
But is A5A2 Q3 pulling the interlock circuit down?
CR14 being so hot can't be healthy.
Manual has Q3 emitter resistor voltage missing, -35V I guess.
-
But is A5A2 Q3 pulling the interlock circuit down?
CR14 being so hot can't be healthy.
Manual has Q3 emitter resistor voltage missing, -35V I guess.
Yep, as far as I can tell Q3 is getting pulled down to about -32V.
-
Somewhere is a description of high voltage crowbar operation but I can't find it anymore.
Section 3-37 describes an operation of A3 capacitor switch, its picture is bottom left of page 8-9.
There is K1 and its closing is delayed.
So A8 K2A is clearly normally open and then A8A1 pin 3 through Q10 is pulled down and A8 K2 energized when that crowbar is operating.
Heat map is showing some action in A8A1 CR23 and R5 to Q10 base, pass element C input R27 also.
A8A1 pin 4, driving bottom of pass element B through Q8 is coming from A5A2 pin 4, left and a bit below of +output.
So A5A2 pin 4 is just a control signal but heat map is showing that R16 for Q5 is participating heavily.
Right of A5A2 page is A7 time delay part, its K2001 and its normally closed K2001B.
Same page and top left, the same K2001B.
So A8 K1A is clearly normally closed but where is its coil.
On top of A5A2 page's interlock chain is clearly A8 K1, but bottom right of A8 page has a normally open K1 switch.
(one of those Fluke schematics again, where relevance of a component has no location privileges)
Then A8 parts page has K1A switch and K1B coil.
For now my prediction is that K1A on A8 picture is a leftover ghost.
-
That confirmation bias is here again.
I'm not familiar with those heat maps so their relative colors are misguiding.
But I'd still use a bit bigger series resistor A5A6 R3 for those range LEDs.
From heat map's A5A2 R16, it is possibly as it should.
If Q3 is pulling the left of +output down R16 is feeding CR3.
But, that A7 K2001B on top should be open and power switch should be also open.
So left +output should be far from -35V, if it is not pulled down elsewhere.
How many swings that power switch has?
I found 5 without searching.
-
Dang, can't say it clearly.
+35V
|
R16 3.9k
|
CR3 12V
|
+output
|
R9 2.4k
|
CR14 4.3V
|
-35V
So voltage over R16 + R9 is 70 - 12 - 4.3 volts.
And current so low it can't heat those components.
Also when one side moves the other side moves along and one component should heat less.
If all tracks are present +output should be far from both voltages to be able to heat both sides.
But if that is the case then probably many other spots would be warm also.
-
Last time I quite clearly had too many thoughts and too few words, happens when you say what you think.
But here it is even worse since all interphases are missing and not even general idioms are coming out.
I must behave better in the future.
And back away if too tired, much more important if you're actually poking something, maybe even high voltages.
Usually single fault is easy and straight forward, but many times one fault take something else with it and make things more complex.
Then when all seems to be ok there can still be those almost gone things and then after a next blink the situation is bad again but in different way, that can be frustrating.
Buuut, I'm sure you can last at least few months and warm nights are coming, our day is already around 15h.
One other thing, when you've been peeking around, have you seen hints of previous repairs.
If yes then keep in mind that if you've not seen this thing working something can be incorrectly connected.
My intuition is against grounding problem, it feels that many other points should also show it, but it's just an intuition.
A5A2 +output on left has that triangle flag #10, that's A5A6 common.
A8A1 has its own common but it's combined through pin 5.
500VAC bridge negative is also combined through orange wire.
A8 HV is not coming out if A8 K1 is open, and its filter C8 is not shorted.
Black wire there goes to A5 DS1, right on page 8-5 and better right on 8-9, and so to -output.
Blue wire from other side of A8 R8 I couldn't find, same with the braid wire.
First I thought the blue is -output but obviously not, if manual pages are what they should be.
More right on page 8-9 and there is A7R8 5.1, but A7 is time delay and its component numbers are 2000 something.
So time delay board is new and A8 used to be A7, and seems that A7R8 is actually A8R8 on bottom right of page 8-7.
Above that A7R8 is also a relay part K1B, pretty much like A8 K1A.
Top right on page 8-9 are A1C1 and A1CR1.
Can't find them from elsewhere anymore but wondered once what that capacitor marking is.
Brighter side is that the other side of old K1 is not missing anymore.
E2,
page
8-9 is old version,
8-8 is new and
8-7 is both.
Page
8-4 has A8 T2 wire as PR purple,
8-5 uses PR,
8-6 uses VIO and
8-7 has A8 T2 wire as VIO violet.
Page 8-16 has A8A2 connected old ways and topmost T1 secondary bridge goes also to K1 & CR1, other side of the page has actual components.
So this K1 can use those K1A and K1B normally closed switches, and K1B marking is then more rational.
K1A of A8 page is also outside of the A8 border.
So few moments for few relays but now my prediction is that K1A is not a ghost.
E,
how I use single screen setup more effectively.
PDF manuals go so that if page points to other page I'll go back and copy that manual and then open that copy to other tab, this continues as long as it takes, here 4 copies have been enough.
With pictures I use old picture viewer Windows used to use as default, that goes so that pictures are open but PDF viewer is on top.
Viewing pictures goes so that mouse pointer is hovered over the taskbar and pictures there and so a peek feature is showing the picture.
Picture viewer uses mouse wheel for zooming and hand for moving, PDF viewer is set to do equally.
Used browser is Firefox and its hover is not working, scrolled down position is not available.
So in case of a browser tab, it must be topmost.
This way I can have quite many active "screens" quite easily swappable.
-
Hey,
thank you for your elaborate postings and sorry for my slow replys. New university semester is starting reducing my spare time to work on this + I moved my whole test equipment to another room as this fluke on my desk dwarfed everything on it and made it very hard to operate the oscilloscope on the table behind it. Now with two desks everything is much easier to work on.
One other thing, when you've been peeking around, have you seen hints of previous repairs.
If yes then keep in mind that if you've not seen this thing working something can be incorrectly connected.
Aside from the wear and tear on the front panel the device is in good condition. I couldn't find a single spot with missing conformal coating. No new components with newer date codes. The only hint that someone might had this open is that only 8/9 bottom cover panel screws were installed.
I've printed the schematics and theory of operation on paper for easier referencing as I also found it hard to keep track over so many pages.
But I'd still use a bit bigger series resistor A5A6 R3 for those range LEDs.
Will do. 95°C seems a bit excessive.
Fyi, the time delay module seems to be operating as it should. I can hear and see the relay engaging after about 2-3 seconds.
Same goes for the capacitor PCB. When downranging from 1000V to 100V I can hear the relay on the pcb engaging
I also tried if increasing/decreasing the lower/upper current adjustment on A5A6 makes any difference but the current limit indicator stays lit.
Edit:
I checked Q5 on the A5A6 current limiter and found some pretty high voltages? Over Base-Collector (I might had the leads reversed, so maybe 120V drop? Still seems pretty high)
Edit2: Scrap that. During moving my equipment this singular scope probe got switched to x1 Attenuation... Doh..
-
Take your time, this is a hobby, or should be.
Boat anchors have boat anchor's habits, usually includes space sucking, that is more than initially estimated.
One underestimated length is diagonal and its regularity of presence.
Maybe somebody had a missing screw elsewhere and though that 8/9 is still more than enough.
Heat map A5A6,
I didn't think much of its other parts earlier.
Smaller Q1 is warmer and R5 and CR5 are not a pair.
Pairs are R2, between power transistors, and CR5 for positive and R5 and CR7, right of C4 on top, for negative.
Heat signature of CR7 is not healthy.
You measured earlier that A5A2 Q3 collector was -32V.
Check now how it is towards A5A6 Q5 base, so measure between A5A6 pin 5 and Q5 base.
It can be high so maybe DMM is a better tool.
+output being nothing Q5 base should be close to common level and so the measurement should be close to same -32V.
R20 is warm so it must be lower than +output, R19, peeking under C1, is also warm so +output there is not pulling up, at least much.
Measured >100V then has a direct path through CR10, so from there it is most likely coming.
Check also how A5A6 input voltages actually are.
One possible solution for those high temps is that A5A6 common is something else.
DC relays have a protection diode, so opposite voltages go through.
So check A5A6 pin 4, where it actually goes.
Interlock circuit of page 8-3 has T1-21, A5A6 common, but page 8-8 has that triangle.
A5A6 common is rational, triangle is not.
Maybe you have a 9/9 screw there.