EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: taco on February 24, 2015, 07:34:06 pm
-
Ok so I'm trying to repair my own tools. I'm learning as I go, by watching all the videos and reading all the beginner materials I can. I have a unit when functioning normally takes mains 110V and goes through an internal transformer to output 17V AC to the board in the picture. The unit has an over voltage light, and reset switch on the front panel. Normally when I set the speed control pot at minimum I get 1.8V DC, speed control at mid point 11.2V DC, and speed control at maximum produces 34.36V DC. Keep in mind also that as resistance is placed on the DC hand piece that the unit will compensate to maintain constant rpm up to and including 45,000 rpm. The unit when turned on was functioning fine, then all of a sudden the hand piece goes to 45,000 rpm and thats all it will do. The over voltage light came on also. The issue I'm having is that the company that repairs these units, says the board is too old, and they won't repair it.
I started by desoldering every component on the board, identifying the component and then researching how to test it after buying a new Fluke 87V. I replaced the BDW93C TO-220 transistor with a NTE54 because it looked overheated on the heat sink side (bluish color). I also replaced the D1047 transistor because I had one in the parts bin. I replaced the LM358N op amp. I checked every other component except the TL431 (because I could not find out how to test it). re soldered everything back in, and when I turned the unit on it outputted 0V DC at min, 10.9V DC at mid, and 12.90V DC at max. When I connected the hand piece, it ran, but when any speed change was introduced, it would cycle up and down. Then as I was testing, it went wide open to 50V DC and the over voltage light came on again.
I want to fix this and learn from it. What did I miss? Any ideas why it's doing this?
Thanks
taco
(Navy Veteran)
-
With a fairly simple board like that^^ the first thing i would do is try to sketch out a rough circuit schematic. As it's all through hole components and just two layers it shouldn't be much work to do that. Then, you can look at the schematic and work out what part is likely to be causing the issue. For example, over voltage on the output is likely to be as a result of just a few components, and you can easily work out where to probe the circuit and what sort of voltages you might expect to see at those points. That way, and abnormality should become a lot easier to spot ;-)
-
What I don't understand is how you can get 34 or even 50V DC from a 17 V AC input. I don't see a DC-DC converter (no inductor) on the pcb, there's just a diode bridge + cap in there... :-//
-
Voltage doubler is my guess.
What I don't understand is how you can get 34 or even 50V DC from a 17 V AC input.
...
-
Is there a way for you to check the wiring, pot's switches' etc? It might just be a bad connection or a noisy pot. Some contact spray in the pot might be a temporary solution.
-
I've tested each component when de soldered from the board. I agree on making a schematic. Funny how I can run a steam generation power plant, but when it comes to little circuit boards I'm an idiot. I'm trying to learn after 50 and I'm definitely behind the curve:)
I never suspected a bad potentiometer..hmm check that in the morning.
Thanks:))
-
What does this 'unit' power? It could just a likely be what you are plugging into it that is causing the problems.
TL431 is a voltage reference controlled by the pot. Provided you are not turning the pot its output voltage should not vary by anything more than a tiny amount.
-
looks to me that C6 (near MCR100) and the big capacitor top left seems both bulky ???
-
I will post more pictures tonight. The hand piece is definitely out of the equation as I have many good ones to interchange with the same results.
-
Ok,
Here are the pictures. Pretty much I tried to show the whole unit from original, to torn down. The pictures are in order of tear down and the reverse sides of the boards. These units are expensive, running around $235-$279 each. My wife and I use them for carving wildlife in wood.
Thanks all for your help.
-
Could you take a full size (say 2048 x 1536 ish), sharp focus, plan view picture of the top and bottom sides?
Tim
-
looks to me that C6 (near MCR100) and the big capacitor top left seems both bulky ???
I would suggest that what krypton is trying to tell you is that the large electrolytic capacitor in the corner near the white connector shows signs of bulging, which in turn implies that it is no longer operating as a capacitor anywhere near the intended value.
The styling on the case of this item, along with the packaging technology suggest that it is 1970s vintage. Electrolytic capacitors are notorious for going bad over time. In your board these are the black cylinders with silver ends with a scribed X, mounted vertically on the board. Others will tell you that all of them should be replaced. I have personally often had success only replacing the obviously bad ones. Look for a minor bulge at the top, or goo leaking out of the bottom. The argument for replacing all of them is not bad. They are all likely to fail sooner or later. You have already got the board out where they are easy to get to. An these components are generally not expensive. Your choice, depends on your finances and access to time and components.
-
Ok,
Here are the pictures. Pretty much I tried to show the whole unit from original, to torn down. The pictures are in order of tear down and the reverse sides of the boards. These units are expensive, running around $235-$279 each. My wife and I use them for carving wildlife in wood.
Thanks all for your help.
If you take a look at the speed setting pot in your last picture, you can see that the black wire has become detached.
Okay, that probably happened during disassembly, but the solder joint on it is poor and an intermittent connection there would explain the symptoms you are observing.
Good luck!
-
+1 for the speed pot with bad soldiers.
from the last pictures, the capacitors look better
but always check (old) capacitors with an esr meter, your fluke 87 can measure capacitance but not esr .
-
I will replace the capacitors and correct the solder joints to the pot. I will also take better pictures of the boards later tomorrow. I apologize for being slow, as I'm having some medical tests done so I work on this in between.
Thanks again all:)
-
Sorry also tested all the caps after taking them off the board with an ESR meter and they tested ok. But I should have just changed them :palm:
-
You may want to start to make some measurements when it is powered, with an oscilloscope. I think you have exhausted most other options :)
-
I wish i owned an oscilloscope and knew how to use it. Anyone in SW Florida (Fort Myers) willing to teach? Just a small plug in..I ordered all new caps, and they should be here hopefully tomorrow.
So here are pictures of both sides of the boards. Hope these help. Once again thanks for everyones help.
-
What is the voltage on pin 3 of the TL431?. (RH pin when viewed from the front)
It should always be around 2.5 volts regardless of the pot adjustment.
Does the voltage on pin 1 of the TL431 vary when you turn the pot?
I am wondering if this is a shunt regulator configuration.
-
No manual. The company won't tell me anything but buy a new unit. I will test those pins of the tl431 tomorrow late. I will not be able to replace the caps until Monday bc they didn't have all of them.
I will post the voltages of the pins late tomorrow.:))
-
Good pictures! They're still hard to see, mainly of course because there's components on top. But this looks plausible? (Ground fill unmarked.)
http://seventransistorlabs.com/Images/MotorControl_Traces.jpg (http://seventransistorlabs.com/Images/MotorControl_Traces.jpg)
Tim
-
How did you find the board? Are they available for purchase? If so where?
Still waiting on the cap kit from digi key. Expect it any day. Then I will post the voltages of tl431. My dust collection system fried (motor), but I've replaced the bearings and she's up and running. Appreciate the help everyone. I'm also tracing and drawing the board schematic.
-
Tesla,
Did You generate the overlay?
Done free hand or with some software tool assistance? Which one?
-
Freehand, old copy of PSP.
Tim
-
can you make pictures of the PCB before you place the new caps? it may help with tracing
-
Yes I definitely will. Thanks
-
Pen tablet? I access the forum via iPad usually. Unfortunately the cap kit is still not here and I will be out of town for a couple days.
-
Good pictures! They're still hard to see, mainly of course because there's components on top. But this looks plausible? (Ground fill unmarked.)
http://seventransistorlabs.com/Images/MotorControl_Traces.jpg (http://seventransistorlabs.com/Images/MotorControl_Traces.jpg)
Tim
you got some kind of pen tablet?
Am I really that good with the mouse?
Or are other people just that bad? :P
Tim
-
Here are the pictures of the board after the components were removed. What is your thoughts after seeing the traces ?
Thanks
-
Nice! That completes the picture:
Tim
-
Nicely done! The pads dont look very damaged :)
-
I will reassembling the board later today, but one question I have is what should the VR1 value be set at? Measured at what points ? Before or after install?
Thanks
-
Ok, got the board back together with all new caps, checked all the resistors and diodes, bridge rectifier, and so on...board powers up fine, but instead of 32V DC at the output, I only am getting 0-15V DC at the output. Also the 1047 on the heatsink is getting pretty hot fast. I could remove the cheap pad and go with CPU heat sink compound, but this board is driving me nutzzzo...Any suggestions?
-
Was hoping someone else might pick up after my excellent tracings, guess I gotta do it myself; ;)
Treat all values as uncertain. You've got the board, so you're a better judge of resistor, capacitor, etc. values. The transistor pinouts seem mostly plausible, though I'm still not real sure what they're trying to do with Q5-Q6. The whole design is kind of retarded, so I suppose one shouldn't ascribe too much to that.
I didn't notice Q1 in any of the photos, or a loose three-pin connector, except for the very first where something is inserted but it's not shown what. So I'm assuming it's a transistor on leads, and that it's connected straight through. And that D1047 is the number (seems reasonable), and etc.
The basic idea seems to be: a high power darlington follower (Q3, Q1), run by an op-amp (U1A, with Q4 serving as level shift), and a hard 5A current limit ("Q2", TL431 -- but it's not a transistor, it's an IC, so they should've called it U2 or something). Output voltage is sensed (R16-VR1-R21) and fed back, with compensation from... well, nothing really, so I'd be fairly surprised if the op-amp ever sits still, but C2 ultimately serves the purpose (it can be pulled down quickly, but rises slowly thanks to R5 being the only current source, so the actual output voltage will always be lesser than or equal to the ideal, stable, quiescent point due to peak rectification effects).
Additionally, U1B provides feedback from the output, which I suspect is actually a differential sense across R1, to provide positive current feedback, which creates a negative output resistance. When matched to the motor's winding resistance, they cancel out, and the motor EMF (and thus RPM) is stabilized. But negative resistance also makes oscillators, so it's dominant-pole filtered by R10-C3.
Seems like Q5-Q6 is wired as a delayed Vbe multiplier, which if U1B is sensing current, then that would mean it's intended to shut down the motor (trigger and latch Q6) if it stays over a few amperes for some R22-C5 time constants.
As for heating up and tripping, seems like excess current would be a problem there. There's absolutely nothing inside this to prevent excessive dissipation; if you want to draw several amps at low voltage, it'll gladly do it, and easily burn out Q1 (even Q3, and then from there, Q2, Q4, Q6, maybe D2, D3 and the power transformer) in the process.
Replacing electrolytic capacitors is always a possibility. There's no shortage of them in this thing, and they all look to be 85C name-brand floor-sweepings from China.
More mundane is probably simply a crappy motor. Have you cleaned it lately? Are the brushes OK? (I presume it's a boring PMDC or series wound type.) Does the commutator need help? Bearings OK? Are you simply using it too harshly?
Tim
-
Hey,
Thanks for all your hard work. To start to try and fill in some blanks, Q1 is a mystery bc it sits next to the lead going to the 1047, so I was assuming that they have that as Q1? The motor is clean, and even paired with new motors ie (brand new hand piece), it produces the same outputs.
Oh by the way I thought I was the only one who ever noticed pi day for the next hundred years...lol
I haven't had time to check the values against your drawing, but will get to that today.
VR1 has me lost as I wouldn't know how to adjust it bc I don't know the specific value being sought. Q3 was replaced . It was originally a BDW93C. I replaced it bc the heat sink was badly dis colored as if it had been overheated. U1 was also replaced.
My thinking here is to start probing voltages. Starting with the RS407 output. The input coming from J2 should be the 36.1V AC, and then just check viable points and record. At some point in this circuit there should be a voltage drop of 50% ? This would be the newb theory? I was even thinking of going to an electronics shop for some help as well? Or do you think this should just be a learning disposable lesson piece?
I have another unit by another manufacturer that looks so simple by design and runs every hand piece I have that I was thinking of just designing (recreating) thier design on this size board and just retro fitting the unit? What your thoughts?
Thanks
BT1 USN Navy retired
-
VR1 has me lost as I wouldn't know how to adjust it bc I don't know the specific value being sought.
VR1 basically sets the scale of the feedback, which compares against what's on J1 pin 2, which I think is from the speed pot? So, its purpose will be to set the maximum speed.
Q3 was replaced . It was originally a BDW93C. I replaced it bc the heat sink was badly dis colored as if it had been overheated. U1 was also replaced.
Yikes, that's a 12A darlington! The NTE54 doesn't have nearly enough gain to match... which explains why R4-R5 seemed so large before. Surely that would've impaired the current (= torque) capacity as well?
That also means that the TL431 is almost entirely useless, because it's connected across an extremely low impedance (high current) node. It would have to shunt over an ampere to do its job under normal conditions! It should be connected cathode Q2 to base Q3, but... clearly, whoever designed this never had a formal review. The whole circuit is just a pile of... you know.
My thinking here is to start probing voltages. Starting with the RS407 output. The input coming from J2 should be the 36.1V AC, and then just check viable points and record. At some point in this circuit there should be a voltage drop of 50% ? This would be the newb theory?
Whoa, when did it get to be 36V? What was the 17V then? 36V AC will make 50VDC even, so that's consistent... but is that normal or what?
Supplying it with way more voltage than the motor needs will definitely cause the transistors some problems: all that extra voltage has to be dropped, while delivering the same load current as ever.
You didn't flip the 120/240V switch, did you? But that shouldn't be able to run, either... the transformer should blow the fuse if it's configured for 120 and operated on 240, and it should simply work poorly (not enough top speed, weak torque near high speeds) for the other way around.
Again, there's nothing in the circuit that provides the least bit of sense, or protection, or current control; if the motor demands some amount of current, this thing is going to dump as much current as required, whether it melts in the process or not.
As far as I can see, it was intended to operate this way. I haven't seen any evidence of a "before-after" failure condition, it's just altogether bad and any unit should operate just as poorly.
I was even thinking of going to an electronics shop for some help as well? Or do you think this should just be a learning disposable lesson piece?
Sure, or -- well, you can do whatever you like of course, but that was kind of my impression, might as well learn something from it. If it ends up fixed (or improved), or (partially or fully) replaced in the process, that's just icing on the cake. I expect your main concern is not to spend $250 breaking it, since you could just buy a new one for that.
I have another unit by another manufacturer that looks so simple by design and runs every hand piece I have that I was thinking of just designing (recreating) thier design on this size board and just retro fitting the unit? What your thoughts?
Heck, and it would probably work, too!
I suppose the main problem will be, connecting your front panel controls to whatever, but since we have the circuit... that probably doesn't matter much, assuming it's similar enough?
Tim
-
I honestly can't rem who told me that the NTE54 was a cross off of their reference sheet bc I couldn't find the original. But that's why I put that in. That might explain the lack of power? I have the original BDW93C, and it checks out fine according to what tests that I followed off YouTube. It's purplish heat sink was concerning, that's why I replaced it.
The unit was originally only putting out 17V AC but that was a bad measurement bc the board still was partially still connected. My other unit (working but newer board) has 36V AC in and roughly 34V DC out. I tested the transformer completely removed from the unit, and got just about the same input of 36V AC. I've tried to find a compatible replacement for the BDW93C but no luck.
-
http://www.digikey.com/short/7rq77f (http://www.digikey.com/short/7rq77f)
TIP102, TIP120, TIP142, [2S]D2081, 2N6045, 2N6388... many good options. The higher voltage (>= 100V) options are probably a good idea, but it'll probably still work starting from 60V, 1A. Could also make your own Darlington from, say, 2N4401 + TIP31.
Tim
-
Ok I valued all the components on the board as original. See if this helps a little. Plus thank you for the alternatives, I will look into them in the morning. If you have any questions with the inventory or values just ask thanks.. :clap:
-
Ah, great.
C8 = 2200uF 50V..?
Otherwise looks pretty good, dang I forgot to draw in C4 (Q5 collector to GND). Only go three resistors wrong, not bad :)
Tim
-
Correct C8 is 50V 2200uf. Looked at the Digikey link.. Is this the same chip I was originally looking for?
http://www.digikey.com/product-detail/en/BDW93C/497-7180-5-ND/1037701 (http://www.digikey.com/product-detail/en/BDW93C/497-7180-5-ND/1037701)
Funny how years ago after I got out of the service I was going to college in Pittsburgh to get my mechanical engineering degree. Found out just how much I slacked off in HS in my math classes. Nevertheless my boss found out I was going to school at night and figured I was after his job somehow and switched my shifts so I couldn't go. Looks like I'm still seeking the same knowledge 20 years later .. :-//
-
And yup, the original transistor is still around too. ^
Never to late to start learning. It gets slower with age, but doesn't go away! :)
Tim
-
Yup I've found my age has now become the "x" factor... :-//
-
Removed the NTE54 and replaced it with the original BDW93C. Turned on the power switch and can hear the transformer buzz on startup ( like the mounting bracket buzzing on the steel plates around the coil) then I took measurements at the input of the RS407 bridge rectifier and get 38V AC. The I take the output of the bridge rectifier and get 50.5V DC ? I was expecting no more than say 34V DC or so based on my ignorance and common thinking that if you only need 32V DC output that it would be close to that. Any other points specific to measure bc I'm still only getting 15V DC at the output with the pot at 100%?
:palm: |O
-
Just to save T3sl4co1l some effort here.
That's normal at the bridge rectifier. AC voltage is measured as RMS (root mean square) which basically means it's DC power equivalent. For example, if I put 3VAC rms into a 1 ohm resistor, it would dissipate 3 watts, the same as 3VDC.
For a pure sine wave, the ratio between the RMS value and the sine wave peaks is sqrt(2), which is to say if you have 38V AC rms, the peaks will be 38 * 1.41 = ~53 volts. The capacitor smooths out the rectified peaks so with no load you see 50 volts. The missing few volts is due to the diode drops in the bridge rectifier (appx .6-1v each depending on load).
T3sl4co1l was surprised the capacitor was rated for 50V knowing this. Normally you want to be well below the maximum voltage of a component.
Personally I would start by measuring the base and emitter voltage of Q1 and Q3 and seeing if what they do is making sense.
PS - Thanks T3sl4co1l for the interesting techniques!
-
In your first post you wrote
I have a unit when functioning normally takes mains 110V and goes through an internal transformer to output 17V AC to the board in the picture.
So finally you answered my question in reply #2 :) It's not 17V but 34V...
-
Sorry about the mixup on the 17V reading. I just grabbed my harbor freight multimeter and checked it. I since getting into this have jumped in with both feet and bought a Fluke 87V, Hakko 888 station, Hakko FR. 300. It doesn't mean I know anything, but now I've got the bug to learn. I could have actually purchased a new unit much cheaper than the learning tools, but I'm sort of refusing to keep buying a throw away $275 every year or so. I'm a fix it by nature but my electrical side of that is the side I've always stayed away from. In this day and age everything is electronic so I jumped off the board to learn how to swim.
Now about the base and emitter testing? I've tested each component out of circuit in diode mode as following you tube vids and they all checked out. What exactly (in newb terms) am I going to measure in Q1 and Q3 base to emitter? I guess to totally hang it out there..tester mode and pos/neg lead orientation? I know this sounds crazy but I just don't want to ruin new equipment with stupid actions.
I was under the impression that the AC comes in from the transformer goes through the bridge and then it's DC? But measuring base and emitter of Q1 in DC gives me no reading, but in AC gives me 14.5V?. That's why I'm confused.
-
Anyone on the Q1, and Q3?
Where in this circuit is it most likely to have a 50% loss of DC voltage?
Sorry for the bump, but I'm at a toal stand still now.
-
Oh, sorry I meant to reply!
Put the speed control all the way up. Set the meter for DC volts, put the black lead on ground (what's connected to the - terminal on the bridge rectifier). Looks like the safest places to pick it up is on J1 pin 6 or ZD1 or perhaps even r24 or R14. Attach the red lead to the base and take a reading. Repeat for the others.
It is really, really hard to ruin a DMM like that, especially a Fluke. Usually the worst you can do is to accidentally try to measure voltage while having the probes in the current range sockets. That'll blow a somewhat pricey fuse. They're made for those kind of accidents. The other thing you can do that's bad is to accidentally short out components on your board while you're trying to test them. That won't harm the meter but it can harm the board or probes.
(As an example, Fluke's are well known for still working after sticking the probes in a plug socket while having the resistance range selected. This will ruin some other, cheaper meters. I unfortunately lost my old Radio Shack one checking a water heater element that turned out to be live.)
-
Thanks I will test and report the findings in the morning.. :-+
-
Tested, and here are the results..
Q1=16.1V
Q2=15.5V
Q3=16.5V
Q4=16.9V
Q5=.3V ??
Q6=19.0V
So, are Q4 and Q5 working in conjunction with only leg Q4 working and Q5 dead? That would be too easy?. :phew:
-
Those were all... pin 3 voltages??
Q5 doesn't normally do anything, it's intended as a current limit trigger I think.
None of the voltages seem out of place. The problem is it's not rising much above 16V though? Even with VR1 all the way towards the R21 side?
Tim
-
I never adjusted VR1, but I will take the same readings with VR1 adjusted fully toward the R21 side.
Thanks
Mike
-
Is there a chance you have a short in the secondary of the transformer?
-
Ok,
Here are the results after turning VR1 completely toward R21.
Q1 - 37.8V
Q2 - 37.8V
Q3 - 38.3V
Q4 - 39.4V
Q5 - 0 leg one, .4 leg two, .3 leg three
Q6 - 38.7V
Results seem normal now? So I adjusted VR1 for an output voltage on Q1 of 32.0V
Plugged in the hand piece, and she purrs like a kitten. I haven't really put any load on it as of yet, but ran it for about ten minutes and Q1 gets warm (normal), and everything seems normal. The caps and re soldering the board must have worked, but VR1 was way out after that, but I think I tried turning the pot during testing it out of circuit, so I'm at fault there.
Thanks for everyone's help!! :clap:
I hope to learn enough in the future to help someone else like you all did for me.
Kudos to all involved...
Mike
-
Ah, cool. :-+ Glad it worked.
Still doesn't fix the fundamental problems we've noted, so... try not to go hard on it, anything that causes a lot of current draw especially at low voltage (low RPM). 8)
Tim
-
Thanks Tim,
Really appreciate your help. :-+