Author Topic: Help finding a replacement for a BFN22 transitor for an oscilloscope  (Read 8125 times)

0 Members and 1 Guest are viewing this topic.

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Hello everyone, this is my first post here.

I am in the process of fixing an Leader 1060 oscilloscope i bought on ebay. I have very little experience with these things so please indulge me. When I received the scope I found out that there was no trace on channel one, but channel 2 worked flawlessly so I decided to hang on to it. After opening it up I saw close to the input a burnt up 100 Ohm resistor. I replaced it but still no luck, I then used my multimeter to look at the resistance on that input and compared it to the same point on channel two. I read 2.6 k on the dead channel and 0 ohm on the working one. I then started probing various components close to the burnt up resistor and found what looks like a transistor (Marked HB) with the same 2.6k across it's collector and base. I did some research I discovered that it could be a BFN22 transistor that has a breakdown voltage of 250 Volts. I removed it and TADA  :), channel one is working again. Funny thing is, it seems as if the emitter of the transistor is not connected to anything. So I am left perplexed as to what it is for.

So my question is, should I replace it considering the scope works very well without and if yes can I have some SMBTA42 that also have a high breakdown voltage. I looked for the BNF22 and it quite hard to find in small quantities.

Thank you all.

Photo of the board with the transistor removed.
https://www.dropbox.com/s/zj3xebhtcus685o/removed_transistor.jpg

Photo of the guilty transistor.
https://www.dropbox.com/s/98206e7sak1u1j7/dead_transistor.jpg

Photo of the burnt up resistor and it's replacement.
https://www.dropbox.com/s/a0j4pd0n57onqdw/_MG_2490.JPG

Photo of the scope displaying channel one before repair.
https://www.dropbox.com/s/c7pum9194j4jnhf/_MG_2485.JPG

BFN22 datasheet
http://alltransistors.com/transistor.php?transistor=27444

smbta42 datasheet
http://pdf.datasheetcatalog.com/datasheet/infineon/1-smbta42.pdf
 

Offline valentinc

  • Regular Contributor
  • *
  • Posts: 119
  • Country: ro
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #1 on: March 05, 2014, 06:24:04 pm »
    Have you checked on the other side of the board ? If that pin is not connected to anything, it could very well be a diode, not a transistor... Do you have a schematic ?

    The fact that the channel works without that component may be because it is only used for input protection and does not pay any role in the signal amplification...
Valentin
 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #2 on: March 05, 2014, 08:10:38 pm »
It's a single sided board, and i' m afraid I have no schematic. There doesn't seem to be very much information on this scope model.

I used this site to identify the component that was marked with the letters HB. Maybe thats my mistake.
http://www.marsport.org.uk/smd/mainframe.htm

Also there is another components with the same markings(now partly erased by my fingers and flux) on the other side of the board where all three pins are used.
« Last Edit: March 05, 2014, 08:13:17 pm by tiltit »
 

Offline valentinc

  • Regular Contributor
  • *
  • Posts: 119
  • Country: ro
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #3 on: March 05, 2014, 09:42:30 pm »
      If it's a diode, it's like this: http://www.mospec.com.tw/img/mas70.gif  (2 diodes in one package, which explains the parts on the other board with all the pins connected: some use both diodes, this one only one diode)...

      Where does this small board fit ? Try to measure with a multimeter, and compare to the other channel... From what I see, the pinout of 2 diodes in a SOT23 package is different from the pinout of a transistor... So in this way you can figure out for sure if it's 2 diodes or a transistor...
Valentin
 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #4 on: March 05, 2014, 10:50:48 pm »
The board is located right against the volts per division knob that make it quite difficult to get to, For channel 1 I had to unsolder it, I then added female headers on the main board and changed the male headers on the defect one so that I can now easily remove it. So to do some measures on channel 2 I would have to do the same thing again. It makes me a little nervous since it is in working condition and I am sometimes quite clumsy.

Here is a picture of the two small board against the knob.
https://www.dropbox.com/s/sqtio6ok4hprsxa/_MG_2536.JPG

And here is top view of the scope.
https://www.dropbox.com/s/pg5rew7unh387c2/_MG_2486.JPG

I just did some quick in circuit measures on the other component that has the same markings and package.

5 K Ohms both ways between pin 1 and 2
14 M Ohm 3 -> 1
14 M Ohm 3 -> 2
0.66 M Omhs 1 -> 3
0.66 M Omhs 2 -> 3

Thanks
« Last Edit: March 06, 2014, 04:46:25 pm by tiltit »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #5 on: March 05, 2014, 11:51:04 pm »
Bipolar transistors may be used as low leakage diodes.  The collector-base junction works up to the collector-base breakdown voltage but has a slow recovery time.  The emitter-base junction breakdown voltage is in the 5 volt range but is very fast.

The failure suggests that the input was overloaded causing the transistor being used as a protection diode to short.
 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #6 on: March 05, 2014, 11:55:30 pm »
So should I try and find a replacement or should a leave it as it is? What coud be the risks?
Thanks
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #7 on: March 06, 2014, 12:41:57 am »
So should I try and find a replacement or should a leave it as it is? What coud be the risks?

If the transistor was being used as a protection diode, then it should be replaced.  Otherwise the JFET or MOSFET that they used for the high impedance buffer will be at risk and that will be more difficult to find a replacement for.
 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #8 on: March 06, 2014, 01:04:27 am »
So am I right to think that replacing it with a transistor with a breakdown voltage higher than the original is probably a bad idea? The smbta42 has a breakdown voltage of 300 volts. Could I maybe use it?
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #9 on: March 06, 2014, 01:12:56 am »
So am I right to think that replacing it with a transistor with a breakdown voltage higher than the original is probably a bad idea? The smbta42 has a breakdown voltage of 300 volts. Could I maybe use it?

Breakdown voltage is not well controlled so a higher voltage part should work fine.  The junction capacitance is more important if the transistor was being used as a protection device.
 

Offline Mr. Coffee

  • Regular Contributor
  • *
  • Posts: 122
  • Country: us
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #10 on: March 06, 2014, 04:01:26 am »
If it was just for input overload protection, could it be replaced with a pair of back-to-back zener diodes to protect the circuit beyond? Just asking.

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #11 on: March 06, 2014, 04:56:47 am »
The reason they would use a transistor junction as a diode for overload protection is that the leakage will be very low.  Low leakage diodes do exist of course but qualified transistors are usually less expensive.  A low leakage JFET with the drain shorted to the source also makes a good low leakage diode.  A zener diode would only be used if a specific breakdown voltage is needed and they are not usually low leakage or low capacitance.

In this case the diode probably clamps the input to one of the power supply rails to protect the gate of the input FET.
 

Offline qno

  • Frequent Contributor
  • **
  • Posts: 422
  • Country: nl
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #12 on: March 06, 2014, 03:55:30 pm »
I think the transistor is used as a low leakage diode.

Check out AN43 Of Linear Tech.
In one of the last pages he discusses a transistor used as a diode..
Can you check if base and collector are connected together?
Why spend money I don't have on things I don't need to impress people I don't like?
 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #13 on: March 06, 2014, 04:56:02 pm »
Quote
Can you check if base and collector are connected together?
No they are not connected together, the collector is connected to the input and the base is connected to a to the pin header, and also to pin 4 on the Ic numbered 4071. I have no idea what that is, maybe an op-amp. I searched the part number but found nothing.
« Last Edit: March 06, 2014, 04:59:25 pm by tiltit »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #14 on: March 06, 2014, 09:22:10 pm »
The burnt up resistor is the input dampening resistor which also acts as a fusible link in case of input overload.  Analog Tektronix oscilloscopes have the same thing.  It looks kludgy for something that should operating into the 10s or 100s of MHz but it works because the impedance is high so the series inductance has little effect.

The connection to pin 4 of the IC is no doubt the negative supply and the other connection of the burnt out transistor is consistent with protecting the gate of the JFET or MOSFET transistor directly above it marked K25.  If that is an N-channel JFET which is the usual case, then pulling the gate high is relatively safe because it will forward bias at about 0.6 volts protecting itself.  When pulled negative however, it will happily go into reverse breakdown at about 20 volts which is likely to damage the JFET.  The protection transistor used as a diode prevents the gate from being pulled more negative than the negative supply.

The IC is a high input impedance operational amplifier and used as part of a servo loop which controls a constant current sink which sets the JFET channel current so that the offset voltage between the input and output of the buffer is constant, probably zero volts, and does not drift.  Later Tektronix analog oscilloscopes used the same design and many active probes do as well.  The alternative is to use a pair of the matched JFETs with one acting as a temperature compensated current sink for the other.

The transistor to the right and even with the upper edge of the IC is the output buffer and the transistor below that is the current sink.

The input current for the buffer needs to be in the 100s of picoamps range or lower to prevent creating offset in the 1 megohm input impedance.  The protection diode needs to have similarly low reverse leakage which rules out common diodes like the 1N4148.  A qualified bipolar transistor is cheaper and more readily available than a low leakage diode.

I do not know why they used that particular high voltage transistor.  I assume Leader qualified it for low leakage but I would have used a qualified 2N3904 or similar.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #15 on: March 06, 2014, 09:26:12 pm »
I am looking at it again and they did not actually compare the input to output offset.  The IC is still an operational amplifier which controls the channel current of the input JFET but the level is set in some other way.  It might just be calibrated once to a fixed level.
 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #16 on: March 06, 2014, 10:52:55 pm »
Thanks a lot for your help and insight.
I guess I will then order 1/8 watt resistor for the input an look out for a low leakage transistor. I also have some SOT23 BC-848bw lying around and looking at the datasheet and i'm not sure witch characteristic I should look out for.
I'm probably not going to use the oscilloscope to look at any high voltages or fix a TV set or anything so I am pretty confident I should wont blow the input.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #17 on: March 07, 2014, 01:54:43 am »
Transistor datasheets are not going to be much use in this case.  When I say "qualified" I mean a part that is tested and qualified for a specific characteristic and often by the user.  Transistors are only tested for leakage down to 50 nanoamps or so because anything lower requires a longer test which takes time making the part cost more.  For similar reasons low leakage diodes are not common and command a premium price.  Low leakage JFETs are a great way to get low leakage diodes with guaranteed specifications.

Testing a transistor for collector-base leakage is pretty easy though.  Connect it in series with your 10 megohm input voltmeter, apply 20 volts or so of reverse bias, and the leakage will be shown as 100 nanoamps per volt so a 3.5 digit meter on a 200mV scale will have a resolution of 10 picoamps which is adequate.

 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #18 on: March 07, 2014, 04:00:11 pm »
Ok, so I've tried a few more thigs with mixed results.
I fist tried to read the leakage current from a smbta42 by connecting 20 volt to the collector the red probe of the multimeter to the base and the black probe to the supply ground. I got the same reading as if the input of the multimeter was disconnected. I then used the method described on this page (http://jeelabs.org/2012/05/27/reverse-diode-current/) on a couple of diodes and read 5mV for a UF4004 and almost 0.5V on a BAT 48. The readings for the smbta42 was 0V.
So I decided to put it in place of the dead transistor in the scope and call it a day. I tested it and noticed that in comparison to channel 2, channel 1 had now slightly round edges on a square wave, very noticeable at 10Mhz.
But that didn't bother me too much.

I then pulled the Volts/Div knob to 10X Magnification and the signal went completely off screen, pushed it back in and noticed that there is now an offset on vertical position :palm:. I couldn't bring the trace down the the bottom of the display anymore. I now removed the component and it's now working properly again, but I am really scared to pull that magnification knob again.

« Last Edit: March 07, 2014, 04:15:33 pm by tiltit »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #19 on: March 07, 2014, 04:35:59 pm »
It would not surprise me if the smbta42 had a leakage so low that your multimeter could not read it even using the method I described.  I regularly see leakage below 10 picoamps on jelly bean 2N3904 transistors.  Where this gets tricky is that the gold doping used to speed transistors up increases their leakage many orders of magnitude and it is not always apparent which transistors are gold doped and which are not even with the same type number so they need to be qualified by manufacturer and process.  Switching diodes are gold doped leading to their high leakage.

Rounded edges imply high capacitance.  The problem with the x10 magnification implies a balance problem somewhere.  Often there is a calibration adjustment for that and there may be additional balance adjustments for the attenuator and variable controls.

I tried but was unable to find schematics or a service manual.
 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #20 on: March 07, 2014, 05:46:41 pm »
Finding information on this scope is a real problem. If I would have known better I think I would have gotten myself a old Tektronics or a Hameg as my first scope. I think I will leave it as it is for the time being and just be careful while using channel 1.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #21 on: March 07, 2014, 06:29:34 pm »
There is a service manual for the similar Leader 1021 oscilloscope freely available but the scans are of such poor quality that they are not as useful as they could be.  From what I could see, the basic circuit design is the same although the 1021 used a JFET for the input protection diode instead of a bipolar transistor.

One reason old Tektronix oscilloscopes are so popular is the availability of excellent service documentation.
 

Offline tiltitTopic starter

  • Regular Contributor
  • *
  • Posts: 56
  • Country: 00
Re: Help finding a replacement for a BFN22 transitor for an oscilloscope
« Reply #22 on: March 07, 2014, 07:13:08 pm »
Thanks again for all your help. I just discovered that I can reproduce the 10x problem by just tapping on the scope. It looks like one of the trimmers is a little flimsy. Again I put smba42 back in. It all seems to work and I feel a little more confident with this scope.
Gosh, I still have a long way to go before I get any good at this, but thats half the fun.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf