Tektronix 2215 Scope repair

[Per Hansson]

Yes it is a big concern Vince, as tautech said if the dip is too big the MOSFET will turn off! EEVblog video #1119 describes this really well!
The graphs in my previous post are comparing different MOSFET's, the type of the MOSFET is the filename of the picture.
So the first one is IRF820PBF, that is like the original MOSFET (I have replaced it with a new one though) and even that has a dip down to around 7.5v.
That's not great but the point it will turn off is from the datasheet: VGS(th) min: 2.0v max: 4.0v, so we are "ok"

The second graph's MOSFET is a IRFBC30, this is actually quite close to what Tektronix changed to in 050-2242-03 as flobydust wrote here.
The dip for this goes right down into the VGS(th) range, so it's not good.
What's interesting is that there is a change for the 2235 in one of it's later service manuals.
At the last pages they add component R909, a 39Ω resistor at the gate of the MOSFET.
Of course that will reduce the inrush into the gate and thus the dip will be much smaller.
Maybe even gone if the problem is with capacitance as I suspect.

The last graph is of the really low RDS(on) part STP15NK50ZFP that I bought, which was a fully insulated part.
It also has some nice features like built in zener diode protection from too high voltages between gate and source.
However it's capacitance is much higher than the original part, and it's gate dip is really pronounced all the way down to 2v!

C921 which I mentioned before sits on the TL494's PWM output, it's an AVX SA205E104MAA general purpose Z5A ceramic capacitor.
It has pretty horrible specifications, significantly dropping in capacitance with increased heat or lower DC bias, or AC component.
That said it's parallel with the 100µF electrolytic cap C913 so it might not be an issue at all.
The one installed on my scope is original, and it measures exactly 0.1µF (103nF) so it's still ok after 35 years.
I noticed it's not even present in the 2215's big brother 2235 schematic, so I took it out: there was a bit more ripple, but the dip was exactly the same.
So it has no effect at all, I even put another small electrolytic cap parallel above it but it did not make any discernible difference either.

Finally I tested to add the 39Ω gate resistor R909 right now, it improved the situation with the original MOSFET IRF820PBF quite dramatically.
I have attached a scope shot of that here, and also a picture showing where I installed the resistor.
As I understand it the 2235 also has a ferrite bead on the MOSFET, so I'll go see if I can find any pictures of that.
The eagle eyed might also notice I've replaced resistor R911 & zener VR917, mine read 161k & 6.0v respectively, so it felt best to replace them since they are 5% components.

[Vince]

Thanks for sharing your experiment son the FET Per, much appreciated because I don't have time nor the inclination to do all that right now... need to move on to the other stuff, can't afford to work on the SMPS for weeks or month on end at the moment...

Call me stupid but I don't understand why it is so bad for the FET to turn off during the rising edge, then turn on back on again to eventually become turned on for good.  I understand (from your best/worse graph comparison) that it about doubles the turn-on time, so twice as much energy/power to dissipate, hence much more losses/heating up issues... but in the 1191 EEVBlog video they just focus on the very fact of turning-off being bad... but don't say why it's bad.  This imply that if, say, the dip was big but the FET didn't turn-off, then they would not care as much about it (other than the increased losses of course).

Does this quick on/off cycle "stress" the FET and reduces its life expectancy ? Its whole purpose is to turn off and on all of its life in the first place... 

Or is it that it would "upset" the "balance" of the control loop/operation of the SMPS.  It would also I guess generate more "crap"/noise at the output, but the slew rate of the "second" rising edge, following the dip, is exactly the same as the first one, no faster, so the harmonic content would be the same, hence the existing filtering downstream, would be just as efficient at getting rid of it. Not saying that as an "excuse", but just saying it would probably be not "deadly"...


I want to understand better 


The video though gave an interesting bit of info as to why this dip even exists : the Miller capacitance that suddenly comes into play at some point as the gate voltage increases past some threshold, and causes extra capacitance to be "seen" from the gate, hence  suddenly putting extra load on the gate driver circuitry, understandably resulting in a voltage drop at the gate.

Anyhow. You got me curious Per (and slightly anxious as well ;-)  so I will have a quick look at the gate on my 2215, out of curiosity... and also just in case I see a horrible dip. If so , will try adding that 39 ohm resistor in series with the gate. Your experiment about this sounds very promising indeed ! I notice that in my 2232, this 39 ohms resistor comes as standard.

In my service manual for the 2215, I see that the waveform they give for the gate (without the 39ohms resistor then), displays no dip whatsoever, bit strange then...

Will go have a quick probe for that and report in a bit...

[Vince]

Here are the results ! 

See pics below. Sorry for them being blurry... was not in the most comfortable of positions : holding the probe in one hand, it hovering scarily close to the 400V cap terminals (had the board upside down, in its normal position, properly secured), and holding the camera in the other hand, trying to press the shutter while making a desperate attempt at remaining steady...   

Anyway. When you look at the general/complete gate waveform, it looks very much like what's pictured in the service manual, i.e. with a good bit of overshoot and less of an undershoot.  Even from this "distance" though, one can "guess"/suspect there is something going on inside the leading edge, as its upper half looks fatter than its bottom half.... begging to zoom on it.  And there you go, a big dip like Per showed.

The dip is about 50% of the pulse height. It dives down to about 7V, which is twice the typical gate turn-on voltage specified in the datasheet (attached) : I am using a IPP60R280P7 , turn on is given for 3,5V typical, and max is 4,0V.  3.0V Min.  So looks like I am well in the comfort zone from this regard.

Still, was quite a big dip, so I tried adding the 39 ohms resistor, hoping it would improve things a lot, like it did with Per.... but as you can see it was very counter productive to say the least ! It made things worse, not better.  And much worse at that : the dip now gets down to about 3Voilts, 3,5V at best. So, pretty much half the gate voltage I had before, hardly an improvement.  And of course, it's now bang on on the turn-on voltage....

Concurrently (consequently ? ), the SMPS starting to emit the dreaded whine, argh !!!

Removed the resistor, whine was gone and the dip back up to 7Volts, phew ! 


So that was an interesting experiment indeed, you can never make too many experiments eh ?! 

So this gate resistor  definitely makes a difference, good or bad apparently depends on what particular FET it is associated with...

... and a possible link to the famous whine as well... so that's yet another lead for when I get round to diving again in the SMPS of my whining 2232 : look at that gate voltage, check if removing removing its gate resistor (mounted as standard on the 2232) helps or not, both with the dip and the whine...


Anyway, that was a cool Saturday evening experiment 

[tautech]

It's all about managing the gate drive current draw so to not exhaust the available gate drive with a fast gate rise time. 39 ohms might be too high resistance so down closer to 10 ohms might be a better choice.
Experimentation would be required. 

My mentor recommends 2-5 ohms for gate resistors.

[Vince]

Yeah would be nice to experiment with different values of gate resistance to see what that does. I do need to build myself a little decade box, indeed.... yet another project ! ...

Will try and do that when I get my head again into my 2232 SMPS for its due recapping.


But for now I have a new more pressing " project " : yesterday I was tasked with saving the life of an old dishwasher !   
But it's a professional bit of kit so the electronics inside is interesting, decent quality. It looks like its most definitely an electronics related issue, so they thought of me for having a look at it... hell maybe I could open a new topic for that ! 

Was built in 2005, they bought it used 10 years later for 2,000 Euros (2500 USD), worked for a year just fine, then one day out of the blue, it just decided not to turn on anymore. Front panel is totally unresponsive and all it does is light solid, all the decimal points from the two 3 digit LED 7 segment displays.

They got the manufacturer (" HOBART ", American), service Tech to come have a look... said it was the main board... which cost 1500 Euros (1800 USD or so) to replace ! They declined the offer ...
That was nearly 2 years ago and has been sitting there unused ever since...

Front panel only carries buttons and LED displays and indicators, and has its own 8 bit micro (a PIC 16FXXX something). A 10 wire cable connects it to the main board that's in the underside of the dishwasher...  half through hole, half SMD. Good 'ol through hole EPROM for the Firmware, micro don't know what it is haven't looked yet. A dozen relays and provision on the PCB for 8 more ! That's one hell of a dishwasher ! LOL

Brought the two boards and interconnecting cable home so I can study them at ease in the lab. I am hoping/thinking it could just be the (much more simple) front panel board that's toast. I mean, even if the main board was dead, say, the front panel being (somewhat at least) autonomous thanks to having its own micro, should at least have SOME life left in it, no matter what the condition of the main board ! Even if it failed to communicate with the main board, I would still expect it to have a "rational" behavior, and still have a degree of responsiveness.

So I need to work on this thing right now, unlike my personal projects, I can't afford to let this one rot on my bench for a year, scattered all over the bench   

Will try contacting the mother ship in the US to ask for a service manual or at least schematics. Not holding my breath though ! Usually these expensive professional gear require you to pay the manufacturer a lot of dosh to get access to their service information. My only hope is that the bugger being not so new, from 2005, 13 year old already, they may be more a little more relaxed about it... will see.... 


Will get back to my scopes ASAP...

[tautech]

This auto-brightness feature in the 2215 is just driving me completely nuts, I am pulling my hair out, losing appetite, losing faith in the ancient Greek God " Elektron ". Spent whole day last Saturday, working/probing the scope... so I made some progress, but still far from a diagnosis... maybe because this scope is acting so erratically that I am still in the process of figuring out WHAT exactly are the true symptoms.... trying to make sense of its weird behavior, trying to see the "signal" in all that "noise" of symptoms.... is not yet fully achieved I am afraid.

I am not giving up just yet, but I think I will not spend too much time on it (other stuff on my plate), because I had a look at the service manual for the 2215A scope, the revised/improved version of this 2215 scope... and came to realize that the only thing they changed on it, spec/feature wise.... the only user visible change they made... was to DROP this auto-intensity "wizardry" altogether ! They didn't even try to fix it, they just gave up completely on it, to the trash !
So I think if Tek themselves gave up on it.. it must be really, really hopeless, and I am basically wasting my time trying to get to work, something that most likely never worked properly even when new back in the day !   

So I will not be losing sleep over this 2215...

I am not out of leads/ideas just yet, and I hate being defeated, so I am still working it for now... but I also know that it will probably never "just work", so I won't try to make it perfect. Just trying to get rid of the most blatant/inexcusable problems with it, as I am hoping that at least these bits I have a decently reasonable chance of fixing. But I will not spend 10 years trying to iron out every tiny little quirk, because they probably existed even when it was brand spanking new ! 

Dunno if this is related Vince, but from another thread:

If you want to make sure the CRT stays nice and bright make sure sup3010 has been done.

SUP3010 instructions attached.

Follow the quote to the SUP3010 link.

[Per Hansson]

Still, was quite a big dip, so I tried adding the 39 ohms resistor, hoping it would improve things a lot, like it did with Per.... but as you can see it was very counter productive to say the least ! It made things worse, not better.  And much worse at that : the dip now gets down to about 3Voilts, 3,5V at best. So, pretty much half the gate voltage I had before, hardly an improvement.  And of course, it's now bang on on the turn-on voltage....

Your post made me redo my tests, the 2235 manual change that adds the gate resistor does it in a point ahead of the point they use to measure the gate voltage.
I looked in the 2215A manual, also attached a screenshot of it and it's the same: gate voltage measure point is before the resistor.
Now if I measure after the resistor, right on the gate as I assume you did Vince I see the same thing as you!
So what where Tektronix trying to convey here, can someone please explain? (Surely the measure point should be on the gate?)

It's all about managing the gate drive current draw so to not exhaust the available gate drive with a fast gate rise time. 39 ohms might be too high resistance so down closer to 10 ohms might be a better choice.
Experimentation would be required. 

My mentor recommends 2-5 ohms for gate resistors.

Yeah would be nice to experiment with different values of gate resistance to see what that does. I do need to build myself a little decade box, indeed.... yet another project ! ...

I did this the easy way: with a multi turn 50ohm pot
Still using the same IRF820PBF the short version is that the best gate response (measured on the gate) is seen with the lowest resistance

Dunno if this is related Vince, but from another thread:

If you want to make sure the CRT stays nice and bright make sure sup3010 has been done.

This is "only" to make sure the CRT does not burn out from too high filament voltage, it does not affect the auto brightness function.

[Vince]

Thanks for the extra info.

Yes I measured right at the gate pin, because it's the only thing that makes sense to me, and because that's where the 2215 service manual measures it...

If they measure it at the resistor end in the 2215 manual, then I am at a loss just like you are... and just to add to the confusion : just looked in my 2232 manual, where the 39 ohm resistor is fitted as standard... they don't even look at the fate anymore, they only give a waveform probe straight at the open collector output of the PWM chip.

My only hypothesis is this : they don't care about the dip.. this is their engineering problem to deal with, it should not be a matter to be addressed by a service tech. Plus, the waveform they give, doesn't show the dip, it's not zoomed in enough for that. Also, they hardly mention it in the "Theory of Operation " pages.

So, what I think it that their only intention with this waveform, is to show us thee PWM output in general, and show us that some overshoot is acceptable/normal.
So, it does not really matter where exactly they probe for the PWM signal.. it will always look the same : still the same voltage levels, still the same square signal, still the same duty cycle, still the same overshoot.

The fact that the probe it a different points, depending on the scope model, maybe simply because mostly likely, the measurements / manuals where made by different people. Their only intention was to document the PWM waveform "in general", so different people probed at different locations, depending on their mood or personal view on where was the best place to probe.

I don't think there is much more to it than this...

Now about the dip, the manual clearly did not mean to address the issue at all... it's only US who decided to look for this thing...so nobody but us can decide where its best to probe for it. Obviously probing right at the gate pin on the FET package, is the only way that seems relevant.. since the FET reacts based on whatever it sees on its gate.. it doesn't care about what the signal looks like elsewhere...

My view on all this any way ! 

[tautech]

Still, was quite a big dip, so I tried adding the 39 ohms resistor, hoping it would improve things a lot, like it did with Per.... but as you can see it was very counter productive to say the least ! It made things worse, not better.  And much worse at that : the dip now gets down to about 3Voilts, 3,5V at best. So, pretty much half the gate voltage I had before, hardly an improvement.  And of course, it's now bang on on the turn-on voltage....

Your post made me redo my tests, the 2235 manual change that adds the gate resistor does it in a point ahead of the point they use to measure the gate voltage.
I looked in the 2215A manual, also attached a screenshot of it and it's the same: gate voltage measure point is before the resistor.
Now if I measure after the resistor, right on the gate as I assume you did Vince I see the same thing as you!
So what where Tektronix trying to convey here, can someone please explain? (Surely the measure point should be on the gate?)

Tek would've not wanted the added confusion of observing the miller dip on the gate, they only needed to show an 'expected' gate drive waveform.
As this recent discussion displays, choice of a low gate capacitance MOSFET is important if the gate drive available is not strong.
Another comment offered by my mentor; hammer gates with sufficient drive !
It's very plain to see why he made this statement. 

Could Tek have added a dedicated gate driver, yes but at increased cost.
Careful choice of components is the alternative.

[Vince]

Tek would've not wanted the added confusion of observing the miller dip on the gate, they only needed to show an 'expected' gate drive waveform.

That's two of us thinking that then.. Per are you joining the band and call it a day ?! 

Another comment offered by my mentor; hammer gates with sufficient drive !
It's very plain to see why he made this statement. 

Yeah sounds good, though IIRC from the Art of Electronics (?), one also needs to be careful with that, it's not impossible to make the FET/ gate oscillate, and one reason why putting a small resistor in series with the gate is always a good idea, to dampen things and keeps them from going completely haywire.
Now I read that book 21 years ago so my memory might be corrupted by now.. will have to dig out the book and read it again...

You keep referring to your mentor, this is all very mysterious, I am getting intrigued... who exactly was your mentor ?.... 

Could Tek have added a dedicated gate driver, yes but at increased cost.

Hardly an expert, but the scope is from 1981 or something, so designed in the late '70's, power FETs were in their infancy or almost, I assume.. so chances are there did not even exist dedicated gate driver chips back in that day, to begin with ?! .... or maybe too experimental and/or expensive, so Tek preferred logically to roll their own driver with this discrete transistor...

[tautech]

Tek would've not wanted the added confusion of observing the miller dip on the gate, they only needed to show an 'expected' gate drive waveform.

That's two of us thinking that then.. Per are you joining the band and call it a day ?! 

Another comment offered by my mentor; hammer gates with sufficient drive !
It's very plain to see why he made this statement. 

Yeah sounds good, though IIRC from the Art of Electronics (?), one also needs to be careful with that, it's not impossible to make the FET/ gate oscillate, and one reason why putting a small resistor in series with the gate is always a good idea, to dampen things and keeps them from going completely haywire.
Now I read that book 21 years ago so my memory might be corrupted by now.. will have to dig out the book and read it again...

You keep referring to your mentor, this is all very mysterious, I am getting intrigued... who exactly was your mentor ?.... 

Could Tek have added a dedicated gate driver, yes but at increased cost.

Hardly an expert, but the scope is from 1981 or something, so designed in the late '70's, power FETs were in their infancy or almost, I assume.. so chances are there did not even exist dedicated gate driver chips back in that day, to begin with ?! .... or maybe too experimental and/or expensive, so Tek preferred logically to roll their own driver with this discrete transistor...

Careful selection of components available at the time is the key to understanding what and why they were used.
Tek engineers did the very best with what they had to work with.....at the time.
Yes today, it would be done differently.

My mentor.....a good friend and Dr of EE and wizard at maths.......I am not ! 

[Per Hansson]

Thanks for the extra info.

Yes I measured right at the gate pin, because it's the only thing that makes sense to me, and because that's where the 2215 service manual measures it...

That's not what my 2215 service manual says, see attached screenshot.
It measures the gate signal on the pin of CR931, one theory I have is that they made it like that to reduce the risk of a probe shorting gate to source, with resulting critical damage.

Tek would've not wanted the added confusion of observing the miller dip on the gate, they only needed to show an 'expected' gate drive waveform.

That's two of us thinking that then.. Per are you joining the band and call it a day ?! 

Well, as I mentioned I have replaced my IRF820 with a newly produced part, I'm not sure it's exactly the same as what was there originally, this is what I used: Elfa artno 171-15-223, a IRF820PBF MOSFET.
Basically I'm trying to make sure I'm not changing the circuit too much, the lower RDS(on) part I bought:  STP15NK50ZFP, that I feel deviates too much from my tests on the previous page. (Likewise for the scavenged IRFBC30).

I also have a theory about the added 39Ω gate resistor R909: that it acts as protection if the MOSFET fails, when my original IRF820 failed it did not take out any other components.
Interestingly it has 42Ω resistance between gate and drain and 13Ω between gate and source now that it's failed...

[Vince]

I also have a theory about the added 39Ω gate resistor R909: that it acts as protection if the MOSFET fails, when my original IRF820 failed it did not take out any other components.

The MOSFET failed in my 2232 and boy that 39ohms resistor sure didn't do much for protection : mains voltage got in there,  blew the gate driver and the entire PWM circuitry behind it, PWM chip literally exploded, was not a pretty sight in there... 

[Per Hansson]

That's a good point of reference Vince
I just read TK5Q65W_application_note_en_20180726.pdf, it made my head hurt a bit
I understand the purpose of the gate resistor a bit better after reading that.
Still from the tests we have shown we just get a worse response for the turn-on.
I mentioned some tests I did with C921 earlier, and found no big difference.
But should the issue not lie there, i.e. C921 is unable to provide the necessary gate drive?

[Per Hansson]

Did some more tests today.
Used the final MOSFET I had bought for this project, a IRF840LCPBF
Out of the low RDS(on) MOSFET's this had the best gate drive characteristics.
I then scavenged a 35v 10µF Epcos tantalum capacitor and added it in parallel with C921: no real difference.

But before calling this project quits I want to ask one question:
Pin 13 on TL494 is tied to ground so both outputs are running in parallel, I have verified this with my scope.
But only pin 10 (the secondary output) is connected to the MOSFET.
Pin 9 (the primary output) is not connected to anything.
If I wired them together would I not get twice the gate drive capability?
If the answer is yes then why did Tek decide to not do it?
I want to ask before I try and blow up my power supply...

[Vince]

Yeah good question ! 

If I understand the datasheet correctly, there is there two selectable mode of operation with regards tot he tow output transistors :

- " Single-ended " where they turn on and off at the same, time, and can therefore be paralleled for extra current.

- " Push-pull " where the two transistors are driven in a mutually exclusive way. When one is conducting, the other one is open.

I guess that since Tek rolled their own gate driver, that there was no need to parallel the two outputs from the PWM chip, as one was already more than enough to drive the external gate driver. If they wanted more current they would have probably modified their own driver rather than paralleling the two outputs ?!

Also, the specs of the output trannies were maybe deemed not fit enough of to achieve the gate driver specs they wanted, other wise they would have not rolled their own driver in the first place, I guess ?
Point is... Tek engineers aren't there to tell us, so let's not torture ourselves with questions that just can't be answered (just guessed at, at best)...   

[Per Hansson]

Yeah good question ! 
<snip>
I guess that since Tek rolled their own gate driver, that there was no need to parallel the two outputs from the PWM chip, as one was already more than enough to drive the external gate driver.

There is no external gate drive Vince, the TL494 drives the MOSFET directly.
The only purpose of Q931 that sits on the gate is to discharge the gate quicker than what R931 the 3K resistor would do on it's own.

The TL494 can supply 200mA on each of it's transistor outputs.
Curiosity killed the cat and I had to try it: I wired the two outputs together (shorting pin 9 & 10 together).
The result in attached screenshot, it's almost as good now as with the original MOSFET so I'm very happy with that!

[Vince]

There is no external gate drive Vince, the TL494 drives the MOSFET directly.
The only purpose of Q931 that sits on the gate is to discharge the gate quicker than what R931 the 3K resistor would do on it's own.

First time anyone points this out in this long discussion, so I am bit surprise !   
But yes... just looked at the schematics again, and the theory of operation pages... I somehow missed this important point since the beginning !   
Scary... 

Still, you would then have to wonder why they rolled their own "gate discharger" (not gate "driver" then, OK ;-), since the 494 as we just said, CAN be configured as a push-pull output precisely to drive the FET AS WELL as discharging it.

So back to square one...

Curiosity killed the cat and I had to try it: I wired the two outputs together (shorting pin 9 & 10 together).
The result in attached screenshot, it's almost as good now as with the original MOSFET so I'm very happy with that!

You now are the official 2200 SMPS experimenter ! 

[floobydust]

The TL494/594 output stage is for driving BJT's, so it's always a hassle to drive mosfets with it. I have never used single-ended mode with pin 13 grounded.
We'd have to estimate the gate-drive requirement to see if that mode would work or if adding an NPN driver+resistor to the TL494 output is needed for high capacitance mosfets.

I tried an LT SPice sim with 2235a circuit but not sure of the transformer winding inductances. With measured in-circuit values, no Miller plateau as large as what is reported, even with 240VAC line, so the sim needs some work. It's not a vanilla flyback converter with C907 at 1uF; I could not find decent mosfet Spice models. IRF840 has a leisurely 0.16Apk gate drive.

I measured:
T906 1,2  3.50mH 2.05 ohms Primary  Winding
T906 4,5  3.75mH 1.57 ohms Reset Winding
T906 9,10 4.50uH 0.22 ohms Tank  Winding
T906 6,7  452uH  2.00 ohms Aux  Winding

Fairchild (On-semi) Semiconductor SuperFET Model Libraries are encrypted and did not work for me.
ST gives six Spice models per mosfet, User Manuals UM1575: Spice model tutorial for Power MOSFETs
But I could not find Spice models on their website 

[Per Hansson]

Well, it's quite difficult to calculate it, at least that was my conclusion after reading up on it.
Thus I just tried it instead, gets more useful data that way anyway

In my previous posts I mentioned I used a 1A load, actually it was two 100Ω resistors so totaling 50Ω and a load of 0.86A if my math is correct.
If I remember right my load meter showed 43w, which would mean the PSU is 86% efficient at 240VAC.
The picture attached to this post is with the reading taken with the scope as load instead of the dummy 1A load...
The dip is a bit smaller: the reason is that the scope presents a bit smaller load: meter shows 35w to 36w when I have a trace on the screen...

[wermundo]

Hi,
i'm not sure if i'm right here, but i currently try to fix my Tek 2215 and so i just found this post.
I managed to find that the Q933 and Q931 is defect, for the Q933 it's no problem to find replacement but for the Q931 i can't find anything. The label says TI114 i guess. If i understand it right, it is a Tektronix "own" component? Do you have any suggestions for replacement part? Or is it not even necessary to replace because the TL494 can operate in push-pull mode like i read here (shorting pin 9 and 10 together)?

[Per Hansson]

Did you read reply #114 to #116 above?
You can't operate the PSU without it, it's purpose is to discharge the gate current.
The TL494 is not setup in push/pull operation mode but in parallel operation mode.
So you either need a new Q931 or if you want to get fancy you could change the TL494 into push/pull operation mode and use it to discharge the gate current.
But since Tektronix did not do this to begin with I would assume they had their reasons for it...

The post here talks about a possible replacement for Q931:
https://www.edaboard.com/showthread.php?180408-Tektronix-2213-preregulator-fail

[Chris56000]

Hi!

Just a courtesy–note for Member Vince to say how much I enjoyed his Tek 2213 writeups – these writeups written clearly and with plenty of pics and circuit diagrams are exactly what less experienced Members need to help with difficult problems!

As to Vince's English – as a native UK English speaker myself I cannot fault it – it's as good as any British person who posts on here!

Please keep your repair tales up Vince, I enjoyed reading them, so much so I'd like a blown up 2213 myself to try, but they've nearly always been only in the USA, and US sellers' sale prices and postage charges are just getting ridiculous!

Chris Williams

[Vince]

Hi Chris, thanks for the kind words, you have me all red faced now !

Sorry for my long absence, I was moving home and it didn't go well at all ! The house builder turned out to be very crooked indeed ! He was supposed to give me the keys in November 2018 but refused to because he didn't want me to notify in writing all the crap work he did here and there....a couple lawyers and over a year later...I finally got the keys no later than last Wednesday ! So your message is incredibly timely !

I have started today doing the final work inside the house that's required to be completed before I can move in : sand down all dry wall all around the house, including the ceiling (a pain in the butt to do !), then I will be able to paint the house, and move in at long last !

I have never done this kind of work before so I expect it to take some time, as I learn as I go...

But I am hoping it could be done in couple weeks, then some time to put the lab all back together again....and hopefully in a month max from now, I should be back in business fixing my old Tek scopes, and therefore posting here again about it ! 

While I was off-line, I had the opportunity of buying quite a few nice old "classic"/ tube Tek scopes...I just love these things, can't help but buy all those that are within reach fuel cost wise ! :-)

So stay tuned ! 

[BiasedAmpere]

>using Tapatalk

What other forums do you use also?