Repair of Tektronix FG502 11 MHz Function Generator

[xrunner]

But why, when you removed "Q540 & Q546" (I assume you meant "Q540 & Q542"), did it work fine without smoking the 10ohm resistors?

Yes I meant those. Because the deed was done by then - the caps shorted out or nearly so burning the new resistors and then they may have changed properties a little or a lot as they cooled down when I turned it off, and then didn't act the same way again. When I turned it back on the second time after pulling the socketed transistors I only flicked the power on very shortly and then back off because I was not confident in the situation.

I can't explain the physics of what went on inside them. But I can tell you that half of the new ones left in the bag are bad and after replacing them with electrolytics it's working just fine.

Did you ever go back and test the original capacitors with a power supply to make sure they were bad?  A little post-repair failure analysis can never hurt for confirmation, especially if it's easy.

With a power supply? No why should I when they tested bad with a capacitance meter and a DMM on the ohms range?  No there is enough evidence they went bad. But as I said it's purring along like a kitten now for several hours, all is well.

If you're up for an experiment, you could try shorting R284 to see if it has any noticeable effect on the signal output waveform (such as increased glitches at the trigger transitions).

Yes I will try that this evening if possible. 

[MarkL]

Here's your waveform. CH 1 is the collector of Q290, CH 2 is the trigger output for reference.

I realize this wasn't directed at me, but I see that the trigger output is about 2.7Vpp.  I assume you have the 50ohm terminator on.  This is the level I'd expect.

And it also shows that the signal driving the Q268 base circuit is about 5.7V, which is close enough to 6.1V considering that there's additional loading on Q268's emitter when the trigger output is terminated.

I'm only pointing it out since you don't usually want the voltage on your trigger output BNC's to be at a level that could cause damage to TTL circuits when they are unterminated.  By some of your previous posts I was concerned about that, which is why I was digging into it.  But everything looks ok.

Did you ever go back and test the original capacitors with a power supply to make sure they were bad?  A little post-repair failure analysis can never hurt for confirmation, especially if it's easy.

With a power supply? No why should I when they tested bad with a capacitance meter and a DMM on the ohms range?  No there is enough evidence they went bad. But as I said it's purring along like a kitten now for several hours, all is well.

You tested them as bad caps, but the test did not show they were shorted or low-resistance under voltage, which would have to be the case in order to fry the 10ohm resistors.

I'm only suggesting confirmation.  I like to do it when possible.  Certainly your call.

[xrunner]

And it also shows that the signal driving the Q268 base circuit is about 5.7V, which is close enough to 6.1V considering that there's additional loading on Q268's emitter when the trigger output is terminated.

I'm only pointing it out since you don't usually want the voltage on your trigger output BNC's to be at a level that could cause damage to TTL circuits when they are unterminated.  By some of your previous posts I was concerned about that, which is why I was digging into it.  But everything looks ok.

Yes you are correct thanks.

You tested them as bad caps, but the test did not show they were shorted or low-resistance under voltage, which would have to be the case in order to fry the 10ohm resistors.

I'm only suggesting confirmation.  I like to do it when possible.  Certainly your call.

I still have all the bad parts in a cup so let me see what I can find out. A little more smoke in the room won't hurt. 

[MarkL]

I have a bin full of brand new BNC jacks and was going to replace the originals since the center pin female are tarnished quite a bit. Problem is I don't have & can't locate the one specialized BNC, the one that has an insulated ground (see pics). I can find ones on Ebay but they are not made the same and go into a "normal" size hole (this hole is much bigger). I suppose I could use washers but does anyone know where to find the original part?

[I forgot to press "Post" on this from yesterday.  Maybe it's still useful...]

This is in the parts list, KC79-67.  A search on "bnc kc79-67" shows that it's commonly available, amazingly enough.

[Johnny10]

Actually, I have just changed that C546 and R546 as they were burned and shorted.
I did not change transistors Q540,542,530 or 532.
With the diode test they checked OK.
Funny seeing that pic of yours, I thought hey that looks familiar.

I rebuilt the Current Switch CR140 etc and new transistors Q140 and Q175. There was a lot of flux in this area from the previous owner and when I touched the connection between the diodes and resistor it just fell apart.


Replaced resistor R284 which was burned and cracked.

Checking Pin 6 output on U140 and U175 varying voltage while moving the Frequency knob.
The timing capacitors check out fine.
Yet, I have never had a Frequency output waveform or an output trigger waveform.
The Trigger Output is dead.

The main output has voltage which can be varied by the amplitude switch of near 0 - 1.3 Vdc.

I am still working my way through since changing those parts.
But I imagine I have some dead transistors somewhere.

I still get no waveform at Q290.

Will have to work on this tonight.

[xrunner]

This is in the parts list, KC79-67.  A search on "bnc kc79-67" shows that it's commonly available, amazingly enough.

Oh cool - I just found that part on Ebay - 2 for $14.99, ordered. Who would have thought the same part was available? Thanks. 

I tested the old 10 uF caps again. One measured a complete open, the other around 45 Mohms. Tested under a voltage the open one didn't break down up to 30V, but the other one tripped the OCP (set to 3 amps) on my supply at around 12 VDC.

I am still working my way through since changing those parts.
But I imagine I have some dead transistors somewhere.

I still get no waveform at Q290. Will have to work on this tonight.

OK let me know if you want a comparison waveform or voltage anywhere. I got a good one now to play with. Going to burn it in all evening. Proceeding here to calibrate and clean up, and replace the BNCs! 

[xrunner]

I just went into the lab to turn the FG502 on for a few hours - and nothing came out of the signal or trigger output. 

I immediately turned it off and put my nose close to the board, and thankfully didn't smell that smell. When I turned it over I saw that a wire on the extender cable had come off. It was one of the 25 VAC wires. I soldered it back on and we're back in business.

I hate it when that happens! 

[xrunner]

Actually, I have just changed that C546 and R546 as they were burned and shorted.
I did not change transistors Q540,542,530 or 532

....
I still get no waveform at Q290. Will have to work on this tonight.

Here's a couple of waveforms for your consideration, might help I don't know.

Had a similar issue with the last repair I did - the FG501A. There were no time-varying signals whatsoever anywhere in the thing. It tuned out that an output from a comparator (which had failed) was required to make the charging circuit switch states. There was nothing wrong with the basic triangle generator in that one.

Although that's not the same circuit as this one has, I think you know that it's just going to sit there until it can switch states. So keep looking in that circuit area. You might try to capture an initial state from as many point as you can (you have a 4 channel scope I assume?) when power is applied up to the point where it stalls, then narrow it down to where a signal is missing that should be there. For example when power is applied it should charge in one direction until it can't go the other way.

In the attached screen shot, the signals are DC coupled centered on Zero. You can see the triangle charge and discharge as the square wave changes.

[MarkL]

I tested the old 10 uF caps again. One measured a complete open, the other around 45 Mohms. Tested under a voltage the open one didn't break down up to 30V, but the other one tripped the OCP (set to 3 amps) on my supply at around 12 VDC.

Well, that certainly fully explains one of the dead resistors.  The other one, not so much.

I threw together a quick simulation of the output amp.  With a 50ohm load and the output fully high (or low), there is 0.257W being dissipated in the 10ohm resistor.  Since the FG502 operates at a 50% duty cycle, the average power dissipation in each of the 10ohm resistors is 1/2 that.  If the high or low time is long in relation to the thermal resistance, it could still get pretty hot.

However, if the output is shorted, the fully-on dissipation is 0.69W.  Even with 50% duty cycle, it's still going to be more than the rating of 1/4W.  Maybe driving a short or very low impedance is what killed it.

I would probably replace those resistors with 1/2W or 1W.  Lab equipment is supposed to be as bullet proof as possible under unexpected conditions.

(Ok, I think I'm done beating this one to death.)

[xrunner]

Well, that certainly fully explains one of the dead resistors.  The other one, not so much.

I threw together a quick simulation of the output amp.  With a 50ohm load and the output fully high (or low), there is 0.257W being dissipated in the 10ohm resistor.  Since the FG502 operates at a 50% duty cycle, the average power dissipation in each of the 10ohm resistors is 1/2 that.  If the high or low time is long in relation to the thermal resistance, it could still get pretty hot.

However, if the output is shorted, the fully-on dissipation is 0.69W.  Even with 50% duty cycle, it's still going to be more than the rating of 1/4W.  Maybe driving a short or very low impedance is what killed it.

I would probably replace those resistors with 1/2W or 1W.  Lab equipment is supposed to be as bullet proof as possible under unexpected conditions.

Well we're thinking the same thing, I guess you missed it but I had decided to upgrade the 10 ohm resistors to 1W a while back. Check the pic out you can see the blue resistors installed, they just fit into the holes, and the transistor sockets and 10 uF electrolytic caps. I'm half way through the service alignment and I haven't noticed any ill effects from using electrolytics.

(Ok, I think I'm done beating this one to death.)

Yea but now we can beat Johnny10's to death!

[Johnny10]

I am taking a methodical path through the parts.
I have found a few out of spec Caps.
Now I need part numbers or substitution numbers for FET Q200, Q202 one is missing leg.

Says matched pair !

 

[xrunner]

I am taking a methodical path through the parts.
I have found a few out of spec Caps.
Now I need part numbers or substitution numbers for FET Q200, Q202 one is missing leg.

Says matched pair !

Try this site out -

Tektronix to Generic Transistor
Cross Reference Database


Transistor Part Number 151-1042-00 has the following equivalent generic types:

2N5454   FET, N-CHANNEL

There's some on Ebay, buy a few and get as close as you can. 

[Kleinstein]

The exact JFET type should not that critical. Also matching is not that critical. One will need to adjust the DC level anyway. A good matching will likely also keep offset drift small.

[Johnny10]

I see that the part number you recommended is available as dual package.
Already matched.

[David Hess]

I am taking a methodical path through the parts.
I have found a few out of spec Caps.
Now I need part numbers or substitution numbers for FET Q200, Q202 one is missing leg.

Says matched pair !

They are Tektronix part number 151-1042-00 which is type 2N4416 and matched for an Idds (1) of 0.5mA between 10 and 15mA at 15 volts which is trivial to test with a power supply and current meter.

The 2N5486 (8 to 20mA) would be best but the 2N5485 (4 to 10mA) is almost as good.  I am sure the 2N5484 (1 to 5mA) will work but the other two will be better at the same cost and are closer to the original specification.

I think a J112 or BF256B (Mouser has them) would be acceptable also at half the cost.

(1) Drain current with the gate shorted to source or Vgs=0.

[bitseeker]

I just went into the lab to turn the FG502 on for a few hours - and nothing came out of the signal or trigger output. 

I immediately turned it off and put my nose close to the board, and thankfully didn't smell that smell. When I turned it over I saw that a wire on the extender cable had come off. It was one of the 25 VAC wires. I soldered it back on and we're back in business.

I hate it when that happens! 

That's a feisty function gen. Glad to see that it's finally behaving. Congrats!

[xrunner]

That's a feisty function gen. Glad to see that it's finally behaving. Congrats!

Yea they can be - Johnny10 has a rascal on his hands too. 

So cool beans - it made it through the entire calibration with flying colors. Just threw in a pic of my bench, I like to use an iPad with the manual PDF on the iCloud right there. Also shown is the test checking the 2nd and 3rd harmonics when set to 11 MHz. Should be -30 dB or better - they are.

[Johnny10]

I guess I know why I have no ramp waveforms.
After finding bad Q200.
And after pulling a few tantulum caps I now have voltage at Trigger Output.

Just need to get those FET transistors back in my unit.

And maybe I can take pictures like xrunner ?

[xrunner]

I guess I know why I have no ramp waveforms.
After finding bad Q200.
And after pulling a few tantulum caps I now have voltage at Trigger Output.

Just need to get those FET transistors back in my unit.

When you get the new tantalums check 'em first. When I got my new parts (the transistors, resistors, and tantalum caps) I checked the transistors and resistors, then figured it all had to be good. It all wasn't good. I'm messaging the seller and trying to get a refund.

And maybe I can take pictures like xrunner ?

Maybe? You gotta post pics dude - we all love pics. Speaking of which I'm not done yet, I'll have more to post, pretty pics with new BNCs and spit-shined knobbies. If she's choochin¹ she's gotta look the part.




1. Chooch - a term used by the vlogger AvE to signify an item is working.

AvE Speak

[xrunner]

I wanted to clean the PCB switch as with some of the other TM500 plug-ins. There are two of them in this unit. When I took the cover off one, I saw a tiny piece of wire just laying there. Just goes to show you can't take anything for granted. 

I gently clean under the contacts with some very fine 600 grit paper soaked in alcohol. Just wipe it a few times is all that's needed. Now I'm proceeding to remove all the controls and BNCs from the front panel so I can take it off, there's a lot of old sticker residue that has a date with some Goo Gone.

[David Hess]

I gently clean under the contacts with some very fine 600 grit paper soaked in alcohol. Just wipe it a few times is all that's needed.

You used sandpaper on gold plated switch contacts?  See photograph below.

Non clay bearing paper like vellum or card stock soaked in alcohol is more than enough and will not destroy the switch's contact surfaces.

[xrunner]

Remain calm - you misunderstand - I didn't use it like I was trying to refinish an old dresser. I just slid it under and back out very gently 2 times wit no pressure soaked in alcohol - nothing is damaged.

[David Hess]

Remain calm - you misunderstand - I didn't use it like I was trying to refinish an old dresser. I just slid it under and back out very gently 2 times wit no pressure soaked in alcohol - nothing is damaged.

[xrunner]

Yea well ... moving on ...

[xrunner]

So not only do I want it to work as new I also like to restore these to as new-looking as possible, because it's a challenge and I enjoy it. 

Removing the front face requires removing the knobs, shaft nuts, and all BNCs. The BNCs will be replaced including the original insulated BNC used for the trigger which I was afraid was unobtanium but MarkL found it was still available (thanks).

Once you get all that off the front face comes off. You can see where the grain of wheat bulb was mounted in the little white holder (will be replaced by an LED). I've also found on these that the corners crack over many years, due to years of pressure by the four corner screws. Some corners even crack completely off (I have a few like that). On this design you have to get the front panel off to remove the top and bottom aluminum frame pieces. The other design they came up with is easier to take apart.

The upper left corner you can see has a small horizontal crack but it's easy to fix with plastic welder (MEK). The front aluminum panel is cleaned with Goo Gone, will need to soak for several hours. It doesn't appear to be scratched so it should look really nice.