Up for another "Help Me Fix" Thread? (IEC F34 Fnct. Gen.)

[corrado33]

Hi all,
I have gotten in my hands a "semi" working IEC F34 Function Generator. The problem is that it does not in any shape or form have a stable waveform. Usually there is a lot of amplitude modulation on the signal.

Law of electronics repair #1: "Thou shalt check voltages."

Luckily, this device is from the early 70s! when they still knew hot to put nice pretty test points in everywhere. (And the manual is awesome.) Manual is attached.

Anyway, it has a split supply +-15 and +- 5 V. Both positive rails are dead on, but the negative 15V rail measures -12.## and the -5 measures -4.5.

So, I checked the manual (attached), and it says "Q31 or Q32 shorted" for the problem of "-5V rail too high." Technically the -5V rail is too "high", but it all depends on what you consider high with negative numbers. The other problem of "-5V rail too low" is simply Q31 or Q32 open". Neither of which is true. I don't know how a transistor can be shorted, but both had 10s of kOhms between the pins. I'm not sure how else to check them.

Regardless, I then checked voltages on the pins of the transistors, there I found my -12.## and -4.5 Vs. I think the -4.5 is due to the -12.##. So, Then I ran into a block. I tried looking into the parts list for voltage regulators... but I don't think they were around back then, cause all they have are zener diodes. Which, sure, can be used as voltage regulators in a way.

Anyway, I'm waiting until tomorrow to print out the schematic, it's next to impossible to look at on the computer (it's split over multiple pages.)

But, if you guys want to take a crack at it, I'd be mighty grateful.

NINJA EDIT: I found a cracked joint on the 2500uF electrolytic (they're wicked old, and look like they have a wooden top?) Maybe I'll replace them anyway, they're probably dead by now anyway right? They have 4 pins, but I think 2 are there for support (aka no traces go to them.)

Attached is the manual for the 30 Series Function Generator from IEC. F34 F33 F32 F31 (For google because I sure had a hell of a time finding the damn thing, I hope that's ok.)

EDIT: Ok it's not attached, but here is the link. It's 10mB so the forums didn't like it ;(
https://drive.google.com/file/d/0B6oqWuYTCiclOUI0d1RzSVRROXM/edit?usp=sharing

EDIT for pictures! Ok so I don't know how to edit and add pictures... I shall make another post.

[corrado33]

Closeup of the one side power supply circuitry Q32 and Q31 are the little TI can and TO 92 right below it. All of the way toward the back of the device (metal part, left side of the picture.) (And by below, I mean aligned to the direction that the text is written.

Other side of the power supply circuitry.



Apologies for the terrible picture taking... I don't have a camera

[Tac Eht Xilef]

I tried looking into the parts list for voltage regulators... but I don't think they were around back then, cause all they have are zener diodes. Which, sure, can be used as voltage regulators in a way.

Young'uns ...   

Zeners are perfectly capable voltage regulators, no "in a way" about it. In some cases they're even better than 3 terminal regulators. But just like 3 terminal regulators you need to make sure they're operating within their parameters, and just like 3 terminal regulators you can extend their abilities by using a series pass transistor. In fact, the circuits are almost identical.

(In other words, check out the -15v regulator circuit; it's unlikely to be anything complicated. It probably looks something like this one: http://www.bristolwatch.com/ele/img/zr5.gif)

EDIT: Ok it's not attached, but here is the link. It's 10mB so the forums didn't like it ;(
https://drive.google.com/file/d/0B6oqWuYTCiclOUI0d1RzSVRROXM/edit?usp=sharing

"You need permission

Want in? Ask the owner for access, or switch to an account with permission. Learn more

You are signed in as yourestupidifyouthinkthisismyrealemailaddress@gmail.com."

[vk6zgo]

What frequency is the AM on your signal?
My guess is either Mains rate,or double that!

If a manufacturer uses the term "too high" for the -5volt supply,they mean "numerically high"as in -6v is higher than -5v !
In the real world there is no room for the silly notion that "more Positive is higher".
Negative & Positive voltages of the same numerical value are equally "high" with respect to zero volts,just different polarities!

[corrado33]

Fixed the link for the manual. I didn't switch it over from "private" to "anyone with link" although it generally happens automatically once you get to the screen where you can copy the link from...

My next step was going to be to check the voltage across the zeners, however it was... hard to do without a viewable schematic.

I'll do that when I get home from work today.

What frequency is the AM on your signal?
My guess is either Mains rate,or double that!

In the real world there is no room for the silly notion that "more Positive is higher".

I will attempt to measure the AM on the signal. There are a TON of things I can do with measuring the signal, the manual lists where the signal comes from, what filters it goes through, what the waveform should look like at each step etc. etc. I just figured the power supply issue should take precedence.

As for "more positive is higher." Unfortunately I have to deal with this daily in my line of work. We tend to avoid ambiguous words like higher or lower and use things like "More positive" and "more negative" etc.

I guess I should have checked voltages more carefully. I missed the RAW output from the transformer (which is supposed to be at 22V.) I thought RAW stood for the raw signal, IE the initial signal before it was processed. I guess I should have thought differently since it's in the "power supply" section of the PCB.

EDIT: I've attached a picture of the power supply schematic.

[corrado33]

I'm going to take a whack at the circuit. Prepare for terrible assumptions and incorrect math.

I'm going to start at TP2, because it that's not correct then there are only a few small parts that could be bad (transformer, caps, or rectifier diodes.)

After TP2 we have a pair of PNPs in a darlington configuration. I'm assuming these are being used in active mode to amplify the current. That current is then fed through a voltage divider which provides a (I'm assuming) stiff current for the base of Q30. From the collector of Q30, the path goes through a 12 V zener. From there it splits off, with another 6.8V zener going to ground, and then a 470 ohm resistor into an op amp. I would make the assumption that the voltage on the collector of Q30 should be around 18.8V. (12 + 6.8 ). The negative voltages are a bit confusing, so I just keep referring to the +15V circuit above which is identical (except reversed polarities) to this point. Anyway, the op amp is where I get confused. It's probably in some sort of negative supply configuration, but I can't understand it. I know that going into pin two you have what should be 0V. There are precision 10k resistors across the +15V line and -15V line that are in parallel with each other, so I think I can assume that they'd add out to 0V. OR, the voltage will be the DIFFERENCE between the voltages, and since the +15V is managed using a pot the op amp can then adjust the negative supply accordingly? (And before you ask, yes, I have adjusted the pot in the +15V circuit to within spec.)

So there, I'm stuck. For the record, I'm probably mostly wrong, but I'm doing this for my own good. Plus I think it's fun to speculate. I'll do a bunch of measurements when I get home then revise my theory.

[SeanB]

Yes you are right, it is a tracking regulator. As it is stuck at basically the 12V zener voltage check the output of the opamp IC11 pin 6, and the voltages at pins 2 and 3 as well. Check the zener VR6 is not shorted as this is the most likely suspect in this case. Then post the voltages. If the zener is not shorted then the IC is the easiest suspect to change.

[KD0RC]

Did you measure the 'raw' voltages of +/- 22.2 VDC?  If the cap is bad, you could have a low voltage condition there.  In that case, replace the big cap and see if you get a better result.

[N2IXK]

Those white plastic cased electrolytics (Mallory?) don't age very well. I would suggest checking them, as well as the large one.

[corrado33]

Well, as it turns out, it was much more simple than I thought, and it was actually something I commented on in my first post.

Here's what I did today.

I measured the RAW voltages, and got +24 and -19. Thought that was weird. There are only a select few components before the RAW testing points, namely the transformer, the rectifier diodes, and the big caps.

The transformer was outputting 19VAC to ground. Maybe that was only half the winding? There was only one wire not green (ground). Anyway. I thought those voltages were weird on the RAW TPs, but I moved on none-the-less. I then went to check the op amp. There are a ton of these op-amps on the board in sockets, so I simply switched two of them. That didn't do anything.

Then I went on to measuring the diodes. One measured 6.8 and one measured 12.##, so those seemed fine.

Then I forgot that one of the capacitor joints was cracked. I looked it over, found the cracked joint, and reflowed it. I had to add a bunch more solder, it didn't seem like much was on there (or it just fell through the giant hole.) Then I looked at the other cap and it had the same cracked joint. Reflowed that one as well. Then I saw a small discolored spot on the ends of one of the large yellow caps (exact center of the board.) So I reflowed that joint as well.

Plugged it in, turned it on and I had a perfectly stable waveform. Everything seems great actually. I would like to somehow either replace or simply clean the legs of the op amps and other socketed pieces. They all have a bit of rust on them (and I managed to break off half of one of the legs of the op amps I pulled out.) I'm not sure how I'd clean the sockets themselves, that'd be a royal pain. Probably would be easier just to replace them, but I'm not sure if it's worth it. Look at the last picture above, on the legs of the DIP 8 op-amp. They're not black because they're reflecting the black color of the package. They're black because of rust.

Anyway, I had fun poking around in this thing, even though it was struck by the EEVBlog lab curse. (I don't consider that a bad thing since I actually want these to work.  )

I also amusingly noticed that there are a pair of "owl eyes" on the solder side of the board. Either that or it spells lolol.