Black Jupiter 2000 Function generator circuit.

[johnmcgivern]

Just a point which has been lost in the discussion about what might be wrong with the generator John, you wrote:

I am just a novice I was using the generator to do rf alignment on a Pilot u650 1936 radio resoration nothing seemed to be happening so hooked up to scope and there was no amplitude on signal I had bought the scope on ebay hardly ever used it sods law I only bought it to try and learn how to use it not repair it ha ha.

Without wishing to sound patronising, that generator is for such tasks as checking the distortion of amplifiers - it isn't the right equipment to align RF stages of radios, and unless you connected it via a dummy load, it isn't the right method. Furthemore, you should never connect a generator or other test gear directly to a radio - only via a 0.1uF high voltage (EG, 630V DC) capacitor or you risk damaging the output circuitry of the test gear.

The U650 is quite a complex receiver, in that it has four wavebands and an RF stage. I'd be extremely reluctant to tamper with the RF alignment unless someone else had tampered with it first. It's highly unlikely that the alignment will be off tune. The service sheet for the U650 states that you should first align the IF stages (via a 0.1uF high voltage isolation cap, then the RF stages via a dummy load connected to the E & E sockets. It outlines thart the sequence should be MW, SW2, SW1 and LQW in that order, and it detials how this should be acomplished.

I'm wondering therefore, whether - if you connected the Jupiter directoly to the front end of the radio without an isolating cap - you may have damaged the output circuitry (TR20, & 21 for example?) Worth checking if you have a square wave as shown on the circuit of the triangle generator and waveform shaping circuit at S10, next to TR11. If you do, at least that narrows it down a bit.

I hope these comments sound helpful rather than critical John.

Best of luck!

Hi David, Thanks for replying.
I used a 0.1uF 1Kv cap to hook up to aerial on u650. I was following the instructions as per trader info as you have said but due to no signal all was abandoned.
The radio plays on all bands I did the IF alignment by ear was quite a way out I was trying the RF alignment to see if there was any more improvement.
I have never done any restoration before this is my first so I am trying to learn as much as possible on this radio. I retired early through medical problems and needed something to occupy myself thats not too physical.
Tried this and got the bug.ha ha.
I bought an I/T Variac, scope and function generator 2 Dmm's other odds and sorts
to learn so I am on a steep learning curve with all equipment and components.
Regards, John.

[Mark R]

John, I have one of these. Any chance of the circuit please.

cheers Mark

[johnmcgivern]

Hi everyone, received circuit.
Checked voltages on IC's all good apart from IC3 AND IC4 should be +20v and =20V both are 3.5v higher +23.5 =23.5.

These supplies are unregulated, so I wouldn't worry too much about that...

Struggling with how circuit actually works any suggestions would be most grateful.

Despite the apparent complexity, taken at a "block diagram" level, it's actually pretty generic...

The way the triangle waves are made involves putting constant currents into a capacitor. First the capacitor is charged, and when it reaches a predetermined threshold, the polarity of the current is reversed, and the voltage heads down towards the negative threshold, at which point the current is once again flipped and the cycle repeats.

The comparator that determines when the positive and negative thresholds are reached outputs a square wave, which controls the current to the capacitor. This square wave is also available as an output.

To make a sine wave, the triangle wave is deliberately distorted, taking advantage of the exponential relationship of VBE and IC in a transistor.

Looking at the diagram, IC3 is the positive current source and IC4 is the negative. The op-amps before these control them, summing the frequency control and the external input. These currents are fed into a diode bridge, which is used to determine whether the positive or negative current reaches the timing capacitor.

You'll note that the frequency decade switches are able to change the capacitor being charged (C1 to 4), and also the value of the resistors in the current sources. By altering the currents, you need fewer capacitors (large ones get expensive, so reduce the charging current instead). IC5 is almost certainly being used as a capacitance multiplier instead of a 10u non-electrolytic device. You'll also note the duty-cycle control, that alters the balance between the currents...

The output from the capacitor goes to the FET (TR5). A high impedance is required here, as any current taken will distort the triangle wave. The following transistors in the CA3086 form a conventional amplifier, the output of which is on TP9. It goes to 3 places; the comparator (IC7), the sine converter (bottom right) and the waveform selector switch.

The comparator is a conventional enough circuit, with the output being amplified and buffered and fed back to the diode bridge (arrow A). The output from the chip is also fed to the TTL buffer and the wavechange switch.

The amplifier takes the "mixsig" from the switch and deals with it in a conventional manner, and has constant-impedance attenuators dealing with the output. Similarly, the PSU is entirely conventional, but perhaps not at glance if you're more familiar with 78xx regulators...

Incidentally, I built one of these some years back - probably with more transistors and discrete circuitry than this one, which is how I'm so familiar with it

I think there has been a new revision on this drawing as it calls for 2 off TL072.
There is only one on the board unless I am reading it wrong and there is a distinct possiblity that I am.

No, definitely two shown on the diagram...

Time to take some measurements to determine where the fault lies. Any sign of the TTL output? Any signal on the capacitors? Is there a DC output - a faulty comparator would probably stick at one rail, so you should have a voltage on the timing capacitor...

As you say the power supplies are OK, this might not be completely straightforward, but look for common-sense stuff first.

Good luck,

Mark

Hi Mark, thanks for the explanation of the circuit It will take me some time to try and get my head round it.
I have just found the other TL072 it was hidden by the 14 pin band connector.
Voltage ok. There are a couple of mods on the circuit board Pin 4on NE527 has been cut off. A 101ohm resistor couples pin6 of ne527 to pin2 of 74ls37 also
2N3958 linked by 47ohm resistor to BC214.
NE527 pin1 =10.27v, pin6 = -10.27v pin8 = 4.94v pin 13 = 4.94v pin14 = 4.94v.
There was a triangle vave on pin3 and a square wave on pin 11.
Will do some more checking tomorrow.
Thanks again, Mark.
I cannot find the test points, I wouldn't know what to expect ha ha.

[Mark Hennessy]

This sounds encouraging, as it suggests that the complicated bits are OK. My next move would be to see if there is a signal at the top of the amplitude control (VR17), and that all 3 (triangle, square, sine) are OK. That rules out the switching. Then check the output of the amplifier (junction of R76/77), and if it's OK there, then the output attenuators are likely to be the issue. I guess the amplifier is most likely to be at fault, assuming you've already tried a bit of switch cleaner on all the switches.

Cheers,

Mark

[Yorkie]

Hi David, Thanks for replying.
I used a 0.1uF 1Kv cap to hook up to aerial on u650. I was following the instructions as per trader info as you have said but due to no signal all was abandoned.
The radio plays on all bands I did the IF alignment by ear was quite a way out I was trying the RF alignment to see if there was any more improvement.
I have never done any restoration before this is my first so I am trying to learn as much as possible on this radio. I retired early through medical problems and needed something to occupy myself thats not too physical.
Tried this and got the bug.ha ha.
I bought an I/T Variac, scope and function generator 2 Dmm's other odds and sorts
to learn so I am on a steep learning curve with all equipment and components.
Regards, John.

Hi again John,

It's good news that you used an isolating cap because if you look at the output(s) of the Jupiter it has no protection and if you squirt high voltage into it, it would be curtains for transistors such as TR20/21, but that said, in a perverse way, even if that happened, they might thus become fast acting three-legged fuses and offer some protection to the rest of the circuitry.

When poking around with test gear it's so easy when for example intending to connect to the grid of a valve, to inadevertantly touch the wrong pin (anode for example), and squirt high voltage into unprotected test gear circuitry, hence, the 0.1uF cap which you wisely used. At least in the Pilot U650, you cant mistake the grid of V1&3 as it's the top caps.

I think that if you first focus on the circuit on 'page 12'- the Triangle and Waveform Shaping Circuits' - you'll see that at S10, S11 and S12 there should be square, sine and triangle waveforms at those points respectively. If you hook up tyour scope and are lucky, and there are those waveforms, the faul lies in the output amplifier circuitry on page 11. If there are no waveforms at those points, you'd need to check for signs of life with your scope at places such as TP9 and other test points, which are there to help you pinpoint the likely area of the fault(s).

Not an easy task for anyone without the manual, but hopefully you'll be able to pick your way through it. Easy to get disheartened, but it's all good experience!

You've got quite a nice bit of test gear - I'd suggest an RF signal generator would be a good addition to your arsenal, and if you haven't already made one, a lamp limiter.

Best of luck John.

[Mark Hennessy]

Personally, I'd use some series resistance in addition to the capacitor - say 10K. Even a relatively small capacitor like 100n is a dead short circuit briefly :-)

Happily, the transistors in this circuit are pretty generic - the small-signal devices are JEDEC BC548/558 as far as I'm concerned. The output pair are pretty common - I've used them a lot over the years - and should be still available, but if not, I'm sure they'll be easy to come by second hand...

My first function generator was a Wavetek 162, and works in much the same way, only it has rather more bells and whistles, such as the ability to stop the generator, and output single cycles or tonebursts. As it goes to 30MHz, the transistors used are somewhat more specific! I think I was 16 when I was given this, and looking back, fixing the amplifier was quite an achievement!

   

Function generators aren't very good at producing low distortion sine waves, but they are so good at everything else that everyone needs one IMHO. For audio work, you really need something that's Wien bridge or state-variable based - although these days it's all gone digital (boo :-()

David, see post #16 - we've got triangle and square, so it's probably the output amp that's at fault. Personally, I'd dive in and change all the transistors (there's not many) and hope for the best. Although if I had plenty of time, I'd attack it more methodically, as it's always more satisfying to chase a fault logically

You're right that there is no protection here beyond the 50 ohm output impedance. I think I'd add 1n4148 diodes connected between output and each rail - these wouldn't offer all that much protection, but it's better than nothing. Bigger diodes like 1n400X would be better, but you'd have to check that the shunt capacitance isn't too high - I guess it would be fine here, but assume nothing .

Good test kit is always worth fixing, and especially if you can learn something along the way

All the best,

Mark

[johnmcgivern]

You've got quite a nice bit of test gear - I'd suggest an RF signal generator would be a good addition to your arsenal, and if you haven't already made one, a lamp limiter.

Hi David, Thanks again for the info as I need as much as I can get.
I have made a lamp limiter been trawling the net for a signal generator but not with any success. I shall try the test points tomorrow.
Thanks everyone for your invaluable help.

[Yorkie]

I hadn't spotted that you do in fact have waveforms at the 527, as Mark has pointed out, (referring to post No 16), which is really good news, because as Mark says, that indicates that the complex part of the generator is working, and narrows it down to the output amp circuitry. It suggests that one or another of transistors TR13 - TR 21 are the culprits. You might like to roam around those transistors with the scope, but then if the transistors are easy to get to, you would probably spend less time changing them than you would trying to chase the fault all over the PCB.

Just a note of caution if the PCB has plated through holes: It's much safer and stands less risk of damage if you first clip the transistors off leaving the leads in the holes, then one by one, melt the solder on the leads and pull the leads gently out with tweezers or forceps. The same would apply to ICs - never try to desolder every pin with a desolder pump with the IC in place. First clip off all the IC pins with snips, leaving the pins in place on the PCB, then unsolder each pin and gently pull it out with tweezers. If you leave the IC or transistor in position and try to desolder the leads, they act as heat sinks and you risk overheating and damaging the PCB. In the case of plated through holes, you risk pulling out the sleeve due to the solder melting on the underside of the PCB, but not on the upper side.

Hope that helps a bit.

[johnmcgivern]

Happily, the transistors in this circuit are pretty generic - the small-signal devices are JEDEC BC548/558 as far as I'm concerned. The output pair are pretty common - I've used them a lot over the years - and should be still available, but if not, I'm sure they'll be easy to come by second hand...

Hi Mark, I can,t see any transistors with BC 548/558.
Regards, John.

[Mark Hennessy]

No, I was saying that the 2N3904/6 transistors used are very generic, and BC548/558 are what I'd use as replacements, as I have lots of them in stock