HP 608C Repair.

[neo]

A connection to ground was ignored, i fixed that and now i get
No load;
Emitter, -7
Collector, -17
Base, -6.84

700R
Base, -6.84
Collector, -16.5
Emitter,-6.74

400R
Emitter, -6.71
Collector, -16.20
Base, -6.85

5R
Base, -3.75
Collector, -12.72
Emitter, -3.35

I don't know about you but i would call that progress. 

With that i bid you adieu, it is 5 am in these parts and i need to be up by my definition of early. 

[neo]

Just a note if the filament is 5.5 ohms, which it was when i measured it, then -3.75V can deliver 750mA to it.

[amspire]

Just a note if the filament is 5.5 ohms, which it was when i measured it, then -3.75V can deliver 750mA to it.

That is the cold resistance. The resistance increases a lot as it gets hot.

[Specmaster]

Someone was messing around in here long before me. This is all kinds of screwed up.

It would really help everyone trying to help you if you could sketch out the circuitry feeding the tube heaters on the affected tubes. What we are currently doing by the look of it is trying to describe a paperclip clip and telling a person who cannot see it, how to draw it. If you have not ever tried this, try it and you will not believe the drawing that the drawer will give you back. To you its simple because you can physically see it, but we can't, so we are trying to draw a picture in our minds based on what your telling us.

I gather from picking through the thread that you have something similar to the schematic amspire provided but what we don't know is how that differs from what you physically have for that part of the circuit?

I attach that schematic with some notes on it that would be useful to know what the volts are at those points, BUT only if you have this circuit, if not then you really will need to draw it for us all to understand what you have there?

[neo]

Someone was messing around in here long before me. This is all kinds of screwed up.

It would really help everyone trying to help you if you could sketch out the circuitry feeding the tube heaters on the affected tubes. What we are currently doing by the look of it is trying to describe a paperclip clip and telling a person who cannot see it, how to draw it. If you have not ever tried this, try it and you will not believe the drawing that the drawer will give you back. To you its simple because you can physically see it, but we can't, so we are trying to draw a picture in our minds based on what your telling us.

I gather from picking through the thread that you have something similar to the schematic amspire provided but what we don't know is how that differs from what you physically have for that part of the circuit?

I attach that schematic with some notes on it that would be useful to know what the volts are at those points, BUT only if you have this circuit, if not then you really will need to draw it for us all to understand what you have there?

What i meant was that there is a resistor missing, a resistor added, a connection to ground missing.  At the time i thought more was missing until i understood what they did, which is in compliance with the schematic.

Voltages on the labeled points,
No load;
Emitter, -7
Collector, -17
Base, -6.84

700R
Base, -6.84
Collector, -16.5
Emitter,-6.74

400R
Emitter, -6.71
Collector, -16.20
Base, -6.85

5R
Base, -3.75
Collector, -12.72
Emitter, -3.35

FL3, 226.4 DC+ 0VAC
FL2 165.44 DC- 0VAC
FL1, RED, Emitter.

All voltages referenced to the chassis.

[Specmaster]

Ok, thanks for this so we can ignore FL3 completely for this section as it plays no part of this equation and is shown unconnected, thus we ignore it altogether.

So essentially you have the same circuit as the one supplied by amspire but you resistor is located in the collector leg, not in the emitter leg and that this resistor is a different value, and also the capacitor in yours is only 1 micro farad as opposed to 1,000 micro farad correct?

So the tube heater is connected to the emitter and the other side of the heater is grounded yes?

What I'm struggling with here is what do the 700R, 400R and 5R relate to? Is the a selection of resistors that you have connected between emitter and ground to simulate the heater load?

[ArthurDent]

It has been many years since I owned any of the HP608/HP606 series signal generators but at the time they were really great pieces of equipment. Where the HP608 was produced for almost two decades, HP made many changes over that long production run. This is why it is very important to have the schematic to the unit you actually have.

Looking at the schematic you will see the filaments for the osc, buffer, and amp tube are in parallel. The two 5675 pencil triodes (6.3V at .135A each) draw .27A combined and the 6BC4 draws .27A for a total of ~.5A. If you have a transistor/zener circuit putting out 9 or so volts into a 7 ohm resistor, the resistor at .5A drops 3.5V and places 5.5V on the filaments of those three tubes. That is why the mod note says to parallel resistors with the 7 ohm to bring the voltage up to about 6.5V. If your version of the mod doesn’t have the 7 ohm resistor to drop 3.5V then you adjusting the circuit to output 9V directly to the filaments will probably be fatal to the tubes, if not instantly, then soon. If your mod doesn’t have a 7 ohm resistor then the zener should be about 6.5V, not 9V. If the gain of the pass transistor has dropped over time that could cause the output to die as well.

When tube filaments are cold and the power is applied, there is a high inrush current that can cause the tube to fail at turn on. The 7 ohm resistor acts like a cheap current regulator and allows the filaments to heat up more slowly and this will increase their life. Initially the filaments are low resistance (they draw more current cold) and most of the 9V will be dropped across the 7 ohm resistor and as the filaments warm up, the drop across the 7 ohm resistor you’ve trimmed drops until there is about 6.3-6.5 volts on the three tubes.

A good simple way to check the tubes out is to use a metered D.C. bench supply and connect 6.5V to the filter (FL) that supplies the filaments and see if the current is about .5A. If it is about .5A then the tubes are probably good. If the current is lower by .135A then one of the 5675 triodes had died. Unplugging the 6BC4 should drop the current to .27A if the two 5675s are good.  I would only remove the pencil triodes as a last resort.

Also the transformers are not oil filled but tar filled. Using the simple diode probe suggested probably won’t work because of loading and detuning. A FET input/bipolar follower probe is a far better choice. I’ve attached a photo of a probe like this I designed years ago that I built into a retractable pen case. The input capacitor voltage rating should be high (1KV) for safety.

[Specmaster]

It has been many years since I owned any of the HP608/HP606 series signal generators but at the time they were really great pieces of equipment. Where the HP608 was produced for almost two decades, HP made many changes over that long production run. This is why it is very important to have the schematic to the unit you actually have.

Looking at the schematic you will see the filaments for the osc, buffer, and amp tube are in parallel. The two 5675 pencil triodes (6.3V at .135A each) draw .27A combined and the 6BC4 draws .27A for a total of ~.5A. If you have a transistor/zener circuit putting out 9 or so volts into a 7 ohm resistor, the resistor at .5A drops 3.5V and places 5.5V on the filaments of those three tubes. That is why the mod note says to parallel resistors with the 7 ohm to bring the voltage up to about 6.5V. If your version of the mod doesn’t have the 7 ohm resistor to drop 3.5V then you adjusting the circuit to output 9V directly to the filaments will probably be fatal to the tubes, if not instantly, then soon. If your mod doesn’t have a 7 ohm resistor then the zener should be about 6.5V, not 9V. If the gain of the pass transistor has dropped over time that could cause the output to die as well.

When tube filaments are cold and the power is applied, there is a high inrush current that can cause the tube to fail at turn on. The 7 ohm resistor acts like a cheap current regulator and allows the filaments to heat up more slowly and this will increase their life. Initially the filaments are low resistance (they draw more current cold) and most of the 9V will be dropped across the 7 ohm resistor and as the filaments warm up, the drop across the 7 ohm resistor you’ve trimmed drops until there is about 6.3-6.5 volts on the three tubes.

A good simple way to check the tubes out is to use a metered D.C. bench supply and connect 6.5V to the filter (FL) that supplies the filaments and see if the current is about .5A. If it is about .5A then the tubes are probably good. If the current is lower by .135A then one of the 5675 triodes had died. Unplugging the 6BC4 should drop the current to .27A if the two 5675s are good.  I would only remove the pencil triodes as a last resort.

Also the transformers are not oil filled but tar filled. Using the simple diode probe suggested probably won’t work because of loading and detuning. A FET input/bipolar follower probe is a far better choice. I’ve attached a photo of a probe like this I designed years ago that I built into a retractable pen case. The input capacitor voltage rating should be high (1KV) for safety.

This seems to make more sense to me at least. Its  always best to get the right schematic if possible but I believe neo has tried and failed to locate one, a free one at least, not sure about paid options.

He does not have the 7 ohm resistor, instead he has a 51 ohm in the collector, I presume that the location is not that critical as it is still in essence in series with the filter but the value of 51 ohms, could this be something to do the gain of his particular pass transistor? also the capacitor is only 1 micro farad not the 1,000 could this be playing a part by not storing sufficient energy to reduce the ripple?   

[ArthurDent]

The function of the 7 ohm resistor is far different than the function of the 51 ohm resistor (which doesn’t seem like a reasonable value). The transistor/zener forms a regulator circuit and could be thought of like a 3-terminal regulator like a 7909 if it uses a 9V zener or as a 7906 if it has a 6V zener. A resistor is often placed in series with the input of a transistor regulator circuit or 3-terminal regulator to reduce dissipation in the regulator by dissipating part of the heat loss in the series resistor. With the resistor before the regulator and the load connected directly to the regulator output, the voltage across the load remains constant.  If the 7 ohm resistor is on the output side of the regulator in series with the load, it is used to control inrush current and the voltage at the load (filaments) will vary, first being quite low, and then rising to the set value once the filaments are up to temp. 

If you had 51 ohms in series with the input to the regulator and you needed an output of 6.5V at .5A for the three tubes, the 51 ohm resistor will not work because the value is much too large. For example, if you wanted a filament voltage of 6.5V at .5A and you just used a resistor to drop the 13V from the bridge down to that value you’d have to drop 6.5V through a resistor and that would require a 13 ohm resistor. I could believe the value being 5.1 ohms, not 51.

The load test neo did with various resistor loads was reasonable except for the last step where he used 5R. The regulator is designed for ~.5A and the 5 ohm resistor was more than double the load the regulator should see. I’d be interested to see what a 10-15 ohm load would look like.

[neo]

Ok, thanks for this so we can ignore FL3 completely for this section as it plays no part of this equation and is shown unconnected, thus we ignore it altogether.

So essentially you have the same circuit as the one supplied by amspire but you resistor is located in the collector leg, not in the emitter leg and that this resistor is a different value, and also the capacitor in yours is only 1 micro farad as opposed to 1,000 micro farad correct?

So the tube heater is connected to the emitter and the other side of the heater is grounded yes?

What I'm struggling with here is what do the 700R, 400R and 5R relate to? Is the a selection of resistors that you have connected between emitter and ground to simulate the heater load?

Those are loads I've selected to test the circuit before reconnecting the heater.

The zener is about 6.5V.

1mF is 1000uF.

[neo]

5R is for a worst case scenario.

Measured the resistor in question, i read the markings at 51 but it is in fact 1 ohm, the marking is 5  1. The 5 is presumably the wattage, i read it wrong and failed to test it, my mistake apologies for the confusion.

As requested,
25R
Emitter, -6.58
Collector, -14.56
Base, -6.84

12.5R
Base, -6.84
Collector, -14.56
Emitter, -6.58

I searched for a manual, all i got was one for a 608D. Couldn't find one for the 608C, paid or otherwise.

[Addicted2AnalogTek]

http://www.ko4bb.com/manuals/73.236.124.51/HP_608C_Operating_and_Service_Manual_Service_Manual.pdf

Is this the manual you need?

[neo]

http://www.ko4bb.com/manuals/73.236.124.51/HP_608C_Operating_and_Service_Manual_Service_Manual.pdf

Is this the manual you need?

 
I probably spent a week looking for that, thanks!

[Addicted2AnalogTek]

If there's nothing else that I'm good at, I'm at least good at finding things on the internet....   

[ArthurDent]

Neo, I see that with both the 25R and 12.5R loads the output of the regulator is constant-a very good sign. It looks like you're going to save it!

[Cyberdragon]

I still think there's something weird going on in this circuit. The base is powered through the -165V rail, whereas in a normal series regulator, it's just powered by the input. It's either a safety shutoff for PSU failure (why would the heaters need this?), or it's using it as a stable reference for the zener, which can be affected by load, and temperature (it's in with tubes after all!). I don't know how the load would change when it's just tube heaters (besides on start-up), but it's definately a sign that it's somehow sensative!

You seem to have it regulating, so if it's at the correct voltage, then leave it alone. (No transistor upgrades or other shenanigans!)

[amspire]

I still think there's something weird going on in this circuit. The base is powered through the -165V rail, whereas in a normal series regulator, it's just powered by the input.

I think there are two reasons for using the -165V for the base/zener bias:

The whole reason of this mod was to eliminate mains ripple noise from the oscillator valve. The -165 is regulated so it will not inject noise into the base. A resistor from the collector will inject main some ripple.

Secondly, they might have done that so that the transistor acts like a constant current source on overload (like on startup). The base current is more or less a constant -165/18K= 9ma due to the high source voltage. Even if the transistor is at the top end of the gain range at 100, the short circuit current is only 900mA or about 14W. Given that it is a 90W transistor, shorts on the emitter will not blow anything up. No need for fuses or current limits components.

[neo]

At 7 volts the filament draws 286mA

[neo]

With the filament connected i get a response from the output volts meter, nothing detectable on the output however. 

[ArthurDent]

When I previously discussed the current drawn by the filaments I assumed there was a 6BC4 buffer between the oscillator and the output amp but after checking the schematic for the C model, this is not the case. There are just the two pencil triodes so the current drawn at 6.5V would be .27A not ~.5A as I first thought. The 6BC4 buffer was added in the D model to eliminate the load pulling the oscillator off frequency. Ampsire is correct that the relatively low current supplied to the 6.5V zener by the -165V line would allow the filament current to be limited if there was a short, thus protecting the solid state devices that were just starting to be used in this mod that came out after the HP 608C was made. 

Now as to the missing output, a fairly common problem with a number of signal generators is that some people use them to check transceivers and accidentally hit the transmit burning out the signal generator's output attenuator resistors. I would start looking in that area for damaged/open resistors.

[neo]

When I previously discussed the current drawn by the filaments I assumed there was a 6BC4 buffer between the oscillator and the output amp but after checking the schematic for the C model, this is not the case. There are just the two pencil triodes so the current drawn at 6.5V would be .27A not ~.5A as I first thought. The 6BC4 buffer was added in the D model to eliminate the load pulling the oscillator off frequency. Ampsire is correct that the relatively low current supplied to the 6.5V zener by the -165V line would allow the filament current to be limited if there was a short, thus protecting the solid state devices that were just starting to be used in this mod that came out after the HP 608C was made. 

Now as to the missing output, a fairly common problem with a number of signal generators is that some people use them to check transceivers and accidentally hit the transmit burning out the signal generator's output attenuator resistors. I would start looking in that area for damaged/open resistors.

The problem being is that there WAS a measurable output when i got it.

Then it just stopped working.

[neo]

Just so that we are all on the same page, at least temporarily,
Obtained free at hamfest
It worked, so was left alone on shelf and occasionally turned on
One day when wanted for an experiment it wouldn't output anything
Took cover off
Saw paper caps
Put it in repair que and ordered new caps to replace paper caps
Sat in repair que until properly bored
Taken off repair que, replaced caps
No change
Lots of head-scratching
Replaced original transistor because it had gone faulty, used 2n665
AHA! A change! the output meter now responds...... but wait there is no output
Back to head-scratching

Also just a note i think the filaments might have taken double the current briefly.

[PA0PBZ]

AHA! A change! the output meter now responds...... but wait there is no output

Since the output meter is an RF indicator there is RF present which makes me question how you concluded that there is no output...?

[neo]

AHA! A change! the output meter now responds...... but wait there is no output

Since the output meter is an RF indicator there is RF present which makes me question how you concluded that there is no output...?

Using a 10X probe i put my scope on the output connector and got nothing.

[Specmaster]

AHA! A change! the output meter now responds...... but wait there is no output

Since the output meter is an RF indicator there is RF present which makes me question how you concluded that there is no output...?

Using a 10X probe i put my scope on the output connector and got nothing.

It sounds like like the output stage is where the problem is then to my mind. Can you detect any sign of a signal prior to the output stage at all?