Please help me fix this HP8601A

[kostas_sv3ora]

This HP8601A (manual here https://bama.edebris.com/manuals/hp/8601a/ ) stays fixed at the maximum frequency on each range, no matter If I move the frequency knob.
The VCO tuning works, when tested with external voltage from the bench power supply.
I suspect the DC amplifier after the discriminator.
I attach all the tests I did in the image (please zoom in).
Greens check ok. Reds do not. Check the voltages. The output ramp voltage to the VCO stays fixed.

Any help towards the right direction, so as not to have to desolder every component to check?

[Radio Wrangler]

The variable frequency oscillator is NOT tuned directly. It is controlled by a Frequency Locked Loop.

There is a frequency discriminator monitoring the output frequency and this is compared to a voltage representing the sum of the tuned frequency pot, the fine tune and any frequency sweep turned on. The result of this comparison is amplified by a high gain amplifier with R-C time constants to keep things stable and this tunes the voltage controlled oscillator.

So, you're right to start looking at that amplifier.

Now look at the input voltages to the amplifier. Use the external supply to tune the VCO around and see if the voltage from the discriminator is sensible. Is the voltage from the sweeping and tuning controls sensible?

If both these seem OK, then the amplifier is suspect. If one voltage or the other is unresponsive or far off, then you're into the tuning pots/sweep etc or into the discriminator and its RF feed.

Any fault in the frequency locked loop will remove the feedback. The feedback is essential to controlling the frequency, and the frequency will go and sit at one extreme of its range or the other.

There is an article in the Hewlett Packard Journal on this instrument which explains it. There is similar in the service manual, and there is a fault-finding diagnostic flowchart in the manual. These are the best resources there are for troubleshooting and repairing a failed 8601A

I can't read those diagrams. Cataracts I'm afraid.

DAvid

[kostas_sv3ora]

Hi David and thanks a lot for your reply!
I am sorry you can't read my small diagrams, some of the things you mention are answered in that picture.
I mention a few just to give a hint:

If I use an external voltage to tune the VCO, it changes frequency. Of course now the feedback loop is broken when I do so.

The discriminator seems to work because I feed an external input RF signal from another generator, at the levels and frequency range expected by the discriminator, and the discriminator output voltage is within the expected range.

However, the amplifier that follows the discriminator, is unable to create the ramp voltage needed for tuning the VCO. Instead this voltage is fix at a single value, hence the VCO cannot change frequency.

The input DC to this amplifier (which is the DC from the frequency tune knob) varies by tuning the frequency knob, as it should.

So both the (simulated) input RF to the discriminator, and the output DC of the discriminator work. Also the input DC to the DC amplifier works.

From the above, the only logical explanation for a non-working output ramp, is a fault in the DC amplifier. This is what I am trying to figure out, but one has to see my measurements in this diagram I posted, so as to have a hint on the problem

None of the voltages in the DC amplifier transistors are close to the values shown in the schematic, these are way off, either too low or too high in other cases.

[Radio Wrangler]

I think you are on the right track.

One of the first things I designed at HP was a frequency locked loop sweeper for an early version of the baseband analyser which never saw production. Getting FLLs to be stable and to accurately track sawtooth sweeps isn't as easy as the usual recipe approach to PLLs

Expect feedback in the amplifier itself to control the forwards gain to get the gain/phase characteristic needed for the FLL outer loop to be closed.

David

[kostas_sv3ora]

Originally Posted by Radio Wrangler

I think you are on the right track.

One of the first things I designed at HP was a frequency locked loop sweeper for an early version of the baseband analyser which never saw production. Getting FLLs to be stable and to accurately track sawtooth sweeps isn't as easy as the usual recipe approach to PLLs

Expect feedback in the amplifier itself to control the forwards gain to get the gain/phase characteristic needed for the FLL outer loop to be closed.

David

Here is a zip file with bigger pictures. Here the voltages I measured are shown.

Shall I replace that Q5 array?
This is equivalent to the 2n4044 but I only have a 2n3807, which is pin compatible and pretty much similar HFE and DC characteristics.

But is this the right move to suspect this Q5, based on the voltage readings?

[Radio Wrangler]

It would be my first suspect. I think it's an 1854-0221 but not available now. a pair of 2N3904 chosen for similar Vbe glued together may be your best option.

1854-0071 was the general purp NPN audio transistor, a TIXS number, later we shifted to 1854-0215 a real 2N3904.

David

[kostas_sv3ora]

Because people were willing to help here, I share my solution to the problem, after quite a lot of hours of debugging and circuits trying.
http://qrp.gr/8601aloop/
Enjoy

[Radio Wrangler]

Good result!

The post-mixer RF buffer amp.

The FET stage below it looks odd. I thing the diagram markes the JFET as 1855-0062 which is HP stock number for the 2N3819 N-channel FET. If so, then the drain and source markings on the diagram are interchanged. However, some JFETs were symmetrical and could be used the reverse way. Is this an error in the documentation, or was someone up to a trick?

David