Hewlett Packard 606A Vintage RF Signal Generator

Okay, This was just a pure coincidence that I was looking for a Component storage shelf and a gentleman was selling it plus this vintage beauty in “as is” condition. This is from 1950s era. He mentioned that he was using this fella as the heater for his shop. I was bit happy and worried about it. Happy coz it is powering on, so not a terribly damaged. Worried coz what if leaky caps have killed the tubes ?


Anyways, got it home. The picture above is taken when this was in the back of my truck. Some more closeup pics. It was in a decent shape, bit of dust around knobs, otherwise for its age, it was really well maintained. Thanks to the person who gave it to me, he was really nice and he kept this toy well as well.

So as first step, time to open it up, check what is inside, condition of parts. Here its all in pics. The unit is heavy.

Rear side. Note the lifting handle.IMG_0062

First part – the EMI/RF  filter. To ensure nothing goes in or comes out to the power line from the Signal Generator.

To remove cover, take off back cover, seat the equipment on its back (rear) and remove two screws under the front panel lift the cover off.


Here you go, full aluminum chassis, RF Shield. Crystal Oscillator

There are different revs of this signal generator. Older revs like this one uses tubes for the crystal oscillator, you can see the crystals in the above pic. later models had transistor. Check service manual for details.

More pics

Someone was in here before, fixed a Cap under the chassis – the Orange thingy 😉

Standard design of the era,diode rectifier,  Tube regulator with VR, Tube oscillator, RF Amp, Modulator

First step as always, cleaning. dismantle the unit completely and clean it. Took a week for me to clean all parts.

And next step is to remove the RF Oscillator module from the main chassis.

I could remove just the cover for the RF Module, but few screws are hard to reach. They supply a long hex wrench with the unit to do this, but of course it was missing

Pic below


So I took the other approach to remove the entire module from the chassis. The screws are in front, 4X screws holding it to the front panel plate. You need to de-solderer wiring going in to the RF Cage.


You can see the main dial indicator and gear box associated with it. The grease on the gear assembly was dry, and it was all stiff to move too. The Frequency Band indicator was not pointing and moving properly either. So first fix was to clean, lubricate and fix it. I removed the entire RF unit and took it to bench for work. Kept rest of the unit on the floor. I have removed the RF Shield of the oscillator unit as well.

Basic cleaning, and inspection, lubricating mechanical assembly. and adjusting band indicator.

Time to look at caps. Lots of discs, few electrolytic. The largest count is a mylar cap 0.1uf 600V.  I was assuming I have to replace all of them. Tested few off the circuit and all of them were good to my surprise.



Also the maximum voltage on this box is +300 or -200 to ground, and they have used 600V caps all over !! :). Its the original HP, engineered for reliability. As always, cut open the cap to confirm its not some weirdo paper.


every single cap I tested worked well. I removed few and tested all the way to 600V using heathkit IT-11 tester too, and all were working well. This is manufactured in 1956-59 and today is 2018 !!!.  Left all caps in place. No need to replace them.

Time for some hardcore porn.

Here is the Turret with RCL to switch frequency band

and the Air gang cap to vary frequency in a band.


Time to get moving. Of course replacing all electrolytic, there are two inside the RF Unit.


Time to get the tubes tested. Standard regulated power supply with 5651. Some pictures from the action.

Specifically wanted to ensure all regulator tubes are good, so is the 5651 VR Tube. Quick Tube Bio – 12B4A Triode as series regulator with 5651 reference. 6AW8A Triode/Pentode tube as error Amp/Comparator.

RF Amp Section is 6CL6 Pentode, 12AT7 Dual triode and 6AW8A Triode/Pentode doing most oscillator sections.  All tubes were good, except one 6AW8A, was slightly leaky. Replaced it even though it was okay to use it. (Leak was ~ 10M). Found aw few Telefuken 12AT7s in the unit.




Now to power supply chassis. Re-cap, test tubes. Watch out for -ve rail on the Power supply to ensure you keep caps at right polarity. On +ve rail you have a voltage doubler.

There are few more caps in the design. One around modulator.  The electrolytic was  faulty, n shorted, replaced it. The Mica one tested good, so left it there.



Here is it – 20uf 25v DC Cap originally installed in the unit.



Pretty tough job to recap the power supply as there is no space. Took almost 6 hours to properly remove, rewire and place caps just on the power supply chassis.

Now other than typical power supply filtering caps, there is a filter for the Filament supply, a 500uF 75V.

This is kinda hidden deep inside the chassis.  Here is how to spot it, from schematic to mechanical drawing to the chassis.


On resistors, I tested few random ones and all tested good, so thought will give it a go, as it is a good quality instrument.

Safety warning :

These instruments contain deadly voltages, powered from a transformer. They can supply sufficient current to kill you. do not attempt to repair or open any of these if you are not well trained/qualified to. There are better ways to die.

Do not poke around with cheap multi meters or probes, as typical voltages can reach 300v+ DC


Time to reassemble the Power supply chassis to RF Module. The circuit is simple with two rails, so I went straight to power up, but monitoring both rails, +300 and -200. You can see the vintage HP3466 doing the job for me and vintage fluke 8020a monitoring input ac. This is not a Tektronix monster with 6-8 power rails, so easy job.


Slowly power up to 30V, watch heaters slowly go red, and increase to a point liker 40v or so, wait for few minutes, observe all is good, then proceed to ramp up input voltage until the 5651 fires up and start regulating. ~80V input AC, the output was reaching +300 and -200V, further increase in input was not affecting output, meaning  regulation was working and all good. Entire Power Supply is good 🙂 You can see the probes going inside power supply to monitor both rails.


Time for the big step, check RF Out, and here you go, just amazing, it is back to life in full form. No errors, totally functional, measured output frequency and that is right to the point on the dial !!!. I am shocked by how it came back to life.

Here is what the frequency counter is telling me at max freqency, 66MHz.


Its all over. Pack it back, test all functions, and everything, including modulation, all bands of frequency are working perfectly !!.

Here is another example for 9Mhz.


The frequency counter is away from the unit, hence two pics.

So, dress him up and let me enjoy the warmness of the tubes and the RF it generates


and view of the 12B4As through the chassis 🙂


and the question now is – Where do I keep it !


There is one RM35 on the floor to fix, one more Tek 475A which I recently restored sitting  in walk in closet, HP3456A and HP34401 on the kitchen counter (new acquisition), and one Tek 547 with Cart sitting in patio being restored,

May  be extend the lab to bed ? 😉

I am sorry, not showing off, I think I am not normal, something wrong with me, needs medical attention!

Update : 2019 – March

I was working on restoring a SX-16 Hallicrafters under guidance from and John Staples and Dennis Monticelli and they both insisted that I pull this out and use as my main signal generator. Of course they are the gods of  this science and I decided to pull this out to my lab to align the SX-16. The key point here is the lowest amplitude it can provide and clean output. However as per them it does fall short of the 606B, which includes additional buffer and full DC power for all filaments. So I tried the spectrum quality of this on my spectrum analyzer. I don’t have another generator which can  do -110dBm/1μV which is good to have on RF side of the game.

So here are some results –

500KHz, 1Khz AM @ 50%, -57dBm. Kind of clean signal


10MHz, 1Khz AM@50%, -57dBm. Weird residual FM, asymmetry. IMG_2296

50MHz, 1Khz AM@50%, -57dBm. Strong 2nd Harmonic of the AM 1KHz signalIMG_2298

I deiced to try higher output levels, now @ -25dBm.

500KHz, 1Khz AM @ 50%, -57dBm. Harmonics  of AM signal visible ~ 50dB down.IMG_2301

10MHz, 1Khz AM@50%, -25dBm. Weird residual FM, asymmetry.


50MHz, 1Khz AM@50%, -25dBm. Harmonics of AM signal visible. ~ 40dB down.


Summary – Expect residual FM at higher frequencies.

Back to Home Page for More Crazy Stuff















2 thoughts on “Hewlett Packard 606A Vintage RF Signal Generator

  1. Great write-up, thanks for sharing your experience in such detail! I found you via your contributions on the Tek Wiki, and it seems we have a few of the same pieces. I’ve had my 606 since around 1990, but haven’t touched it in years. It will (someday) make it into the restore queue, and wondering if you can recommend the grease you used for the mechanism.


    1. Hey Paul, Thanks and interesting to know that have similar toys 🙂 , Even I am a bit surprised that our Blogs looks the a lot similar, including pictures and the way we mark components in schematics :). Reg Grease, I used silicon grease, however I would assume any general grease should work too as there are no rubber or plastic parts in it.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s