Citation II oscillating, Uh oh

[kff322]

Hey I am nearly finished building a "modern" (and modified) citation II Amplifier.
BUT...
The rig oscillates at a VERY high frequency. This problem has been keeping this thing from getting off my bench and onto my audio rack.
A little Background
- Its not the tubes.
- Unlikely to be the power-supply
- Oscillations can be heard from both sides which makes it difficult to tell where its coming from in the circuit.
- I checked for bad caps and wiring errors.
- Moving wires around makes it sound like your "Tuning" a radio.

I modified its design a little bit.
Schematic from McShane's Site
-The output is in Triode mode with 330 ohm resistors between pins 3 and 4
-Removed Feed-Back loop consisting of R8, R17, R43, and R52.
- 12BY7A has DC Filaments
- .1 uf 600V caps on all the wipers of all the POTS.
Got these mods from
this site.

And I have a scope, the oscillation is detectable from any point in the amp.

Thanks
Kff322

 

[korneluk]

I suspect the removal of the feedback loop is what's making the amp unstable. Put the feedback back in temporarily and see if the feedback goes away. If it is, then you can play with reducing the amount of feedback little by little until it pleases your ear and it no longer oscillates.

BTW, do not plug in your speakers while it is oscillating - it will blow your tweeters!

-- josé k.

 

[kff322]

Sounds like a plan but.
What do you think the minimum values would be for putting the feedback loop back in should be?

 

[TubeHead Johnny]

0 NFB causing oscillation?

I never heard of taking OUT NFB and causing oscillations. It's the other way around. TOO MUCH NFB will cause problems. It's probably the lead dressing on your grid resistors or something like that. Check your wiring. Take a pencil or wood stick and gently move your wiring around as it's powered up. Do the oscillations stop when you touch a particular wire? Also there's a possibilty your OPT primary is wired out of phase. Switch the two primary wires around and it should go away.

~~~Johnny~~~

 

[Giaime]

Re: 0 NFB causing oscillation?

I never heard of taking OUT NFB and causing oscillations.

Try to remove feedback from a big and bad SS amp: instant blowing


That happens because amps have exagerate open loop gain (to get high NFB factors).

 

[kff322]

Its not the OPT polarity, I tried switching the leads and I made the floor rumbling sound that these reversed OPTs make. Moving wires around will vary the pitch of the very high pitched oscillations but not stop it.
Although I made an observation.
Adjusting the bias pot changes the oscillations the most and sometimes even stops it, but it starts again when music plays.

 

[korneluk]

I had another idea, the 330 ohm resistor between pins 3 and 4 seems kind of low - for triode-strapped pentodes, the value is usually about 1k. Also, make sure you disconnected the ultralinear connection when you triode strapped it.

Keep us posted,

-- josé k.

 

[Eli Duttman]

Time out please!

The H/K Cit. 2 uses a sophisticated set of nested NFB loops. By eliminating an inner loop, the net gain was raised. That increased the otherwise minimal amout of global NFB significantly and oscillation was the result.

Triode strapping the "finals" is fine. Put everything else back the way Stu Hegeman designed. None of us on this site, STARTING WITH ME, can dare to think lifting Stu's jockstrap is possible. In this day and age, we have the advantage of better passive parts and SS stuff to support tube circuitry. The mind boggles at the thought of what Hegeman, Recklinghausen, et. al. would have achieved with access to the modern parts inventory.

Please guys, show some respect for the GIANTS of the past. Their memory is revered for good reason.

 

[kff322]

I only removed the feed-back because I thought it was part of triode strapping the power tubes. You can talk to the man who goes by kevinkr on this site for having that recommendation on his site. (www.Kta-hifi.net under OTS Articles)
Increasing the resistor values to 1.2K on the Powertubes just make the sound worse. Tomorrow I am gonna see if I can put those Feed-Back loops back in and see what happens.

 

[Eli Duttman]

I only removed the feed-back because I thought it was part of triode strapping the power tubes. You can talk to the man who goes by kevinkr on this site for having that recommendation on his site. (www.Kta-hifi.net under OTS Articles)
Increasing the resistor values to 1.2K on the Powertubes just make the sound worse. Tomorrow I am gonna see if I can put those Feed-Back loops back in and see what happens.

Dude,

When it comes to the "care and feeding" of a Cit. 2, Jim McShane is THE man.

 

[HollowState]

I tend to agree with Eli when it comes to feedback in the Citation II amplifier. (I own several) This amplifier has a total of 30+ dB of feedback in no less then three loops, and was a critical part of the design. Removing any feedback will alter the gain/bandwidth product of the circuit.

Also, this amp was designed and built with carbon comp resistors and paper and ceramic capacitors. When one substitutes fast capacitors like polycarbonate, polypropylene or even polyester for paper, and replaces the resistors with film types, the high frequency response of the circuit is raised. This plus the feedback change, and possibly lead dress, is the cause of your problem.

Victor

 

[Giaime]

Although I made an observation.
Adjusting the bias pot changes the oscillations the most and sometimes even stops it, but it starts again when music plays.

That's a clue. I would replace those bias pots, probably they have loose contact, which lift the grids of the power tubes from ground and start oscillation.

 

[kevinkr]

I've built quite a few amplifiers based on Citation II iron and repaired/modified a couple of dozen without difficulty over the last 20 some odd years. I have nothing but admiration for Stewart Hegeman, and early on recognized the merit of this design. I might add this was at time when a Citation II could be purchased in working order for just a couple of hundred $$, and was NOT well regarded by the local audiophile community.

I'd start by opening all loops and making sure that you have no local oscillating stages. 12BY7 are quite notorious for oscillating if insufficient care is applied in the design of the layout. Note that rather large grid stoppers can be a good idea with this tube as well.

Can you share a few pix of your layout, perhaps there will be a couple of clues there.

Note as well if you are using an output transformer that is significantly different than any of the Citation II variants that all feedback loops will need to be recalculated.

My mods previously disparaged elsewhere in this thread were heavily tested against the stock Citation II and are guaranteed only to be stable with an original amplifier.

Kevin

 

[kff322]

Here are some pictures of Its layout and the amp its self. When I was ordering its parts the only two things that were hard to match was the OPTs and the main PowerTransformer

The OPTs I decided on getting was the hammond 1650N specs as follows
60Watt Output
Primary Impedance - 4300 ct

Thanks For looking Guys
Download Pictures via YouSendIt
Thanks for all your help kevinkr

 

[Eli Duttman]

My mods previously disparaged elsewhere in this thread were heavily tested against the stock Citation II and are guaranteed only to be stable with an original amplifier.

Kevin,

I was not disparaging, merely questioning. You have a SOLID reputation and, undoubtedly, have forgotten things I'll never learn. Except to correct a flaw, I'm cautious about "monkeying" with a proven superior design.

You are spot on about the O/P "iron". OEM Cit. 2 O/P trafos have a phenomenal bandwidth. Hegeman relied on that bandwidth to achieve his end result.

 

[Eli Duttman]

Here are some pictures of Its layout and the amp its self. When I was ordering its parts the only two things that were hard to match was the OPTs and the main PowerTransformer

The OPTs I decided on getting was the hammond 1650N specs as follows
60Watt Output
Primary Impedance - 4300 ct

Thanks For looking Guys
Download Pictures via YouSendIt
Thanks for all your help kevinkr

Dude,

You may be in DEEP do-do. Hammond O/P "iron" is woefully inadequate, from a bandwidth perspective, for use in the Cit. 2 circuit.

 

[kff322]

Thanks for your optimism

 

[Eli Duttman]

Thanks for your optimism

Sorry, but I had to tell the truth. Keep on "trucking".

Make sure the global NFB loop is handling only a few dB. Large error correction signals in the global loop will cause you GRIEF. Roll infrasonic noise off at the amp's I/P. Insert a cap. in the lines between the I/P RCA jacks and the 1 MegOhm grid leak resistors. Select a cap. that places the resulting high pass pole's "corner" freq. just below 20 Hz. If you can tolerate the limitation, 25 Hz. might be better with the Hammond "iron".

 

[TubeHead Johnny]

One Observation

One observation I made in your photos was that your hook up wires leading to the perf boards are WAY too long!

~~~Johnny~~~

 

[kff322]

I emailed Jim McShane over this issue. Lets see what he says

kff322