Pass Citation 12

Audio has been my hobby since early 70´s, but I have been tinkering last 5 years mostly with tube gear. Last weekend I bought HK Citation 12, as it was amplifier I could not afford in mid 70´s. I also remember Nelson Pass Citation 12 mods from 1981 AA.

However I have some questions about this topic. Maybe Nelson or other fellow members have useful information available?

It seems that here are two Pass Citation 12 versions available. I usually think that simpler is better, but now I have wondered if the original or later complementary version is way to go? Do complementary FETs have so much lower capacitance that driver stage will become unnecessary? Also modern day replacements would be appriciated for MPSL51, MPSL01, IFR100, IFR130, IFR9130.

Nelson specified usual passive components to be used. However when I did Quad 22/II combo rebuild few years ago, I replaced all carbon comp and two precision carbon film resistors from 22 line stage with better spec metal film resistors as suggested. This produced inferior sound quality. When I later reassembled 22 with original parts, pair by pair, I discovered that original carbon comp and carbon films produced better sound at anode and grid circuits. However MF:s were OK in low impedance cathode and feedback circuits.

I do not believe in snake oil, but it seems that also passive parts do have effect on sound quality. It would be nice to know which resistor types Nelson would use today in different Citation 12 locations. This would be interesting, as there is so much high spec close tolerance hype here. It was so refressing to read Nelsons remark about 10% wirewounds to be perfect for L300 crossovers.

Best Regards

Kimmo Salminen
 
Well hello everybody, I've been waiting for almost 2 years and today is the day I will do this Citation 12 MOSFET. I've bought all the stuff at leist a year ago and it was sitting in my shelf waiting for me.

Just to be a bit rude it is a "deluxe" edition with the wood case and faceplate...in very nice shape.

My plans were to use the first plan I've red from the "audio amateur" review. Reading some thread here, I've found R8 being unnecessary, should be direct to gnd. Are there other things from the original plan I should change? I have IRF130 as MOSFET, BD139/140 as pre-fet and 2N5551 and 2SA916 for differential. Should I use BD139/140 for diff too?

Thanks gentlemens!
 
second oups: K135 & J50...
 

Attachments

  • IMG_1673small.jpg
    IMG_1673small.jpg
    666.8 KB · Views: 1,653
...but I have been tinkering last 5 years mostly with tube gear.

...I usually think that simpler is better, but now I have wondered if the original or later complementary version is way to go? Do complementary FETs have so much lower capacitance that driver stage will become unnecessary?

...I discovered that original carbon comp and carbon films produced better sound at anode and grid circuits. However MF:s were OK in low impedance cathode and feedback circuits.

...but it seems that also passive parts do have effect on sound quality

After tinkering with tubes you should jump to Pass' class-A instead of Citation. The Citation is the worst amp associated with NP that I know of. Well, high capacitance mosfet in class-B, that's even against NP philosophy I guess. And then there is that splitter because it is a quasi one (For quasi one I prefer that amp from Alex Nikitin, based on JLH, but with output cap). And then, say who simpler is better in class-B mosfet.

I haven't seen (I guess) a later complementary Citation. If there is so, it shouldn't be called Citation 12. I know only the original BJT Citation that is modified to use Mosfet with one driver stage removed.

If you look at oscilloscope, or simulation or whatever numbers, you may find that omitting the driver is fine. But soundwise? It's subjective but I think it is required in class-B mosfet, even with low capacitance lateral mosfets.

You have been tinkering with tube amps. Not surprising that you can "hear" the carbon resistor sound. And I agree with your finding regarding resistor positions. For carbon composition (Allen Bradley), I like to use it in input (series) but not in FB.
 
I would like to thank everybody for your kind response.

Nelson´s comment about nulling distortion in Quad 22 line stage is one possible answer. However, I measured all original parts and they actually were quite close to target values. Actually much better than original 20% tolerance would allow for carbon comps, carbon films were even closer to target. So, horror stories about 50 year old carbon comp being most often off spec, are not 100% true. Most likely there has always been good and bad quality parts.

Jays question about why bother to build Mosfet Citation is sound, if I only sound quality is issue. But I have collected and rebuilt some extremly rare audio gear like Conrad Johnson Premier 7b, Quad 240 and KEF P60 with Pro Kube. If only 220, 180 or 300 pairs in case of P60 in total, but propably less than 50 pairs with Prokube EQ, were made they should be preserved.

On the other hand I have collected, rebuilt and in some cases upgraded gear like Quad 22/II/FM1/MPX, EAR509, ARC Classic 60, Conrad Johnson Premier 11a, Garrard 301, Kef Cantata, JBL L36, Tannoy LRM which were made in ample quantities, but I have somehow felt that they had been pretty important products when they were introduced. I think that Pass Citation 12 fits in this category pretty nicely. Especially as the original HK Citation 12 was the amplifier that I could no afford, instead I had to buy HK930 receiver.

Anyway suggestions for IFR130 and IFR9130 replacements would be appriciated.

Best Regards

Kimmo Salminen
 
6BQ5 amp with Tannoy

When you use different amplifiers with negative feedback, sound quality of different speakers seems to be quite unpredictable.

Crossover network and loudspeaker seems to form some kind of energy storage that feeds feedback loop of the power amp. As time constants of this energy storage are different at different frequencies, amplifier or loudspeaker reviews can be quite misleading if you do not use exactly the same amp and speaker combination.

Best Regards

Kimmo Salminen
 
Well I've completed my MF Citation 12 but have a problem with biasing.... Can't get lower then 350ma.

I'm using the K135 and J50 old Hitachi TO3s and BC139 BC140 everywhere else.

Cant find transconductance neither bias info on the datasheet for the Hitachis.

Playing with the BIAS pot don't seems to have any effects. Even removing the 4.7K in parallel don't have an effect.

Can anyone help me?

I've verified the schematics on my original "modified" boards and everything is right, as the "Revised Mosfet 12 schematic, complementary version" shows.

Can K135, J50 be wrong for this application?

I've done all verifications, got the amp to clipping around 20 RMS volts on both channels. Both channels have the exact same problem; bias around 350ma, even higher at start up going a bit lower after.

Thanks anyone!
 
Well I've completed my MF Citation 12 but have a problem with biasing.... Can't get lower then 350ma.

I'm using the K135 and J50 old Hitachi TO3s and BC139 BC140 everywhere else.

Cant find transconductance neither bias info on the datasheet for the Hitachis.

Playing with the BIAS pot don't seems to have any effects. Even removing the 4.7K in parallel don't have an effect.

Can anyone help me?

I've verified the schematics on my original "modified" boards and everything is right, as the "Revised Mosfet 12 schematic, complementary version" shows.

Can K135, J50 be wrong for this application?

I've done all verifications, got the amp to clipping around 20 RMS volts on both channels. Both channels have the exact same problem; bias around 350ma, even higher at start up going a bit lower after.

Thanks anyone!

If you want to use lateral mosfet for citation 12, you must modify the VBE multiplier resistor value since they are intended for Vfets which has around 4v for Vgs. You can also omit the VBE multiplier and replace with just a low value resistor since laterals don't need thermal compensation.
 
Gate to source voltage is at +/- 14 volts.

Hmm... what's wrong with 350ma of bias voltage... gets hot but I've seen worse. I was just trying to follow the original plan. I should get both clipping starting at the same voltage hey? Means I should replace 5K pot by a higher value as this one seems to have no effects on the current...

I've been trying to work on that Citation12 project for a few years now. I even bought 4 IRF130 at the time ($15ea!). I've never noticed it was an article you wrote at the time Mr Pass. Really interesting project to convert quite regular architecture for a more modern plan...

What do you all think should sound better: IRF130, Hitachi J50 and K135 or IRF240 and 9240?

Thanks all!
 
The one and only
Joined 2001
Paid Member
The Hitachis ought to be conducting at around 1V Vgs, so the complementary
pair of followers would have a bias circuit around 2V. At the voltages you
mention, they shouldn't work at all. What are we missing here?

How about the pinouts? Aren't these laterals with Drain and Source pins reversed?

:cool:
 
Dooohhh!

I knew about Drain/Source reverse pin out and swapped these 2. And verified, verified, verified,...

I've had this Citation 12 Mosfet project for a few years now and never had a chance to get into it until a few weeks ago. I then noticed it was one of your (N. Pass) projects and thought "that is why it is interesting; taking a quite regular architecture audio amp plan and change it to a way more modern one using more up to date discrete components". So a little note here to thank you giving so many interesting projects using none conservatives thinkings.

OK. I did bought 4 IRF130 ($15ea!) at the beginning with some regular 5% resistor at $0.25 for 5 pieces at the corner electronics store...

When I've started to work on it, noticed there were supposed to be 2% metal resistors 1/4 watts. SO I did order some 1% metal resistors. I also thought I had some of these old Hitachi Mosfet that has a quite good reputation (they were on one side of a H&H 800 watts amp).

Anyhow... I'll try to add a few pictures of my project to this post. If that can help a bit... You'll notice one red jumper that will on one the front side of the board on one side and on the back for the other... Didn't noticed there were wholes...

Well thanks for helping!
 

Attachments

  • backboard.jpg
    backboard.jpg
    383.4 KB · Views: 884
  • closeup.jpg
    closeup.jpg
    186.9 KB · Views: 861
  • frontboard.jpg
    frontboard.jpg
    286.7 KB · Views: 828
  • under.jpg
    under.jpg
    414.5 KB · Views: 817
Got it!

Finally I'm now listening to this CT12 MOSFET using K135/J50. Amazing sound. I'm using Mark Levinson ML1 and Tracer speakers. I did try Bryston 2B amp; very detailed but too dry. Tried Dynaco ST70; not bad. Warmer but less details. Tried my Heathkit EA-1; around the same, nicer mids & highs,nice female voices. But now, this CT12 is really the total all of them; very well detailed in the mids/highs with full bass. Note that I've replaced the 4 6kuf by 8200.

And better soundstage as well!

I'm really happy with the sound!!!

Only odd thing I would like to do is decrease the bias voltage a bit. It gets really hot after 1 hour or 2. Can't leave my fingers on the heathsink...

I don't know how to calculate this VBE multiplier or how to replace by a single resistor....? Help anyone?

I can measure anything... I'm an telecom tech fooling around in vintage audio.

God it sounds good... If I lower the current, will it sound different?

I have +/-45 volts as rails.

If needed, I can measure at what voltage each FET starts conducting...