• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Marantz-9

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Would any can tell what is the funtion of "C38" near the V8(EL34) ?
Thanks a lot !
The schematic is attached !
 

Attachments

  • m9.jpg
    m9.jpg
    150.2 KB · Views: 486
I don't know. There are several other unusual things in that circuit:
1. C8 appears to feed PSU noise into the phase splitter - hum cancellation?
2. C20 etc. network only in one phase - which means the forward path is not minimum phase?
3. Anti-parasitic (?) inductors in only two EL34 anode circuits.

Also no safety resistors in case a bias pot wiper leaves the track.
 
Administrator
Joined 2004
Paid Member
In normal operation (not setting the bias) all of the cathodes are tied together and that capacitor appears in parallel with the 3.3 ohm cathode resistor eliminating cathode feedback in the ultrasonic region. Probably something to do the stability margin or ?

IIRC C7/C8 are part of a ripple cancelling network..

The bias pots used in these amps were locking milspec hermetically sealed units, I've never seen a bad one in 30 some odd years. (Mostly Ohmite) I use them or the Canadian made equivalent in my own amps and have never had one fail, I usually recommend safety resistors in my published designs, but almost never use them in my own gear.. Perhaps I'm playing Russian Roulette.. :D Seriously their inclusion is a very good idea.

I've refurbished a few of these amplifiers in the past, and will say they were amongst the best sounding PP UL amps I have ever heard.
 
Part of the confusion is that there's stuff not obvious on the schematic. Chief among these is layout and wiring (these were point to point and some wires were unavoidably long). This was all made more critical by the high feedback and high transconductance tubes.

Caps and inductors were added to stabilize the circuit as actually laid out. PPP with high gm output tubes can be a bear as a practical matter for production amps with expected variations.

edit: cross posted with Kevin and fully agree- these are great amps, I've used them and repaired them in the past and was in "awe and admiration" mode.
 
Administrator
Joined 2004
Paid Member
In keeping with SY's comments I would say they are quite special, sound quite special, and the build quality was superlative, certainly the best of the best made here in America. I coveted a pair for quite a long time, but good vintage pairs in running condition command more than $10K here in the US - I believe the Japanese are generally willing to spend considerably more for a mint pair..
 
kevinkr said:
In normal operation (not setting the bias) all of the cathodes are tied together and that capacitor appears in parallel with the 3.3 ohm cathode resistor eliminating cathode feedback in the ultrasonic region. Probably something to do the stability margin or ?
OK, I didn't try to trace the switch wiring. Bypassing 3.3 seems a waste of time until it is realised that it is in series with 24/4=6 - still only a small proportion, though.
 
Lots of interesting details in this schematic. Is ST9 an invert switch? If so, does V1A work as cf? There are two more cf s to drive the output tubes, but this one is not inside the feedback loop.
And what is this business whith Ra and Rb? Do they add positive feedback to compensate for wire resistances? Similar to IxR compensation in motor drives.
Slightly off topic, but it seems funny to me that native speakers will call this circuit ppp. If you say " parallel push pull" in Germany people will think you are referring to a circlotron.
 
Would any can tell what is the funtion of "C38" near the V8(EL34) ?
Thanks a lot !
The schematic is attached !

I can't see the schematic very well, but if the capacitor in question is the cathode bypass cap on the last pair it is to make the lower tube have a higher output signal due to the nonlinearity in the previous stages. (the non inverted signal (cathode) is less than the inverted signal(plate))
 
Thanks SY, of course it says lower damping factor in the corner, not higher. R73 is confusing me though, it connects to the speaker side of Ra and it adds something to the NFB. So I thought it had to do with cancelling the effect of Ra.
I had never seen a bridging option on a tube amp before. Two model 9s, one switched to inverting, connect your 16 Ohm speaker to the 4 Ohm posts and off you got with 140 Watts.
Marantz added the CFs to drive the grids of the 2+2 output EL34s. As DF96 says, a CF can add low impedance and distortion to a signal. But the grids a driven through a 68k+1k. Ok, the 68k is shunted with 150p. A bit of treble boost? So why the CFs and the series resistors?
A shame no one showed me this schematic before!!
 
grometeer said:
As DF96 says, a CF can add low impedance and distortion to a signal.
Only when a CF is unnecessary. Where a CF is appropriate it can reduce distortion by acting as a genuine buffer.

The 150pF will give a little phase advance, to aid loop stability. Because of the overall NFB they won't boost treble in the output. The CF is probably needed to drive the output stage Miller capacitance when in triode mode. Probably not necessary for UL mode, but simpler to leave in place for that too.
 
Hello DF96, I hope you don´t mind me quoting you less than precisely.
I was just wondering why they decided to use the relatively high 68k resistor to drive the grids. I cannot just be a grid stopper, the 150p makes that impossible. Then I thought it might be used to create a dominant pole, but with the different (EL34´s) input capacitance of UL and triode mode this is not likely, either.
Does the phase splitter work better in any way if it is loaded with pure resistance, compared to a complex RC load?
 
Yes, without the 68k the CF would be capable of storing quite a lot of charge in the coupling caps if the output hits grid current on peaks. A conventional (grounded cathode) driver can't do this so much because of the anode load resistors which are the only pull-up available. An interesting example of how a design decision (to use CF) has consequences which then have to be foreseen (or found in testing) and dealt with.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.