Fender AB165 Bassman

[Stevem]

Would some of you out there take a guess as to why with this amp only Fender chose to make it only a 25 some odd watt RMS amp by installing those 220K resistors from the output tube plates back to the PI?

I mean the combo Bassman in tweed dress was a higher output power amp as was the blackface AA165 amp for God's sake!

I kind of want to chalk it up to a blunder that got out of the factory, or maybe more to the point Fender got feed up with blown up 12 inch drivers that they started to see a ton of when dropping down from 4-10s sapping up the power to the 2-12" set up!
As I recall Jensen never made a driver that could really handle Bass until they came out with the C15L driver if you ask me!

It is a easy way to knock a amp down to about half power though!?

[pdf64]

It just looks like local feedback around the power tubes; it will reduce the gain but should not restrict the max signal swing / power output.

[Stevem]

Well it does, with those in the circuit the amp does a clean 17.5 watts into a 4 ohm load with the used output tubes I have in it, remove those connections and the amp puts out 36 clean watts and 79 watts peak, so it's not a tube issue even with the used 6l6 I have in the amp!

Thanks for the replay though!

[martin manning]

It just looks like local feedback around the power tubes; it will reduce the gain but should not restrict the max signal swing / power output.

Yes local feedback, probably to increase bandwidth, but at the expense of power output. Probably so much reduction in power that the amp doesn't cut it, so to speak, and the idea was dropped.

Negative feedback will always reduce gain, and therefore output voltage swing for a given input voltage swing, so output power goes down. In this case the feedback divider is 220k and 100k//PI Zo, maybe 0.09. Open loop gain might be ~8, so the closed loop gain would be 8/(1+0.09*8)=4.65, a reduction of 4.7dB.

Steve, you're seeing about -3dB, so thats at least close.

(Edited to correct my P's and V's for dB's)

[ToneMerc]

It just looks like local feedback around the power tubes; it will reduce the gain but should not restrict the max signal swing / power output.

Also, the feedback circuit is wired to the opposite side of the PI and there's an additional FB loop at the 3rd gain stage. In addition, look the non-traditional bias circuit.

IMHO, this was all an attempt by CBS engineers to improve distortion properties and at the same time save some pennies by allowing a wider mismatching range of power tubes to be used.

In the end we all know that "hi-fi" theory and guitar amp tone generally doesn't play well together.

An interesting experiment would be to compare the power output my early 65' AA165 to it's 66' AB165 brother.

TM

[martin manning]

Yes, it has both local and global NFB. Perhaps Steve is thinking that the LNFB might be a way to do a half-power mod? Sure, it could be done that way.

[pdf64]

I don't see why the local NFB on the power tubes should reduce / restrict power output, provided that there's sufficient gain from earlier stages to make up the loss, and voltage swing available from the phase splitter?
Schematic http://bmamps.com/Schematics/fender/bas ... _schem.pdf

[martin manning]

I see your point, Pete. Steve is measuring max clean power though, and the thing we don't know is where the distortion first appears. If it's in the preamp, then the max undistorted signal into the PI is fixed. Also, since he's measuring at a particular frequency the increase in bandwidth may get into it, as well as the percent of the global feedback being different with the reduction in power tube gain, so the place where the distortion first appears could change with the removal of the LNFB.

[Stevem]

I will do a closer test out of the amp tonight and report back, I also had no good 12at7s left to use in the PI so I had to use a ax7.
The power transformer is also one for a JCM 45 so the amp only has some 409 volts on the OT center tap, and yea that use of the bias balance in a blackface amp was unexpected when I first saw that ages ago!

CBS had a strange thought process going on there, cutting a amps clean RMS power in half to try a get less distortion, thank God they gave up that quest after the whole UL OT thing and went back to making guitar amps and not competing with Dynaco or such in making stereo gear!

Thanks for the replys gang!

[pdf64]

The higher output impedance of the 12AX7 in the LTP slot may have resulted in the inverting amp gain being reduced more than intended.
Can you try and assess where the clipping is occuring first?

[Stevem]

Yes, I will O scope that out with both tubes just for the hell of it!

[martin manning]

I believe the gain will be about the same, but the 3x higher ra of the 12AX7 will increase the FB ratio.

[Stevem]

Well here is what a quick check out did!
The output stage clips first with those local feedback 220k resistors in or out, but what I did notice was that without those resistors in that the drive signal coming into the PI was misshapen, almost like it had a form of crossover distortion taking place, with those resistors then back in circuit that misshaped wave form was gone, and the wave form was pristine !

Now this test was done with a Ax7 in the PI slot and without those resistors in the output stage had very little crossover taking place, it did not show up until the output voltage into a 4 ohm load was 14.7.
If I then put a AT7 in the PI with those resistors still out of the circuit the crossover distortion started at a output voltage of 13.8 and the shoulders of that crossover distortion where far greater of a right angle and a guess off that top of my head would be that the crossover distortion was also 20%greater!
Does any of this make sence?

[pdf64]

Without the 220k plate feedback resistors, the greater gain of a 12AX7 in the LTP slot may increase the degree of correction that the global NFB can effect, thereby reducing the apparant crossover distortion.
To check this idea, maybe try the 2 tube types in the amp with the global NFB loop open.

[Stevem]

I also noticed that this amp as in 5 or 6 others I have worked on (AA165) does not have the resistor from pin 6 to pin 7 on that 3rd gain stage on the bass channel, has anyone ever seen this resistor in a amp or is it a change that never made it out the door on these amps?

I also worked on a 67 Bandmaster yesterday and the use in those heads and many others of the .047 PI coupling caps makes for a huge increase in full power crossover distortion as compared to the Bassmans .022 uf caps, it's been a long time since I did a near back to back test on that!