The purpose of 6.6k Primary on Wreck OT's

[funkmeblue]

I had the understanding that when one increases the primary impedence there is a decrease in volume but a wider bandwidth.

[JamesHealey]

I had the understanding that when one increases the primary impedence there is an decrease in volume but a wider bandwidth.

Now that is interesting!

[Zippy]

I had the understanding that when one increases the primary impedence there is a decrease in volume but a wider bandwidth.

Why and to what limit?

[funkmeblue]

actually I am wrong, you get a lower thd and less bandwidth unless the transformer is wound to make up for the differences. A stancor 3801 has more bandwidth than a guitar amp really needs. So increasing the primary impedence would lower the volume and distortion from the transformer while making a hifi transformer more useful for a guitar amp.

[Richie]

think or my opinion is, he used that because the Express could run EL34s or 6v6s..and the low voltage for EL34s If you look at the Komet site,one of the things he said was he designed the transformers to run one specific tube, the EL34. And Higher plate voltage.

here is what they said..

This new Trainwreck circuit, which includes Ken Fischer's specially designed power and output transformers, offers an exceptional sound. The Concorde is designed to operate exclusively on EL 34 power tubes.

[funkmeblue]

well, ya, and that too....

[funkmeblue]

he also said the 6v6's would sound more hi fi, the concorde also runs at higher voltages than 6v6's could handle. Almost everyone chooses el34s over 6v6s in the express circuit. I believe ken would of chose a different output transformer, not so much a different impedence, for a true 6v6 amp.

[JamesHealey]

So I'm sat here still trying to understand the "reasoning" behind the choices of the power section in a Wreck.

Low B+, Low Screens, 6.6k OT etc.. not really your typical Marshall 50w!

So I decided to look into operation classes, so to get a pair of EL34s to run in Ultra Linear A/B1 You use the following:

Anode: 430v
Cathode Biased with 470ohm Resistor
with a Primary Impedance of 6600ohms
and you get about 37watts.

now to run in 'guitar standard' Class A/B with a pair of EL34 most people use something like this
Anode: 400
Screens: 425
Grid: -38
with a Primary Impedance of ideally 3400ohms
and you end up with about 55watts

there are variations on both ideas to get around something similar for each one, obviously some are run on higher plates etc.. but the general ideal is not to over dissapate the EL34s.

So ken mentioned in the latest video interview posted by Glen Kuykendall, that what made Wrecks special was the combination of HiFi and Guitar amp technology, eg using Solidcore wire in the pre-amp etc. obviously he didnt get into operation classes etc. etc.

But isn't it pretty obvious he drew influence from HiFi spec operation classes and the 'Class A/B' standard used in most guitar amps?

Or am I just being stupid?

[Jackie Treehorn]

With a pentode, there is an impedance match which gives the most power. Say with an EL34, it's something like 3.5k (I'm approximating). There is also a very large amount of odd order distortion at this impedance, but not much second/even order. Increasing the impedance from this point gradually reduces the power output and the odd order distortion, but the even order distortion begins to increase. I think what Ken was going for was both reduction in power and odd order distortion (crunch). Even order distortion would be mostly cancelled out in the output transformer since the amp is push pull.

Using the above example, most hifi amps would decrease the impedance slightly, say 3.3k-3.4k where you get almost all the power, but much less distortion.

There's a great graph illustrating impedance vs power/2nd/3rd distortions in a book I have. I'll have to dig it out.

Also, the impedance affects the current in the tube, or the incline of the loadline. Higher matches make the loadline more level, lower matches are steeper.

Ultralinear, the screens are attached to the output transformer to get the pentode to act more like a triode, so it's probably not relevent to the express output section.

[j-po]

I don't think even order distortion is cancelled in a push pull amp when it is driven into clipoff/saturation. How could the tube that is cut off completely reduce distortion? It is just sitting silent.

[dave g]

I don't think even order distortion is cancelled in a push pull amp when it is driven into clipoff/saturation. How could the tube that is cut off completely reduce distortion? It is just sitting silent.

It doesn't work that way, and that's backwards intuition and reasoning. Take the Fourier series of a square wave (i.e., symmetrical clipping) and you will find that the even order terms end up being out of phase and canceling.

I did some measurements a while back on my spectrum analyzer showing how both the preamp and poweramp distortion in a plexi work together to produce the characteristic sound, and why you can't claim that just one part of the amp is responsible for the distortion you hear.

The plexi pre image is the output of the preamp only in response to a 200 Hz 200 mV input signal with the volume cranked. The plexi output image is the spectrum of the signal delivered to the speaker, but with the preamp disconnected and a 2 volt clean sine wave being fed into the PI. The "plexi" image is the spectrum of the output but with the preamp connected to the power amp with a 200 Hz 200 mV input signal.

Notice how the output section alone cancels the even order harmonics (they're -50 dB down), but by introducing asymmetrical clipping in the preamp we can nicely fill in the missing parts of the spectrum.

This is also the biggest reason why blackface fenders don't sound nearly as good distorted as old marshalls - the blackface topology with the tonestack in between the 2 preamp stages prevents distortion in the preamp, causing you to miss out on the even harmonics that give the fatness to the sound. Notice that the TW preamp looks a lot like a blackface fender preamp, but with a single additional gain stage for asymmetrical clipping. Hmm....

[paulster]

It would be really interesting to see the plot from an Express output stage to compare since it's almost identical to the plexi output stage save for the operating points.

[dave g]

I have some plots of Expresses, Liverpools, and Rockets (not to mention a few Dclones as well, along with a ton of other amps). Unfortunately I don't have them on my computer here at college. I'll be heading up to chicago next week so I'll try to post them then.

[Zippy]

This is also the biggest reason why blackface fenders don't sound nearly as good distorted as old marshalls - the blackface topology with the tonestack in between the 2 preamp stages prevents distortion in the preamp, causing you to miss out on the even harmonics that give the fatness to the sound. Notice that the TW preamp looks a lot like a blackface fender preamp, but with a single additional gain stage for asymmetrical clipping. Hmm....

Make that "Express/Liverpool" - the topology of the Rocket is more closely associated with a tweed Fender Bassman.

[dave g]

Make that "Express/Liverpool" - the topology of the Rocket is more closely associated with a tweed Fender Bassman.

I'm not sure what your point is. I think that most people here generally understand the phrase "TW preamp" to refer to that of the Express/Liverpool. I still maintain that the Rocket is nothing but an AC-30 clone.

Regardless, notice that both the AC-30 and Tweed Bassman preamp is identical in topology to the Plexi's. The lack of an attenuating tonestack between the two preamp stages allows the second stage to undergo a good amount of clipping, so my claim regarding the interaction of preamp and poweramp distortion is also valid for the Rocket.

What I was trying to point out is that the Express/Liverpool preamp is essentially a modified Blackface preamp in that it fixes the lack of preamp distortion by introducing the 3rd stage. It's really an elegant design, because it gives you the asymmetrical clipping in the preamp necessary to restore the lost even harmonics in the push pull output section (akin to a plexi), but unlike a plexi it doesn't have an attenuating tonestack in between the preamp and poweramp. This means that you also get a larger voltage swing hitting the poweramp, allowing you to clip the output tubes even harder than a plexi allows.

The result is a "high gain" amplifier that doesn't get a disproportionate amount of its distortion from the preamp - it's the perfect balance. Add to that the fact that like others have already mentioned, pushing the output tubes further into saturation than a plexi means that they'll draw a lot more current...so you'll get more power supply sag and therefore more compression under clipping, which brings about the clean-to-mean phenomenon.

In my opinion, it's a far better design than the JCM 800 for getting the plexi sound taken to the next level. The problem with the 800 is that you can't get any larger of a swing hitting the poweramp than you could with a plexi due to the fact that you've still got the tonestack stuck right before the PI. Additionally, at low gain settings the amp behaves quite like a plexi (except worse), mainly clipping the 3rd stage/CF pair...but turn the gain up and you start clipping the 2nd stage. Since each stage is inverting, all of a sudden you're getting symmetrical clipping in the preamp as well as the power amp - goodbye even harmonics, hello fizz city!