Manzamp!

[M Fowler]

Great build and it sounds fantastic.

[andyhardy]

Very impressive build,playing, sound and investigative research well done!

[Synchu]

Sounds great this last one as well as the previous ones!
Niki

[ic-racer]

Thanks again for all the kind comments.

[ic-racer]

Today I 'discovered' two more Manzamp photographs. The front picture is really nice to find, looks like we guessed correctly on many aspects of the front panel, though not exact. Thanks to Taylor, again, for suggesting the front-screw corners!

[img:618:462]http://img.photobucket.com/albums/v670/ ... nzamp3.jpg[/img]

[img:618:465]http://img.photobucket.com/albums/v670/ ... 1233_n.jpg[/img]

[ic-racer]

This is what we came up with with only the rear picture to go by. The perspective of the rear view made it look like the top and bottom front panel were the same size. But I see now the bottom front panel is smaller on the original and the baffle may be mounted differently than I predicted.


[img768]http://img.photobucket.com/albums/v670/ ... CF7141.jpg[/img]

[amplifiednation]

Oooh and there's black piping on the baffle too!!!

[ic-racer]

Oooh and there's black piping on the baffle too!!!

The good eye of a amp cabinet builder! I missed that.

Also, there is no way we would have seen those newer pictures of the original as it looks like they were posted on the faux "Dumble" facebook page about 3 months after my cabinet was finished.

[ic-racer]

This thing sounds so darn good, I can see myself putting more time into the project. I might try to fit some piping like this around the baffleboard:
What do you guys think, is it brown or black?
[img:822:747]http://img.photobucket.com/albums/v670/ ... piping.jpg[/img]

[ic-racer]

There was a question on capacitor orientation. This picture shows how I oriented the capacitors, but I did not check for outside foil orientation. When placing the capacitors I was at a stage of not even knowing if the amp would work, so I skipped that detail.

[img768]http://img.photobucket.com/albums/v670/ ... ile-27.jpg[/img]

[10thTx]

I think your amp sounds so amazingly incredibly good, that I am not sure I would change anything about it IF it were mine!


With respect, 10thtx

['67_Plexi]

I've not actually worked on one of these amps, but I can't imagine Dumble running first stage regulation with 110uF from a GZ34. That's a pretty fundamental design faux paux. Maximum design limit is 60uF, in practice even less on new prod. rectifiers, I typically don't dare go above 40uF. This is due to ripple current, the higher the capacitance, the more current it will pull from the rectifier every half cycle pulse. A fully drain capacitor is like a an instantaneous short circuit to the rectifier. The larger the capacitor, the longer that short circuit is 'in place'. A rectifier failing can quickly cost you a power transformer. Above the design limit of 'input or reservoir capacitance' , as its correctly called, and you will almost always get arcing inside the rectifier and early failure. I actually had a new production GZ34 fail on me at a gig in an original JTM50 with stock capacitance and it took the PT with it. It was also pretty common back in the 80's when guys repaired AC30's and used too high a value of capacitor. I've had to rebuild a few of those back in the day.
The schematic seems generally over-regulated for what is essentially a modded JTM50, which only has a 32uF reservoir on the first stage and is perfectly happy without undue ripple. In general tube rectified circuits respond better to less overall power supply regulation than SS.

A second point is the low valued cap in parallel with the high value cap on the second gain stage cathode bypass. The two paralleled caps are summed together to give the total value of 0.1% increase over the 10uF???? It would make more sense if there was a series switch for the 10uF to take it out of circuit, removing a lot of gain in the lower frequencies . Even then 0.1uF still seems a low value and will chew in to the critical mid range gain band. The range 0.5 to 1uF with the 10uF switched out would allow the amp to hold together better when cranked while keeping the gain up in the important ranges.

I would also be tempted to go with a more traditional early Plexi PI and feedback circuit. The one in the drawing is not best suited for driving EL34s in this style of amp, regardless if Dumble used it or not, that's just my feelings on it.


oh...and there's every reason to add grid stoppers to the power tubes and no reasons not to.

Apologies if any of this has been brought up before, I didn't read all the posts and I'm not knocking anyone's efforts or designs. Just some random thoughts.


Alan Phillips
Carol-Ann Amps.

[ic-racer]

I can't imagine Dumble running first stage regulation with 110uF from a GZ34. ...


Alan Phillips
Carol-Ann Amps.

Thanks for having a look and I appreciate the input.

The schematic may need to be updated to 5U4 which is what I am using. I have 'helper' diodes on the 5U4 socket.

Thanks for the comments on the PI. That is something worth looking into.
I used values similar to the PI values in this EL34 modded bassman ( http://www.blueguitar.org/new/schem/mis ... er_sch.gif )

As you know Dumble usually puts the grid stoppers up in the air but I don't see that in the pictures.

A second point is the low valued cap in parallel with the high value cap on the second gain stage cathode bypass. The two paralleled caps are summed together to give the total value of 0.1% increase over the 10uF????

Two caps in parallel on the cathode bypass is odd for sure, but does show up in other amplifier schematics.
[img719]http://img.photobucket.com/albums/v670/ ... 011077.gif[/img]

The mindset I try to keep with this project is that this is not my schematic or my amplifier design.

['67_Plexi]

I can't imagine Dumble running first stage regulation with 110uF from a GZ34. ...


Alan Phillips
Carol-Ann Amps.

Thanks for having a look and I appreciate the input.

The schematic may need to be updated to 5U4 which is what I am using. I have 'helper' diodes on the 5U4 socket.

Thanks for the comments on the PI. That is something worth looking into.
I used values similar to the PI values in this EL34 modded bassman ( http://www.blueguitar.org/new/schem/mis ... er_sch.gif )

As you know Dumble usually puts the grid stoppers up in the air but I don't see that in the pictures.

A second point is the low valued cap in parallel with the high value cap on the second gain stage cathode bypass. The two paralleled caps are summed together to give the total value of 0.1% increase over the 10uF????

Two caps in parallel on the cathode bypass is odd for sure, but does show up in other amplifier schematics.
[img719]http://img.photobucket.com/albums/v670/ ... 011077.gif[/img]

The mindset I try to keep with this project is that this is not my schematic or my amplifier design.

No problem. I came across this thread as part of some research. I have a high profile client that mentioned the Manzamp as part of his description.
I didn't know anything about it. Turns out to be basically a modded Marshall....but your work was definately helpful. In the true spirit of how I remember TAG, providing input is the right thing to do as opposed to just taking information another person has taken the time to research. I make money from this, many of you guys don't, I'm not going to take without giving something back. That said, I'm not building a Manzamp clone, I just wondered what it was.

On that note:

The max capacitance of a 5U4 is actually half of what a GZ34 can safely supply at only 32uF. Look at the pdf data sheet:
http://www.r-type.org/exhib/aaa0726.htm
Looking at the sheet, I'm sure you know already, but condenser is the old word for capacitor. It's only ever really used now in relation to points type ignition systems on cars.
The 5U4 has a higher volt drop too, which can be a good or bad thing depending on the rest of the circuit.
You really should stick within(or as close as possible to) the max operating ranges of the components you use. It doesn't matter who's schematic it is, its a pretty serious design flaw if that schematic is correct. Kind of like having drum brakes on a 600hp car. Using diodes is like using ceramic brake shoes in the drums. Sure it will improve things, but there's a fundamentally better way. I guess I think in the way that every design can be improved, innovation as opposed to invention or blind cloning. Isn't innovation exactly what Dumble is doing with this amp, when he clearly took an early JTM50 as the basis of this project? Try dropping that res. cap string down to a couple of 68uF in series and I bet you like the feel more. If he did use such a high value of reservoir, the only thing I can imagine is that it was a way to tighten the low end of the amp. Weird its identical to the higher powered SS rectified ODS models. Is this schematic an assumed schematic or actually taken from an original amp ?
THe JTM50 is a little spongy on the low end at high volumes. A safer way would be to drop that reservoir value and put a high pass filter in the power amp feedback loop. 150K in parallel with a 0.01uF before the feedback resistor works nice on a 50W Marshall type design. If that's too much, decrease the resistor and increase the cap to values that keeps the Q frequency the same.

Yeah, I'm not sure of the rationale behind those two bypass caps. Tonally. it should makes no difference. I'll make a point of researching it though. There could be a reason I don't know about. The fact the lower value has such a high voltage rating may be a clue. Does it maybe come in to play in a failure mode? Would make more sense on a pentode like an EF86 than a triode though.

Alan Phillips
Carol-Ann Amps.

['67_Plexi]

The use of the small bypass cap in parallel with the large bypass cap on a cathode isn't new, or novel. Similar to how they're used in hifi circles putting a small value film cap in parallel with the electrolytics in the power supply. Different responses/different frequencies, de-harshing the 'lytic, better clarity.

I don't accept those type of explanations its like saying 'it just happens' I need to see the math. If something happens in a circuit, it happens because of some pre-determined set of rules. I don't pretend I know them all, but here's what I do know.

They are not at different frequencies. You look at the network in it's entirety, R1//(C1+C2), not independently R1//C1 and R1//C2.
The paralleling of two capacitors increases the total dielectric area and consequently the amount of capacitance and in the network that is seen as one equivalent capacitance.

C1=Q1/V and C2=Q2/V where Q is the charge. Q=Q1+Q2. The voltage across each capacitor is the same, therefore C = C1+C2

The cutoff frequency point of the network is 1/(2*Pi*R1*(C1+C2)) where C1 and C2 are in Farads and R is in ohms. There isn't two different cutoff points.

The only difference is the charge time for each capacitor in this case is vastly different.

The audio effects of which I'm struggling to visualize. I do know that I would discount it as insignificant in any circuit design calculations....but that doesn't mean I'm right to do so.