X-chassis. Flexible Amp Design Platform

[silverfox]

Above are some images of the power amp section, a Mission Amps PPI-MV schematic and the X-chassis I've begun construction on. The project is an amp design and test platform to allow for rapid prototyping of designs. This is not original idea, just my version of it and I would really appreciate suggestions to make it work smoothly. It's early in the construction phase so ample opportunity to help work out the bugs. From the images you can see the intent is to solder spade connectors onto various components and use the barrier strips to wire the curcuits up and then onto the desired sockets, connectors, pots, ect.. Signal runs, lead dress will follow logical paths and I'll use shielded cable and copper clad board over top of sensitive circuits for shielding. The completed prototypes, if done properly should be very close to the final layout.

The power supply is stock from a Peavey Road Master and so is the PI and power amp section. I need some assistance and input regarding modding it with the Mission Amps MV circuit. Is this going to be a good fit with this phase inverter and power amp? The other question relates to where to wire it into the circuit. The sch. Shown isn't an exact fit for the version of PI in the power amp circuit. For example I don't see the 100k resistor that matches with the 82k.

Thanks for considerations and assistance,
Rants, Raves??

Silverfox.

[soundguruman]

[ATTACH=CONFIG]27145[/ATTACH] [ATTACH=CONFIG]27146[/ATTACH] [ATTACH=CONFIG]27144[/ATTACH]

Above are some images of the power amp section, a Mission Amps PPI-MV schematic and the X-chassis I've begun construction on. The project is an amp design and test platform to allow for rapid prototyping of designs. This is not original idea, just my version of it and I would really appreciate suggestions to make it work smoothly. It's early in the construction phase so ample opportunity to help work out the bugs. From the images you can see the intent is to solder spade connectors onto various components and use the barrier strips to wire the curcuits up and then onto the desired sockets, connectors, pots, ect.. Signal runs, lead dress will follow logical paths and I'll use shielded cable and copper clad board over top of sensitive circuits for shielding. The completed prototypes, if done properly should be very close to the final layout.

The power supply is stock from a Peavey Road Master and so is the PI and power amp section. I need some assistance and input regarding modding it with the Mission Amps MV circuit. Is this going to be a good fit with this phase inverter and power amp? The other question relates to where to wire it into the circuit. The sch. Shown isn't an exact fit for the version of PI in the power amp circuit. For example I don't see the 100k resistor that matches with the 82k.

Thanks for considerations and assistance,
Rants, Raves??

Silverfox.

Yes, I had envisioned it long ago.
I called it the: "tweak-master"

Very Cool!

[Alan0354]

I think the standard boards with holes that you can put two rolls of solder post works just as well if not better than screw terminal block. They have kits for those boards with with screw on solder post. Time it takes to solder components on the post is faster than screw the components onto the terminal blocks. Not to mention you have to put the lug onto every components before you can use them. The leads might break off close to the lug after you use it for a few times and you have to make it again.

I think it will be better to turn the terminal blocks 90 degree so components propagate along with the tubes like most of the layout of the existing amps. That is if you look at the tubes all in one straight line, the components should be perpendicular to the line of the tubes just like Fender and Marshall Plexi.

[kg]

Thanks for considerations and assistance,
Rants, Raves??

Silverfox.

if you're going through the trouble of using a tube as a CF driver for output stage, why not use it to direct couple to the grids? for example, generate a high potental low impedance bias rail, say -100vdc. tie the cathodes of v6 to that rail through appropriately sized cathode resistors. tie the plates of that tube to a lower b+ rail, say +150vdc. inject your bias voltage through r69 and r74 (can be via two pots for side by side adjustable biasing).

for comparison, see http://ken-gilbert.com/images/pdf/BA...iverschema.pdf which is the driver circuit for the 600w baga.

the cf drivers in this case are loaded with a constant current sink set to 15ma, and the tubes are el84s connected in pentode mode--two complications you would not need to incorporate. the input tube is a 12at7. though as shown the circuit is fully differential, i actually drive this circuit single ended, using only one paraphase input. the relatively long tail resistor splits the phase for me. this would allow you the same flexibility for testing/prototyping--see if you prefer using a concertina immediately prior, for example. or you could use the other diff amp input as a loop NFB injection point. or you could use it as a second power amp input from an entirely different signal path/channel.

note the use of the different ps rails, principally the -300v rail. that allows higher value resistors to be used, improving headroom and linearity, and in the case of the LTP paraphase performance.

the great advantage of this directly coupled CF driver is that grid current operation is now possible (ie increased output power due to increased peak Ik), and there is absolutely no blocking distortion (feed it as hot a signal as you want without fear of farting out). additionally the bias is ROCK SOLID. a tube gassy enough to produce enough ionization current to cause immediate runaway and component death in a traditional high Z bias supply can be used with perfect success.

i see no great advantage to incorporating a differential master volume control in this section. i'd just feed the LTP diff amp with a variable signal level via a pot. one potential issue with bruce's placement is that it is inside the feedback loop, which will cause diminished effect.

ken

[silverfox]

Thanks for the suggestions-

i see no great advantage to incorporating a differential master volume control in this section. i'd just feed the LTP diff amp with a variable signal level via a pot. one potential issue with bruce's placement is that it is inside the feedback loop, which will cause diminished effect.

ken

This part I understand and have reservations regarding the use of a PI Master Volume, (do I really need it?), but wanted other thoughts on the matter.

The rest... , I'm going to have to study. The existing Fender based power amp circuit sounds fantastic. I have not driven it to full output power so I have no experience with blocking distortion. So far existing preamp circuits also sounded good with the AB763 design.

However, I need to better understand what you are proposing and if there are some audio clips-examples of a PI design or an amp that uses this PI design I'd like to hear it. And as I said, I need to better understand why I would use this design. I'll look into it, I'll try to find the 600w baga. (The name KG sounded familiar and it wasn't until I went to the web page that I realized I had been there before.)

Another challange I would have would be coming up with the exact component values to make that driver cicuit work wit the 6L6 output section.

Alan0354- "I think it will be better to turn the terminal blocks 90 degree"

Yes that is how I originally was going to arrange things and will likely rotate them back. Wire clearance was a bit of a problem but I'll figure that out. I may even be able to mount the barriers on a copper clad board and place them over the tube sockets for unlimited space...

Thanks for the feedback,

Silverfox

[Alan0354]

Yes that is how I originally was going to arrange things and will likely rotate them back. Wire clearance was a bit of a problem but I'll figure that out. I may even be able to mount the barriers on a copper clad board and place them over the tube sockets for unlimited space...

Thanks for the feedback,

Silverfox

I was thinking about that, you can wire the socket connection with wires so you can put the boards over the socket. But I think you should have room, Fender and Marshalls found room.

[kg]

The rest... , I'm going to have to study. The existing Fender based power amp circuit sounds fantastic. I have not driven it to full output power so I have no experience with blocking distortion. So far existing preamp circuits also sounded good with the AB763 design.

However, I need to better understand what you are proposing and if there are some audio clips-examples of a PI design or an amp that uses this PI design I'd like to hear it. And as I said, I need to better understand why I would use this design. I'll look into it, I'll try to find the 600w baga. (The name KG sounded familiar and it wasn't until I went to the web page that I realized I had been there before.)

basically instead of those last two cathode followers being self biased across r70 and r75, they would run in fixed bias. because you now have "control" over their cathode voltages (via their fixed bias grids) you can directly couple them to the output tube grids. feeding their cathodes from a negative rail means you can set that node at your desired output tube bias voltage, reducing the DCR of your output stage grid circuit by a factor of 1 or 2 orders of magnitude.

as for examples of it, i'm afraid there aren't too many in MI use. the ampeg svt circuit is similar, however they hobbled the circuit by using enormous grid stoppers ampeg svt and grid current

the fender ps400 monster uses an interstage tranny driven by a 6l6 power tube as a driver circuit, with about 1k of DCR in the loop. it's the closest thing to ab2 operation that i've seen in a commercial all tube MI amp. some common grid music man amps can push into grid current but they use BJTs in the cathode circuit to do it.

[silverfox]

Conclusions and Questions Regarding AB763 Power Amp



Thanks for the explanation KG. At this point in time I would have to study this circuit out substantially more to understand it and I don't want to put too many irons in the fire in the start so for now I'll have to keep the power amp I had and consider changing that when “X” is up and running. I'll likely build the PA on barrier strips for easy mods or stripping out in the future. As you can see from the following questions, I'm perhaps a little beyond novice level.

Regarding the AB763 circuit, last night I was transcribing parts from the schematic to a parts box. I ran into two questions.

First the PCB doesn't match the schematic in one part- The power amp coupling cap is .068 instead of .047 (C35). It was also of the Tombstone variety. Should I change that to .047 tubular or leave it as found?

C46 is a 39 pf Disc cap. I don't understand the function of this cap, (speculate it is forming a high frequency pass filter or bias for V6a cathode? Should that be changed to a Black Beauty, (Mica) of 47 pf or just leave it alone?

General question on the 2.2M and 220k parallel resistors R57-58. Is the 2.2M being used as a cap or is this an attempt to obtain an off stock value resistor?

Thanks for your considerations of these questions,

Silverfox.

[silverfox]

Project Status

I'm getting a little closer to powering up the build. At this point the only goal is to get the origianally designed phase inverter back up and running and a scaled down version of the power tubes- Two instead of the original six. I've hacked the original Peavey schematic for those portions and had my mind in a puzzle trying to untangle all the grounds and rewire them back into the power board.

I'll use a Local Star and then ground the various Stars to the chassis near the components they group with or bring a ground for each back to the supply cap they draw power from. What's best? The wire towards the front is a ground bus. Don't know if I'll use it in the future or not or perhaps isolate it from the chassis first... Here is a pic of what I've got done so far.



Comments welcome,

Silverfox.

[silverfox]

Phase Inverter Input Levels

I do not understand the terms being used here and if the effects box I will be using will provide a strong enough output signal to drive the power amp to 100 percent output. Will the effects unit drive the phase inverter into power amp distortion region? I freely admit to knowing very little about these concepts dB and the associated per volt ratios.

The phase inverter circuit I'm putting together is based on the Fender AB763. In Peaveys documentation it states:

Preamp Out: 6 dBV 2 V RMS out. Maximun output 18 dB, 8 V RMS

Power Amp In: +6 dBV, 2 V RMS.

Initially I want to drive the power amp with an effects box that produces up to -10 dB line levels.

How is -10 dB relate to +6 dB input levels?

Silverfox.

[silverfox]

Oscillations on Startup

Schematic at end for Referrence

Last night I powered up the X-chassis and it doesn't work. I need some suggestions and here is what I've got and steps performed so far:

AB763 design. On power up it begins high pitched squeal around 90 VAC input to power transformer. I'm bringing it up on a Variac. I sounds like the crossed primary wires problem on the output transformer but I already swapped those leads.

I then tried swapping the leads feeding the power tubes- the drives; Having found an old post by Enzo suggesting that could be a problem. Still squeals.

Pulled the 12AX7 driver tube and the squeal stops.

A cursory check of the voltages indicates they are very close to the schematic values as listed above.

The bias at the 47K resistors is reading slightly low but that is due to a different power transformer: 53 volts (-).

There is a very low signal that is making it through to the speakers at the same time as the squeal.

I've got to get this amp running before the tube embargo hits.... (For those of you that followed that thread).

Thanks for any assistance,

Suggestions??

Silverfox.

http://music-electronics-forum.com/a...seinverter.gif

[g1]

How about with the phase splitter in and the other preamp tubes removed?

[Enzo]

If you want to be certain there is no output phase problem, simply disconnect the NFB line.

The variac is a powerful tool. To me its main value is that I can bring up a circuit slowly and partially powered to see if excess current is going to be drawn. Once I know the circuit is not going to smoke or blow fuses, then I ditch the variac, or just set it to mains voltage. Amps are amazingly flexible, but unless you are up to full voltage, they can be unstable. So unless you still need to watch current for some condition, stop using the variac.

Your schematic shows a pair of cathode follower as drivers after the phase inverter. That doesn't seem very AB763 to me. You have a 12AX7 phase inverter feeding the 12AT7 drivers. The AB763 has a 12AT7 phase inverter and no drivers.

Now we need to make sure we are all referring to the same things. You said you pulled the 12AX7 driver and the noise stopped. Well there is a 12AX7 and there is a driver, but they are not the same tube. SO which did you pull? Also, your circuit has the two tube types, so do you indeed have a 12AX7 in the first socket and a 12AT7 in the second?

[silverfox]

I just got back. I won't be able to get to this until Sunday. I picked up a gas engine at a sale today and the carb is probably gummed up. Consequently I've been trying to pull start a gas engine to the point I can hardly think or stand straight at the moment. WOW. Not 18 anymore.

As soon as I get to this, Sunday?? I'll fill in all the blanks.

Thanks for the suggestions I'll disconnect the NFB line first and try again.

Silverfox.

[silverfox]

Found the Problem

I just got it figured out. During the testing process I inadvertantly plugged the output of my effects unit into the input of the effects unit. This is a stereo effects box so I still had one output going into the power amp. Of course the effects box was oscillating and feeding that into the amp.

For what it is worth, I only built the power amp portion of the amp as indicated in the schematic.

So it's working.

I do wonder if the input capacitor to the power amp, .0068 uf, (682 code) isn't too small a value for the input thus cutting the lows? The schematic shows .047 but on the PCB of the original circuit the .0068 was the value used. Hmm... And Good Night,

Silverfox.