New 2 channel KT88 based P/P design. Check my work?

[Gainzilla]

Hi all,

I'm embarking on a new project. It'll be a 2 channel affair (high gain/fender clean) into a pair of KT88s. The problem is I've never used KT88s before, let alone designed a power supply that will power it.

I read a post on another forum that said essentially a typical Marshall style power supply getting somewhere near 500v is ok, as long as there are 200+ ma current as well. I've tried to simulate this in Duncan's PSU Designer, but get -0v warnings. I'm not sure I'm setting it up properly. The Pwr Supply and Power Stage designs are kind of a mashup of the Naylor Superdrive 60 and the Deliverance 60 (LTPI as per the Naylor, but with P/P KT88 values from the Deli).

Anyway, does this look like it'll work? I'm specifically looking at are the PSU values look ok, and do the PI/PA values look ok for the KT88s?

Any info would be most appreciated.

Thanks!

Output Stage:


Power Supply:


Preamp:


PS Sim (as close as I could get without generating warnings or errors):

[Alan0354]

You have some mistake on the preamp page. You connect the input to the second stage to "E"!!! Also, I would put the 0.0022 DC blocking cap right after the plate of the first stage so you don't have the two RC network floating at high voltage.

I don't think you can have a switch to ground the cathode of the second gain stage.

[potatofarmer]

Off the top of my head:

if you're running DC heaters, you don't want that artificial center tap before the heater rectifier.
Also check to make sure those grid leaks aren't too big for KT88s in fixed bias.
If that 60V is a separate winding, go for full wave rectification over half wave. Easier to filter ripple.

Also, bridge rectifying a 500V secondary is going to give you over 700V peak. Probably not a whole lot less than that at idle. You're simulating a choke-input rectifier, but your schematic is for a capacitor-input.

[Dave H]

The outputs from the first three stages of the high gain channel should be taken from the plates not 'E' and similarly on the low gain channel take the outputs of the first two stages from the plates not 'D'

The 'More' switch is connected to ground. It should connect to the other end of the 470k.

V1a and V2a grids are floating when not selected also bias 25k and power amp 5k pot wipers are floating.

Bias supply 10u caps are reversed. They should be +ve to ground.

[loudthud]

PSUD2 often generates errors when using a choke input.

[MarkusBass]

On every drawing there are mistakes. On the power amp drawing the slider of the PRESENCE pot is not connected. The output transformer is not connected to ground. There is only NFB connection but without ground connection it will not work.
On the preamp drawing the signal is taken from power supply instead of anodes of the tubes. Look at V2a, V2b, V1a, V1b and you will understand.
On the power supply drawing the slider of the bias pot is not connected.
This is just from a quick look at the drawings. Other mistakes are also possible.

EDIT: other mistakes: MORE switch connects the grid of V6 to the ground - you will not hear any sound in this case. Missing grid stoppers resistors on V1a and V2b can be discussed.
Switches designed in this way may cause unpleasant sounds when switched. This is not a good project, or the author wasn't focused enough on making the drawings. The amp build as it is will not work.

Mark

[Gainzilla]

Hi all,

Wow, so many mistakes. So thankful for your help!

I think I've addressed most of the feedback I've heard so far, with a couple exceptions:

I kept the dc blocking cap at the end of the v1a r/c network to stay closer to the original Naylor design. Still working through some of the Pwr supply related issues.

potatofarmer - With the elevated heater "design" I'm trying to only elevate the preamp tubes. Would it be better to elevate all tubes? The 60v supply is indeed a separate tap. I'll have a look at the full bridge rectification there.

Here are the updated drawings, hopefully with no more mistakes. I've updated some of the values around the KT88s to be closer to the Marshall Jcm800 Kerry King signature model, which uses a quarted of KT88s (lol). Hopefully Marshall has done their homework so I can stand on their shoulders.

Does the channel switching look better now? Originally I wanted to try and keep the signal completely out of each circuit that isn't currently switched in. Is there any downside to having the switching as shown (channel bleed, excessive tube wear, popping when switching, etc.)?

Thanks again for all the feedback! Any new feedback?

Updated Supply:

Updated Output Section:


Updated Preamp:


Updated Supply Sim:

[MarkusBass]

Some of the problems still haven't been addressed. For example speaker is not connected to the ground. NFB will not work.
Channel switching is corrected so it will work but you will get popping when the channels are switched. I would add at least a resistor after 0.047uF capacitor from anode of V6. 680k or a little bit more.

Mark

[Alan0354]

Use 1N5408 for the hv rectifiers, 4007 is 1A diode, 1N5408 is 3A 1000V.

You need to logically go over the schematic, don't just copy one part with another. A lot of the mistake is so obvious that lead me to suspect you are just cut and paste.

[Gainzilla]

Thanks Mark!

Updated scheme. I'm assuming that if I still get popping I could bump up to 1M?

FWIW, the NFB does work. I built it in another amp, but it is a little unconventional. It's based on the Naylor SuperDrive 60 (attached, in case you're not familiar with it). I did have to mount some of the parts on the pot to avoid oscillation, though.

UPDATED:

Output Stage:


Preamp:


Naylor Schem:

[potatofarmer]

If your NFB works without an explicit connection of the OT secondary to ground, then it works because the OT secondary is grounded through the jack contacting the chassis.

I still really don't like the elevated and rectified heaters, but I've never messed with DC heaters, so I'll leave that for someone else. At the very least, you shouldn't be referencing the power tube heaters to half of B+.


You sure you don't want to start designing with a project that's a little simpler?

[J M Fahey]

6 gross errors on the preamp, another 2 on the power amp and another 6 on the power supply must set some kind of record, and show you have no clue.

I suggest you build a tried and true circuit instead of butchering other people's designs.

[J M Fahey]

Use 1N5408 for the hv rectifiers, 4007 is 1A diode, 1N5408 is 3A 1000V.

You need to logically go over the schematic, don't just copy one part with another. A lot of the mistake is so obvious that lead me to suspect you are just cut and paste.

Not even 1N5408 will be enough, PIV there is 1400V , way above 1N5408 specs, so at least 2 in series each are needed (8 to complete a bridge).

[MarkusBass]

I'm assuming that if I still get popping I could bump up to 1M?

No, it's not the value of the resistors that stops the switch from popping. In your schematic the resistor was simply missing and this (the lack of resistor) would cause popping .
I was wondering where you are going to buy dual 560k pot but it looks like it should be rather 2x500k.

Other component values are a little bit high but they may work. For example capacitors and resistors in the bias circuit seem to me slightly too high. I would also decrease the capacitors in the power amp (0.047uF). People increase these values in order to have more low frequencies but they get also blocking distortion and as a result the amp sounds worse than it could with other component values. But this can be discussed elsewhere.

Mark

[Gainzilla]

[EDIT] I see what you're saying now. After reading JMFahey's post, it seems like I need some more thought here. The high voltage is complicating things in the build in ways I hadn't anticipated. Thx