How to determine filter caps values?

[Patrik]

Ok, this might just be it. In the third revision the amplification circuit hasn't changed much, I added those grid return resistors that were bugging me for all that time. Most notable change is in the power supply circuit that lost its EZ81 tube, being replaced by more powerful GZ34 so I can utilize 5V transformer winding and reduce load on heater winding. 

[pdf64]

The HT voltage is way too high, 350-0-350 Vac with a GZ34 will be more like 450Vdc.
Where are you getting your numbers from? The load line is horrendous, plate dissipation way over the limit!

A 100pF bypass cap won't have much effect.

The preamp gain is very high, some provision for attenuating the signal level between stages should be considered.

[sluckey]

The LED will not work as drawn. Replace the bottom diode with a 1K resistor for current limiting (use smaller resistor for brighter LED). Be sure the cathode of the top diode is connected to the anode of the LED.

[ac427v]

You were considering the Hammond 372BX 300-0-300 in your first post. How would that work?

[Patrik]

So, at first I was discussing two possible power transformers, Hammond 372BX and 373X. At first I wanted to use the 372BX due to its lower output (cca. 600V) but when I was notified that EZ81 can handle up to 350V and that this might improve the amplifier I switched to 373X (cca. 700V). Now I still had 372BX in my mind and that is where the false input in to rectifier tube comes from. Also I didn't  expect that with GZ34 I might get such a boost in output voltage, therefore I will probably change the transformer once more, so I get between 350V and 360V on the reservoir cap becouse as I can see GZ34 at 300V input will put out somewhere about 330V and if possible I would like to stick to current voltages so I don't have to completely redesign power supply, that might be a good idea becouse as you said the load line is rather poor and would benefit from redesign. Also how do I fix plate dissipation? I didn't check the characteristics for GZ34 (it was a mistake) and just assumed it will respond in the same way EZ81. Before that I also considered using 5Y3GT but abandoned the idea due to its lower capacitance rating for reservoir cap, 37uf. Thanks also for noticing the problem with LED driver circuit and that bypass cap is poorly written, it is actually 100uf.

[pdf64]

... the load line is rather poor and would benefit from redesign. Also how do I fix plate dissipation?...

Draw it onto the plate chart, see http://www.valvewizard.co.uk/se.html

This is a nice online tool, once you understand how to use it http://bmamps.com/ivds.html

[Patrik]

But is the amplification part of the amplifier ok? Just so I know if I have to fix something or if I can use it as a starting point.

[sluckey]

But is the amplification part of the amplifier ok? Just so I know if I have to fix something or if I can use it as a starting point.

pdf64 said, "The preamp gain is very high, some provision for attenuating the signal level between stages should be considered."

4 gain stages without any attenuation (voltage dividers) between stages seems excessive to me. The amp will probably be a squealer. Be prepared to put a voltage divider between every stage. Here's an example of a 4 stage preamp with voltage dividers to tame the gain...

[Patrik]

Thanks for advice.

[Patrik]

And what if I used some lower gain tubes on third and fourth stage? Something like 12AY7 or maybe even 12AU7.

[sluckey]

How 'bout if you only use one tube?

What's your goal in using four triodes? Gonna play some big hair music?

[Patrik]

I want to have enough gain to push power tube into fuzz, over the overdrive. I am not sure if one 12AX7 will do.

[pdf64]

I want to have enough gain to push power tube into fuzz, over the overdrive. I am not sure if one 12AX7 will do.

With a simple vol / tone arrangement, I think it probably would, see

Likely schematic https://el34world.com/charts/Schematics/files/Fender/Fender_princeton.pdf

[sluckey]

I want to have enough gain to push power tube into fuzz, over the overdrive. I am not sure if one 12AX7 will do.

You're on the right track. 4 triodes will probably put out square waves for any guitar you plug in. Definitely gonna do fuzz. Controlling all that gain is gonna be a challenge.

[shooter]

you have an SE amp, which by design wants to be Class A, A1.
They can fuzz out, but they can also scream, make treble totally annoying and leave Bass farting all over the stage.
This just a guess-tim-it, the cathode VDC will land somewhere 15-25vdc (your bias).  Once you know that number "design" the pre to give the PA tube somewhere between 30-50 volts AC signal with gain n volume at 10.
then play it, then modify it closer, then repeat.


or
take a known good amp that you like the fuzz and "build around that design" adding your twist

[Patrik]

For now I found out that I enjoy the sound of any amp without negative feedback much more than with negative feedback. Maybe doing removing NG and using that Maggie amp circuit that was mentioned previously on this thread would be a good starting point, eliminating the need to do from the scratch designing and give some foundations that hopefully work, to build up from there becouse as I see it I can always take the amp apart and modify it. What I want to say is, that I will try to find some suitable 10W amplifier circuit, SE or PP, doesn't really matter, and start from there and we will se how it goes. Now I have in mind previously mentioned Weber Maggie and Fender Princeton Blackface (with that I will have to find a way to remove vibrato and reverb circuit). Thanks a lot to all of you who responded.

Update: I went to have a little look and seems I am not the only one with the idea of reverbless and vibratoless princeton. Rob Robinette made such a modification and can be found in the link bellow. He used a dummy load to immitate the effect of reverb on the rest of the circuit. I think I read somewhere that reducing that load would increase the drive emmensly so replacing it with some lower resistance resistor might be viable option to get more drive, or maybe even putting in a potentiometer. Modifying the tone stack to volume tone arrangement and loosing negative feedback and this could just be a winner for me. I have always loved the sound of overdriven 6V6 tubes so this is a very intriguing option. I also have stacks of them lying around. Loosing the bright cap would also be nice I think becouse I intend to use Jensen P12R speaker which is rather bright so balancing it with a bit darker amp would de a good idea I suppose.

The amp explanation page: https://robrobinette.com/AA1164_Princeton_Reverb_With_Reverb_and_Tremolo_Deleted.htm

[Patrik]

Hello to all once more. After I did some research I found out that, as someone indicated before, weight isn't the SE amps strong point, kinda counter-intuitive but anyway, after I posted that diagram of a modified Fender Princeton yesterday I made a little research and found out that the weight of the output transformer for that amp is one third that of a SE 10W amp. Therefore I made a decision to switch po PP amplifier, so I wave quite some weight and also cost, transformers are cheaper by quite some margin. Now I settled on diagram by Rob Robinette for Fender Princeton but without reverb and vibrato. I already made some modifications and I think that for now at least that is what I will go with. The changes I made are, modified tone stack, so I have only Tone control, tone and volume control bypass, reverb load was replaced by a potentiometer (it turned out I was correct on that that smaller load will cause more gain, therefore I decided to replace resistor with a 500k pot), I also lost bright cap that was originally there and put on a pre-phaseinverter master volume. For transformers I settled on Hammond 376X power transformer and 1750E output transformer, which is a direct replacement for original Princeton one. I was just wondering if I could replace GZ34 rectifier with 5Y3GT, since it still should be powerful enough for this amp.
For now this is all from me and I would like to hear some response from someone with some more experience than I have. Thanks to you all.

[sluckey]

Your LED and diode will both fail immediately when you apply power.

[ac427v]

I think the new schematic is getting close. But the PT is still way too big. The 376X is rated to provide 640v CT @173ma. The amperage rating for your design should be around 100ma or the resulting DC voltage on the power tube plates will be way too high. I reread your first post where you noted that your Dynacord KV10 was the close you what you want. Very interesting amp--I think I would like it a lot. It is rated at 10 watts and has a B+ of only 220 dc volts. Amps with B+ over 400 volts don't get power tube distortion without being really loud. I think a push pull amp with B+ about 350 dc volts might be closer to what you want. The Hammond 290AEX is rated 550v CT @ 100ma using the low voltage secondarily taps. I use the USA version with a 5Y3 and a 1750E to get B+ of 350 in my Princeton Reverb clone. When I use the high voltage secondaries I get 390v with a 5Y3. Varying rectifier tubes and secondary taps allows great leeway in finding the right B+. I'm assuming that going from 220v B+ in the Dynacord to 350v in a Princeton will give a more crunchy distortion sound without losing your ability to adjust it with the guitar volume control. I think it would be close enough to allow parts-tweaking to get your ideal performance--until you "need" another amp.

[Patrik]

In this case I can probably replace 376X transformer with something of lower voltage and power rating.
And for now I will let go of LED and just use usual wolfram bulb. Probably my best option then is a Hammond 372BX transformer, which is a bit lower voltage, but if it was wired as for 220Vacc or maybe even 200Vac with modern 230Vac it is probably the closest I can get to original specs, without actually using Princceton replacement transformer (I do't like Fender replacement transformers since they need a huge hole in the chassis to be installed, much prefer the upright type).

[Patrik]

Hallo, to you all. Just a little update on my project. Atter a bit of thinking i decided that ordering a precut chassis would be much easier than modifying and using a general purpose chassis. So I designed a small chassis, so all components can be wired point to point, plans are attached bellow. Don't bother reading becouse it is written in slovenian, so just if someone else might be interested. I am having chassis laser cut and costwise this is nearly cheaper than using universal chassis.