New build: Magnatone 213 (Tonemaster/Titano/Evil Robot) + mods --- advice wanted!

[klooon]

Version 6. I believe it’s getting there… Colored circles resemble the ground points. Same colors will go to the same ground “node”.



Please let me know if there are still some modifications/improvements to be made. Your comments are welcome!

[klooon]

Some small changes; here is version 7:


Questions:
- What is the OT primary impedance value in the Magnatone? I have chosen a 6.6kOhm:8Ohm OT (Hammond; Fender Deluxe Reverb). Is this fine, also considering the extra parallel speaker output??
- What kind of grid stoppers are needed for the power tubes? In my Princeton build I used 1k5, but I see that also higher values, e.g. 4k7 appearing. What is preferable?
- Regarding the second fuse: what rating should I use here?
- For a possible OT protection during speaker disconnect, is a 820R - 2 Watt parallel resistor fine? (based on the previously suggested link)

Then I first version for a possible layout:


Quite hard to wire it in a neat way, so your comments and suggestions are welcome!

[tubeswell]

I'm picking the grid leak for the cathodyne is meant to be 1.0M on the Magnatone 213 schematic. (10M is way too much grid leak for a 12AX7.)

[klooon]

I'm picking the grid leak for the cathodyne is meant to be 1.0M on the Magnatone 213 schematic. (10M is way too much grid leak for a 12AX7.)

Thanks. I will adapt that!

What about the general layout. Is there a better way to position the parts? Especially as regards the ground nodes and maybe some additional filtering/decoupling?

[tubeswell]

I would not choose to do a common ground buss like that myself. You may find it noisy.

If you are aiming at galactic grounding, you should do it as per Merlin's article http://www.valvewizard.co.uk/Grounding.pdf

[klooon]

I would not choose to do a common ground buss like that myself. You may find it noisy.

If you are aiming at galactic grounding, you should do it as per Merlin's article http://www.valvewizard.co.uk/Grounding.pdf

Although it possibly doesn't look like it, I actually tried to do a galactic grounding. There are 4 local stars/nodes (indicated by the 4 colored circles in my schematic) and the corresponding numbers 1-4 in my layout. All the parts that share the same supply and filtering are going to the same ground node. All the 4 nodes are connected (indicated by the thick black line on the layout)

One of the 'problems' is that most parts share the same supply as node no. 2. All the blue circles ("2") go to Node "2" (e.g. LFO circuit, modulation circuit, 2nd preamp , PI). All share the "A" B+ on the schematic.

What to do?

Should I make more local stars/nodes? If yes, how to do this? E.g. also add more (decoupled) supplies/filtering??

[trobbins]

Mark, it looks like you are doing pretty well in preparing the details - if you don't get any comments from any posters then I suggest that means that what you are planning is 'ok'. Amp circuits can cope with a lot of variation in most values, and there is no 'best fit' values in many situations (eg. grid stoppers). If you have taken the time to read Merlins article, and prepared your own wiring scheme, and generally appreciate what is being aimed at with distributed star grounding, then you are managing the grounding well, and are unlikely to have any issues or concerns. I can't see any glaring issues, although I would personally use a standard electro for decoupling the B power supply node.

The value of grid-leak needed for a valve can vary widely - so best to check the anode voltage in circuit, and adjust the grid-leak to get near a target anode voltage. I'm pretty sure I've seen photos of 10M grid leaks in maggie amps.

Cloning the varistors with a simple 3-zener-R type circuit, as per Martin's scheme, was discussed in that thread, and from memory used 33V, 56V, and 82V zeners (of 400mW to 1W rating). The Amp Garage :: View topic - Varistor Characteristics for Magnatone Vibrato Circuits

[tubeswell]

I would never argue with Tim about tube amps. ;-) Having said that I still think 10M is too high for a grid leak resistor on a 12AX7 unless you are trying for grid-leak biasing. The input impedance with 1M (or even 470k) will be very high for that cathodyne anyway because it is bootstrapped. Anyway its just a little resistor so if you don't like it later you can swap it around. Over to you.

When I was talking about the galactic grounding in my earlier post, it seemed like you had it grounding at the output end whereas its better to ground it at the input end. Mind you I could just be following your layout incorrectly.

[klooon]

Mark, it looks like you are doing pretty well in preparing the details - if you don't get any comments from any posters then I suggest that means that what you are planning is 'ok'. Amp circuits can cope with a lot of variation in most values, and there is no 'best fit' values in many situations (eg. grid stoppers). If you have taken the time to read Merlins article, and prepared your own wiring scheme, and generally appreciate what is being aimed at with distributed star grounding, then you are managing the grounding well, and are unlikely to have any issues or concerns.

Thanks! I have the feeling that I am just a bit up from doing a 'painting by the numbers'-job. And I learn a lot by the responses I got so far, also in my previous build threads.

I can't see any glaring issues, although I would personally use a standard electro for decoupling the B power supply node.

Do you mean, instead of the 0.047uF cap use a electrolytic, e.g. also a 22uF/450V? Could I just replace this? What is the function of the 0.047 exactly in this supply part?

The value of grid-leak needed for a valve can vary widely - so best to check the anode voltage in circuit, and adjust the grid-leak to get near a target anode voltage. I'm pretty sure I've seen photos of 10M grid leaks in maggie amps.

I will try them both... ;-)

Cloning the varistors with a simple 3-zener-R type circuit, as per Martin's scheme, was discussed in that thread, and from memory used 33V, 56V, and 82V zeners (of 400mW to 1W rating). The Amp Garage :: View topic - Varistor Characteristics for Magnatone Vibrato Circuits

Thanks, I found it. I have to get used to the ampage forum: you need to log-in to see all the attachments... ;-)

[trobbins]

The maggie amp uses grid-leak biasing for that particular 12AX7 with a 10Meg grid-leak

[klooon]

The maggie amp uses grid-leak biasing for that particular 12AX7 with a 10Meg grid-leak

Then I will stick to this! Thanks.

[klooon]

OK after a bit of reading and thinking... A few changes to the schematic and the layout.

- Maybe two or three 5W 10V zeners will do the job of bringing the B+ down a bit? The B+ with the current PT is estimated about 400-410VDC, target is about 360VDC-370VDC. Now the Hammond transformer has 240VAC tap primary, but here in the Netherlands we have 225V-230V, so about 5% less. So that should also be 5% less of 410VDC and gives about 20VDC less: 380-390VDC...

- I wasn't too happy with a second channel being much the same as the first one, other that a bit of change in the bright/mellow settings or a different cathode bypass cap. Why not try something more extreme and I found an example of a cascaded input (here)

Here is the adapted schematic and layout v8:

[trobbins]

Your zener direction is correct. Be mindfull that it is almost impossible for a 10W zener to dissipate 10W unless you have amazing heatsinking to 25C ambient. You will certainly need good heatsinking, and even then may only safely be able to dissipate eg. 5W. The zener in that position passes high current pulses, so estimating rms power dissipated in the zener is harder, and the ability of the zener to dissipate that power is not as good as a 'continuous' form of power dissipation in the device.

[klooon]

Your zener direction is correct. Be mindfull that it is almost impossible for a 10W zener to dissipate 10W unless you have amazing heatsinking to 25C ambient. You will certainly need good heatsinking, and even then may only safely be able to dissipate eg. 5W. The zener in that position passes high current pulses, so estimating rms power dissipated in the zener is harder, and the ability of the zener to dissipate that power is not as good as a 'continuous' form of power dissipation in the device.

Thanks, maybe I should apply a larger safety factor then! My idea is to have them up in the air, without further heatsinking.

I assume the B+ current draw to be around 100mA, then with a 10V zener it dissipates 1W. Should I go to safer values, e.g. 5.6V 5Watt zeners, if I want them up in the air? Just like here: link...

[trobbins]

PSUD2 will help you identify the peak current and rms current in a zener - use an RC first stage filter and make R very low and use as the sensing part for current through the zener.

Look through the zener datasheet. It may have some guiding graphs - the aim should be to keep Tj below 110-120C, even though it may have a max spec of say 175C. Having the zener in free air with a longish lead, will give a high thermal resistance. Mounting on a tag board certainly is much better than soldering the leads in series in free air.

Try and gather the data above and do a thermal calculation as to what Tj may get up to - you will need to estimate the maximum ambient air temp - which could be quite high if under a chassis near to other hot parts, in a head unit, butted against a wall in a heated room.