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

[klooon]

Time for a new build! The plan is to build a Magnatone Troubadour 213 / ToneMaster 214 / Titano Custom 314 amp. Actually two, for a friend and myself

I will report here my progress. Like previous builds your comments and advice are very welcome again!!

Inspired again by several builders showing their work on the net, e.g. some great stuff from Kevin O'Reilly found here.

I will use this schematic as a base:

found here

I would like to include two additional options:
1) Add next to the pitch shift vibrato, a nice bias tremolo.
2) Power scaling.

First question: regarding the addition of the (bias) tremolo, would it be possible to somehow bypass the vibrato (varistor) part and shortcut the LFO part in the 'rest' of the circuit directly????

[trobbins]

You could introduce a spdt switch to direct the LFO depth output to either the vibrato circuit, or to say channel B cathode for tremolo. You will need to manage LFO beating from the tremolo, either by RC filtering the LFO going to the channel B cathode, and/or CR filtering the anode output or lowering the coupling cap, and/or lowering some of the output stage coupling/bypass caps.

With no LFO signal in to the vibrato circuit then that circuitry just acts as a fairly benign amp stage, and I suggest not worth trying to bypass in any way, but do short-circuit that input (as if the depth pot was min).

Do you have NOS maggie varistors, or are you going to try a cloned varistor using zener-R or MOV-R strings (of which I can vouch are pretty damn good).

Ciao, Tim

[klooon]

You could introduce a spdt switch to direct the LFO depth output to either the vibrato circuit, or to say channel B cathode for tremolo. You will need to manage LFO beating from the tremolo, either by RC filtering the LFO going to the channel B cathode, and/or CR filtering the anode output or lowering the coupling cap, and/or lowering some of the output stage coupling/bypass caps.

With no LFO signal in to the vibrato circuit then that circuitry just acts as a fairly benign amp stage, and I suggest not worth trying to bypass in any way, but do short-circuit that input (as if the depth pot was min).

Thanks, Tim! That switching looks like a fine idea. For the tremolo part, then it's similar to e.g. a Vibro-Champ I assume, which is a fine sounding tremolo. Is it also possible to insert it to the cathode of the triode just after the tone control? Otherwise, it's only working for Channel B, I assume?

Do you have NOS maggie varistors, or are you going to try a cloned varistor using zener-R or MOV-R strings (of which I can vouch are pretty damn good).

I have them on order from eBay (warbler_muse), but I just read the progress on the ampage proposing alternatives. Is this the alternative that is now proposed (from here):
. Right? I will try to order the parts for comparison!

[trobbins]

The tone triode may be as good - it will require some experimentation, as the residual thumping will determine how much filtering you need to apply, and that may then get more difficult as you move the modulation point closer to the output. The following vibrato and output stages may be enough, or enough with a bit of LFO extra RC filtering.

Yes Martin is pushing the MOV-R path on that thread, although it may be ok with just the 3x zener-R scheme as the difference is likely to be benign - it depends how long a piece of string is - if you've got time and resources (and no NOS maggie varistors) then the 33 zener-R scheme gave pretty damn good results.

[klooon]

The tone triode may be as good - it will require some experimentation, as the residual thumping will determine how much filtering you need to apply, and that may then get more difficult as you move the modulation point closer to the output. The following vibrato and output stages may be enough, or enough with a bit of LFO extra RC filtering.

Thanks. Do you mean something like this? (I will probably get rid of one channel anyway...)


If so, what (range of) values should I try for R and C for the filtering? And, why exactly do mean by 'residual' thumping?

Yes Martin is pushing the MOV-R path on that thread, although it may be ok with just the 3x zener-R scheme as the difference is likely to be benign - it depends how long a piece of string is - if you've got time and resources (and no NOS maggie varistors) then the 33 zener-R scheme gave pretty damn good results.

I will try to get these two variants to compare and see if they work fine as well...

[trobbins]

Residual thumping is the LFO and harmonic signals added to the instrument signal and amplified and sent to speaker - the speaker cone noticeably vibrates like a subwoofer cone, and depending on the speaker and enclosure then you may 'hear' that infrasonic signal.

The tremolo technique needs the LFO fundamental signal passed to the modulation point (ie. cathode resistor), but doesn't need any harmonics, so low-pass RC filtering is added to attenuate frequencies above say 15Hz filter corner frequency (ie. the harmonics of the LFO signal). Sometimes you will notice multiple RC networks laddered one after the other (as per the vibrato circuit). The RC filter needs to couple with the cathode resistor such that the C is not directly across the cathode resistor - so add a buffer series resistance (which acts as a voltage divider with the cathode resistance), which would then be used to 'normalise' the tremolo versus the vibrato effects for the same depth pot setting.

Once the LFO has modulated the input signal, then the wet signal passes through subsequent stages to the speaker - those subsequent stages usually have high-pass CR filtering inherent in the circuit design (eg. coupling cap and the following grid-leak), and should provide good attenuation of any LFO and residual harmonics. Some amps have coupling circuits that achieve excellent bass response, but that is not what is wanted here - so aim for RC corner frequencies of at least about 80Hz, as a compromise with lowest guitar note.

[klooon]

Thanks a lot for this great info!!

The RC filter needs to couple with the cathode resistor such that the C is not directly across the cathode resistor - so add a buffer series resistance (which acts as a voltage divider with the cathode resistance), which would then be used to 'normalise' the tremolo versus the vibrato effects for the same depth pot setting.

How do you mean this exactly? Something like this:


or like this:

[trobbins]

The second schematic would be my recommendation :-)

However, as that circuit is prior to the modulation point, the RC low-pass filter corner frequency should be down towards 15Hz (ie. something like the max LFO frequency aimed for), as you are wanting to attenuate harmonics of the LFO that extend higher than circa 15Hz.

[klooon]

The second schematic would be my recommendation :-)

However, as that circuit is prior to the modulation point, the RC low-pass filter corner frequency should be down towards 15Hz (ie. something like the max LFO frequency aimed for), as you are wanting to attenuate harmonics of the LFO that extend higher than circa 15Hz.

Great!

I adapted the schematic. Should it be like this? The corner frequency is now about 15Hz.
[EDIT: I did not take into account the 47k before, now the combination of 180k +47k and 0.047uF gives about 15Hz]


Further, what values should I take/stock for the R_buffer?

[trobbins]

The RC filter on the output of the LFO also includes the depth pot, so perhaps better to use a higher R and smaller C to get the corner frequency, so as not to lower the corner frequency hugely when depth pot is lowered away from max. If the depth pot is too aggressive, and drops the modulation too quickly away from max, then raising that RC corner frequency may help. I think the 47k avoids loading the LFO too much when the footswitch is on, as that could stop the oscillator.

R_buffer is 'SOT' (ie. select on test), as the aim would be to have the same 'intensity' from each effect when toggling between vibrato and tremolo. If you have an oscilloscope to look at the tremolo modulation depth, then note that the maggie vibrato includes a noticeable level of tremolo as well which will show up on a slow sweep oscilloscope plot. But its your ear that would be the best tool to use.

[klooon]

The RC filter on the output of the LFO also includes the depth pot, so perhaps better to use a higher R and smaller C to get the corner frequency, so as not to lower the corner frequency hugely when depth pot is lowered away from max. If the depth pot is too aggressive, and drops the modulation too quickly away from max, then raising that RC corner frequency may help. I think the 47k avoids loading the LFO too much when the footswitch is on, as that could stop the oscillator.

R_buffer is 'SOT' (ie. select on test), as the aim would be to have the same 'intensity' from each effect when toggling between vibrato and tremolo. If you have an oscilloscope to look at the tremolo modulation depth, then note that the maggie vibrato includes a noticeable level of tremolo as well which will show up on a slow sweep oscilloscope plot. But its your ear that would be the best tool to use.

Thanks for pointing that out! I could use a 2.2M (+47k) with a .0047uF to get to around 15Hz. Then with the depth pot lowered away from max, it could rise to a maximum of 3.25M and then the corner frequency is still around 10Hz. I could go to a .0022uF, but then the R should be even higher and with the R_buffer that I expect to be in the same order of magnitude (?), it's better to stock lower value resitors... So, it should be fine - I think...

Further, I would like to apply some kind of power scaling to bring the volume of the amp a bit down. I am intending to use a Hammond Deluxe Reverb power transformer, which is possibly giving to much B+, so I would like to bring it down with this power scaling as well, setting a maximum value for the B+.

I have not done this before, but would this be the idea??

[trobbins]

As you indicate, aim for resistors you have for starters, and perhaps RC resistors up to about 1-2Meg value. You may need to alter RC values a few times, so no real need to be pedantic at this stage.

As it is a clone, then perhaps add some future proofing protection items like a PT secondary CT fuse, and some output valve cathode current sense resistors. I think you have a typo with the output valve screen resistor values.

Power scaling all the preamp rails may get you in to some problems, especially with the vibrato circuit and LFO - perhaps start off with power scaling the output and PI stages, and have a diode steered and resistor dropped and loaded feed for the preamps and LFO. You will also need a bleed on the VVR output, as a backup.

[klooon]

As you indicate, aim for resistors you have for starters, and perhaps RC resistors up to about 1-2Meg value. You may need to alter RC values a few times, so no real need to be pedantic at this stage.

Thanks, I will stock various values to try out!

As it is a clone, then perhaps add some future proofing protection items like a PT secondary CT fuse, and some output valve cathode current sense resistors. I think you have a typo with the output valve screen resistor values.

- PT Secondary CT fuse, sounds like a good idea. I will add one.
- Biasing, I will do that by taking the resistance readings from the OT with respect to the CT and calculate back the currents.[EDIT: this is for fixed bias of course...In this case, cathode bias, I take the average cathode current and with the plate voltage minus the cathode voltage that gives the plate dissipation... good enough for me. I don't like the extra resistors there so much...].
- For the output valve screen resistors I took the values on the schematic I found. Better to take the 470R (2Watt) or so values for the screens that I am used to in my Fender builds?! 4.7k's for the grids are fine I believe?
- Also I see a 220R 20W resistor appearing on the speaker jack as well. This is possibly for protection in case there is no speaker load?

Power scaling all the preamp rails may get you in to some problems, especially with the vibrato circuit and LFO - perhaps start off with power scaling the output and PI stages, and have a diode steered and resistor dropped and loaded feed for the preamps and LFO. You will also need a bleed on the VVR output, as a backup.

OK, I have to work it out further then...

[trobbins]

Adding something like a 1R or 10R resistor in series with each 6V6 cathode is the standard method for measuring idle bias current, and for very good reason, as it avoids measuring across points at high DC voltage, and does not require specific info to be known of the OT primary DCRs. I often take five wires to easily accessible terminals (eg. an octal socket on a rear panel) to allow simple safe measuring after the amp chassis is installed in its cabinet (0V, common cathode node, 6V6 cathode 1, 6V6 cathode 2, 100:1 divided HT).

Your schematic shows 750k screen stoppers - yes perhaps start with 470R 2-5W.

There are a few ways to protect the output transformer form un-loaded condition - Output transformer protection

[klooon]

Adding something like a 1R or 10R resistor in series with each 6V6 cathode is the standard method for measuring idle bias current, and for very good reason, as it avoids measuring across points at high DC voltage, and does not require specific info to be known of the OT primary DCRs. I often take five wires to easily accessible terminals (eg. an octal socket on a rear panel) to allow simple safe measuring after the amp chassis is installed in its cabinet (0V, common cathode node, 6V6 cathode 1, 6V6 cathode 2, 100:1 divided HT).

That makes good sense! I will add them. I didn't do this in previous builds but it's indeed much safer.

Your schematic shows 750k screen stoppers - yes perhaps start with 470R 2-5W.

I took a schematic which I didn't study too well before posting as 'my reference' (e.g. it used an LFO circuit with a LED, possibly needed some different power supply values, etc, etc.). In the meantime I have been comparing a lot of 213/214/314 schematics and have come to the following version, which I think I should be fine...?



I have added the 470R screen stoppers (just as I did with my previous Princeton build), also the grid stopper on the cathodyne PI.
Further, I will not do the VVR as it will possibly mess up to much (with the sound) and requires too much experimenting as well. First have the basic circuit right...