MkIIC+ inspired preamp.. DIYLC layout drafting

[owlexifry]

full credit to the author of the schematic and build thread that inspired me to consider the build and draw a DIYLC layout.
(schematic based on lead channel)

http://diy-fever.com/amps/mesa-mark-iic-preamp/

my first attempt at using DIYLC. (took me all day).

i would be so grateful if anyone could review my layout against the schematics provided above below and below.

- planning to use hammond 369BX PT.
- adjusted the filter cap values (from the schematic above)
- a few other differences taking into account schematic below.
- not sure what to do with centre taps.. (or if i even need them)
- unsure if the ground for the bridge rectifier should go elsewhere..

would appreciate any feedback
thanks!



(other schematic)

 

[owlexifry]

additional images (it didnt look like the resolution uploaded very well in the image above)

 

[Lowerleftcoast]

The 50V tap is unused
The high voltage CT is unused

You need R, a 220k resistor to complete the RC filter with the C 0.022uF coupling cap. (One schematic shows a MV in this spot. The MV is the R of the RC filter. The other schematic does not show R. Depending on what this plugs into, this could be important.) See below:



The negative side of the Bridge Rectifier should be as close as practicable to the neg of the reservoir cap. You may be close enough.

I traced most of the layout, but not all. I will have a look at the rest later.

 

[Lowerleftcoast]

OK, I went through the layout.

I color coded the filter nodes.

The resistor I mentioned above will terminate with the purple node. To have separate nodes, there must be dropping resistors. I would suggest no less than 3.3k for excellent filtering.

There was a missing ground from the drive pot (noted in blue).

The *out* jack should be isolated from the chassis. (IDK if that is what you had intended yet.)

Gold star for keeping the nodes together.

 

[printer2]

Looks like an error in it, do not care all that much to check for others. the 3.3M resistor and cap. Might make more sense reading the article.

 

[owlexifry]

The 50V tap is unused
The high voltage CT is unused

You need R, a 220k resistor to complete the RC filter with the C 0.022uF coupling cap. (One schematic shows a MV in this spot. The MV is the R of the RC filter. The other schematic does not show R. Depending on what this plugs into, this could be important.)

Thanks so much for taking the time to look at all this!

the 'original' schematic (supposedly from M/B) has x2 'master volume' controls, the 'lead master' (250K), and then 'master' (100K).
the 'diy schematic' by Bancika (diy-fever.com), that I've followed primarily for this build/design, includes the 'lead master' in the same location. the 'master' is absent in his design...... perhaps it isn't really needed?

he mentions that the final stage in his design is intended to be an ac coupled cathode follower:

"Voltage divider between 5th and 6th stage reduces signal level to just few volts to make it effect friendly.
Since I’m building the preamp only I don’t need the last stage to boost the signal back.
Instead, I converted the last stage to AC coupled cathode follower.
It’s supposed to be a transparent buffer that will provide nice low impedance output that should drive any effect and long cables if needed.
Since it doesn’t cut any bass, it will essentially have similar response like the original stage with deep pulled out".

OK, so that output coupling cap (22nF) needs a resistor to ground?
i would assume that would shunt signal to ground? (and reduce output volume?)
220K? any reason for that value? (just for my understanding, cheers)
what if I used a larger value (eg. 470K, 1M), for more output volume?

i wonder why he didn't include an R there? (quite puzzled about this one).

OK, I went through the layout.

I color coded the filter nodes.

The resistor I mentioned above will terminate with the purple node. To have separate nodes, there must be dropping resistors. I would suggest no less than 3.3k for excellent filtering.

There was a missing ground from the drive pot (noted in blue).

The *out* jack should be isolated from the chassis. (IDK if that is what you had intended yet.)

Gold star for keeping the nodes together.

View attachment 921571

thanks again! changes made as suggested.



quite right about the dropping resistors. i completely forgot about the last two...
the 'diy schematic' called for 1K dropping resistors, but I have put in 3.3K as per your recommendation.
(planning to run these under the board, depending on space)

with regards to the target B+4/3 voltages, I would be concerned that 3.3K might end up being too high, and to get the voltages closer to target, I might have to lower these dropping resistor values, perhaps down to 1K or somewhere between... would this be an issue?
alternatively, i guess i could first try lowering the initial 10K dropping resistor to a lower value and see how that goes...

perhaps a cliff jack would be better for the output? where should this be grounded? or is the sleeve/ground on this jack supposed to be completely unhooked/isolated?

ill be honest, im not sure how i earned the gold star, this concept of grounding 'nodes' you presented here isn't something ive seen before, i'm still a newbie....
really appreciate the feedback.

 

[owlexifry]

Looks like an error in it, do not care all that much to check for others. the 3.3M resistor and cap. Might make more sense reading the article.

thanks for looking!
from what i can tell, appears the 3.3M resistor and cap are there for a reason..
looks like it allows the signal to bypass the x2 'lead stages', and then this 'clean signal' mixes back in after it skips those x2 stages.

 

[printer2]

thanks for looking!
from what i can tell, appears the 3.3M resistor and cap are there for a reason..
looks like it allows the signal to bypass the x2 'lead stages', and then this 'clean signal' mixes back in after it skips those x2 stages.
View attachment 921678

View attachment 921676

Makes more sense when you know there was two A's that were connected. And is there a switch to bypass the stages or is the stages bypassed all the time? I don't see any. Seems the schematic is missing some information.
.

 

[owlexifry]

Makes more sense when you know there was two A's that were connected. And is there a switch to bypass the stages or is the stages bypassed all the time? I don't see any. Seems the schematic is missing some information.
.

certainly went over my head when i first looked at it a few months ago..
the intention for this particular design is to have this connected permanently. ie. lead channel engaged permanently / lead stages not bypassed.

if you read the article in the link, the schematic should make more sense:
http://diy-fever.com/amps/mesa-mark-iic-preamp/


the author of the 'diy schematic' above, has also designed a subsequent build where he did make it switchable between 'rhythm' and 'lead' modes.
http://diy-fever.com/amps/mesa-mark-iic-pcb/

(switching to rhythm mode (lead stages bypassed) doesn't interest me. i only want the lead channel)

 

[owlexifry]

did some more. hopefully i've got the primary leads arranged the right way...

 

[Lowerleftcoast]

the 'diy schematic' by Bancika (diy-fever.com), that I've followed primarily for this build/design, includes the 'lead master' in the same location. the 'master' is absent in his design...... perhaps it isn't really needed?

Designing amplifiers is all about choices. Nothing can be patented because most everything that can be done has been done. It is too basic. It is like making a pizza, making it a different shape, trying different toppings, changing the type of flour, anything that can possibly be changed can't be patented. The choice to use a MV is just a choice. Neither MV is *needed* but you may choose to have one or both.

he mentions that the final stage in his design is intended to be an ac coupled cathode follower

A CF will provide low impedance. One of the reasons this is wanted is less noise will be picked up. It is very handy when placing before a tone circuit, efx loop, or output to the next stage. Iirc, there was a EQ tone circuit after this CF in the original design. As mentioned, it is good to have for efx etc so this is an expected addition to this design. This preamp can be plugged into anything, and it will be able to do so.

OK, so that output coupling cap (22nF) needs a resistor to ground?
i would assume that would shunt signal to ground? (and reduce output volume?)
220K? any reason for that value? (just for my understanding, cheers)
what if I used a larger value (eg. 470K, 1M), for more output volume? i wonder why he didn't include an R there? (quite puzzled about this one).

You are at a point where some education is wanted. Internet search for *RC filter calculator* (it will save you from doing the math). Every coupling cap is part of an RC filter. Every power filter cap is as well. If you know the frequencies of the guitar signal (and 50Hz VAC wall voltage), you will be able to manipulate the cutoff frequencies with the RC filter to reduce unwanted frequencies.

So, if the .022uF cap at the output of this preamp did not have a resistor. The input of the next thing plugged in would complete the RC filter. We don't know what that is. Let's imagine the next device has a coupling cap (high pass RC filter) as the first thing the signal encounters. The two caps would be in series. Caps in series reduce in value. The resulting RC filter would change the cutoff frequency from what was intended. You may have less bass than desired/designed. Completing the RC filter at the output of this preamp will keep that from happening.

Yes an RC filter shunts signal to ground. It gets rid of volume of selected frequencies. Not all frequencies.
The R of 220k with the C of .022F = a first order high pass cutoff frequency of 32Hz. The signal will be 3dB down at 32Hz and frequencies below 32Hz will reduce by 6dB per octave. Lows are being reduced but the highs pass to the next part of the circuit. Try different R and C values with the calculator. (Low E is 82Hz.) (low values of R will lose very high frequencies, so choice here is important depending on what you want. You may want vey high frequencies gone, or maybe not.)
He may have forgotten the R or most likely he assumed the next device would complete the RC filter with 1M to ground. Who knows? The choice of 220k. If the next device had a 1M resistor to ground as the first thing the signal *sees* the 220k in parallel with 1M = 180k. The RC filter 180k with .022uF = a cutoff frequency of 40Hz. Come to think of it, I would suggest a higher value for R and C because some amplifiers have around 40k as the first resistance to ground the signal may see... so how about 1M and .1uF? hah. As you can envision the next device plays a role here so choose this RC filter accordingly.

with regards to the target B+4/3 voltages, I would be concerned that 3.3K might end up being too high, and to get the voltages closer to target, I might have to lower these dropping resistor values, perhaps down to 1K or somewhere between... would this be an issue?

Again we are looking at RC filters. This time these are low pass filters. We want to get rid of the rectified 50Hz waveform so 100Hz is the ripple current buzz. The 100Hz is attenuated at 6dB per octave. 50Hz is one octave down. 25 Hz is the next octave down. 12.5Hz is next and it continues halving on the way down. R of 1.1k and C of 22uF = a cutoff frequency of 6.6Hz. So roughly a 24dB drop of the 100Hz hum/buzz which may be enough. I suggested 3.3k for a cutoff frequency of 2.2Hz so a drop of over 30dB. The multiple stages will help keep this preamp quiet, so 24dB drop may be enough. As you can see it is all choices/tradeoffs but you want to calculate these things so you know what you are doing. Guessing the vales will bite you.

cliff jack would be better for the output? where should this be grounded? or is the sleeve/ground on this jack supposed to be completely unhooked/isolated?

I would ground it with the purple section of the layout I showed above.

ill be honest, im not sure how i earned the gold star, this concept of grounding 'nodes' you presented here isn't something ive seen before, i'm still a newbie

Read through the *Valve Wizard Grounding* Chapter 15 a hundred times or so and you will know why you received the gold star. hah.

 

[owlexifry]

Designing amplifiers is all about choices. Nothing can be patented because most everything that can be done has been done. It is too basic. It is like making a pizza, making it a different shape, trying different toppings, changing the type of flour, anything that can possibly be changed can't be patented. The choice to use a MV is just a choice. Neither MV is *needed* but you may choose to have one or both.

A CF will provide low impedance. One of the reasons this is wanted is less noise will be picked up. It is very handy when placing before a tone circuit, efx loop, or output to the next stage. Iirc, there was a EQ tone circuit after this CF in the original design. As mentioned, it is good to have for efx etc so this is an expected addition to this design. This preamp can be plugged into anything, and it will be able to do so.

You are at a point where some education is wanted. Internet search for *RC filter calculator* (it will save you from doing the math). Every coupling cap is part of an RC filter. Every power filter cap is as well. If you know the frequencies of the guitar signal (and 50Hz VAC wall voltage), you will be able to manipulate the cutoff frequencies with the RC filter to reduce unwanted frequencies.

So, if the .022uF cap at the output of this preamp did not have a resistor. The input of the next thing plugged in would complete the RC filter. We don't know what that is. Let's imagine the next device has a coupling cap (high pass RC filter) as the first thing the signal encounters. The two caps would be in series. Caps in series reduce in value. The resulting RC filter would change the cutoff frequency from what was intended. You may have less bass than desired/designed. Completing the RC filter at the output of this preamp will keep that from happening.

Yes an RC filter shunts signal to ground. It gets rid of volume of selected frequencies. Not all frequencies.
The R of 220k with the C of .022F = a first order high pass cutoff frequency of 32Hz. The signal will be 3dB down at 32Hz and frequencies below 32Hz will reduce by 6dB per octave. Lows are being reduced but the highs pass to the next part of the circuit. Try different R and C values with the calculator. (Low E is 82Hz.) (low values of R will lose very high frequencies, so choice here is important depending on what you want. You may want vey high frequencies gone, or maybe not.)
He may have forgotten the R or most likely he assumed the next device would complete the RC filter with 1M to ground. Who knows? The choice of 220k. If the next device had a 1M resistor to ground as the first thing the signal *sees* the 220k in parallel with 1M = 180k. The RC filter 180k with .022uF = a cutoff frequency of 40Hz. Come to think of it, I would suggest a higher value for R and C because some amplifiers have around 40k as the first resistance to ground the signal may see... so how about 1M and .1uF? hah. As you can envision the next device plays a role here so choose this RC filter accordingly.

Again we are looking at RC filters. This time these are low pass filters. We want to get rid of the rectified 50Hz waveform so 100Hz is the ripple current buzz. The 100Hz is attenuated at 6dB per octave. 50Hz is one octave down. 25 Hz is the next octave down. 12.5Hz is next and it continues halving on the way down. R of 1.1k and C of 22uF = a cutoff frequency of 6.6Hz. So roughly a 24dB drop of the 100Hz hum/buzz which may be enough. I suggested 3.3k for a cutoff frequency of 2.2Hz so a drop of over 30dB. The multiple stages will help keep this preamp quiet, so 24dB drop may be enough. As you can see it is all choices/tradeoffs but you want to calculate these things so you know what you are doing. Guessing the vales will bite you.

I would ground it with the purple section of the layout I showed above.

Read through the *Valve Wizard Grounding* Chapter 15 a hundred times or so and you will know why you received the gold star. hah.

many thanks for answering all my questions so thoroughly. you explained it very well.

decided to go with 0.1uF/1M for that output combo.

updated layout (changed input jack to switching type)



anything else here that needs attention?

 

[owlexifry]

parts arrived.

started drilling the chassis.

not real sure how i should mount the PT....

(on my previous builds the PT position was already given, and decision was all about OT and choke position with the headphone method)..

i was hoping someone here can confirm how the bells / iron laminations should be positioned with respect to the tubes and other parts of the circuit (eg. output jack)...

it's quite small, ~2.5"x2.5" mounting base. upright-style PT. (chassis is 13.5"x5")

should the laminations be parallel or 90' to the preamp tubes/ board?
closer to front panel or back panel?

some reading tells me the coils (bells) should be 90’ to the tubes and as far as possible.

i would then assume ‘C’ would be best placement.
correct?

A

B

C

D

 

[dan40]

Pic "C" is the way I usually go. I like for my wire bundles to enter the amp with one set facing the controls and the other set faces the tubes. It's up to you which way you want them to enter. Try to think about where your wires need to go and what is the shortest path for them to take. Mount your OT with the lams facing the opposite direction. The headphone trick is nice but I have found that once I get at least 4" of separation, the level of hum drops off considerably.

 

[owlexifry]

Pic "C" is the way I usually go. I like for my wire bundles to enter the amp with one set facing the controls and the other set faces the tubes. It's up to you which way you want them to enter. Try to think about where your wires need to go and what is the shortest path for them to take. Mount your OT with the lams facing the opposite direction. The headphone trick is nice but I have found that once I get at least 4" of separation, the level of hum drops off considerably.

thanks @dan40 - appreciate your input

please note this is a preamp build only, so no OT, just PT.

with the PT mounted as per "C", it looks like the primary leads will end up close to power switch, with the secondaries opposite side and close to the bridge rectifier, so that's a win.

only concern is the PT 'shield' wire would end running across to the back panel side, to the grounding lug shared with the mains earth..

I always assumed it was logical to share the PT shield wire with the earth mains (since the PT shield carries essentially no current).

does it really matter? could just ground these separately? (with shield wire grounded to PT mounting bolt instead?)

 

[dan40]

I don't see why that wouldn't work fine for the shield wire. If it helps to clean up the layout and prevent long runs of wire, I would go for it.

 

[owlexifry]

I don't see why that wouldn't work fine for the shield wire. If it helps to clean up the layout and prevent long runs of wire, I would go for it.

for sure, agreed. thanks again.

 

[andrewRneumann]

Come to think of it, I would suggest a higher value for R and C because some amplifiers have around 40k as the first resistance to ground the signal may see... so how about 1M and .1uF? hah. As you can envision the next device plays a role here so choose this RC filter accordingly.

Did you decide to muck around with output coupling or just leave it as designed? IMHO, I think you’re fine without a resistor.

A small 10R resistor on the ground lug of the output jack might be enough to kill any ground loop hum (through the other device) you encounter. Or a switch that lifts the ground going to the output jack completely.

 

[owlexifry]

Did you decide to muck around with output coupling or just leave it as designed? IMHO, I think you’re fine without a resistor.

A small 10R resistor on the ground lug of the output jack might be enough to kill any ground loop hum (through the other device) you encounter. Or a switch that lifts the ground going to the output jack completely.

yep, given LLCs advice, I decided to change it to 0.1uF/1M.



that's an interesting idea - so you're suggesting that inserting a 10R in between output jack ground lug and ground connection (back to the node) will prevent ground loop hum?
I guess if i have any issues, that would be fairly easy to 'retrofit' when it's finished.
thanks for the tip!

 

[andrewRneumann]

that's an interesting idea - so you're suggesting that inserting a 10R in between speaker ground lug and ground connection (back to the node) will prevent ground loop hum?

That’s the idea. Just a little impedance in a near zero impedance circuit should be enough to break the loop.

I was reading an earlier post in this thread asking about MV. The 250KA pot before the AC coupled cathode follower is your MV. The cathode follower is designed to be a transparent stage that provides low output impedance. If you put a pot after the CF, you muck up the low output impedance and that was the whole point of the CF.