JW
Well-known member
Hey folks,
This is concerning squeezing more gain out of channel strips that have been racked up and wired to go directly out of C37, the output cap to the 1:1 transformer.
I came across this interesting post on the gearslutz forum from someone named mikeboyvcr, that is an indepth look at the various gain stages. I pm'd him, but I don't think he frequents that forum much, and neither do I really.
Anyway, I thought it might interest some folks here as there has been a little bit of talk about whether a single channel of this amp can drive a 600 ohm transformer well. I have a couple questions of my own that follow his post that maybe you all can help with:
From mikeboyvcr:
"PM-1000 channel strip gain structure, as I understand it, and change it…
There may be some errors in this, and I’m SURE someone will have issues with my calculations, and changes, AND being human I may have made a few mistakes…
Input transformer:
The transformers is 300 ohm to 2.7 K. Therefore… 2700/300=9 so 1 to 9 is the impedance ratio. The square root of the impedance ratio, or 9, is the turns ratio of the transformer, that’s 3, so the voltage gain of the transformer is 3 and a voltage gain of three = + 9.542425 DB.
I have not taken the time to calculate the exact insertion loss of this transformer, but -1 to -1.5DB would be my guess. So for this exercise we shall give the transformer a total voltage gain of 8 DB.
First stage amplifier:
There are two main amplifier sections in each channel, along with the twq echo sends. I’ll be talking about the main amplifiers not the Echo’s.
The gain, (which is set by the amount of negative feedback), is from 21 to 33 DB in main amplifier sections. Yamaha includes the transformer voltage gain in the calculation for the first stage gain, they call it 40 DB. I figure the amp is set at about 32 DB gain + the 8DB from the transformer and you get 40 DB.
No one is most likely interested in the padding, bypass and blending of the input with the out put of the first stage amplifier, however it is understandable, if, you want to take the time and take one switch at a time. For now let’s just say it works.
However, let’s take a look at the settings at -38 DB and up on the input selector switch.
With the input gain selector set at -38 the first stage output section still has R29, 30, & 31 for a total of… 10K from the output to ground, or a 10K load. That should pose no level reduction to the output of section one, as it is a very low z output.
At the input switch your straight to the transformer…at -38 to -60.
The only thing that controls the gain of the first stage from -38 to -60 is the feed back circuit. At -38 you have R19 at 330K in parallel with R21 at 5.6K, that works out to 5.5066K, or 5.5K.
-38 uses 5.5k in the feed back loop.
For another 6DB of gain switch to -44. At that setting R19, 330K & R22, 12K =11.5K.
For another 6DB of gain switch to -50. at that setting 330K & R23, 27K = 25K.
For another 10DB of gain switch to -60. At that setting 330K & R24, 100K =76.7K
So as the feed back resistance goes up the feedback goes down and the gain goes up…
Ok… I know most of you already new this feedback thing and a lot more, but some may not and that’s why I went through it.
From -60 to -38 the maximum output of this amplifier section is limited to -20DB by the input switching system. With the input switch set to -32 you can get up to -14 out. With it set at -26 you get -8 out. At -20 you’re at unity, zero gain. Everything above that gets padded down, and or, bypasses the first stage completely or in part.
So a maximum of 40 DB gain from the input to the EQ, and a maximum of -20 to the EQ.
The EQ is unity. According to Yamaha the gain of Q5 is consumed by the load of the EQ section which phase adds or cancels at a certain frequency based on the inductive capacitive filtering selected.
I’m not sure I agree with them on the gain thing. It may be that the circuit it self is totally unity because Q5 looks like a normal phase splitter to this guy. The gain of Q5 seams to just hold the positive phase above ground so that the center taped pots work as a voltage divider with the center tap to ground. When the boost or cut pot wiper is centered the inductive capacitive circuit is effectively grounded and the + & - phase are equally grounded through each half of the 50K pot. The reactance of the inductors and their series capacitors works out to around 1200 ohms for each of the three mid frequencies.
(At least that’s what I cam up with.)
You also have two resistors in each circuit and all that adds to the total load to ground of the circuit.
The EQ it self is a great design, but the frequency points are even harmonics of one another and basically trash the thing. To make it work right you need to change not only the caps but the inductors. Changing the caps makes it somewhat better but to me it’s not worth the time and money. Thing needs a total makeover!
Bypassing it does change the sound of the preamplifier.
Enough of that already….
Again these are my conclusions and I may be very wrong….
So we will go on to the channel fader, it cost the EQ section 6DB of gain but only because it’s on the emitter output of Q5. That’s the phase inverting high impedance end of the transistor. So if you keep the EQ and the fader you loose 6 DB.
So… first stage gain with transformer +40DB, Fader loss after the EQ is -6DB= +34 add booster amp at +21DB=56DB
That’s it you have a 56DB amplifier with a channel strip. Now add the output transformer at 600 to 600 with an insertion loss of 1DB and you end up with a preamplifier with 55 DB of voltage gain and 600 ohm output impedance.
55 db total!!!!!!!!!!!!
And… You will be running the preamp way beyond the sweet spot and it will sound weak… The higher the gain setting used the lower the headroom and the thinner the sound.
So if you add direct outs to you PM-1000 that’s what you’re going to get.
A weak sounding preamplifier.
Your typical microphone preamp has around 68 to 75DB of total gain so this thing will not make a very good preamp as it sits. As for direct outs…. Again, you will be running the preamp way beyond the sweet spot and it will sound week…
When building a microphone preamp out of a channel strip….
Here is what I do…
I scrap everything after C37, the last cap in the output booster amplifier; this reduces the load on the final amp.
I scrap the fader… That gives me an additional 6DB of gain. Then set the booster amp feedback loop for +33DB instead of the +21. That adds another 12DB for a total increase of 18DB.
55+18=73DB total
Now you have a preamp with 73DB total gain with just one level control….
Want more gain?
Scrap the phantom power section that is fed through the center tap of the input transformer. Rebuild the phantom power on the primary like it’s done on most modern gear. Make the center tap the + in and keep the – as is. Now you have 150 to 2.7k. 2700/150= 18 the square root of 18 is 4.2426 which = a voltage gain of 12.55 or 12.5, take away 1.5 db for insertion loss and you have a total of +11DB, which gives you another 3 db and brings your total to…76DB!!!!!!!!!!
Now it sound great… "
My questions:
I actually have 8 of them in a little mini mixer that I built, and am happy to hear some explanation about the various amplification stages.
I just had a couple questions to clarify. Maybe someone else can answer them as well.
My questions are:
1. What resistor/s do you change to alter the booster amp feedback? Does this effect the frequency response at all? (Big gains available here . . . )
2. Bypassing the fader should be super simple. (Just from the hi pass filter out, straight to C18?) But I don't understand why this would take additional load off of Q5 with the fader opened all the way up (no resistance?)
3. And getting rid of the phantom power input. This is interesting. Very easy to remove that resistor going to the center tap. And just verifying, but you're talking about taking the hot primary lead where it comes from the center of the input switch and then connecting that to the center tap?
Part of me is curious too if there is an interesting sweet spot that is attained with louder sources (with the input switch dialed down a bit) that makes people like these strips BECAUSE of the straining to drive a 600 ohm transformer? Ribbon mic on fingerpicked guitar is definitely not cool (thin) from my experience, because of the boost required, but a snare drum is super fat sounding as it is. Still, I'd like to get these into all around use territory.
Also, maybe someone can comment, but could it be a better idea to simply build out the additional amplification from the master channels and insert that before the transformer direct out? And if so, exactly which part of the master module?
I attached schematics for the master module as well as the input module (2 parts)
This is concerning squeezing more gain out of channel strips that have been racked up and wired to go directly out of C37, the output cap to the 1:1 transformer.
I came across this interesting post on the gearslutz forum from someone named mikeboyvcr, that is an indepth look at the various gain stages. I pm'd him, but I don't think he frequents that forum much, and neither do I really.
Anyway, I thought it might interest some folks here as there has been a little bit of talk about whether a single channel of this amp can drive a 600 ohm transformer well. I have a couple questions of my own that follow his post that maybe you all can help with:
From mikeboyvcr:
"PM-1000 channel strip gain structure, as I understand it, and change it…
There may be some errors in this, and I’m SURE someone will have issues with my calculations, and changes, AND being human I may have made a few mistakes…
Input transformer:
The transformers is 300 ohm to 2.7 K. Therefore… 2700/300=9 so 1 to 9 is the impedance ratio. The square root of the impedance ratio, or 9, is the turns ratio of the transformer, that’s 3, so the voltage gain of the transformer is 3 and a voltage gain of three = + 9.542425 DB.
I have not taken the time to calculate the exact insertion loss of this transformer, but -1 to -1.5DB would be my guess. So for this exercise we shall give the transformer a total voltage gain of 8 DB.
First stage amplifier:
There are two main amplifier sections in each channel, along with the twq echo sends. I’ll be talking about the main amplifiers not the Echo’s.
The gain, (which is set by the amount of negative feedback), is from 21 to 33 DB in main amplifier sections. Yamaha includes the transformer voltage gain in the calculation for the first stage gain, they call it 40 DB. I figure the amp is set at about 32 DB gain + the 8DB from the transformer and you get 40 DB.
No one is most likely interested in the padding, bypass and blending of the input with the out put of the first stage amplifier, however it is understandable, if, you want to take the time and take one switch at a time. For now let’s just say it works.
However, let’s take a look at the settings at -38 DB and up on the input selector switch.
With the input gain selector set at -38 the first stage output section still has R29, 30, & 31 for a total of… 10K from the output to ground, or a 10K load. That should pose no level reduction to the output of section one, as it is a very low z output.
At the input switch your straight to the transformer…at -38 to -60.
The only thing that controls the gain of the first stage from -38 to -60 is the feed back circuit. At -38 you have R19 at 330K in parallel with R21 at 5.6K, that works out to 5.5066K, or 5.5K.
-38 uses 5.5k in the feed back loop.
For another 6DB of gain switch to -44. At that setting R19, 330K & R22, 12K =11.5K.
For another 6DB of gain switch to -50. at that setting 330K & R23, 27K = 25K.
For another 10DB of gain switch to -60. At that setting 330K & R24, 100K =76.7K
So as the feed back resistance goes up the feedback goes down and the gain goes up…
Ok… I know most of you already new this feedback thing and a lot more, but some may not and that’s why I went through it.
From -60 to -38 the maximum output of this amplifier section is limited to -20DB by the input switching system. With the input switch set to -32 you can get up to -14 out. With it set at -26 you get -8 out. At -20 you’re at unity, zero gain. Everything above that gets padded down, and or, bypasses the first stage completely or in part.
So a maximum of 40 DB gain from the input to the EQ, and a maximum of -20 to the EQ.
The EQ is unity. According to Yamaha the gain of Q5 is consumed by the load of the EQ section which phase adds or cancels at a certain frequency based on the inductive capacitive filtering selected.
I’m not sure I agree with them on the gain thing. It may be that the circuit it self is totally unity because Q5 looks like a normal phase splitter to this guy. The gain of Q5 seams to just hold the positive phase above ground so that the center taped pots work as a voltage divider with the center tap to ground. When the boost or cut pot wiper is centered the inductive capacitive circuit is effectively grounded and the + & - phase are equally grounded through each half of the 50K pot. The reactance of the inductors and their series capacitors works out to around 1200 ohms for each of the three mid frequencies.
(At least that’s what I cam up with.)
You also have two resistors in each circuit and all that adds to the total load to ground of the circuit.
The EQ it self is a great design, but the frequency points are even harmonics of one another and basically trash the thing. To make it work right you need to change not only the caps but the inductors. Changing the caps makes it somewhat better but to me it’s not worth the time and money. Thing needs a total makeover!
Bypassing it does change the sound of the preamplifier.
Enough of that already….
Again these are my conclusions and I may be very wrong….
So we will go on to the channel fader, it cost the EQ section 6DB of gain but only because it’s on the emitter output of Q5. That’s the phase inverting high impedance end of the transistor. So if you keep the EQ and the fader you loose 6 DB.
So… first stage gain with transformer +40DB, Fader loss after the EQ is -6DB= +34 add booster amp at +21DB=56DB
That’s it you have a 56DB amplifier with a channel strip. Now add the output transformer at 600 to 600 with an insertion loss of 1DB and you end up with a preamplifier with 55 DB of voltage gain and 600 ohm output impedance.
55 db total!!!!!!!!!!!!
And… You will be running the preamp way beyond the sweet spot and it will sound weak… The higher the gain setting used the lower the headroom and the thinner the sound.
So if you add direct outs to you PM-1000 that’s what you’re going to get.
A weak sounding preamplifier.
Your typical microphone preamp has around 68 to 75DB of total gain so this thing will not make a very good preamp as it sits. As for direct outs…. Again, you will be running the preamp way beyond the sweet spot and it will sound week…
When building a microphone preamp out of a channel strip….
Here is what I do…
I scrap everything after C37, the last cap in the output booster amplifier; this reduces the load on the final amp.
I scrap the fader… That gives me an additional 6DB of gain. Then set the booster amp feedback loop for +33DB instead of the +21. That adds another 12DB for a total increase of 18DB.
55+18=73DB total
Now you have a preamp with 73DB total gain with just one level control….
Want more gain?
Scrap the phantom power section that is fed through the center tap of the input transformer. Rebuild the phantom power on the primary like it’s done on most modern gear. Make the center tap the + in and keep the – as is. Now you have 150 to 2.7k. 2700/150= 18 the square root of 18 is 4.2426 which = a voltage gain of 12.55 or 12.5, take away 1.5 db for insertion loss and you have a total of +11DB, which gives you another 3 db and brings your total to…76DB!!!!!!!!!!
Now it sound great… "
My questions:
I actually have 8 of them in a little mini mixer that I built, and am happy to hear some explanation about the various amplification stages.
I just had a couple questions to clarify. Maybe someone else can answer them as well.
My questions are:
1. What resistor/s do you change to alter the booster amp feedback? Does this effect the frequency response at all? (Big gains available here . . . )
2. Bypassing the fader should be super simple. (Just from the hi pass filter out, straight to C18?) But I don't understand why this would take additional load off of Q5 with the fader opened all the way up (no resistance?)
3. And getting rid of the phantom power input. This is interesting. Very easy to remove that resistor going to the center tap. And just verifying, but you're talking about taking the hot primary lead where it comes from the center of the input switch and then connecting that to the center tap?
Part of me is curious too if there is an interesting sweet spot that is attained with louder sources (with the input switch dialed down a bit) that makes people like these strips BECAUSE of the straining to drive a 600 ohm transformer? Ribbon mic on fingerpicked guitar is definitely not cool (thin) from my experience, because of the boost required, but a snare drum is super fat sounding as it is. Still, I'd like to get these into all around use territory.
Also, maybe someone can comment, but could it be a better idea to simply build out the additional amplification from the master channels and insert that before the transformer direct out? And if so, exactly which part of the master module?
I attached schematics for the master module as well as the input module (2 parts)