Question about the tc integrated pre / fortin grind design

[nonoxxx]

Hello,
I am actualy experimenting with the Tc integrated Pre / fortin circuit and theres is two things on this circuit that I don't understand.

If you look at this schematic : 


What is the purpose of the first Bc550 transtor on the left
and why V ref is not made with 2 resistors whith equal value ? it seem to be higher than 18/2

Thanks

[electrosonic]

I thought it provides signal limiting to prevent overdriving the opamp. I simulated it but it doesn't seem to do anything, so I'm not sure what it does.

Also, why all the series resistors? They could be combined easily with no sonic change.

ie R18,R12, and R13 could be replaced by a single resistor.

same for R6 and R16
also R2 and R15
and R11 and R4

Andrew.

[Fancy Lime]

There seem to be loads of things wrong with this schematic.

First of all: The component numbering is all over the place, which is rarely a good sign.

Q2, R9, R10, D1 and D2 do nothing that I can see. Looking at the schematic of the TC Electronic Integrated Preamplifier, it seems that the input is connected wrong. The lower end of R10 should go to the input instead of ground. And the input should not be connected to the drain of Q2.

The series resistors are indeed unnecessary.

The VRef bias network is almost certainly wrong. I would say R101 should be the same as R103 (33k to 100k), R102 and R1 should not be there at all.

The method of voltage trippling is unnecessarily complicated and inefficient. The same 27V can be had much simpler. See the Klon Centaur for how that's done. In fact the "Triangulum method" only gives you about 27-(6*0.7) = 22.8V. This is just really bad design.

The whole design as it is in the schematic is a bit strange, but so is the TC Preamp. There are easier ways to reach exactly the same result with less parts, less hassle and probably less noise. I could not find a schematic of the Fortin Grind.

All in all I would say, that this schematic is almost certainly complete bollocks. Either someone who had no idea what they were doing pulled a "design" from where the sun don't shine and just put it out there, hoping to sell a few PCBs, or someone who had no idea what they were doing traced an existing pedal and got everything wrong. Or, third choice, I'm stupid and this is the work of a genius, and I sincerely mean that this is also a realistic option, no sarcasm intended. But the unnecessary parts and series resistors are very suggestive of confusion tactics to throw off copycats. Which suggests that the effect does exist but the board was gooped and someone traced it badly. My guess is, this is supposed to be a clone of the Fortin Grind with a bunch of errors.

Finally, I could not unearth any evidence that anyone has ever successfully build this thing (only a brief google search but that still should have brought up at least one photo or even a demo video of a finished device). The only references to "Triangulum boost" are from the PCB manufacturer and sellers and quite some confused questions posted on forums. Not a good sign either.

TL;DR I wouldn't waste my time trying to build this. Sorry. There are better, simpler boosters out there. Like, lots and lots. You can design one from scratch that is better with minimum technical skills.

Cheers,
Andy

[Fancy Lime]

Well, I just found this thread:

https://www.diystompboxes.com/smfforum/index.php?topic=89712.0

It seems that the Fortin Grind / Fortin 33 is just a TC Integrated Preamp with the tone controls fixed. And the "Trinagulum" is a copy of that with loads of errors. If you compare the schematic of the Triangulum to the one of the TC posted by dano12 in the thread I linked, you can see what is wrong. Now this thing suddenly looks a lot more interesting. I might even give it a try. Although, as I said, with modern opamps there are easier ways to get the same result.

Andy

[ElectricDruid]

+1 what Lime said. That is not a nice schematic. The kind that makes you shudder and go "Urrrghh!".

It could be fixed, but it's easier to throw it away and start again from something decent.

T.

[teemuk]

Vref biases the common emitter amp, so it needs to be Something else than typical vcc/2 "half supply".

I believe the shunt transistor is somekind of mute circuit for the input when guitar coord is disconnected.

[nonoxxx]

Thanks, so I am not the only one disapointed with this schematic lol,

I built some boost with the chuggah-pre board and the design is better (except for the v ref that is also not the vc/2)

I am currently making my own pcb with some mods (based on the chuggah pre) , so for you guys adding another Vref  whith two equal resistor should be  better (on my design I also add an Overdrive stage with a blend at the end of the circuit) ?
I noticed on my proto board that adding another opamp to the current vref was bugging.

[antonis]

adding another Vref  whith two equal resistor should be  better (on my design I also add an Overdrive stage with a blend at the end of the circuit) ?
I noticed on my proto board that adding another opamp to the current vref was bugging.

A diagram of what we're are talking about should be helpfull..

[nonoxxx]

for sure,

Here is a picture of the schematic I am currently working on in eagle (it's not yet totally finished and there is surely some errors )

I added a Vref2 (VA) with 1/2 of VCC (18V)     R17=R18 = 100K   
for the Vref 1  R3 = 100K and R5 = 1M (stock from the original schematic)




I started from the schematic of the chuggah pre (I build severals that work really nicely)
https://www.rullywow.com/wp-content/uploads/2015/05/ChuggaPre-v1.0.pdf

[antonis]

>  I added a Vref2 (VA) with 1/2 of VCC (18V)     R17=R18 = 100K   <

That results in 3.86V (considering R16=10k) which sits far away from 1/2 V_CC
(with stock value resistors V_R stays at 15.32V - in case of my brain calculator isn't out of power supply..)

As you can realize, you alter enough swinging margins..

[nonoxxx]

I think I forgot to take consideration that it's not connected to ground but on -9V
so what will be the best voltage for Vref 2 ?

[Rob Strand]

What is the purpose of the first Bc550 transtor on the left

Transistors are usually put on the input of opamps to reduce noise.  For guitar/bass preamps the transistor needs to operate at a fairly low current.

Preamps like this have been used in on-board bass preamps.   Ibanez comes to mind.   Perhaps Peavey also.


EDIT:    My calculations here assume a single supply.  I realized that circuit has funky supply rails.   While the calculations aren't right for that circuit the idea of getting maximum swing and making the transistor operate correctly is still valid.

and why V ref is not made with 2 resistors whith equal value ? it seem to be higher than 18/2

Ultimately you tweak the Vref for maximum signal swing.  However in this circuit there is a need to make Vref higher than vcc/2 so the transistor operates correctly.

Suppose maximum swing occurs when the opamp sits at vcc/2.   Since the transistor has an emitter resistor, and emitter current, the emitter voltage must sit higher than vcc/2.   Because the opamp's input is connected to Vref the collector is sits at Vref. So the voltage drop across the collector resistor is (Vcc-Vref).  Since the emitter resistor is twice the collector resistor the emitter must sit at Ve = Vcc/2 + 2*(Vcc - Vref).  In order for the transistor to operate the collector must be higher than the emitter,
       Vc  > Ve
       Vref  >  Vcc/2 + 2*(Vcc - Vref)
So,     3Vref >  2.5 Vcc
or          Vref > (2.5/3) Vcc
or          Vref  > 0.83 Vcc

The other thing we need to be certain about is that a base current can flow when we account for the Vbe drop and the drop across the base resistor.   I'm not going through all the calculations.

The main point is Vref needs to be pretty high for the circuit to work; more specifically the transistor.

In bass preamps I've seen the 220k is much smaller like 10k which lets Vref be smaller.  However a small value like 10 might introduce stability issues - which would take a fair amount of work to check for sure.

EDIT: added a caveat, see above

[nonoxxx]

Thanks Rob,

So it's seem to be biased correctly,
Am I wrong to do a vref 2 on my pcb? or I should leave it as is and connect the two opamps input at the original Vref?

[Rob Strand]

Thanks Rob,

So it's seem to be biased correctly,
Am I wrong to do a vref 2 on my pcb? or I should leave it as is and connect the two opamps input at the original Vref?

It's OK to have one funky Vref1 for this part of the circuit and a standard Vref2=vcc/2 for other parts of the circuit.   It would be wise to have cap between the output of this circuit and the circuit using Vref/2.


You can see an example here of a circuit with two Vrefs,

http://guitarwork.ru/electronic/firm/Distortion/Ibanez%20Power%20Lead%20PL-5.gif

[nonoxxx]

I was assuming that c16 anc c 17 was doing this job  (drangon ball names lol )   

[Rob Strand]

So for this circuit,


The caps that isolate the DC in the circuit around IC2A from the DC in the circuit for IC2B are
C10 and C7.  With those in place the DC in the VA referenced circuit is separated from the DC in the VR referenced circuit.

I think the opamp should be running from -9V and +18V, yes?


Just spotted a problem, the bass circuit to ground is stuffing up the biasing on IC2B.

[nonoxxx]

scuse me I read to fast and I was thinking you were talking about isolating dc after the second opamp

So with C10 and C7 all is fine?

[Fancy Lime]

Ah, that's better! Both the ChuggaPre and your schematic make a *LOT* more sense than the Triangulum thing. And it does indeed use the same voltage boosting technique (-9/+18V) as the Klon Centaur. Electrosmash have, once again, an excellent article on the Centaur:
https://www.electrosmash.com/klon-centaur-analysis

One more thing that I would change, is the opamp. The TL072 is popular in charge pumped systems because it draws very little current (2.5mA maximum per opamp, so 5mA total if you use both halves) and has a high input impedance. But it has the problem of sounding extremely awful when pushed into clipping, which may happen in this booster circuit. The input impedance need not concern us, because the transistor stage infront of the opamp takes care of that. So I would go with something with higher headroom and nicer clipping charactaristics if it does clip like an NJM2068. That one has a maximum draw of 8mA per opamp, so 16mA total, which is pushing a MAX1044 a bit but anything under 20mA should still be ok. A NE5532 with max 16mA per opamp would probably need a more powerful charge pump like an LT1054 to keep voltage-loss through loading reasonable. But the opamp change would be a minor optimization based on my personal taste, by no means necessary for operation.

Please let us know how it went when you finished it. This thing starts to interest me, especially because of the interesting architecture of the gain stage. I would really like to know how this is called. Does anyone know? Paur R. (PRR) replied with a similar thing implemented with a JFET instead of a BJT to a question of mine some time ago:
https://www.diystompboxes.com/smfforum/index.php?topic=122279.msg1152716#msg1152716
I was meaning to experiment with that for some time but you know... time constraints. However, due to using a JFET instead of a BJT, you ommit the strange bisasing. What I am still wondering is: Where do we want to bias the +input ot the opamp to? V=0.5*supply or 0.83*supply?

Cheers,
Andy

[Rob Strand]

So with C10 and C7 all is fine?

Those sort out the IC2A side of things but you still have a problem with the bass circuit.

There is a path to ground R8, R9, Bass pot which will pull down the bias voltage from Va + R15.
A simple solution is to put a cap between R8 and R9.

[nonoxxx]

Please let us know how it went when you finished it. This thing starts to interest me, especially because of the interesting architecture of the gain stage. I would really like to know how this is called. Does anyone know? Paur R. (PRR) replied with a similar thing implemented with a JFET instead of a BJT to a question of mine some time ago:
https://www.diystompboxes.com/smfforum/index.php?topic=122279.msg1152716#msg1152716
I was meaning to experiment with that for some time but you know... time constraints. However, due to using a JFET instead of a BJT, you ommit the strange bisasing. What I am still wondering is: Where do we want to bias the +input ot the opamp to? V=0.5*supply or 0.83*supply?

Cheers,
Andy

I fact I choose Max1044 in my eagle library because I didn't fint LT1054 , I use only 1054 for these circuits , concerning the opamp I was hesitating between TL072 and burr brown 2134  i am not still decided due to power consumption.

I still have finished a prototype by heavily modifiying a chugga pre pcb and adding a strip board for the last gain /blend stage, it's look like noodles inside but it's sound awesome 

I take the opportunity of tracing it on eagle to do optimisations , and all of you guys are a really great help.
Thanks a lot