GGG Brassmaster: Does anyone know the correct resistor values?

[DavidRavenMoon]

I posted a re-engineered version of the circuit at GEO - The Bass Ackwards.

Simulation says it runs, but simulators sometimes lie. I think it's a fruitful place to dig. In this version, the transformer/diode ring seems to give a fairly smooth doubling. We'll see.

I replaced the 3PDT with a 4053, used an LM386 for a bias generator, did a few other cleanup things. As soon as I can locate the breadboard, I'll hack one together.

Cool!   I'll have to try this one out soon.  (scrounging though my parts bin)

[airplanehuh]

Ah, so I was having trouble with my brassmaster clone and I stumbled on this really awesome thread you guys have going here. I, too, discovered the reverse pin arrangement for Q6 and reversed it promptly. I am having a bit of a rookie problem which led me to that discovery. I was given this RG's version of the board about a year ago perhaps...maybe more than that...by a friend of mine. I just now got around to getting all the parts and began putting it together. For some reason the clean bass works fine but there is nothing coming through the octave/fuzz section at all. I've checked all of the transistors and diodes. Would you suggest the next culprit to be T1? I've been over the schematic probably a few dozen times and all the other parts look correct (except the ones that you guys are finding I should change...definitely swapping the R29 location to be in front of C13). I've checked what I have the equipment to test on the transformer (which is basically just the resistance levels of the individual sides) although I have very little experience with these at this point. Any tips on things I should check out? I suppose it couldn't hurt to buy a new T1 and try it, although I can't get those anywhere nearby me that I know of. Also, were there perhaps any issues with an older version of the board? Mine says TUBA KING on it.

Keep up the awesome detective work. I'm learning a lot by reading your discussions/discoveries.

[R.G.]

Ah, so I was having trouble with my brassmaster clone and I stumbled on this really awesome thread you guys have going here. I, too, discovered the reverse pin arrangement for Q6 and reversed it promptly. I am having a bit of a rookie problem which led me to that discovery. I was given this RG's version of the board about a year ago perhaps...maybe more than that...by a friend of mine. I just now got around to getting all the parts and began putting it together. For some reason the clean bass works fine but there is nothing coming through the octave/fuzz section at all. I've checked all of the transistors and diodes. Would you suggest the next culprit to be T1?

It's far more likely that there is some issue with soldering or wiring. Measure and post your transistor pin voltages. That's about the fastest way to find the bugs.

Also, were there perhaps any issues with an older version of the board? Mine says TUBA KING on it.

An "R.G.'s" board with "Tuba King" on it? Hmmm... I'd sure like to see photos of the top and bottom of that board if that's at all possible.

[George Giblet]

You can get more drive out of the filter with the following mods:
- reduce on Q6 (RG layout) to say 68k
- increasing the cap 50nF that connects to this resistor to 100n
- reduce 1k resistor on the emitter of Q6 (RG layout) to 680R

Set the brass pot to 12 O'Clock to make it sound like full on the original.

[George Giblet]

A good example of a vintage Brassmaster is the section in Close to the Edge about 6:01 right after one of the choruses.  That's a typical Brassmaster setting... a thin envelopy type of thing. Close to the Edge (excerpt)

I've always loved Chris Squire's sound and playing on that album - especially that part of the track.

[fogwolf]

Sorry to ask, I'm sure the answer is here, but there are so many suggestions and what not I'm a little overwhelmed - basically if I want to build this based on the re-worked "Bass Ackwards" schematic, could I still use the PCB at GGG? If so, what changes are made to how it's assembled? Again, I know the answers are probably here (though not sure if the old PCB image will work) but would be nice, if the PCB would work, to have it summarized in one posting how what you would change when assembling from that image.

Thanks!

I posted a re-engineered version of the circuit at GEO - The Bass Ackwards.

Simulation says it runs, but simulators sometimes lie. I think it's a fruitful place to dig. In this version, the transformer/diode ring seems to give a fairly smooth doubling. We'll see.

I replaced the 3PDT with a 4053, used an LM386 for a bias generator, did a few other cleanup things. As soon as I can locate the breadboard, I'll hack one together.

[R.G.]

if I want to build this based on the re-worked "Bass Ackwards" schematic, could I still use the PCB at GGG?

No. It's not practical IMHO. Sorry. Too many changes.

Of course, I just finished some touch ups to the new Bass Ackwards PCB layout. 

would be nice, if the PCB would work, to have it summarized in one posting how what you would change when assembling from that image.

I'll do that when we get to a conclusion. Right now, there's things still remaining to be checked out.

[fogwolf]

Gotcha - thanks!

if I want to build this based on the re-worked "Bass Ackwards" schematic, could I still use the PCB at GGG?

No. It's not practical IMHO. Sorry. Too many changes.

Of course, I just finished some touch ups to the new Bass Ackwards PCB layout. 

would be nice, if the PCB would work, to have it summarized in one posting how what you would change when assembling from that image.

I'll do that when we get to a conclusion. Right now, there's things still remaining to be checked out.

[DavidRavenMoon]

I've always loved Chris Squire's sound and playing on that album - especially that part of the track.

Yeah, me too!   That album made buy the Brassmaster!  Listening to that album made me want to build a new one. 

It's exciting that just when I got around to building one, and then searching for why mine wasn't working right, is happening just when some progress to that answer is being made.

[gtrgeek1]

Hey Guys,

I finally got the original Brassmaster on my bench and have the following info for your reference:

Transistors
The transistor designations follow the original factory schematic.
Q1,2,3,4,6,7 are 2N3392 factory marked 83930. All have an orange dot on the top, possibly matched or selected.
Q5 is 2N5308 factory marked 135 43C. Pinout matches 2N3392 not schematic pinout.

Maestro Bass Brassmaster Voltage Chart 01/03/09
V1 = 9.3V
V2 + 8.0V
                 E         C         B
Q1         193 mv 2.107V 0.793V
Q2          63.9mv 4.68V 0.656V
Q3           4.04V 5.16V 4.68V
Q4 H1     1.319V 9.34V 1.918V
Q4 H2     3.195V 9.34V 3.796V
Q5 H1B1 28.2mv 7.58V .907V
Q5 H1B2   61mv 5.88V .980V
Q5 H2B1   6.5mv 8.7V .816V
Q5 H2B2 40.8mv 6.92V .937V
Q6          267mv 2.986V .870V
Q7          2.391V 8.04V 2.987V

H1 = SW2 closed / 82K connected
H2 = SW2 open
B1 = 47K resistors selected in filter
B2 = 6.8K resistors selected in filter

Diodes
All Diodes 1N457A

Transformer (measured in circuit)
Primary = 680 ohm CT
Secondary 572 ohm
500 mv 1K sinewave into secondary, measured 500 mv on primary
The transformer on this particular unit is much smaller than others I have seen. Mounting tabs did not line up with pads in pcb and were folded flat.

Resistors
All 1/2 watt carbon comp
All pots 10k CTS with 1971 date code

Capacitors
.01, .005 (x2), .05 in filter are ceramic disc
.05 on base of Q3 is ceramic disc
Electrolytic as marked on schematic
All others mylar 100V

PCB
C.M.I. 980 018260

Since the owner paid a boatload for this thing, I am a little hesitant to unsolder anything for a more accurate measurement, but I did compare all voltages to a clone built on RGs layout and all voltages match.

Since I will have to return this treasure soon, please let me know if there is any other info needed.

Regards,
gtrgeek1

[R.G.]

From listening to the posted excerpts, I think the Brassmaster can be de-composed and re-formulated.

The BM is a high gain preamp and less-distortion-s frequency doubler that feeds a resonant filter which can be set to one of two frequencies. The frequencies are about 700Hz and about 3kHz, and you only get to use one at a time. It wouldn't be all that hard to ditch the Twin-T setup and sub in a multiple feedback filter or a state variable and be able to set the resonance wherever you want it. Or shoot, put in two filters and mix to your heart's content.

The transformer based doubler is kinda nice from what I've seen so far, and that may be the big advance value of the design.

[DavidRavenMoon]

OK, I got it working!   

gtrgeek1 is correct... there seems to be nothing wrong with the original schematic (except for Q1's supply voltage being 8V and not 9V), and RG's PCB is correct.

So here's what I did... first there was the issue where on RG's layout sheet R22 and R27 are swapped.  So I removed them and put them in the correct locations.  That did little to help the Brass switch, but I at least heard something when I switched it.  I had tried moving where R29 connected to C13, but I switched that back to the original location.

I decided to check Q6, since that was responsible for amplifying the output of the filter and sending it back to the circuit.  I looked up the pin out for the Fairchild 2N5308.. voilà!  I had it in backwards.  It did match the layout as far as where the flat side was, but it had the emitter and collector swapped!

So I carefully unsoldered it and flipped it around.

Here's the results.  First you hear the two settings of the Brass switch, then I flip the Harmonic switch, and then switch the Brass switch again.  You go from very full to very thin, just like the real thing.  The very thin setting is what Squire used a lot, as in the example I play from Close to the Edge, which has the dry signal mixed back in.

Brassmaster Fixed

I don't have a real one in front of me to compare, but going from memory, this is pretty much nails the tones I remember using.   

[airplanehuh]

Ok, great...I got mine working! It sounded great in the stock form (maybe I got lucky with the values). I tried comparing the new placement of R29 to the original schematic location and I definitely like the old (less obvious) harmonic switch because the tones in both positions, although slight, are packed with good tone and are more responsive to me than putting R29 before the cap. With it before the cap the tone seems much too brittle. Although I was only testing the differences through a small combo amp in my room and have yet to hook it up to the big svt rig or the traynor bassmaster. Brassmaster into the bassmaster running 2x15's could sound pretty good. What other mods to the original pcb layout would you guys suggest?
I like the possibility of the new design you are working on R.G. I may have to try that one sooner or later. Has everyone on here been using the millenium bypass with their bb-1s? Is anyone satisfied with the old-school true-bypass version? Are there any noticeable differences aside from the led option?

[R.G.]

All right, I'll get the documentation straight.
So, using the layout as a reference, R27 as marked is 22K and R22 = 6.8K, correct?

And the actual Fairchild 2N5308s have B and E reversed from the pinout picture on the original schematic, which I then propagated?
The datasheet at both Fairchild http://www.fairchildsemi.com/ds/2N/2N5308.pdf and Central Semiconductor shows to match the 2N3392 pinout.
So the second bug was that the pinout on the 5308 was backwards.

Other than that, the evidence is that the layout works. Is that the net net, David (since you actually had one and remember it, as well as having recorded examples and a working board now)?

[gtrgeek1]

Hey RG,


Don't forget to connect Q1 to the 8V supply. I have the original on my bench right now and soundwise everything jives between the original and the clone on your board. If there is any other info you need to finalize this, just email me gtrgeek1@yahoo.com

Regards.

[George Giblet]

gtrgeek1, that's awesome!!!

Thanks for all your efforts deciphering the bm - no more speculation.

What is interesting is:
- the circuit was correct all this time (except for the minor 8V/9V thing)
- the suggested transformer MOUSER 42TM018 should be spot on
- Everyone should have *already* been building bm's cones (apart from those who had the wrong pinout).
  Nothing has changed other than the knowledge in our heads.

[R.G.]

Don't forget to connect Q1 to the 8V supply. I have the original on my bench right now and soundwise everything jives between the original and the clone on your board. If there is any other info you need to finalize this, just email me gtrgeek1@yahoo.com

Yep, got it. Here's the changes to the board:
1 - Changed Q1 to +8V (although this is viewed as unnecessary; 820R can be replaced by a jumper to make it all 9V)
2 - swapped nomenclature on R22 and R27 on the schematic
3 - swapped the direction of the darlington to match Fairchild's datasheet.
4 - added pads for separate resistors as in the original Twin T filter, as opposed to adding a resistor in parallel; no big difference, just a bit of "originality" for those who can't see far enough to calculate paralleled resistors
5 - added pulldown resistors on input and output; added polarity protection diode
6 - general cleanup; used standard sized box film caps instead of generic poly outlines; rotated all the transistors to face one direction; rotated all polarized electros to face the same direction; used new footprints with better accuracy; changed to a larger and more easily found outline for the sensitivity trimmer; moved every cotton-picking component at least once; kept the same outside dimensions and mounting dimensions.

Who's gonna do the checking pass? 

Gtrgeek, you and David are PERSONALLY responsible for this getting ironed out to its current state of flatness. You realize that don't you? 

[DavidRavenMoon]

2 - swapped nomenclature on R22 and R27 on the schematic

The schematic is correct, the parts layout and wiring diagram is wrong.  Since people follow that to populate the board, the mixup happens.  If they are building from scratch using just the schematic, that wont happen.

3 - swapped the direction of the darlington to match Fairchild's datasheet.

Same as the last comment... the schematic is correct, but the build instructions have the transistor facing the wrong way.  Might be good to add a "B, C, and E" to the drawing.

4 - added pads for separate resistors as in the original Twin T filter, as opposed to adding a resistor in parallel; no big difference, just a bit of "originality" for those who can't see far enough to calculate paralleled resistors
5 - added pulldown resistors on input and output; added polarity protection diode
6 - general cleanup; used standard sized box film caps instead of generic poly outlines; rotated all the transistors to face one direction; rotated all polarized electros to face the same direction; used new footprints with better accuracy; changed to a larger and more easily found outline for the sensitivity trimmer; moved every cotton-picking component at least once; kept the same outside dimensions and mounting dimensions.

Nice touches!  I used box caps too, and I had to bend the contacts before I could fit them in the board.

One last thing, the Sensitivity knob would never be a trim pot.

Gtrgeek, you and David are PERSONALLY responsible for this getting ironed out to its current state of flatness. You realize that don't you? 

Cool!  RG thanks for your effort to have this board out there in the first place.  I just wanted the damn thing to work, no matter who and how often I had to bug people!     After I built mine and then went through the schematic several times comparing it to the PCB, and didn't find a problem, I figured let me see what others were saying.  You probably only hear from people who built it and never checked which way to put Q6 in (like me) and can't get it to work.  And since we didn't realize that the main issue with the filter amp was an improperly installed Q6, where do you look? Then once the idea is out there that the schematic is wrong, everyone starts looking in the wrong place.

And thanks to Gtrgeek for insisting the schematic was correct.  We had a long conversation about his experiences with the circuit, which made me decide to open mine back up and take another look.  He was very thorough in his comparison against the real thing.

You had to see me jumping up and down today when I made the change and plugged it in for the first time!  It was a real eureka moment!

[gtrgeek1]

Hey RG,

I can double check the pcb if you like.

I'm glad that David had a problem and brought this issue back up!  I didn't want any more people in that state of misery - you know, removing parts so much the traces are breaking, transistor leads breaking, bad solder joints, cutting traces, second guessing, burning fingertips and still not knowing why it isn't working.

Feel free to use/post all of the data and voltages I pulled from the original as a troubleshooting tool.

[Gus]

I do have some questions about this design:however they are really for the original designer(s) of the circuit.  I will post them anyway most are why some of the transistor operating points?

Q1 first gain stage was that designed for .193VDC at the emitter and 2.1VDC at the collector?

Q4 the emitter follower to the filter  Why not offset the emitter more to +9VDC, for a few reasons.  With an emitter follower with a resistor in the emitter leg the transistor can turn on and"pull" up harder than the resistor can "pull" down" for more equal drive that plus the bias shifting with the harmonic switch would seem to make more sense to have the emitter around 5VDC or so.

The output amp 2.98VDC at Q6 collector and 2.391VDC at Q7 emitter?  Q6 is both a gain stage and a summing amp the two 10Ks from the volume controls are summed at the base junction of Q6 because of the DC bias and feedback gain control resistor 470K from collector to base forming a inverting summing amp.

Are the offsets on purpose or are they a result of different transistor hfe etc changing the operating points and this was just accepted?

I would guess Q1 is on purpose because that biasing is more predictable. Q2 Q6 that bias is not as predicable IMO.

8VDC looks like it was done as a power supply RC filter for the "cleaner" gain stage sections input and output circuits to help remove noise that might be generated from the other sections.

Q6 is both a gain stage and a summing amp the two 10Ks from the volume controls are summed at the base junction of Q6 because of the DC bias and feedback gain control resistor 470K from collector to base forming a inverting summing amp.
  Maybe change the stock E 470 ohm and cap bypass it and/or the C 10K value it this can increase the open loop gain and let the 470k and other parts set the gain with a different collector voltage?.