Tech21 - Character Series.  [schematic]

All about modern commercial stompbox circuits from Electro Harmonix over MXR, Boss and Ibanez into the nineties.
User avatar
Fix_Metal
Information
Posts: 20
Joined: 11 Jan 2010, 09:18
Been thanked: 15 times

Post by Fix_Metal »

I ended up throwing in the bin that Joyo PCB. Actually it is possible to just modify that but you gotta be EXTREMELY careful during soldering process.
I ordered various samples from TI for SMD 226x ICs and changed up the parts a few times. In the end I was getting bad bias over opamps because of bad contacts which I couldn't resolve. I.e. I tried to isolate one feedback opamp for voltage measure because I couldn't figure out WHY it was not biased. Even if ALL the traces where isolated from other parts I was reading some real Volts. Plus, I had 2 pot chassis sweeped by voltage too, even if they were isolated on the board!!
I think the main problem is the real poor PCB job done on these, that gets worse and worse as soon as you heat it.
Just for having a complete reference about this pedal modification (Joyo over Tech21), here are the missing parts:
-R106 is missing in the British pedal (220K). It is located near U2 and it's R16.
-Bias voltages are all V/2 except for IC2a which is a bit higher (keep in mind that Vcc pass through R37//R38, and ground is elevated by R39//R40, so that what should be +9V will actually be kinda lower, ~7,5V)
-Switching part is shown in this schematic. It is not true bypass and the buffer alters frequencies, so I'd suggest just to get rid of that. There's no reference for diode/transistor as there's nothing written on them, but I'd suggest they are mosfet+zener protection diode (if it was JFET, a diode wouldn't be useful in that place)
-All pots are 100KB

I ended up making my own pcb for the V-Bass and it worked fine at first time. Oh, and I got rid of those Vcc/GND resistors. I think it's just pointless using rail-to-rail technology if you don't run them at full available voltage. The resulting voltage range is actually identical to what you'd expect from straight TL07x/TL06x.
Attachments
Switching.png
Switching.png (3.8 KiB) Viewed 3970 times

User avatar
Fix_Metal
Information
Posts: 20
Joined: 11 Jan 2010, 09:18
Been thanked: 15 times

Post by Fix_Metal »

And btw those switching 2,2uF decoupling caps are SMD ceramics.

User avatar
solarplexus311
Information
Posts: 1
Joined: 13 Mar 2013, 14:27

Post by solarplexus311 »

Any idea what the parts differences are for the Oxford?

User avatar
GodSaveMetal
Resistor Ronker
Information
Posts: 301
Joined: 19 Sep 2009, 16:11
Has thanked: 139 times
Been thanked: 54 times

Post by GodSaveMetal »

rcustoms wrote:I read the whole post and make the pcb, this is now being tested.if everything is working well,i will post it
You made a new one PCB for that you post??? it is working???? 8) please man post it!!! thanks!!! great job!!!!!!! :applause: :applause:

User avatar
rcustoms
Resistor Ronker
Information
Posts: 417
Joined: 01 Oct 2009, 14:12
Location: COLOMBIA
Has thanked: 357 times
Been thanked: 503 times
Contact:

Post by rcustoms »

GodSaveMetal wrote:
rcustoms wrote:I read the whole post and make the pcb, this is now being tested.if everything is working well,i will post it
You made a new one PCB for that you post??? it is working???? 8) please man post it!!! thanks!!! great job!!!!!!! :applause: :applause:
soon as i can ,i share the last pcb
http://www.rcustomspedals.blogspot.com
guitars,pedals and amps.
@zuleta.rafael(Instagram)

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

Thought I would revisit this thread and offer some new insight into the Character series gain module :wink:
I decided to simulate just the module (CH34-4) in National Instruments Multisim and examine it's clipping characteristics.
This module is the "guts" of the distortion in these pedals and from what I have seen is identical across the complete series.
There are two stages before the module - first a small buffer or gain stage (depending on whether the unit is in bypass mode or active)
The second stage "character" is a tone shaping pre distortion module stage - this is different for each of the models in the series and is basically what differentiates them all sound wise.
After the CH34-4 module, we have a two stage high frequency rolloff or "speaker simulator" circuit and output level control, which is followed by two fairly conventional post filtering EQ active baxandall Treble Middle and Bass stages.
The final stage is a small op amp buffer.

Anyway - back to the CH34-4 module :wink:
This module uses the LINCMos TLC2262 dual op amp - internally entirely constructed with mosfets - it gain swing the signal full rail before it clips and Tech 21 have used this same device in all of their pedals from the original Sans Amp onward. It has a very soft clipping characteristic and can sound a bit "mushy' in the bass region in my opinion. However it's implementation in the character series pedal is a little different :wink: With reference to the schematics earlier in this thread you may notice two back to back 3.3v silicon zener diodes in the feedback loop of the second stage of the CH34-4 module.
It is THESE zener diodes that clip FIRST in this module according to my simulations NOT the op amp as in the original Sans Amp pedals. These zeners also roll off the high frequency response of this stage - the signal is -3dB down at 5khz and drops to over -22db at 100khz. Without the zeners, the response is flat to 100khz and beyond :shock:
I must emphasize that the signal is attenuated at high frequencies by these zeners EVEN when the signal is NOT being clipped by them :!:
The TLC2262 NEVER has the chance to influence the clipping behaviour in these pedals - all the clipping and high frequency rolloff is performed by the two back to back zener diodes - food for thought :secret:
Cheers
bajaman
be kind to all animals - especially human beings

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

Character series Oxford schematic 150413.gif
not sure about the cap values - resistor values and cabsim switching is correct though.
A few more surprises compared to the earlier models before the cabsim switch was fitted :wink:
enjoy :hug:
bajaman
be kind to all animals - especially human beings

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

I would suggest trying a value of 1n (possibly more) for the capacitor marked with a ?
This sets the high frequency rolloff when the speaker simulator cabsim switch is disengaged.
For the suggested value the high frequency response will be approximately -3db down at 8KHz.
Double the value (2n2) and the response will be -3db at 4Khz.
cheers
bajaman
be kind to all animals - especially human beings

User avatar
artaommahe
Information
Posts: 22
Joined: 16 Sep 2010, 13:32
Has thanked: 18 times
Been thanked: 2 times

Post by artaommahe »

bajaman, oh.. thanks ) speakersim bypass is from real pcb?

User avatar
artaommahe
Information
Posts: 22
Joined: 16 Sep 2010, 13:32
Has thanked: 18 times
Been thanked: 2 times

Post by artaommahe »

bajaman, any chances on us steel? =)

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

bajaman, oh.. thanks ) speakersim bypass is from real pcb?
yes :D
bajaman, any chances on us steel? =)
sorry - I have not seen that one yet - someone did post some gut shot pictures in another thread on this forum, but I have been unable to view them - sometimes images just seem to disappear :cry:
If anyone has these pictures, would they please repost them in this thread. :thumbsup
cheers
bajaman
be kind to all animals - especially human beings

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

Oxford composite.gif
be kind to all animals - especially human beings

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

Oxford bottom.gif
be kind to all animals - especially human beings

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

Oxford component side.gif
be kind to all animals - especially human beings

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

:oops: I left R14 off yesterdays schematic - here is the corrected version revision 1
Baja Tech 21 Character series Oxford schematic 160413.gif
be kind to all animals - especially human beings

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

Baja Tech 21 Character series Oxford schematic 160413.gif
I corrected the cabsim filter capacitor values (C14 and C15) - these are the same as the British pedal which emulates the Celestion greenback response. it seems reasonable to assume these will be the same in the Oxford which according to the builders has the same loudspeaker simulator.
Here is revision 2 :D
be kind to all animals - especially human beings

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

an interesting observation - C13 and R16 introduce a -6db gradual rolloff in the bass response EVEN when the cabsim is bypassed. At 900Hz this filter has little effect on the response curve, but by 100Hz the output level has been gradually reduced so that at low frequencies the response is -6db down. If this filter is removed, the response is flat from 900Hz down to 100Hz :wink:
be kind to all animals - especially human beings

User avatar
artaommahe
Information
Posts: 22
Joined: 16 Sep 2010, 13:32
Has thanked: 18 times
Been thanked: 2 times

Post by artaommahe »

bajaman, at pdf schematic from first pages(liverpool) we see that there is no one part of circuit - after speakersim (10k -> 10k_gnd -> 1u ->10k). It's from original pcb?
Also what do you think about Blend pot realization at VT Bass DI?

User avatar
Fix_Metal
Information
Posts: 20
Joined: 11 Jan 2010, 09:18
Been thanked: 15 times

Post by Fix_Metal »

I forgot to update the topic. First, there's a errata in the E1 value, it's definitly 10nF. Let's start from this.

Code: Select all

        +------------------------------------------------------+-------------------------------------------+
        |                    T e c h   2 1                     |                  J O Y O                  |
        +---------+---------+------------+-----------+---------+----------+---------+------------+---------+
        | Blonde  | British | California | Liverpool | VTBass  | American | British | California | AC Tone |
--------+---------+---------+------------+-----------+---------+----------+---------+------------+---------+
R4      | *3.3k   |   10k   |            |   3.3k    |  3.3k   |   *10k   | 10k(R4) |            |   10k   |
E1      |  open   |   18n   |            |    22n    |  10nF  |   open   | 18n(C3) |            |   22n   |   Input &
Cin     |   22n   |   22n   |            |    22n    |  22n    |    22n ? | 22n(C1) |            |   22n   |  filtering
C2      |   47n   |   47n   |            |    47n    |  open   |    47n ? | 47n(C2) |            |   47n   |
--------+---------+---------+------------+-----------+---------+----------+---------+------------+---------+
C5      |   22n   |   10n   |            |    22n    |   10n   |    22n   | 3n (C8) |            |   22n   |
C6      |   22n   |   10n   |            |    22n    |   10n   |    15n   | 82p(C7) |            |   22n   |
R6      |    1k   |    1k   |            |     1k    |    1k   |     1k ? | 1k(R6)  |            |    1k   |  Character
R7      |  100k   |  100k   |            |   open    |  100k   |   100k   | 100k(R7)|            |  100k   |   pre-EQ
R9      |    1k   |  3.3k   |            |   3.3k    |  3.3k   |   3.3k   |3.3k(R10)|            |  3.3k   |
--------+---------+---------+------------+-----------+---------+----------+---------+------------+---------+
R14     |   47k   |   33k   |            |    47k    |   33k   |    33k   | 33k(R24)|            |   33k   |
C14     |  470p   |  470p   |            |   180p    |   470p  |     ?    |470p(C17)|            |    ?    |  HP sim
C15     |  2.2n   |  2.2n   |            |   2.7n    |   2.2n  |   2.2n   |2.2n(C18)|            |  4.7n ? |
--------+---------+---------+------------+-----------+---------+----------+---------+------------+---------+
C20     |   10n   |   22n   |            |    18n    |   10n   |    15n   |15n(C25) |            |   18n   |   MID
C21     |   10n   |   22n   |            |    22n    |   10n   |    15n   |22n(C24) |            |   22n   |    EQ
--------+---------+---------+------------+-----------+---------+----------+---------+------------+---------+
*Not used
Then I followed the advices of 12bass guy, found in this topic:
1) 22K resistor accross high pot, between lugs 1 and 2 (or maybe 3 and 2? who remembers :mrgreen: )
2) DPDT simcab with pre sim on R14, R15 etc., the other wired to level pot and put a 10nF over ground when the switch is engaged
3) Put a 100K log instead of linear pot in volume (odd enough, the pedal had so much volume already at 1/10 with linear)

Great work bajaman, sorry for the interruption :applause:

User avatar
bajaman
Old Solderhand
Information
Posts: 4385
Joined: 26 Jun 2007, 21:18
Location: New Brighton, Christchurch, NZ
Has thanked: 476 times
Been thanked: 1737 times

Post by bajaman »

I just bought a JOYO California sound pedal - couldn't resist the $1.00 price tag (plus US$32.00 for shipping - hah ah :wink: )
First impressions were very favourable - I changed the dual op amp (TL062) - put a TLC2262 in it's place - i left the other two quad op amps (TL064).
May get around to tracing the component values but a this price - just buy it and do the mod :wink:
bajaman
be kind to all animals - especially human beings

Post Reply