Two Thousand Six Hundred (TTSH) Ver.1 General build thread

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, luketeaford, Joe.

Post Reply
User avatar
Altitude909
Super Deluxe Wiggler
Posts: 3851
Joined: Wed Aug 24, 2011 5:13 pm
Location: Meesheegan

Post by Altitude909 » Tue May 13, 2014 9:19 pm

sduck wrote:I used an old plain BC547. I just compared the specs between it and the A and B kind on mouser and I don't see any difference. Use whichever.
different hfe (DC current gain)

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Tue May 13, 2014 10:24 pm

Does that matter? I really don't know what factors will make a difference for noise production duties. It seems there are several other small differences I missed before - here's what I found - there are also several other varieties of BC547s available, you could most likely find one that's even closer to the original specs, if that's important. Most likely just getting what's cheap and available will work fine. It's just noise, you know...

Image

User avatar
Altitude909
Super Deluxe Wiggler
Posts: 3851
Joined: Wed Aug 24, 2011 5:13 pm
Location: Meesheegan

Post by Altitude909 » Tue May 13, 2014 10:57 pm

I seriously doubt it. Things tend to get squirrely with super high gain parts so A or B would be a safer bet. I had the nosorted parts on hand and those worked out fine. didnt measure the hfe but whatever ended up there sounds fine
Last edited by Altitude909 on Wed May 14, 2014 8:35 am, edited 2 times in total.

guillaume10
Learning to Wiggle
Posts: 29
Joined: Sun May 26, 2013 4:57 am
Location: France

Post by guillaume10 » Wed May 14, 2014 3:25 am

thanks for all the informations for the BC547.

best regards

User avatar
Nordcore
Veteran Wiggler
Posts: 583
Joined: Fri May 10, 2013 2:05 pm
Location: Germany

Post by Nordcore » Wed May 14, 2014 6:04 pm

I think it is more useful here to evaluate transistors in the target circuit than to discuss their data sheet.
(NOT a general advice, just here for this circuit. The transistor is abused. )

Original:
Image
And no, its not hum in the bass area.


As most of us will have plenty of them: 2N3904, hfe 200
Image

Sorry, no BC547 here ... but in most applications very similar: BC550B
(... but this application here is very special, so they might be different ... )
Image

... and a BC550C (same type, higher Hfe):
Image

... first one which looks really better. But don't bet on the high hfe for this behavior...

BC337-16 (hfe around 200)
Image

and the same type with more hfe: 337-40 (Hfe around 500)
Image

and the linearization with the 1k series resistor as mentioned before:
Image
(BC337-40 looks nearly the same with 1k series resistor )

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Wed May 14, 2014 11:21 pm

OK, I'm going to ask for some help again. I'm getting really frustrated with this, and while I got some initial help both here and from zthee, I still don't know what the problem is.

Symptoms: my ADSR section does not work correctly. While the AR section works fine, the other part doesn't. I've tried using the built in clock/S+H gate/trigger, and the button, and external gate and trigger combinations. The Attack fader has no effect. The Release fader is working as an offset - as you raise it, a positive voltage is added to the overall ADSR output, although when the thing is triggered/gated the Decay and Sustain settings override this offset during their cycle.

What I've done: double/triple checked power, all the parts placements, orientations, and correct values. Double/triple checked with a magnifying glass and continuity checker for solder bridges and cold solders. Swapped out the LM301 with another working one. All resistors are correct, the diodes are placed correctly and are tested to be working, the caps are correct and oriented properly. This leaves the transistors, which are about all that's left that can be a problem. And there are a LOT of them. I've changed out the 2N4392 as per zthees suggestion. I put in a new 2N5460 tonight, but I got these from UTSource, and I suspect they may be a problem, I've ordered some from mouser. I changed out the 2N3904 up between the faders because it's directly associated with the release fader. This leaves a bunch of 2N3906's and 2N5172's. These things are a pain to switch out, I'd rather be a little more focused on what to approach if possible.

So? Any ideas? Should I just wait for a different 2N5460? Should I just keep on switching transistors until I go blind?

This is especially frustrating as the first one I built just worked fine.

User avatar
elmegil
Super Deluxe Wiggler
Posts: 3063
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Thu May 15, 2014 12:08 am

Looking at the schematics in the absence of an actual board, I'm quite confused. The scans you posted clearly label the ADSR pots as 1M/1M/100K/1M, but the schematic only shows four 1M pots. It's entirely unclear to me which of the four is what. And I also see in your scans that it connects to a 390R resistor, and again I see nothing near any of the ADSR pots in the schematic that has a 390R.

If I just make a guess that R180 is the release, I would see how the 2N5460 at Q4 being bad could be causing issues. I'd think you might want to see what happens at the junction of R180, R9, and the drain of Q4 to confirm that.... But I'm shooting entirely in the dark here...

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Thu May 15, 2014 12:45 am

The 390R is a current limiting resistor for the LEDs (I think, typing from my iPad). And I just got an email from zthee suggesting that indeed the 2N5460 might be the culprit. So I guess hanging tight for my mouser order is the thing to do.

User avatar
elmegil
Super Deluxe Wiggler
Posts: 3063
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Thu May 15, 2014 12:57 am

Ah, and here I figured the LEDs were the pins at the other end. Serves me right for not checking a datasheet :). Good luck....at least Mouser ships pretty promptly, right?

User avatar
Nordcore
Veteran Wiggler
Posts: 583
Joined: Fri May 10, 2013 2:05 pm
Location: Germany

Post by Nordcore » Thu May 15, 2014 1:10 am

elmegil wrote:Looking at the schematics in the absence of an actual board, I'm quite confused. The scans you posted clearly label the ADSR pots as 1M/1M/100K/1M, but the schematic only shows four 1M pots.
Obviously R181 is labeled wrong (in the schematics), should be 100k (lin). (Sustain Level).
R180 is decay time, R182 is release time, R179 is attack time.

sduck, seems that something is really messed up, we'll have to check what works and what doesn't.
Have you got a scope, or, even better, a digital scope? A multimeter would be required at minimum.

First: check voltage at collector of Q2 (2N5172, going to three diodes and a 15k Resistor to +15V). Here you got the "conditioned" gate input signal.
Should be around +14,3V for "no gate" and -14,9V while gate on (button pressed).

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Thu May 15, 2014 2:10 am

Thanks nordcore, I'll check on that, but give me some time - it'll be late tomorrow US time that I can do this. Yes, have scope and dmm.

User avatar
fuzzbass
Super Deluxe Wiggler
Posts: 2089
Joined: Thu Feb 06, 2014 6:08 pm
Location: Florida, USA

Post by fuzzbass » Thu May 15, 2014 6:54 am

Here is a variation of the gate/trigger interface I used on my auxiliary panel. I borrowed liberally from Ray Wilson and Ken Stone here.

This was tested using my Arturia Beatstep, which puts out 7V gates. The Beatstep alone cannot get a rise out of the TTSH EGs, via either Keyboard Gate, or Sample and Hold Gate inputs. Using this circuit, it works fine.

This will generate steep 11V gates from any positive going signal at around +2.67V. The fall time is fast too, and will make it back down to near zero using 50% gate pulses up to around 2-3khZ.

The triggers generated are short and sharp, 11V, duration around 100uSec.

The outputs are buffered and have enough juice to drive multiple destinations.

The ratio of R2 and R3 sets the threshold. The value of C1 determines the trigger duration.

The circuit draws around 10mA alone, and a wee bit more when driving the TTSH. It can be powered from the unused +/-15V access on the TTSH main board, and it should not break the power budget.

If you intend to normal the output of this to the Keyboard Gate and Trigger inputs on the TTSH main board, I strongly recommend you use RG174 coax to run those lines in, and connect the screen to ground at one end of the connection only. 11V triggers in particular will bleed into audio paths if they are exposed for any significant distance. Yes it's a bloody PIA to terminate RG174 into a header connection, but worth it in this case.

Image
Wired for weird

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Thu May 15, 2014 4:23 pm

Nordcore wrote: First: check voltage at collector of Q2 (2N5172, going to three diodes and a 15k Resistor to +15V). Here you got the "conditioned" gate input signal.
Should be around +14,3V for "no gate" and -14,9V while gate on (button pressed).
Ok, I've got +1.385v at that location with no gate, and -15.4v with the button pressed. Hmmmm.

User avatar
Nordcore
Veteran Wiggler
Posts: 583
Joined: Fri May 10, 2013 2:05 pm
Location: Germany

Post by Nordcore » Thu May 15, 2014 4:38 pm

Check if you have fitted the 1,2MOhm Resistor R3

Check Cathode CR23/Emitter Q1: 0V button released, 9V envelope button pressed.

ADSR image with check points

ualslosar
Wiggling with Experience
Posts: 336
Joined: Wed Mar 24, 2010 5:42 pm
Location: West Coast, USA

Post by ualslosar » Thu May 15, 2014 7:43 pm

That's a really cool circuit. Thanks for posting.

- Can it be made to accept separate gate & trigger inputs, please?

Thank you
Larry

User avatar
elmegil
Super Deluxe Wiggler
Posts: 3063
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Thu May 15, 2014 8:47 pm

Split the circuit between the base of Q1 and C1. Front half is for gate. Run trigger into C1, back half is for trigger.

If you want to be fancy about it, run the base of Q1 to the switch lug of the trigger in jack.

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Thu May 15, 2014 9:46 pm

Nordcore wrote:Check if you have fitted the 1,2MOhm Resistor R3
It's there!
Nordcore wrote: Check Cathode CR23/Emitter Q1: 0V button released, 9V envelope button pressed.

ADSR image with check points
I'm getting those voltages correctly. And thanks for the scan with the component designations - extremely helpful! And thanks for your help in working on this with me.

User avatar
Nordcore
Veteran Wiggler
Posts: 583
Joined: Fri May 10, 2013 2:05 pm
Location: Germany

Post by Nordcore » Fri May 16, 2014 12:06 am

So either there is a fault in Q1/Q2 and the surrounding resistors.
Or some of the circuitry right of the diodes CR1, CR2, CR3 is drawing to much current as soon as it should go above 0V.
I guess the latter - and assuming proper diode direction and function the fault is either 'behind' CR2 or CR3.

So: unsolder (pull out) one leg of CR3. Do you get now next to supply level when button released at M1? (M1 =Cathode of CR1 =collector of Q2 ..)
If not: unsolder one leg of CR2. Now approximately +15V at M1?

User avatar
fuzzbass
Super Deluxe Wiggler
Posts: 2089
Joined: Thu Feb 06, 2014 6:08 pm
Location: Florida, USA

Post by fuzzbass » Fri May 16, 2014 7:52 am

elmegil wrote:Split the circuit between the base of Q1 and C1. Front half is for gate. Run trigger into C1, back half is for trigger.

If you want to be fancy about it, run the base of Q1 to the switch lug of the trigger in jack.
My thoughts exactly, minus the fancy part. I am incapable of fancy thinking.

Yesterday, I compared this to some similar circuits by people who actually know what they are doing, and realized I probably should have put something like a 1M resistor from U1 pin 3 to ground. This will change the threshold voltage a little, but is probably an overall stability improvement.

If you split this circuit so you could run in either parallel or cascade mode, I would put a 100K load resister in front of C1. The triggers might be longer in duration.

Running in parallel is a consideration if you want to drive this really fast (in the audio range). In cascade mode at high speed, you might begin to appreciate the inherent delay caused by generating the trigger from the generated gate. Otherwise I would not worry about it, I checked for this. The delay is insignificant compared to the effects of slew rate in the op amps. And both factors are insignificant at the low frequencies this circuit is designed work with.
Wired for weird

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Fri May 16, 2014 10:53 am

Nordcore wrote:So either there is a fault in Q1/Q2 and the surrounding resistors.
Or some of the circuitry right of the diodes CR1, CR2, CR3 is drawing to much current as soon as it should go above 0V.
I guess the latter - and assuming proper diode direction and function the fault is either 'behind' CR2 or CR3.

So: unsolder (pull out) one leg of CR3. Do you get now next to supply level when button released at M1? (M1 =Cathode of CR1 =collector of Q2 ..)
If not: unsolder one leg of CR2. Now approximately +15V at M1?
Ok! I pulled CR3, and then I'm getting the +14.7v at M1. This is good? What does this mean? (thanks for bearing with me!) (edit: replaced CR3 with a new one, back to the +1.4v at M1, yuck)

User avatar
Nordcore
Veteran Wiggler
Posts: 583
Joined: Fri May 10, 2013 2:05 pm
Location: Germany

Post by Nordcore » Fri May 16, 2014 11:13 am

This narrows down the problem to the path "behind" CR3.

Check CR7 for proper direction.
Check R9 for proper value (1M= brown black black yellow brown)

Check Q4 for proper type and orientation.

Check for solder bridges at these components.

... I guess everything is OK up to here.

Than probably Q4 is broken. Unsolder it.
Put CR3 back in. Check if you still get aprox. +/-15V at M1.
If yes, a new 2N5460 is a good bet.

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Fri May 16, 2014 10:58 pm

Yes , it all comes back to the 2N5460s I got from UTSource instead of mouser like the first build. I don't know if they're bad, or just maybe pinned out differently. Maybe tomorrow I'll try putting it in backwards and see if it makes a difference. Anyway, I've got an order from mouser with the kind I used before, should arrive tomorrow, so hopefully I'll get it going then. Thanks for your help in solving this mystery!

User avatar
Kipling
Super Deluxe Wiggler
Posts: 1054
Joined: Tue Dec 10, 2013 3:31 pm
Location: South-East England

Post by Kipling » Sat May 17, 2014 2:29 am

sduck wrote:Yes , it all comes back to the 2N5460s I got from UTSource instead of mouser like the first build. I don't know if they're bad, or just maybe pinned out differently. Maybe tomorrow I'll try putting it in backwards and see if it makes a difference. Anyway, I've got an order from mouser with the kind I used before, should arrive tomorrow, so hopefully I'll get it going then. Thanks for your help in solving this mystery!
I'm glad I invested in a Peak DCA75 (Semiconductor) and LCR45 (Inductor, Capacitor & Ressistor) tester before I started building. Takes considerably longer but I test every single component (bar the Op-Amps, TDA2030 and DC converter) before fitting it so it checks the pin-out as well as functionality and values. The DCA75 is great for matching trannies too.

Note that many J-Fets have no distinguishable difference between Drain & Source so could be swapped in theory, though I don't recommend it. The Gate terminal is crucial though.

Still room for human error though - easy to put a 47k in a 4k7 position etc. Luckily spotted one such error before soldering.

Pfurmel
Super Deluxe Wiggler
Posts: 1570
Joined: Wed Mar 31, 2010 12:57 pm
Location: Ireland
Contact:

Post by Pfurmel » Sat May 17, 2014 8:04 am

sduck wrote:Yes , it all comes back to the 2N5460s I got from UTSource instead of mouser like the first build.
This won't help you directly but I got some bad transistors from UTSource for my builds, too. I will not be buying from there again, they are hit and miss at best.

User avatar
sduck
experimental use of gravity
Posts: 14448
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Sat May 17, 2014 9:44 am

Pfurmel wrote:
sduck wrote:Yes , it all comes back to the 2N5460s I got from UTSource instead of mouser like the first build.
This won't help you directly but I got some bad transistors from UTSource for my builds, too. I will not be buying from there again, they are hit and miss at best.
This was actually the fourth or fifth time I've ordered from them, and most of the really hard things to find have worked. All the other TTSH rare parts I got from them worked. It's actually odd that this was the most common part I've ordered from them, and also the first thing that didn't work, which is why I'm suspicious that they're just pinned out differently. So I wouldn't dismiss them just based on this one experience. I've also gotten stuff that didn't work from mouser!

Post Reply

Return to “Music Tech DIY”