You are here

Transcendent 2000 Powertran PCB diagram

Customising, building or repairing your own gear? Need help with acoustic treatment or soundproofing? Ask away…

Re: Transcendent 2000 Powertran PCB diagram

Postby midierror » Sat Jan 24, 2015 10:33 am

Thanks Folderol, I hope you got your door sorted! So the readings post-IC1 and solder checking are...

IC10

Pin 3: -5.99
Pin 6: -0.9

IC11
Pin 6: 6.2

The end of R39 reads -11.99
User avatar
midierror
Regular
Posts: 217
Joined: Tue Apr 26, 2005 12:00 am

Re: Transcendent 2000 Powertran PCB diagram

Postby Folderol » Sat Jan 24, 2015 11:52 am

OK, pin 3 tells the story.

That voltage is exactly half the negative rail voltage. This tells me that Q9 is drawing no current at all, and your meter internal resistance is 1M :)
So... why is is Q9 drawing no current? It could be O/C, or there could be a bad connection - especially the base O/C would mean the transistor was 'off'.

This is in a x5 transistor package so might be quite difficult to work out which pin is which, but again you need to go on the actual chip pins to find out what's going on. If the emitter is at 0V then there is a break (possibly bad socket connection) between this and R39/IC10-3. If either the base or collector is not at 0V then that's where to problem is.

Finally, if the base and collector are both at 0V and the emitter is at around -6V the transistor is indeed dead. These packaged transistors are still available but getting less common.
User avatar
Folderol
Jedi Poster
Posts: 8123
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.

Re: Transcendent 2000 Powertran PCB diagram

Postby Hugh Robjohns » Sat Jan 24, 2015 12:17 pm

Folderol wrote:This is in a x5 transistor package so might be quite difficult to work out which pin is which, but again you need to go on the actual chip pins to find out what's going on.

It's a CA3046. The published PCB layout suggests that the pin connections to the transistors are as follows:

Q9 collector pin 14, emitter pin 13, base pin 12

For reference, the connections to Q7 are:

Q7 collector pin 11, emitter pin 10, base pin 9

I think the OP's PCB layout linked in the first post misses off a lot of tracks hidden under the ICs.

H
User avatar
Hugh Robjohns
Moderator
Posts: 23205
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Transcendent 2000 Powertran PCB diagram

Postby midierror » Sun Jan 25, 2015 10:30 am

Good morning chaps.

Q9 Readings are

Pin 14: 0
Pin 13: -5.99
Pin 12: 0

Folderol wrote:if the base and collector are both at 0V and the emitter is at around -6V the transistor is indeed dead.

Looks about right Folderol!

Hugh Robjohns wrote:I think the OP's PCB layout linked in the first post misses off a lot of tracks hidden under the ICs.

Yeah, you`re right

And for the record...Q7

Pin 11: 0
Pin 10: 0.9
Pin 9: 0.04
User avatar
midierror
Regular
Posts: 217
Joined: Tue Apr 26, 2005 12:00 am

Re: Transcendent 2000 Powertran PCB diagram

Postby Wonks » Sun Jan 25, 2015 2:45 pm

While we're at it, can anyone recommend any component changes or easy mods to give a higher and less noisy output? It's not got the strongest output signal in the world - fine into an instrument amp but less so into a line input, and it does have quite a lot of background hiss.
User avatar
Wonks
Jedi Poster
Posts: 8612
Joined: Thu May 29, 2003 12:00 am
Location: Reading, UK
Correcting mistakes on the internet since 1853

Re: Transcendent 2000 Powertran PCB diagram

Postby Folderol » Sun Jan 25, 2015 4:18 pm

midierror wrote:Good morning chaps.

Q9 Readings are

Pin 14: 0
Pin 13: -5.99
Pin 12: 0

Folderol wrote:if the base and collector are both at 0V and the emitter is at around -6V the transistor is indeed dead.


Looks about right Folderol!

Pretty conclusive then.

As a point of interest, the entire purpose of Q9 and IC10 is to compensate for the base-emitter pedestal voltage of Q7. This voltage is very temperature sensitive - not the ideal situation for a voltage controlled oscillator.

With Q7 and Q9 on the same bit of silicon you can be pretty certain that not only will they be subject to the same temperature fluctuations, but their general characteristics will match pretty closely too.

Ideally, for best match both transistors should have as low a base current as possible. Hence, in this circuit they've supplied Q9 from a 1M resistor. Q7 current will be very low as it is used just to charge C22 via the 'see-saw' action of IC11. IC10 acts as a buffer so that Q7 emitter current can't starve Q9.
User avatar
Folderol
Jedi Poster
Posts: 8123
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.

Re: Transcendent 2000 Powertran PCB diagram

Postby Hugh Robjohns » Mon Jan 26, 2015 2:32 pm

Wonks wrote:...can anyone recommend any component changes or easy mods to give a higher and less noisy output?


I suspect you can't do much about the hiss, as it is largely down to the overall design and all the preceding circuitry! However, the output buffer (IC21) is a pretty nasty and noisy (by modern standards) 741. You might well get a quieter output if you swapped that for something more modern, such as an AD711 or TLO71 which I think should just drop straight in without any other changes.

An NE5534 isn't a drop-in replacement in that circuit configuration, but could be used if you want to modify the design into a low gain stage (see below).

As for output level, you could reconfigure the output buffer to provide some gain... but that would involve hacking the PCB (since the pin 6 output is wired directly back to the pin 2 negative input via a track under the chip to provide unity gain at present). You'd also have to find some way to fit two extra resistors to set the gain. Do-able... but not pretty.

Another possibility would be to increase the control current flowing into the preceding VCA stage (IC22) from the ADSR section. The VCA gain is controlled by the current on pin 5. However, whether that is practical, from a headroom and clipping perspective, would require some testing and/or trial and error. It's not at all clear how much headroom there is in that stage.

If it was my synth (and assuming there's physical space), I think I'd make up a little piggy-back PCB which could plug into the IC21 socket, but with a more modern IC configured as a +10dB non-inverting gain stage. The piggy-back board would need a separate wire connection to the output ground, for the bottom of the lower gain stage resistor, as there is no local ground available on any of IC21's pins, but other than that very minor addition (to the nearby output ground connection point), there would be no hacking necessary.

That way, the piggy-back board (and wire link) could be taken out and a 741 (etc) replaced to restore the system to its original design without any damage should hat be desirable.

H
User avatar
Hugh Robjohns
Moderator
Posts: 23205
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Transcendent 2000 Powertran PCB diagram

Postby Wonks » Mon Jan 26, 2015 2:47 pm

Thanks, Hugh. I'll try the chip change first.
User avatar
Wonks
Jedi Poster
Posts: 8612
Joined: Thu May 29, 2003 12:00 am
Location: Reading, UK
Correcting mistakes on the internet since 1853

Re: Transcendent 2000 Powertran PCB diagram

Postby Folderol » Mon Jan 26, 2015 6:30 pm

I see Hugh beat me to it, sneaking in there while I was concentrating on the unwell oscillator :tongue:

TL071 is definitely a drop-in replacement in this situation and they're cheap as... er... chips.

There's also a 748 in the oscillator circuit that you might get away with swapping out (not that it'll make much difference here). The original needs the 4p7 cap on pin 8 to maintain stability, but the 071 can usually handle the (virtually) open loop condition without problems.

Another thing worth looking at is the type of resistors, especially the high value ones in the signal path. If they're carbon then changing them for metal film can make a difference, but that's starting to get a bit serious!


P.S.
If you're getting chips, don't forget the vinegar :lol:
User avatar
Folderol
Jedi Poster
Posts: 8123
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.

Re: Transcendent 2000 Powertran PCB diagram

Postby Wonks » Mon Jan 26, 2015 6:58 pm

Don't worry. I've already filled the case up with curry sauce.
User avatar
Wonks
Jedi Poster
Posts: 8612
Joined: Thu May 29, 2003 12:00 am
Location: Reading, UK
Correcting mistakes on the internet since 1853

Re: Transcendent 2000 Powertran PCB diagram

Postby midierror » Thu Jan 29, 2015 10:00 am

Hello,
Wonks your sigs. are hilarious! Makes me laugh while I cry all over my broken synthesizer.

Noise reduction advice is welcome here too..post functioning oscillator.

The new 3046 chip has arrived and is installed.
Pins on it read

14: 0
13: -0.62
12: 0

11: 0.02
10: -0.63
9: 0.16

Is this good?
User avatar
midierror
Regular
Posts: 217
Joined: Tue Apr 26, 2005 12:00 am

Re: Transcendent 2000 Powertran PCB diagram

Postby Folderol » Thu Jan 29, 2015 11:03 am

These figures are much more sensible. Q9 is absolutely correct, and Q7 is probably right (emitter certainly is).

What are you now getting on IC11 pin 6? If the oscillator is actually running it should be around 2.5V.
User avatar
Folderol
Jedi Poster
Posts: 8123
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.

Re: Transcendent 2000 Powertran PCB diagram

Postby midierror » Thu Jan 29, 2015 12:36 pm

Excellent news, thank you once again Folderol. IC11 pin 6 is at 4.1V
User avatar
midierror
Regular
Posts: 217
Joined: Tue Apr 26, 2005 12:00 am

Re: Transcendent 2000 Powertran PCB diagram

Postby Hugh Robjohns » Thu Jan 29, 2015 2:35 pm

At this point you may need to switch to a scope, rather than a multimeter, but a static 4V at the output of IC11 isn't good -- although it's now two volts lower than the first time you measured it! ;)

What is the voltage on the output (pin 7) of IC12?

I think it might be worth while swapping out IC12 for a new one since that LM311 chip has a maximum rail voltage rating of +/-15V, and you've already found the unit was running on +/-17 for some time. It could well be dead, which won't help the oscillator oscillate.
User avatar
Hugh Robjohns
Moderator
Posts: 23205
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Transcendent 2000 Powertran PCB diagram

Postby Folderol » Thu Jan 29, 2015 4:53 pm

You may be right, and certainly the chip will have been stressed, but I'm concerned that IC11 is apparently holding at 4V, which is a bit more than a volt below the IC12s trigger voltage. I agree a scope would help here, just in case it is oscillating but with a weird waveshape at an inaudible frequency. It might be worth putting your meter on AC and seeing if it shows anything on IC11-6

If it's definitely a DC voltage, and it remains the same with IC12 removed, then I would suspect Q8 as either having a disconnected gate, or having gone leaky.
User avatar
Folderol
Jedi Poster
Posts: 8123
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.

Re: Transcendent 2000 Powertran PCB diagram

Postby Folderol » Thu Jan 29, 2015 9:10 pm

Had time for another look at the drawings, so a few more thoughts.

First, the obvious. Have you actually connected up the output to see if there is any sound now? A sawtooth wave should have an average value halfway between it's max and min levels. I'm assuming max is just over 5V and min is 0V, but be warned I haven't used up this months quota of mistakes :)

Some more nitty gritty.

IC12 is simply acting as a switch to turn on Q8 - to be pedantic, it actually holds it off and R45 lets it come on again. The output of IC12 is an open collector, so it either pulls the voltage down or does nothing. This being the case it should have no direct effect on IC11.

However, looking at it's internal schematic, there is one failure mode where it could have a small amount of 'pull-up' capability, and the and the gate-source junction of Q8 could possibly pass that through to IC11-6. If that was happening I would also expect IC11-2 to be quite high.

Disregarding that, if Q8 was shorting then I would expect IC11-6 to be the same as IC11-2, i.e. zero. If it was O/C then IC11-6 should climb to almost the positive rail.

With the usual caveats about bad trackwork etc. I wondering if it is IC11 itself that is misbehaving.
User avatar
Folderol
Jedi Poster
Posts: 8123
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.

Re: Transcendent 2000 Powertran PCB diagram

Postby midierror » Fri Jan 30, 2015 3:10 pm

Okay then...

IC11 has been replaced and now read 2.74 on pin 6.

Pin 7 of IC12 now reads -11.83

Looking at the PCB, of all the sodlering done Q8 seems like the most iffy. It has a lot of brown (soldering?) residue on each point.

Should have some more replacement parts in a few days.
User avatar
midierror
Regular
Posts: 217
Joined: Tue Apr 26, 2005 12:00 am

Re: Transcendent 2000 Powertran PCB diagram

Postby midierror » Fri Jan 30, 2015 3:32 pm

I thought I`d give it a try once more, and could hear a faint oscillation in the background....

then suddenly! https://soundcloud.com/midi-error/sets/first-noises-of-the-transcendent/s-JbF0y

WOOOOOOOAAAAHHHHHHH!!! It sounds like its working!!!!!!!!!! Will test over the weekend.

:D
User avatar
midierror
Regular
Posts: 217
Joined: Tue Apr 26, 2005 12:00 am

Re: Transcendent 2000 Powertran PCB diagram

Postby Folderol » Fri Jan 30, 2015 5:50 pm

As soon as I saw those voltages I was pretty sure it would be :)

The reason IC12-7 is reading almost -12V is because it is 'on' most of the time and only switches off very briefly while Q8 discharges C22.

Now have a good play with it :bouncy:
User avatar
Folderol
Jedi Poster
Posts: 8123
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.

Re: Transcendent 2000 Powertran PCB diagram

Postby Hugh Robjohns » Fri Jan 30, 2015 6:14 pm

Hurrah! Be prepared to go through the alignment process (detailed at the end of the construction article) to make sure the oscillator tracks the keyboard voltage properly.
User avatar
Hugh Robjohns
Moderator
Posts: 23205
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

PreviousNext

Who is online

Users browsing this forum: No registered users