Maxon PT-909 True Bypass

[Rob Strand]

Don't think I saw any star washers but may have some laying around. thanks Rob

Actually I can't see any in the pics so maybe they didn't use them.

FYI, I finished tracing the unit.   There's no weird JFET switching or anything like that.   The switching is very straight forward like the old MXR pedals.   It should be very easy to do a true-bypass mod.   Everything is like we first thought.

It's not 100% like any of the other Ibanez units.   It's kind of a transition between the early PT-900 (pcb marking PT999) and the latest PT-909 schematic you posted at the top of the thread.
As a very brief summary:
- Input buffer is JFET not an opamp
- The output mixer is a plain vanilla inverting mixer.    That means overall the pedal inverts the signal.
- The phase stages are more like the later PT-909.   No 100pF cap from what I can see.
   Feedback circuit very similar except no trimmer.
- The LFO is more like the PT-900 (which is similar to the MXR Phase 90).  It's not an integrator type like the later PT-909.
- The LFO has some transistor buffers added.  That's where the two 2SC1815 transistors come in.
  The LFO output is buffered in order that the Width control can be added.

On "the other" forum there's a thread called "Ibanez PhaseTone PT-999".    Dirk Hendrik's pdf  schematic is the for actual PT-999 - yet another unit.  This one has JFET switching but a normal switch (probably round head).   However, Bernard Duur's  schematic at the end is actually for the PT-900.   It could be for an earlier PT-999 as well [Actually I'm sure it is].  The confusion arises because the PCB for PT-900 is marked PT-999.   The PT-900 is the earliest and is the one before the PT-909.

The PT-900 is a flat box unit (approx 1975)
http://www.effectsdatabase.com/model/ibanez/first/pt900

The early PT-999 which is like the PT-900, (approx 1976)
https://mirosol.kapsi.fi/2015/02/

[DJPsychic]

Awesome Rob!

I'm not as well versed as you, so a little lost in all of the "P's" and "9's" I have a lot of homework to do tonight!

I'll try and get all the resistors mapped out asap

FWIW As far as vintage phasers, I've really only had my hands on a P90, P100 and now the PT-909. And by far I dig the PT-909 over the other two. I love the sound/functionality. Still need to dig deep on what magic ingredients make up all of the different phase flavors.

I've spent a lot of time reading up on Mods for the MXR's, to basically have the functionality (Depth, Feedback) as the PT-909. That's why I think the PT-909 is worth mapping out and cloning.

[DJPsychic]

So went nuts and just rehoused it. (I know I'm going to vintage pedal hell  )

Stock output jack was wonky, and the new jacks were unusable in the existing enclosure unfortunately. Replaced the 9v jack and battery snap. Repurposed the pots and knobs. Doesn't look that cool but it'll do for now, needed to be boxed up!

Might spend some more time and make it fancier one day, add an LED, who knows

Pedal sounds fantastic imo, it might actually replace my long standing vibe. It's got all the psychedelic swirl I'm after with great control. 

(Thank you Rob for your infinite knowledge)

[Rob Strand]

I'm not as well versed as you, so a little lost in all of the "P's" and "9's" I have a lot of homework to do tonight!

This site has some very good info.   
https://www.tonehome.de/ibanez/narrow-box-series-w-round-switch/pt-909-phase-tone/

The only thing I don't like about the site is he gives versions from the first phaser instead of versions *within* the same model.
I think I've got 5 or 6 versions of PT-909 alone.

The site resolved a few weird versions I've seen but I still have some others that look like they shouldn't exist!  Examples where
there's two features that appear together at the wrong time.

I'll try and get all the resistors mapped out asap

I'll do some final checks and start drawing it up tonight.   I wanted to avoid drawing it up in Ltspice.

So went nuts and just rehoused it. (I know I'm going to vintage pedal hell  )

Yeah pretty extreme.  I'd probably just source some new jacks which were like the existing ones.

Pedal sounds fantastic imo, it might actually replace my long standing vibe. It's got all the psychedelic swirl I'm after with great control.

Cool.  Do you use the feedback at all?  It's useful for copying sounds off some records but in general I tend not to use it.  It does add some extra swirl if you like that.

FWIW, my old PT900 sounded a little weak.  I have heard a few better ones on Youtube.   It could very well depend on thre JFETs that ended up in the units.   I modded mine somewhat about 25 years ago and its a much deeper phasing than it originally was.   Interestingly I removed the cap between the LFO and the JFETs.  Looks like that's what Maxon have done on the PT-909.

[Rob Strand]

Well I've got the bulk of the schematic drawn up.    I did some checks and my resistor count doesn't match up.   I'll need to check it after I've had some sleep.

[DJPsychic]

Well I've got the bulk of the schematic drawn up.    I did some checks and my resistor count doesn't match up.   I'll need to check it after I've had some sleep.

Sorry havin a busy week over here. I will have those resistors mapped out hopefully by the weekend. Get some sleep!

[Rob Strand]

Sorry havin a busy week over here. I will have those resistors mapped out hopefully by the weekend. Get some sleep!

No problem.  Whenever you get time.   I've found a few things to clean up on the schematic anyway (always is).  On the whole it's not looking too bad.

A couple of things popped-up which need checking:
- value of the big cap with the red dot.
- values on the tantalum caps
- the C1815 should have a letter like C1815GR or C1815Y

[Rob Strand]

This might save some time identifying the parts.   You can just write down the values for each part designator in a text list.
(Last Resistor R43, Last cap C17)

[DJPsychic]

EDIT

[DJPsychic]

C1  473z 25V
C2  4.7uf 25V
C3  473z 25V
C4  473z 25V
C5  473z 25V
C6  473z 25V
C7  473z 25V
C8  473z 25V
C9  473z 25V
C10 473z 25V
C11 4.7uf 25v
C12 4.7uf 25v
C13 10uf 16V
C14 33uf 10v
C15 47-35
C16 103
C17 6.8-10

[Rob Strand]

Way to go!  We are nearly there.

I mapped this out with the assumption the color band in attachment is "brown" - i.e. "brown black orange gold"

Any resistor with a (?) was difficult to read but fairly sure they are correct.

Yes, that's brown.  It's hard to read.

Some of the ? values I am able to fix up as well as a few others where I wrote "pretty sure"
However, there's a few which need rechecking.  I put the colors next to the value if it helps.

If possible, on the 2SC1815 transistor could you check if it's 2SC1815Y, 2SC1815GR, or 2SC1815BL

Part      Prev Value    Possible Value
R1        4K            1K Brown, Black, Red (pretty sure)
R2        4M            1M Brown Black Green (pretty sure)
R3        45K           15k Brown, Green, Orange (pretty sure)
R4        9.2K          Recheck. 12k Brown, Red, Orange or 8.2k Grey, Red, Red
R5        350K          Recheck.  maybe 10k Brown Black Orange?  or  15k Brown, Green, Orange, or 3.9k Orange, White, Red
R6        40K           10k Brown, Black, Orange (pretty sure)
R30       1.3K          Recheck.  maybe 30k Orange, Black, Orange ; Actually it could be 3.3k  Orange, Orange, Red
R31       10K           Recheck.  maybe 5.1k Green, Brown, Red
R32       1K            Recheck. maybe 2k Red, Black, Red
R34       9.1K          Recheck maybe 910 ohm  White, Brown, Brown;  9.1k White, Brown, Red still a possibility. 
R40       10K           Recheck  Could be 10k Brown, Black, Orange but maybe 20k Red Black Orange or even 30k Orange, Black Orange

[DJPsychic]

Yep some of those are wrong, not sure what happened. I'm going to redo the list today. Sorry for the delays!

[Rob Strand]

Yep some of those are wrong, not sure what happened. I'm going to redo the list today. Sorry for the delays!

Don't stress out about any delays, no problems at all.

It's good to check all the values but I'm fairly confident the other parts not in that second list.

[DJPsychic]

Resistors:

R1  1K
R2  1M
R3  15K
R4 (white/grey) red red
R5  10k
R6  10K
R7  10K
R8  150K
R9  30K
R10 100K
R11 10K
R12 10K
R13 150K
R14 30K
R15 100K
R16 10K
R17 10K
R18 150K
R19 30K
R20 300K
R21 10K
R22 10K
R23 150K
R24 30K
R25 100K
R26 15K
R27 10K
R28 10K
R29 10K
R30 3.3K
R31 5.1K
R32 2K
R33 100K
R34 (white/grey) brown red
R35 150k
R36 470k
R37 470K
R38 4.7K
R39 10K
R40 20K
R41 1K
R42 10K
R43 100K 

[DJPsychic]

If possible, on the 2SC1815 transistor could you check if it's 2SC1815Y, 2SC1815GR, or 2SC1815BL

Hard to read but 99% sure it says "BL"


Side note: I ended up purchasing an additional unit as a backup (in case I blew up the first). There are slight variances in values as well as RED painted dots on the K30A - but that's for another day. We'll just focus on the original 

[Rob Strand]

Resistors:

Awesome work.

The one value I'm suspect of out of that is R20 300k.   Since the all-pass filters are all the same
it's highly likely to be 100k.

Hard to read but 99% sure it says "BL"

Thanks for checking as I was struggling to see how they would get this thing working reliably with a GR (and no way a Y).

Side note: I ended up purchasing an additional unit as a backup (in case I blew up the first). There are slight variances in values as well as RED painted dots on the K30A - but that's for another day. We'll just focus on the original 

Wow, that's dedication.   I don't think you will lose out since it's a good unit!

Regarding the 2SK30A dots.    Hard to believe but I've managed to work out a pattern between the position of the bias trimpot and the color of the dots.   That means Maxon are sorting the JFET into four categories and putting paint dots on the JFETs.  Moreover I've managed to work out that each color corresponds to a 100mV change in the JFETs VP (Vgs_off) parameter.  That's the parameter which is measured when sorting JFETs.

So I think we have gone far enough to put up a schematic.

Schematic:


Schematic PDF:
See attachment.

Part Desiginators:


So a couple of points.

C15   I've made it 470nF (0u47) 35V as that makes a lot more sense.

C17 is probably the only thing causing grief on the schematic.
With C17 at 6.8uF 10V  the speed of the LFO seems like it won't match-up with the published LFO speed of 0.2s to 14s.

If C17 is increase tod 10uF then it nearly matches.   I can't imagine you reading reading 6.8uF 10V instead 10uF 6.3V and even so a 6.3V cap is quite running quite close to the limit.

One idea is you could set the unit to the slowest speed.  Wait for the sweep to hit the top of the swirl then start timing/counting while it falls.  Then wait for the sweep the go down then back up to the top of the swirl again and stop.   If the time is 14s we have a problem if it's like 8 or 9 seconds then it's likely 6.8uF is correct.

Please check over the schematic and see if it makes sense or if you find anything wrong.

[DJPsychic]

Wow Rob amazing work. I have (literally) no idea how you did that! 

Concerning the caps:

C15



C17

[Rob Strand]

Concerning the caps:

For C15 I can see a small dot in front so .47uF  = 470nF.     That's what I put on the schematic.  So all looks good.

For C17 I'm reading exactly what you read 6.8uF 10V.

So the values on the schematic match the board.   I'm happy it's all good .

It's not a big deal that the unit doesn't match the published specs.  The published specs of pedals don't always match the final units, maybe they did at some point but things got changed.

I guess it would be more puzzling if the unit could dial the speed out to 14 secs but the schematics look like it only goes to 8 or 9 secs.  That would cause some head scratching for sure.

FYI, anyone can trace a PCB it just takes a bit of perseverance.   The hard thing is not having any errors (values and connections).   That takes a lot of checking.    Thanks for sticking with it, this thing wouldn't be what it is without your help.

[DJPsychic]

Thanks Rob for all your hard work, super impressive. Hopefully others  will get some good use out of it.

I've got a couple other older pedals I'll be uploading in the near future. I'll give you a heads up  might be something you'd dig. 

[Rob Strand]

I've got a couple other older pedals I'll be uploading in the near future. I'll give you a heads up  might be something you'd dig. 

Yep, no problem.   It might be worth checking "the other forum" as many of the older pedals have been traced.

In the late 70's Ibanez/Maxon made a lot of changes over a short period so it's easy for some variants, like your  PT-909, to fall between the cracks.  From time to time I've found obscure schematics for a few models in this era.  They tend to get lost on the web since people think there's one or two versions but there's really two or three.  Dod's another one with lots of variants.