Tascam M-___ Story...

  • Thread starter Thread starter sweetbeats
  • Start date Start date
Wow!

Until now I've assumed you came to this expertise from DIY tinkering in your workshop and some self study.

More or less as a weekend warrior on this stuff.

Your grasp of understanding and working with circuits is incredible!

How much formal electronics training have you had?

:spank::eek:;)
 
None.

I've had a lot of help from a lot of people.

And don't think that I knew to do what I'm doing right now without guidance. My friend to me to start drawing out schematics around some likely suspects which I did, which then led to more "do this/do that". Now with that I've traced the source of the oscillation back to the eq board...it's generating from there, and then I was able to do some testing to see what other conditions might effect the problem while seeing the oscillation on the scope. Looks like pushing and releasing the function that dumps the eq to the monitor buss also kills the problem until the console is power cycled, and furthermore engaging the LPF immediately kills the problem (until the console is power cycled).

So now I'll draw the LPF circuit out and also see if I can identify any physical clues in that area, but I suspect it is a failed component and won't be readily apparent to the eye.

Anyway, point is I can't do this on my own in the least. I'm learning and have learned a lot and can progressively do more and more on my own, but I continue to be challenged by issues beyond me and solicit help.
 
Wow!

I'm very impressed.

You've done some excellent work on that incredible board!

You'll find the offending component.

That board is already much improved from original.

I can't wait til you get all the bugs and nits out and do some real production with it!

:spank::eek:;)
 
Well after a lot of chasing around I'm back to what I feared: I believe the problem is somewhere in switching logic on the main board in the module. Brain is tired and I'm a bit disillusioned so it's time to call it a night.

The problem can be killed temporarily in a number of ways...and there's no signal oscillation coming into the eq board...only going out...so that would indicate a problem with the eq board...nope...swap in one from another module without the problem and it exhibits the same problem. I won't list all the troubleshooting and testing that has been done, but it's hours and hours.
 
Its lame when you get down to the very end of a major project to encounter a gremlin like you have found. Im confidnt you will cure its ills, but maybe a break from it and a fresh look in a day or two?
 
mdainsd said:
Its lame when you get down to the very end of a major project to encounter a gremlin like you have found. Im confidnt you will cure its ills, but maybe a break from it and a fresh look in a day or two?
Click to expand...

Took your advice and stepped away from it for about a week.

Now starting on the process of mapping out the signal flow...all 26 amp stages and how they stack up and interconnect. From that a strategic plan to hunt and destroy can be developed.
 
One problem I have seen in mixers before is too much capacative loading on an opamp, and the opamp will oscillate. Somewhat random in the mixer (Soundtrax). When you operated switches in the area the oscillation would stop. The cure is a small resistance on the output of the opamp to isolate it from the capacitave load. 100 ohms should do it.
 
Yep-yep...

Excellent point. That's one of the things we're potentially looking for which will be aided by the mapping of the signal flow.

Gonna add your comment/suggestion to the mix...feel free to suggest anything else you'd like. :D
 
New to the forum but this looks a lot like the prototype maybe to the M15 or M16. I currently own an M16 that i'm looking for a power cord to fire it up. Wondering if you ever got this running? Looks pretty damn cool!
 
Thanks!

If you read through the thread (not that you or anybody would actually have that kind of time, considering its size...), you'll find that it works and works well...still teasing out some minor issues, but I'd say it is at least 97% functional.

Its not a prototype for the M15 or M16. Those mixers were already on the market when this prototype was developed. It is a prototype for the M-50 and subsequently the M-500 series.

The backplane and motherboard design were taken from the existing M15/M16 products.
 
Hi,

This thread piqued my interest just over a week ago, been reading it on and off since then, and I've Just finished reading the last post.

A mammoth task you had, really well done. Judging by the photos in posts 565 & 566 it looks a really beautiful piece of equipment, all credit due.

Looking forward to reading further updates in this story.

The noise you are hearing is the appreciative cheers of the Tascam development engineers.:)
 
Hah! Thanks! :D

Glad you enjoyed it. :)

Been busy on a bunch of other minor projects as well as eBaying stuff for family, and getting my motorcycle project finished up, but the M-__ keeps looking at me from it's new perch (finally got it off the floor), and I'm getting ready to dive back into it. A mentor has been having me study equalizer circuit theory and history, so that will help me navigate the M-__'s architecture to get the oscillation problem tackled.

I'm actually planning to take a trip up to Seattle to pick up some gear and supplies soon and am arranging to take a couple modules to that mentor's place, put them on his bench and together to get the problem licked.

image.webp
 
Well, I made that trip up North...and my friend and I poked and prodded and conjectured...and we couldn't come to a conclusion as to the cause of the oscillation. And he's of the sort I figured if he can't determine the issue then I just need to let it go. He did give me a couple more ideas of things to try and/or check, and I set myself to do those things this evening but I couldn't...because it's no longer oscillating. I don't like it when problems "fix themselves", but for now I'll take it. So the M-__ is all back together, and soon I'll finish up where I left off, which was doing the comprehensive function test. I think there are 4 I/O modules left to test.

image.webp
 
Was it discussed if the oscillation issue is possibly related to the location of where it was previously, in close proximity to that radio station transmitter you've mentioned? Or is it still in that same room?

I ask because the room it's pictured in now is a different room.

Cheers! :)
 
...

I love that old board! You did such a great job restoring it. I can't wait to hear some projects coming off it.

PS: I love when problems resolve themselves!

:spank::eek:;)
 
The Ghost of FM said:
Was it discussed if the oscillation issue is possibly related to the location of where it was previously, in close proximity to that radio station transmitter you've mentioned? Or is it still in that same room?

I ask because the room it's pictured in now is a different room.

Cheers! :)
Click to expand...

No the M-__ is at my apartment, so nowhere near that AM transmitter. And the oscillation is most definitely an internal active component going into regular and measurable oscillation. Its an opamp oscillating at around 546kHz.

@ Dave: well, the problem I have with issues that resolve themselves is that often they return, and the cause remains unknown...its suspenseful.
 
Oooo I'll have to check the sign on my way out tonight. I don't recall off-hand.

As an aside I realized today I neglected to do the "shock" test. See, the module oscillates upon powering the module if the EQ is bypassed. I discovered some time ago that, after killing the oscillation (by either enabling the EQ or switching the LPF into the circuit), I could get the module to oscillate again by bypassing the EQ and overloading the input (like by tapping a mic with the trim at the upper end). So last night I found the module is no longer oscillating when the module is powered up, right? But this evening I did the "shock test" and I can still get the module to oscillate. So the problem isn't gone. Note I can only get the module to oscillate again if the EQ is bypassed and I overload the input. So the reality is the module won't go into oscillation if the EQ is enabled, and so maybe in a real-world sense this is a non-issue (i.e. maybe much of the time I'm using the console the EQ will be enabled), but it's not *right*. So I'm going to go ahead and try the other measures my friend suggested and see if I can get any further. Some of you might be thinking "DUDE let it GO!" But for me I've done SO much to this device and come SO far and it's important to me to turn over every rock rather than give up because I want to be done. I want it to be right. I will eventually reach a point where I go "okay...enough", but I'm not there yet. I will reach a point where a quirk, if it still exists, is accepted as part of the device...but I'm not there yet.
 
Well I sort of had a breakthrough...sort of...

So I did the "shock" test to the oscillating module that mysteriously seemed to have fixed itself, okay? The thump the mic test...and sure enough it sent the module into oscillation again. I wasn't surprised, wasn't downtrodden...actually relieved because I just don't like problems that go away on their own...they're just hiding.

Then I thought "waiiiiiit a minute...most of the modules were doing this oscillation thing at one point, and then it was just 3 of them, and then just this 1...but suddenly this 1 isn't doing it...*unless I make it do it*." Do you see where this is going? I learned about the ability to force the oscillation without power-cycling the unit AFTER all the other modules "fixed themselves". So, right...you guessed it...I hooked the mic up to every other module and did the thump test and SURE ENOUGH THEY ALL OSCILLATE! Except for TWO of them...why not those two?? BECAUSE I SOCKETED AND UPGRADED MOST OF THE OPAMPS ON THOSE TWO MODULES AND BYPASSED THE POWER RAILS AT EACH CHIP on those two modules.

I think I'm are dealing with a design flaw here. It may have existed before and I just never noticed it because frankly I was focused on refurbishing and repairing other things on it, and only just this Spring did I engage in doing the in-depth full-function testing which is how I caught this issue. If it didn't do it before maybe it is only now manifesting because I recapped the modules with lower ESR caps and maybe the old caps were providing just enough HF damping to keep this from happening. SO...I guess I've got some experimenting to do.

I think how I'll proceed is to get enough new TI branded TL072 parts to replace all that are currently installed on one module (12 per module), as well as sockets. I'll start with JUST the EQ board, and one at a time replace the input amp, and then the output amp (the two that are still in the circuit when the EQ section is bypassed). If that doesn't stop it then I'll bypass the power rails at those opamps. If that doesn't do it I'll try OPA2134 parts in there (which is what's installed on the two modules that don't oscillate). If that doesn't do it then I'll (ONE AT A TIME) do the same process to the next amp in the signal path. If that doesn't do it after replacing all the opamps on the EQ board I'll work upstream in the same fashion, and finally downstream.

I'm hoping new TL072 parts and (if necessary) bypassing the power rails does the trick. I like how the stock modules sound better than the ones I already hot-rodded and I'd rather keep things relatively stock if I can. Once I rectify the issue I'll purchase the parts necessary to do the rest of the modules.

The only parts I'm really interested in maybe "upgrading" would be the 4556 parts (two per module) and the one 2041 opamp. I think I can probably use OPA2134 parts there...I believe the 4556 and 2041 parts were used because, at the time, they had the necessary output current potential to fit the applications (i.e. fader booster for instance)...70mA and 25mA respectively. The OPA2134 handily trumps the 2041 for sure at 35mA, but is only half that of the 4556...the 4556's 70mA max output current is uncommonly huge and I'm hoping 35mA will do it...gambling that Teac didn't really NEED 70mA but there really wasn't anything in between the 2041 and the 4556. I'm not sure of that but I have yet to run into anything that, at the time, was in between.

Anyway, I'm actually relieved that the issue is universal across all the unmodified modules...I'm much more comfortable knowing that this is likely a simple design limitation, and rather than searching for a failed culprit the focus can be how to manage the issue, which may be very simple.
 
sweetbeats said:
Its an opamp oscillating at around 546kHz.
Click to expand...

Forgive me if I sound ignorant. But as this is well above human range of hearing, and well above the sample rates of most digital media that will eventually be delivering your content to end users... why is it a problem? Does it cause sympathetic oscillations or other problems to the recorded audible range?
 
You must log in or register to reply here.
Back
Top