Tascam M-___ Story...

[sweetbeats]

Yes, knowing what's wrong is well over half the battle I think, but there's no gaurantee the ugly -15V rail is causing the noise in the headphone circuit. And who knows what's going on with the peak LEDs. So at this point it's just three problems instead of one.

Not sure what I'm going to do with the M-500 modules. I don't own an M-512 or M-520 anymore. Nearly all of the knobs and color caps are there so that's a goldmine in spares to sell or keep...some of the pots are common with the M-__...so more than likely I will scavenge some spares from some of the sections, and sell the rest. I get excited when I get spares of those color knob caps because they are common to so many vintage pieces of Tascam gear and they *always* go missing.

 

[sweetbeats]

Heeeeere crappy part....c'mon little crappy part...

 

[sweetbeats]

Okay. Some progress.

So, to refresh, the problem I encountered last Friday eve with the power supply was this: the regulated +15VDC rail was throwing about 1.4V ugly AC. There should be essentially no AC voltage on a regulated DC power rail. The other thing I was concerned about was the -15VDC rail was >10% above spec...I think it was pushing about -17VDC. I'll get to that one further on in this post. First, a little history...detailed earlier in this Story, but I don't want to go looking for it:

With some guidance I upgraded a number of things on this power supply (which is a modified PS-520 power supply from an M-520 mixer) particularly around the +/-15V rails, since they power the audio circuits. One of the the things I did was replaced/upgraded the +/-15VDC regulators U001 and U002 with new low-noise TI parts, and relocated them from the PCB to that nice big heat sink. In stock form those poor regulators sit on the PCB with nothing to help keep them cool. OK I suppose...I mean, really, I have yet to run across any glaring under-engineering in Teac/Tascam gear from this era...but I didn't like it...cool parts are happy parts, and the M-__ has a fair number more active devices than an M-520. So my guide suggested I bypass the input and ground of the 7815 and 7915 regulators with a small value tantalum cap...this was suggested to ensure stability (to keep the regulators from oscillating since the regulators would be mounted distal to their filter capacitors). Okay. Well, it appears one leg of the bypass cap on the 7815 regulator must have been touching the body of the part. I did a little tracing around, and went with my gut after a little bit and bent the little suckers back away from the regulator bodies...see here...one of the little orange tant blob caps is right about dead center in the picture...that's the way it was before I bent it back:

Shazam. Both + and - 15V rails were now nicely and quickly resolving to about 1mVAC, whereas before the +15V rail was about 1,400x that. That should help.

And as far as the DC output, the +15V rail is sitting nicely at about +14.94V. The -15V rail is another story. It's at about -18.45V.

So now I need to figure out why the -15V rail is running over 20% high.

Maybe I need to check the voltage with the supply under load. That would make sense to me but I'm apprehensive about using the mixer as the load when it's running that high above spec.

 

[RFR]

Good deal. progress is being made. Funny how that tiny cap was causing so many problems.

Hey, I thought that looked like a 520 power supply.

Don't know what to tell you on the higher voltage, but you're a few steps closer. Congrats.

 

[sweetbeats]

I appreciate a chassis that can easily be partially disassembled for access.

 

[sweetbeats]

New 7915 -15VDC regulator installed...no change...-15VDC power rail still running greater than 20% out of spec at -18.71VDC.

New series pass transistors are soldered up and I'm about to mount them to the heatsink.

 

[RFR]

Is there a fingers crossed icon?

 

[sweetbeats]

That is a very nice thought...

And I wish I had a success to report, but not yet.

The original issues with the M-__'s power supply were:

1. -15VDC rail producing about 1.4VAC

2. -15VDC rail running greater than 20% higher than spec

I noticed the 7915 regulator bypass cap legs were touching the body of the regulator. I adjusted the position of the part which mitigated the AC voltage.

Tonight I replaced the 7915 regulator U002, no change...-15V rail still running at -18.81V.

I replaced the series pass transistors Q003 and Q004.

-15V rail now -18.4V, and suddenly the +15V rail which *was* about spot-on is now at about +14.1V, and that rail is producing about 1.3VAC.

Grrr...

My multimeter has been acting up...I determined the battery was almost dead.

Replaced the battery. Well, not really. Borrowed is more like it since every spare 9V battery in the kitchen drawer is DEAD, and I had to rob a battery from a smoke detector. The irony would be thick if my electronics project caught on fire.

-15V rail is actually running at -15.12V as measured by meter with borrowed battery. Yay.

BUT...

+15V rail is still throwing about 1.3VAC.

Grrr again.

I double checked the voltage at the bridge rectifier and it's appropriate (49VDC + to - output, and about 34mVAC across those same terminals).

I checked voltage across the input of the regulator U001 and then the output:

Input about 25VDC (good), and about 200mVAC...mmmm not good.

Output 14.12VDC (meh...bet it would be better were it not doing double duty regulating AC), and there's my 1.3VAC.

Because there is an inappropriate amount of AC voltage at the input of the regulator I'm looking upstream...I've already replaced the series pass transistor Q003...looking at the schematic and wondering about Q001, the current limiting transistor. Is it possible it is toast and oscillating I wonder?

Going to bed now after I yoink the battery outta the meter and stuff it back into the smoke detector so it will stop yelling (beeping) at me.

 

[sweetbeats]

I've spent many hours the last couple days on this...too much to be able to type right now, but I'm setting the power supply issues aside for now. The issue with AC in the +15VDC rail spontaneously went away...I've spent an hour looking over the PCB in excellent light with a jeweler's loupe and find no bad solder joints, fractures, etc...I can't get the problem to reappear so I'm letting it go for now because its working.

The mixer...

Good news: the stereo buss peak LEDs being stuck on issue is gone. I don't know where that problem came from, but with the mixer back in the frame its not happening.

Good news or at least relieving news, the master meters weren't working...they can source the stereo buss or the monitor buss, so I switched the source and they were working when sourced to the monitor buss...some quick thinking and I'd realized I installed a small buss PCB backwards...stereo buss meter activity back in business.

So I'm back to baseline which is sputtering crackly noise in the headphones, and a generally high noise floor.

Long story short I have determined it *isn't* just the headphones, its global to every output from the Control Module (the "master" module), and since the noise can by controlled by each buss' level control, the noise is upstream of the level each level control...it *has* to be coming from somewhere upstream of the Control Module...I re-do some tests I've done in the past as I'm now getting my head around maybe what's going on, and I verify the noise is NOT coming from the Control Module. So, good news, I can actually finally put the Control Module back together with all its cosmetic bits and call it "done".

So back to the noise...With the Control Module out of the frame and all the Input/output modules installed, I probe and monitor the motherboard connector terminals that bring the monitor, stereo, program and aux busses to the control module...there's the noise!!!

So it is coming from the motherboard itself or from the Input/Output modules. So I do some tests to see if it is one, or more, or all of the Input/Output modules generating the noise. I quickly deduce each Input/Output module is blasting out relatively high amplitude white noise from whatever program or aux buss slot it is plugged into...with them all plugged in they are blasting white noise across all program and aux busses, and I think the noise is bleeding across other busses and culminating in sputtering and crackling...I've got a global issue here.

My gut is the noise is not actually being generated within each Input/Output module, but rather there is some problem with the motherboard that manifests when a module is plugged in.

Here is a YouTube video I captured to memorialize the problem and convey to evm1024...it was late...my brain was dragging a bit, so I'm not sure how much sense it will make to viewers, but here it is nonetheless:



I'm suspicious of my grounding mods I did in 2014. Its a bit whacky on this console...normally you want to isolate audio power rail grounds in the power supply, from each other, and then bond them together with the mixer frame as close as possible to the entry point of the umbilical to the frame...then make sure all metal elements of the mixer frame are bonded together. Then any point of the mixer frame becomes a low impedance path to ground for any audio ground connection coming into or going out of the mixer...and the power supply chassis gets bonded to the mixer frame through the power umbilical. In this way induced noise interference is shunted as close as possible to ingress/egress points, AND you avoid ground loops. And then any NON-audio power rail grounds stay isolated from the audio ground plane. Well...here's the wrinkle...on the M-__ there is a +6VDC power rail that powers the meter lamps and all the LED indicators...but it also powers the headphone amp. I think....I think I bonded that power rail's ground with all the other audio rail grounds in the mixer simply because it powers audio circuitry (the headphone amp)...If I'm recalling correctly, I think that was a mistake...I'll need to check. I don't know if that is causing this problem, but I have a hunch it is causing *some* problem(s) because I think there are separate signal ground paths for the headphone circuit, and I think I might be creating a global ground-loop issue...not sure...need to test and experiment.

 

[sweetbeats]

Well, it's nothing to do with the +6V power rail or any grounding mods.

This is very interesting.

I have determined the noise issue does indeed propagate from each I/O module.

Maybe a video would do a better job of explaining this, but there is 40mV of DC noise coming from one of the negative *inputs* of the PGM MASTER dual opamp.

I don't know if the problem is the result of a prototypical design flaw and has always been there, or if something has failed across all channels.

All channels have been recapped. I might have installed a polar cap incorrectly across all channels, but last night I verified each of the related caps were installed correctly according to the orientation printed on the PCB, and I verified the printed orientation is consistent with how the original caps were installed (via pictures I took of the PCB prior to recapping).

I don't think it is a wholesale failure of the opamp, because I have replaced the original 2041 opamp on two of the modules with (I think) OPA2134, which does indeed make the noise more clear, bright and prominent , but the noise is still there.

These are true inline modules; the PGM MASTER controls and circuitry or on the I/O module. Signal assigned to a PGM BUSS goes to the motherboard through the module card edge connector. The motherboard distributes the signal to the summing input BUSS IN on another contact on the card edge connectors if the I/O modules. The motherboard traces ensure signals assigned to PGM BUSS 1 (for instance) only go to the BUSS IN terminal for a module connected to slot #1; PGM BUSS 2 to the module in slot #2, etc. From the BUSS IN contact signal goes to pin 6 of the 2041 opamp, which is a unity gain configuration buffer amp. Pin 5 is connected to ground. Pin 7 output goes to the BUSS MASTER level pot, the output of which goes to pin 2, the input of the BUSS OUT driver...the pin 1 output goes to the BUSS OUT jack which involves discrete level boost circuitry (switchable -10dBV or +4dBu), and also to the MONITOR buss.

I isolated the amp circuit from its destinations; there are connectors to disconnect the backplane and MONITOR buss from the output of the opamp. The noise persists on pin 6.

It's a pretty simple circuit...so I'm kind of baffled as to what is happening here.

Anyway, my plan is to draw up the circuit and get some help to analyze the circuit.

 

[sweetbeats]

BUSS MASTER amp schematic completed:




I'm pretty sure there's a resistor missing upstream of pin 6 of U1-2...there's supposed to be an input resistor there. I think it should be 22K...not sure yet if it should be in series with the input or strap to ground...I'm thinking it should be in series.

 

[sweetbeats]

Wow...

The M-__ Story hits 90,000 views. Now, I realize that could be one person looking at it 90,000 times during the course of the past nearly 8 years...but I am always humbled and a little amazed people care to follow these Stories that often seem to have no end, and the subject of which are often pieces of audio gear almost all the world cares nothing about.

This is a shout out to you all who check in on the progress and for the periodic posts. It means a lot to me and the project. Thanks!

 

[sweetbeats]

I just zorched the +/-35V power rails...a wayward test lead...

 

[sweetbeats]

Aaaaaannnnnnd that's why we have fuse protection.

Back in business.

 

[sweetbeats]

I hate it when something fixes itself...

I entirely can't explain it, but the noise problem is fixed...every bit of it...and I did nothing except zorch the "+/-35V" unregulated DC power rails. Those rails power the discrete level-boosting circuitry (i.e. switchable -10dBV or +4dBu nominal output) for the BUSS OUT jacks on the I/O modules, and on a number of outputs on the Control Module.

This is what happened: I was setting up to do some more tests and accidentally had a lead clipped to the -35V rail contact the +35V rail. As I indicated in my last post, it turned out I only fried the fuses; replaced them and proceeded with testing, but now the DC noise on pin 6 of U1 was gone...and gone from everywhere effected by it when it was present. I spent an hour or more chasing my tail thinking I didn't have something hooked up right. I ended up plugging the modules back in and taking it for a spin thinking "maybe...its...fixed...?"

No more high amplitude noise floor from the PGM BUSS groups, no more sputtering/popping/fizzing in the headphones and other master busses...its just simply working.

I don't know how triggering the fuse protection for those two rails could fix it...but that's the only thing that was executed (even if unintentionally)...no other changes were made.

ANYway...

 

[bachelorb]

.... I know you aren't a EE student, but this may apply here......

 

[sweetbeats]

Very true, Brad...very true.

Okay.

So this whole sputtering noise thing has been plaguing me for literally years now. I'm too tired to go into detail at the moment, but it's fixed now. I found the problem. And I killed it. And ate it for a late night snack. Raw.

The umbilical from the power supply to the mixer plugs into the back of the mixer frame. The wires corresponding to power rails on the back side of that connector go to another connector at the motherboard. It's redundant. And the type of connector and the pin and socket combination was a poor choice in my opinion. It is probably what was handy on the shelf; this is a prototype after all. I could just slightly wiggle the wire for the +15V rail for instance and make all kinds of fuzzy furry noise come out of the monitor speakers. Bye connector, hello wires soldered directly to the motherboard.

Time for sleep.

 

[Muckelroy]

So you... wait you fried the +- 35V rail, replaced the fuse, fixed the problem, then reproduced the problem again? And fixed it by directly soldering the aforementioned internal connector?

 

[sweetbeats]

Correct.

The problem has been a real bug-a-boo because it is intermittent, and I got rabbit-trailed several times with erroneous leads and other misadventures (failure of some components in the power supply, faulty readings from my meter because the battery was low, etc.). I never suspected the noise issue was caused by a lousy connector bringing power to the motherboard, but that is what it has been all along. And I rectified the issue by taking the connector (actually two of them) out of the equation and directly soldering the wires to the motherboard.

 

[Muckelroy]

"bug-a-boo". That's a pretty Rated G term for what I would call this issue if I had to troubleshoot it!

Happy to hear the progress was made!