Tascam M-___ Story...

[evm1024]

No smoke dump the light

The Light tests for hard shorts and limits the current to the DUT(desk under test ) should it be shorted.

Nothing blew up and the smoke stayed inthe wires so you can lost the light and let the desk have its full share of power.

Peak lights staying on - Osc comes to mind but you would have to check dc values as well.

--Ethan

PS good work. I'm snow bound (hah, not really) and I've decided that I need to finish off my Great River preamp project now!)

 

[sweetbeats]

ponderingponderingpondering...

Maybe I was wrong about the 0V rails...remember how I noticed that one of the 0V rails in the PS-520 (the one associated with the 6V rail) seemed isolated from the others even though it connects to the common rail? That you didn't get 6V on the voltmeter unless you put one probe on the 6V rail and the other on that one 0V rail associated with the 6V (I think I called that 0V #3...)? I made the assumption that that was the right match for the 0V rail in question that goes to the meter bridge and the master Control Module...

I wonder what would happen if I used 0V #1 or #2 instead...totally fishing here...the dangerous thing is I don't know enough about electrical circuitry to truly understand what is happening here in order to hone in on isolated probable causes, and therefore I have to go fishing in a very big sea and possibly miss the cause altogether.

Where's Ethan when I need 'im??

I triple-checked every cap that I put in for value and polarity orientation...I was very careful, but that doesn't mean it woudn't be worthwhile checking again...

And I guess I'll desolder the cross-connections between the meter amp PCB's and try just one at a time and see if they all exhibit the problem or maybe it is one bad apple...hate to pull them apart because some of the solder joints and pads are looking pretty ragged, this being a test-bed/prototype board...some joints appear to have been soldered/desoldered/resoldered maybe even a few times...

WHY DIDN'T I GO FOR AN EE DEGREE?!?

 

[sweetbeats]

sheesh...that's like the third time that's happened

where I'm hoping for input from you Ethan and you've already given it...its like putting instant coffee in the microwave and going back in time.

So I can lose the test light for now even though it glows steady??

And even if I can lose the light, where do I look for oscillation...the IC's on the meter amp PCB's? I imagine I'll need a scope to find it, or narrow down the PCB by trying them one at a time and then swapping the IC's out for new in a blind test?

huh...

The other thing that is making me concerned is that the VU needles peg hard on power up...they don't stay up there but you can hear them all *clink* when you turn it on and I'm not used to that...

The issue does seem to be a conflict between the 6V/0V#3 connection (on one connector from the motherboard), and the +/-15V/0V#1 connector. If I power up with only the latter connected, the meters do their peg thing but naturally the peak LED's and meter lamps stay off (since they depend on the 6V to operate). If I keep only the former connected the lamps light but the LED's are dark and of course the needles don't move (since they depend on the +/-15V to operate). In either of these cases the test light stays totally dark...if both connectors are connected then the meters peg at power up, meter lamps are lit as are the peak LED's and the test light glows constantly but dimly.

I'll shoot a short video and put it up later to demonstrate...probably will make more sense.

 

[evm1024]

out shopping in the snow storm.

first thing to remember is that 0 volts is 0 volts because we say so. It is the reference point. Now it happens that the reference is so common that we tend to tie it to the chassis and to the ground and to just about everything else. This might be obscure or so obvious as to not be worth speaking but it is fundamental and should be remembered.

I had forgotten about the isolated ground. As I recall from the PS520 the 6 volt was isolated to power the peak LED so as to keep noise off the audio circuits.... I had thought that the 6 volt ground and the analog ground wold be tied together at one point.... I'll have to check some day.

Have you tied all the grounds together? It may be that this ground needs to be floating from the rest of the grounds (Ha, the real ground) Too bad there is not a schematic.

I'll take a look at the m520 schematic to see what they do and make some suggestions from there. I specifically want to look at the meter section power and what the 6 volt is doing.

As for cap polarity. You would know if you reversed any of the electrolytics. They tend to blowup when reversed biased. Think loud pop to small bang.

No need to think about oscillations as that this is a ground reference problem most likely.

--Ethan

 

[evm1024]

Ok

I'm making the assumption that the m520 and your board have like electronics. Your mileage ....

Take a look at schematic 4-15 Input amp. This is abouut the only place that +6 volts is used. There are 2 usages on the page. One is to light the indicator LED. Just a simple switch and LED such as the EQ on and AUX Mute. The other use is for the Overload LED.

Look at U4. This comparator has its - power pin (4) connected to the 6v ground and its + power pin (8) connected to +15. WTF you might say. However this appears to be a little engineering trick to cause the output of the comparator to be pulled to the 6v ground.... Keeps some switching noise away from the mic preamp circuits.

THe 6 volt ground 0V(d) is tied to the system ground 0V(c) in the PS.

Not sure if this gets you anywhere. I would look to see if there is a LM393 or other such device driving the LED and then check what the voltages are on pin 4 (0v) and pin 8 (+15) and if the ground is the 6 volts ground. You might lucky and have the circuit much the same as the m520.

-Ethan

 

[sweetbeats]

Thanks, Ethan! I was gonna say that all grounds tie up in the PS-520, but you beat me to it. I put up tht hard-to-read PCB layout of the Meter Ampl. PCB for the M -___ some time ago...I'll take a look at that again. I believe it does utilize an LM393 as well as a 4558...IIRC tho' the LM393 functions in the same role as the same chip in the M-500 meter amps...I'll be able to tell from the PCB layout what is doing what and I'll report back. BTW, in your last post in the second to last sentence you said " check what the voltages are on pin 4 (0v)"...did you mean pin 4 (6v)?

 

[evm1024]

the lm393 has a + and a = supply. pin 8 is the + at +5 and pin 4 is the - PS at 0v(d) If they are the same then something else is wrong if they are different them we need to look further (Assuming a m520 type design)

--Ethan

 

[sweetbeats]

Ah...it is an LM339 with the 4558, not the 393...I'm trying to map it out now ...

 

[sweetbeats]

Meter Amplifier Schematic

I drew up a schematic for the Meter Amplifier...Like the M-500 mixers each Meter Amplifier PCB serves 2 meters...siamese amplifier circuits...I only drew one side. No guarantees that it is not error free.

Below is a jpeg. If you need a better version there is a full-size .pdf here.

I haven't studied it yet. Might seem funny but I was concentrating on understanding what was connected to what, not so much what that meant or what was happening.

I labeled the power rails relative to what they are called in the M-500 schematics, and they are labeled as they are presently connected to the M-___ meter amp, not as they neccessarily should be. Make sense?

 

[evm1024]

Same card

Darn close to the M520 meter amp circuit. Take a look at page 185 (Diagram 4-10) on your 520.

You could compare voltages between your 520 and mystry mixer to help you figure out what is going on.

First thing is to look at the voltages on U1. Pin 12 should be at ground and pin 3 at +15. Both 0V(c) and 0V(d) should be the same (measure the voltage between them on both DC scale and an AC scale). If there is any voltage between them find out why. Trace back the wiring.

Diode D1 keeps the input of U1 from going negative. R3 and R2 form a voltage divider and set the trigger point for the comparator. Any voltage about that and the LED should light.

R6 and R7 set the second reference voltage - the exact voltage is not very important (not like the voltage set by R3,R2 which is the overload trigger voltage). Cap c1 is charged to +15 volts by r5. R5 limits the current available to charge C1 so it takes a small moment for C1 to come up to +15 volts.

When a peak is detected the output transistor of U1-1 is turned on (the output is a transistor with the collector tied to the output pin - look at the datasheet to see the chips schematic) This "open-collector" transistor when off does not conduct. Its emitter is grounded. When the output transistor is on the collector is pulled to ground. So when the transistor is on pin 2 is grounded and C1 is discharged (quickly) through r4. Even a microsecond (WAG = wild ass guess) of a peak causes C1 to discharge. this means that the voltage across c1 is near 0 and presents low voltage to U1-2 input on pin 7. The input on pin 7 is less than the reference on pin 6 so the output transistor on this comparator is ON and the output pin 1 is pulled to ground 0v(d) and the LED lights.

Remember that C1/R5 have a time constant. What this means is that it takes some time for the voltage on pin 7 to go above the reference voltage and thus the LED stays lit for enough time for our eyes to see it - even for the michrsecond peak.

Sorry for the long winded thinking out loud. I'm hoping that this gives you enough to go on to figure out what is causing the LED to light.

Regards, Ethan

 

[sweetbeats]

let the wind blow and blow and blow...

Ethan, as always, your time and advisement is greatly valued. AFAIK you can be be as 'long-winded' as you please.

I will make good on your time investment with that last post and study, learn and apply in the next 24 hours...I'm sure I'll have questions...I'm sure I'll learn a lot too, and ultimately I hope to get my meters to work.

Anybody have any comments about the meters momentarily pegging on power-up? Is that as abnormal as my limited experience says it is?

 

[Muckelroy]

Anybody have any comments about the meters momentarily pegging on power-up? Is that as abnormal as my limited experience says it is?

I recall seeing that happen on an MCI-JH24. Granted, the machine was out-of-service as 7 out of 24 tracks had issues, but I did notice that when it was powered on, the meters pegged up real quick and came back down.

Whether that's expected behavior or the JH-24 had other issues causing that is beyond me. Again, the machine was out-of-service, and it was in somebody's office serving as a paper-weight

On that tangent, when I was at that school, I tried to revive a more-promising MCI JH-110, but unfortunately the project was never finished...

 

[evm1024]

Most meters jump

Most of the meters on my "old" hardware jump on startup. Sounds like the power supplies coming into regulation at differing rates. None of them look like they are going to bend the needles though.

Side note: I always wanted a JH-110 (most likely a B) and still look at craigs list for one... Perhaps an irrational desire but real none the less.

--Ethan

 

[evm1024]

Most meters jump

And how to delete this double post....

 

[sweetbeats]

Video

Here is a video I captured about the whole meter pegging/peak LED thingy...I also cover the power rails on the M-___ and where they connect, etc:

YouTube

Somewhat anticlimactic at this point since I came up with a schematic and Ethan responded as above, but I had indicated I would put up a video so there ya go.

One thing to highlight here is that I confirmed that what I have been calling 0V#3 (that is equivalent to 0V(D) in the M-500 schematics) is indeed completely isolated to the 6V rail. 0V#3 will not form a complete circuit for the 48V, 35V or 15V rails, and neither will 0V#'s 1 & 2 (equivalent to 0V(C) in the M-500 schematics) form a complete circuit for the 6V rail. So I actually studied my own schematic last night to see first if there was something illogical with the flow of things at the present (i.e. +/- rails trying to resolve to an incorrect 0V rail), and I didn't see anything obvious, though I did keep nodding off at the time...

So that's good in a way and not so good in another...It sorta means my assumptions were right, that the actual circuitry matches what I hooked up to each terminal power-wise. What is not so good is that it is definitely not working and therefore more investigating is required.

So it is high-time to take a close look at Ethan's post and see if I can figure this sucker out.

I must say, however, that I continue to be completely baffled as to how 0V#3 is isolated from #'s 1 & 2 when they are all connected to the same trace in the PSU...???

Furthermore, since I'm understanding things better, I think I need to do away with the whole idea of labeling the 0V rails in the M-___ as #'s 1, 2 & 3...I'll stick with 0V(C) and 0V(D)...I originally went with the former not knowing whether or not they were 3 distinct and isolated 0V rails, but it looks as though the M-___ thirsts for only 2 of them, just like the M-500 mixers so we'll stick with Tascam's nomenclature of 0V(C) and 0V(D).

 

[evm1024]

I must say, however, that I continue to be completely baffled as to how 0V#3 is isolated from #'s 1 & 2 when they are all connected to the same trace in the PSU...???

-----------------
Isolated as in spikes on that ground don't cause audible clicks. Common practice to have one point to tie all grounds together. That one pint is the 0v reference. Once you move away from it Ohms law comes into play and there is an offset in the ground due to the resistance of the traces, connectors and wires.

Here is something to try - just jumper the 2 grounds together right on the meter board and see what happens. (Might want to measure the voltage difference first just to be sure.)

This will force a common reference point and perhaps cause the LED to not light. I'm not remembering seeing the grounds tied together on the PS PCB.

I might go check on mine.

---Ethan

 

[sweetbeats]

Wait...

Isn't it an issue that pin 3 of U1 on the M-___ is connected to +15V (okay no problem there) and pin 12 is connected to 0V(D)???? Remember, 0V(D) doesn't complete a circuit with the +/-15V rails......huh?

Okay, also measured AC and DC voltage between 0V(C) and 0V(D) and there is about 0.68VDC and about 20~30mVDC...that is measured at the connections for the meter bridge as well as coming straight off the PS-520. Problem? So in terms of tracing the wires back to find out where the problem is, it is in the PS-520...

I'm going to go try a different DMM and test voltage between 0V(C) and 0V(D) on my M-520 PS-520...

Maybe unimportant or unrelated, but why would VAC or VDC drop slightly when the DMM probes are in the mixer end of the power umbilical (so not touching the mixer at all) and I touch the mixer frame??

I'm not remembering seeing the grounds tied together on the PS PCB.

The 0V(C) and 0V(D) rails are tied together on the schematic for sure, and IIRC on the PCB too.

 

[evm1024]

As designed, As built

Ah, I do remember correctly - Take a look at your photo and count the PS output pads up fro the bottom. Pad 6 is the +6 volt and pad is 0V(d). It is not tied to the other grounds here! You should tie them together at a logical common ground point. Which would be the ground reference for everything.

+- 15 are referenced to 0V(c) and +6 is referenced to 0V(d) and 0V(c) and 0V(d) should be tied together so that they do complete the circuit taking a longer route that does not go through the analog section ground.

Tie them together with a jumper and see if the LED goes out. The photo is the key.

--Ethan

 

[evm1024]

Think of the comparator as a....

Think of the comparators as some analog circuit that compares the input voltage to a reference voltage and turns a relay on or off. The relay is the output transistor in the comparator of course.

The analog section is powered by the +15 volt rail and the return to the +15 volt supply is through 0V(d) along the trace until 0V(d) it connected to 0V(c) where it travels along that trace until it makes its way back to the +15 volt supply.

The output transistor is on or off. When on is shorts pin 1 to pin 12 which is connected to 0V(d) and the +6 v goes through R?, the LED , the output transistor and returns to the +6 volt supply via 0V(d).

--Ethan

 

[sweetbeats]

I used my Fluke DMM that, unfortunately, has a flukey LCD screen, but both my modded M-___ PS-520 and the unmodded M-520 PS-520 measure about 0.7VAC and 0VDC at the power supply output on the chassis of the PSU...Is this by design, or are both my PS-520's messed up?

+- 15 are referenced to 0V(c) and +6 is referenced to 0V(d) and 0V(c) and 0V(d) should be tied together so that they do complete the circuit taking a longer route that does not go through the analog section ground.

Tie them together with a jumper and see if the LED goes out. The photo is the key.

But-but-but...they don't tie together in the M-520...do they???

Okay...regardless, where do I tie them together??? What is a logical point? Inside the PSU since that is what the schematic shows? Run a jumper between the two where they connect to the motherboard?

I don't totally understand why there are two separate 0V rails if they tie together anyway. I think this is beyond me.