Tascam M-___ Story...

[sweetbeats]

So, random post. But maybe not so random.

The M-__ has mostly been just sitting under a custom LeCover for since my last post, but recently I’ve been prepping to capture a series of videos detailing the features and functions of the M-__. This is something I’ve been wanting to do for a long, long time…a video operations manual of sorts. I’ve shared that previously in this long forum topic. But anyway, because I’d like to finally capture those videos I’ve been taking the time to *finally* tease out details of the M-__’s functions and signal flow…I know it’s been, like, 18 years, but life has been crazy, so forgive me. I’ve never fully gone through nailing down ALL of the details of the functions and signal flow aspects of the M-__.

What I mean by that is detail stuff I can’t remember because I don’t use it enough, like the five different SOLO buss sources on each input/output module…where does each specifically tap the signal chain? The MONI TO STEREO switch that sums the monitor buss source on each module to the main buss…is it pre or post channel fader? The CHANNEL MUTE kill switch…what all does it actually mute? There are a total of 19 output channels from each input/output module…I knew off the top of my head the CHANNEL MUTE switch did NOT mute the DIRECT OUT jack. What else? Anything? Now I know. I have a couple other things I want to verify yet, but I’m nearly done going through the whole console with a fine-tooth functionality comb.

In the process, however, I became sidelined with finally clearing up some annoying issues…the signal flow to the STEREO buss has been problematic…both to the audio output as well as to the meters…intermittent…really annoying. And there has also been an annoying problem with the headphone output related to higher than normal noise level and DC popping in the headphones when any of the monitor buss global REMOTE source switches are pressed on the Control Module, which is what Tascam named the module that contains the master fader, oscillator, talkback functions and studio and control room functions as well as the master monitor and “remote” control functions. BUT I’m happy to say I’ve tackled those issues…some reflowing of solder joints as well as reinforcing the wire management took care of the intermittent STEREO buss signal flow and meter monitoring. I’m also finding it’s been enough years now it’s needing some maintenance on the edge connectors that interface the modules to the mother board, so checking the contacts in the sockets on the motherboard as well as a little gentle polishing of the contacts on the module edge connectors and then treating with DeoxIT D100…exercising, etc. It wasn’t bad, but it was time to do something and I want to take that variable out of the mix if I am troubleshooting a problem. Anyway, STEREO buss problem fixed. The headphone issue…I used to have a quote at the bottom of all my posts: “Never underestimate your capacity to be a complete ignoramous”…something like that. Well, I’ve done it again. Many years ago I cleaned up a VERY kludgey abomination of a circuit block on the “Phono Amp PCB” assembly in the Control Module. That’s the right-most PCB assembly of the four main boards in the module. I can’t remember what the function of that circuit block is at the moment, but I detailed it in this thread years ago…stacked opamps, one in the board and the other upside-on top of the first, just hanging in free space held only in place by the point-to-point wiring and components soldered to its pins…no judgement…it is a prototype device after all, and sometimes these kinds of things are even incorporated into production models by any manufacturer. But it was messy and I wanted better access to both opamps, so I modified things, moved one opamp into a vacant DIP8 site on the board, incorporated some jumper wires, etc. In that process I noticed the signal ground pins of the headphone level pot were referenced to signal ground via the shield of one of the output wires. So let me explain the level pot and the source switches for the headphone amp are on the right-most of the four PCBs, but the amp itself is on the left-most of the four PCBs. So there are wires from the two output pins of the level pot on the right-most PCB that go the amplifier input solder joints on the left-most board. The shields of those wires are also terminated on the PCB with headphone amp circuit block. But they also terminated the shield of one of those wires on the right-most board on a hand-added solder pad in order to connect the signal grounds of the level pot to signal ground on the PCB with the amp. Maybe that’s fine. But I wanted both channels’ shields to float at the pot so they could both properly shunt any induced noise interference to ground on the board with the amp…float at the source, shunt at the load. So I did that. And then found the closest place on the PCB with the level pot to strap the signal ground to the appropriate ground path…the headphone amp is powered by a +6VDC power supply, which also powers the meter lamps and LED indicators, so the ground reference for that supply is on almost every PCB in the console. So it made sense to me to just jumper the headphone level pot signal grounds to the closest spot where that ground was available on the same PCB as the level pot. Nope. What I didn’t understand at the time or think through at the time is the potential for DC offset by doing what I did. The original reference to ground through the shield of the one signal conductor strapped the signal ground of the level pot close to the amplifier circuit block ground path, as well as very close to where the PCB with the amp references system ground at the motherboard…and more importantly the +6V rail reaches the LM386 opamps in the headphone amp directly…no dropping resistor. However the +6V rail goes through a series dropping resistor on the PCB with the level pot. So that enabled some amount of DC offset in the power rail for the headphone amp between the pot and the amp (enabling the *pop* when I would latch any of the monitor buss global REMOTE source switches that are on the same PCB as the level pot, and the switches also power an LED indicator when latched), and beyond that actually increased the risk of noise interference in the headphone amp circuit…and in fact I think raised the noise floor of the headphone amp. I think the reason I didn’t notice after I did those mods years ago is because I wasn’t really focusing on the headphone amp after the mods (more focused on the modified amplifier circuit), and then probably didn’t do anything operationally with the console after that…I think I moved shortly thereafter and then the console was in storage for awhile. By the time I noticed the new problems I bet I didn’t associate them with the mods I’d done because substantial time had passed. Anyway, ground issue fixed…I ran a separate connectorized ground wire from the level pot signal ground pins to the left-most PCB on an available solder pad close to the headphone amp circuit block…no more pop, noise floor better, and proper attenuation. Fixed!

And here is the Control Module ready for final reassembly and reinstallation in the console frame, so I can tease out my last couple question regarding functionality…

Next step is to write out all the details of the control functions on pages I’ve printed with sections of each module…number each and document narrative like in an operations manual…and then capture a video series covering all of this and get it up on YouTube.

 

[sweetbeats]

Well dangit…not fixed…two new issues…one of which was likely there all the time, and the other probably the result of my reflowing of solder joints…

Problem #1: the right channel of the STEREO buss allows bleed from all sources assigned to it even when the master fader is all the way down…from the I/O modules, from the oscillator, etc. It happens regardless of how the main buss is monitored…at the buss’ output jacks, or via the from control room or studio monitoring banks…in the headphones, everywhere. I’m thinking this will be pretty easy to sort out.

Problem #2: some weird phase cancellation between the CR monitor and STUDIO monitor banks. If the same source is selected in both CR and STUDIO source select switchracks, and CR is also sourced in the STUDIO bank, the signal cancels if the selected buss’ level control (like the STEREO master fader if monitoring the STEREO buss, or the monitor buss level pot if monitoring the MONI buss) is swept to a certain level. This has likely been there all along and is the result of all the hackery that’s been present in the Control Module ever since I’ve had it. I’ll have to start investigating.

And this is how it goes. It’s a prototype.

 

[sweetbeats]

Fixed it…sort of.

So with the first problem with the bleed-through in the right channel of the master buss, I figured either it’s reaching the outputs in some way outside of the fader and fader booster amp, or it’s somehow a problem with the master fader. So I figured one way to test that is just remove the master fader. If the problem stops it’s either upstream of the fader or a problem with the fader. So I did that and the problem was gone. Okay. So then I figured I could test if the problem is the fader itself by swapping the L and R connections…if the problem was upstream of the fader it wouldn’t matter which fader element was connected to which amplifier PCB. Conversely if the problem followed the swap (if there was now bleed in the left channel), then the issue resides in the fader assembly. The problem followed the swap. So I pulled the fader assembly out again to open it up and take a look. These faders are older ALPS assemblies, like pre-K fader type…tall vertical conductive plastic element board, wipers are very fine filaments, polished slide rails, etc. Pretty nice…I couldn’t guess what the problem might be. And you know what? I’m still not sure. The wiper carriage is an assembly of cast aluminum and molded nylon. The nylon tends to shrink over time and crack, so I did see that, but I couldn’t identify what or how that was causing an issue. I did install some mini zip ties to help hold the carriage together. So, I still don’t get it…but everything is now working correctly. I don’t usually sit comfortably with “I don’t know how or why, but it’s fixed”, but in this case I’m pretty okay with it because at least I know it’s the fader. If the problem resurfaces I’ll pull it back out and dig deeper. The good news is with the modular plugin construction, I can have the fader accessed and removed in about 20 seconds.

Now for the second problem, the wierd phase cancellation between the control room and studio monitor banks, well, I chose the very easy path…I did nothing. But this was after taking some time to just mentally deliberate the issue…time to tease out a question I had when I first discovered the issue. The problem is this…the console is all analog and unbalanced internally right? So it’s not like I have hot and cold conductors to swap as in a balanced line phase inversion…and that was my question when I was just discovering the problem…if there is only a single AC signal conductor, how could it possibly be inverted when sourced both in the control room monitor and studio monitor switchracks and summed? Short answer is it can’t, at least not at the signal conductor level…if I’ve lost my mind on this and am overlooking audio signal 101 somebody please tell me…anyway that led me to realize it is likely an issue with signal paths in those two monitoring banks, and a design error where there is a misalignment in the number of inverting and non-inverting opamp stages. For those that understand opamp theory hopefully what I’m saying makes. I know enough to be dangerous. But opamps have inverting and non-inverting inputs, and care has to be taken so the outputs of each circuit block are aligned either inverted or non-inverted, or you run into issues just like I’m experiencing. So I think that’s the issue. And I’m not going to dive in and try to identify the specific issue needing redesigned, because it is highly likely there will be a domino effect from that. I’m going to need to chalk it up to prototype “charm”…I could even probably make lemons out of lemonade and find a way to use the flaw to my advantage when using the null technique to set tape machine azimuth. But the reality is the problem only arises when monitoring the same exact source between both monitoring banks, and then summing them…and there is never a functional reason to do that…for instance to source the main buss in the control room monitor bank and also in the studio monitor bank and then sum those together and listen to them both in the headphones or whatever. No. Certainly might be sourcing the same thing in both, but for the purpose of sending playback to the recording room as well as listening to playback on the control room speakers, but not summed. And there’s not a general functional problem with one being inverted compared to the other because the outputs of those monitoring banks are not used for tracking. They are used for monitoring. But I think I want to do a little further testing to verify my theory and check if the problem exists at the output jacks as well without summing using the “CR” source switch in the studio monitor switchrack. It would matter to me if it was the control room out inverted compared to the rest of the console vs the studio monitor out, and I think I know how to test that. Sorry this all probably doesn’t make a lot of sense…it would be easier to demonstrate in a video I think…selfishly part of the reason I’m info-dumping this here is for my own reference down the road.

I had one more little thing happen…the LED indicator for the global REMOTE source switching for the MONI buss on the Control Module wasn’t making good contact with its socket anymore…wasn’t lighting up. Nothing I was doing as far as tweaking the pins was working. I looked closer and the position of the little daughter board the socket housing is soldered to is further away from the control panel, so the LED when flush with the control panel isn’t making good contact. To fix it right the entire Control Module assembly would have to be completely disassembled in order to desolder, reposition, and resolder the daughter board to the Phono Amp PCB. Not doing it. I consider the Control Module assembly to be relatively fragile. The phenolic resin PCBs don’t help, but neither does all the prototyping activities, cut traces, jumpers, resoldering and probably countless insertions and extractions, both preceding my acquisition of the console as well as all the repairs and discovery I’ve done since I’ve had it. Every time I have to take it apart I worry what else is going to break. When I do discover something else that’s failed I re-do it in a way that will last longer, but it’s increasingly important I think to be mindful of minimizing disassembly and handling. So instead I was able to just desolder the socket housing itself, tweak the position, and resolder. Its good now. Not “right” as much as I’d normally like to accomplish, but in this case weighing the pros and cons I felt it better to avoid full disassembly for something I could address more locally without any disassembly.

SO…now…maybe…after a few more tests regarding the control room and studio monitor bank phase inversion issue, I can get to capturing that video series detailing the features, functions and construction of the Tascam M-__.

 

[sweetbeats]

Ohhhhhhkayyyy…so I verified the same phase cancellation occurs at the STUDIO OUT jacks just like in the headphones if I select the same source in both the studio and control room monitor select switchracks, and then sum those in the studio monitor bank by also latching the “CR” source switch in the studio monitor switchrack. This means it’s not some weird thing isolated to the headphone pathway, which is kind of what I figured. The next thing after that was to determine if the phase was inverted at either the STUDIO OUT or CONTROL ROOM OUT jacks, which I was expecting would be the case for one of them, but was hoping not. I tested this by connecting the same source to the line input of an I/O module as well as one of the EXT inputs, assigning the line input to a PGM group, sourcing the EXT input to first the studio monitor bank and then then subsequently to the control room monitor bank, and then in each case taking the output of each bank at its respective output jack and connecting that to the same PGM group BUSS IN jack to sum the signals back together…basically taking a solitary signal source, splitting it with a ‘Y’ cable and inputting one side to an I/O module line input, and the other to first the studio monitor path (via the EXT line input on the Control Module), and then summing them back together via a PGM buss, listening to the result, and then repeating the same except through the control room monitor path. And you know what? In each case the summed signals did not cancel (no phase interference)…they did what the should which is increase by about 6dB. SO…this means the problem ONLY manifests when sourcing a common signal in both the studio monitor and control room monitor switchracks and then summing them via the “CR” or control room source select switch in the studio monitor source select switchrack, which, again, there is zero circumstance I can think of where that would be useful or necessary from a music production standpoint. I actually still think it could be useful in some way as a diagnostic tool. But regardless I believe it to be a design flaw. And I don’t care. At this point, since the problem only manifests when doing something procedurally that is totally impractical (again, from a music production perspective), it doesn’t bother me that it’s there, and I rather consider it to be endearing.

Thinking about it technically, and doing some theorizing about why it is the way it is, IIRC the crazy point-to-point glob kludge of piggy-backed opamp and components I cleaned up was on the Control Room PCB…and IIRC the circuit block was related to the MONO function in the control room monitor switchrack…the mono function sums the stereo sources of the control monitor bank to mono to check mixes for mono compatibility. This affects both the headphone mix when sourcing the control room bank, as well as the CONTROL ROOM OUT jacks. I’m wondering if the phase issue was identified and the kludge-glob added to correct the issue at the output of the control room bank to the jacks and headphone amp…but maybe maybe the “CR” source switch in the studio monitor bank sniffs the signal before the kludge-glob…I think this is correct because, IIRC, the “CR” source switch in the studio monitor switchrack is pre-MONO function in the control room pathway…but…that’s something I’m not sure I’ve tested…anyway, FWIW this is what it looked like before I modified it and cleaned it up…this cluster is on the back-side of an opamp mounted in a socket on the component side of the Control Room PCB:

And now I think I can move on…except…no…I want to verify if the “CR” source switch is pre-MONO summing function. It would make sense for it to be.

 

[sweetbeats]

Well that’s annoying…the MONO function is indeed pre-“CR” source switch, which makes sense, but the headphone amp source is also upstream of the MONO switch. What that means is that if you want to audition your stereo program in mono, it has to be done on your control room monitors. I don’t know if this by design or its a design flaw simply due to fact the console is a prototype…normally I’d expect a function like that to apply to both headphone and control room outputs, and in fact that’s the way it was implemented in the M-50 and M-512/520. But it is what it is. I may be wrong about this but in a dusty corner of my mind I think I recall looking at the signal flow in the control room switchrack more closely some years back and finding that the MONO switch was actually post DIM and MUTE controls. Which still doesn’t make sense to me if that’s the case, but certainly would reinforce why the MONO function doesn’t apply to the headphone circuit, because it makes logical sense the DIM and MUTE functions would not, and if the MONO function is post DIM and MUTE controls then naturally it’s not going to be heard in the cans. Moving on.

 

[BeatleFred]

Hi SB,

Thanks for the latest update on your project, interesting info. I can see you have invested a bit of time in your troubleshooting & analysis efforts. When it comes to those vintage mixers (and recorders) were they all designed by engineers in Japan, or by the personnel in California, just curious.

After spending some time getting acquainted with the M-30, I spent the past weekend doing the same with my M-1516. Although its not the same complexity as what you were investigating, I find myself in a similar situation when it comes to questioning why certain things were done back then in terms of signal routing options.


Overall, I like the M-1516. Everything is laid out logically from what I can see. I like certain things such as the Mono button and the EXT. EXT makes it easy to connect a 2 track such as the 22-2 to it, hit one button (EXT) and easy to hear the recorder.

The setup I have for now is a vintage Roland SIP-300 preamp that I plug my electric guitar into, which gives a preamp boost to connect to my ART VLA II compressor, which is connected to mixer channel 1's Insert. I also have a Lexicon Alex effects unit and a Roland DEP-5 connected to the Aux 1 & 2 stereo jacks of the M-1516.

An example of something I can question would be: since I am using a compressor, its best placed in a channel insert. But since I am using the Roland preamp, it does not need to go thru the Trim control, so I figured connecting the Roland preamp output to the mixer would be best put into a Line B jack, and the M-1516 has many of them on channels 9-16, but there are no channel inserts on 9-16, so I cant use the compressor with the guitar signal. Its not a dealbreaker, I simply just connected the Roland preamp to channel 1's Tape input which bypasses the trim. And since the compressor and Roland preamp are on the same channel strip, I can use them together. Its just that to me, its great that the mixer offers an option that would be useful, (Line B- no trim), but you cant use it with another device, because the other device requires an insert.

Another example would be the effects in the Aux Send/Returns. The M-1516 offers a set of Dual outputs (1/4") which is nice, BUT.... the dual outs DONT allow for the effects to be heard wet, only the clean, dry signal. Again, not a dealbreaker, I can use the Monitor outs instead which can play the wet effects. So, in this case, my thought is: nice to have dual outs but wont work for wet effects. Looking at the op manual- it mentions the dual outs can be used as an ADDITIONAL effects send, but for my purposes for now, it would have been preferred if the Dual outs could have the same capability as the Monitor outs in terms of ability to play wet effects.

I also find it interesting that the Monitor outs are rca jacks but the Dual outs are 1/4". I would think by the early 90's when the M-1516 was new, that self powered monitor speakers were not unusual to have. Thus, I was thinking that using xlr balanced jacks on the M-1516 monitor outs to connect to the monitor speakers xlr jacks would be better to keep any noise away, vs unbalanced rca cables, but M-1516 has no XLR outs.

 

[ecc83]

Re the "glob" "Oooo! Nasty!" I wonder if anyone here can give me the source of that phrase? Robby A is probably the only one if he is old enough!

 

[sweetbeats]

What an interesting journey this thing is…

So a bunch of stuff has happened in the last week, and I’m going to ramble about it right now.

I decided I was just going to leave the MONO function thing alone…like not even look at refreshing my memory about where it sat in the signal chain. I had the Control Module all back together, and I decided to leave well-enough alone and move on. But when I loaded the module back in the frame to run function verification testing, I identified something wasn’t working right with the REMOTE switching functions…like some of the functions worked, and some didn’t. Like maybe everything would work except the LINE 1 source select in the REMOTE A bank and the BUSS source in the MONITOR bank. To make a week-long story (somewhat) short I now know EXACTLY how those switch banks operate. Each source switch passes -15V to specific channels of the 4069 logic chips in each I/O module. There are three 4069 chips in each module…one for bank A, one for bank B, and one for the MONITOR bank. The 4069s act as analog switches for their respective functions whether controlled locally or globally from the Control Module. There are also a series of JFETs associated with each 4069. I’m not sure if their function is to defeat the local source controls when global control is active, or if they do the signal switching and the logic chip provides voltage control…there are resistor network banks with each 4069 as well, so I need to spend a little more time staring at that to understand how they implemented the logic. But suffice it to say I now understand better why I don’t experience a lot of skritchies from many of the source switching functions on the M-__, because the switches control voltage, not signal flow, and the electronic switches do the actual passing of signal. Pretty cool for early 1980s “budget” mixing console tech. Tascam did put this tech in the M-50 and M-500 consoles in a much more limited way (focused auto-switching of monitor source for instance, vs. multiple banks of per-channel selectable global source switching) as well as in the M-300 series and 388…others as well I’m sure. Anyway, I now have a very in-depth understanding of the control signal flow for all 12 of the control busses in the Control Module…why? Because I spent the last week here and there about losing my mind chasing my tail around failing card edge connections and mini Molex connections. And I had to rule out switch issues or other mechanical (solder joint) issues. Every time I’d pull the module and investigate and/or repair something, when I’d reinstall one thing would be fixed and another new thing not working…losing my ever-living mind. In the end what it all was came down to some bad mini-Molex socket contact crimps (rectified by cinching and soldering…guess the folks in R&D were moving a bit too quickly that day), and those card edge connections. The format used on the M-__ is the same as a number of preceding Teac/Tascam consoles…the module backplane connectors on the Teac Model 5 series and its descendant the Tascam M-35, the Teac Model 15 and its descendant the Tascam M-16…the 80-8 and 85-16 series tape machines…others I’m sure, like maybe even back to the Model 10 console, 70 series and 90-16 tape machines I don’t know…hey even my Ampex MM-1000 2” tape machine uses the same connector format for the amp card backplane in the card cage of the signal electronics modules (same as AG-440)…vintage video game machines commonly used them…so, super, super common format spanning decades…0.156”/3.96mm pitch, single and dual row, I think I’ve seen as small as 4-pin and as much as 72-pin. Anyway the problem is, as with many things, not all parts are created equal, and there are good quality connectors, and not so good quality connectors. A key distinction in the quality is the contacts, the quality of the metal and plating. The connectors used in the Ampex stuff, those contacts are often gold-plated and high quality metal and they last. I’ve never had an issue with the card edge connectors in an Ampex AG-440/MM-1000 signal electronics module (and I bet the same holds true for the MM-1100, MM-1200, ATR-100 series, etc…same connector format…). The connectors in the M-__? 45 years in and the contacts have deformed and are starting to not make good contact. I’m sure it’s worse with modules where the contacts have been cycled more with installation/removal actions, and the Control Module surely has seen the most action…who knows how many times that thing was cycled during R&D, and I bet I’ve done it hundreds of times by now over the years…not sure what the cycle spec is on the particular connectors installed on the M-__ motherboards, but what I do know is multiple contacts are not connecting to the edge connectors on the Control Module…there are a total of 50…two 18-pin single-row single-contact connectors and one 14-pin single-row single-contact connector. This is almost the same as for the I/O modules, but in the I/O modules the 14-pin connector is replaced by a 10-pin connector. The smaller connector propagates the power rails, and that connector is larger for the Control Module because of an additional power rail (+6V) that feeds the Control Module, as well as carrying the L/R meter signal since two contacts are needed for the Control Module (L/R meter) vs 1 for the I/O modules. The larger connectors carry control and audio signal bussing; one of them (the middle connector) carries the audio signal bussing; the I/O meter signal, PGM BUSS IN signal, AUX sends 1~4, PGM groups 1~8, the ST L/R buss and the MONI L/R buss in that order pins 1~18. The other connector (the lower or first connector closest to the arm rest) mostly carries control bussing, though the SOLO buss audio signal is also included on this connector, so it’s SOLO L/R audio buss, and then control signal busses for the SOLO buss, and then the control signals for the REMOTE banks (MONI INPUT, LINE 1, LINE 2 and BUSS, and then bank A MIC, LINE 1, LINE 2 and MUTE, and then bank B MIC, LINE 1, LINE 2 and MUTE). That takes us through pins 1~15. Pins 16~18 are not connected, but there are contact “fingers” on each module and a 3-pin header installed on the PCB connected to each contact, so you could do anything with that. Anyway, the connectors are all the same make/model, except for that one 14-pin connector for the Control Module…the contacts are higher quality in that one and it has fared better. Me being me I would typically just find the parts and replace everything with new, but the problem I am having is it’s hard to find 18-pin single-row, single-contact connectors with solder pins…best I can find are single row double-contact connectors with solder eyelets for cable mounting vs PCB mounting. The double-row is fine, it just means there are contacts on both sides interfacing the PCB where there are only contact “fingers” on one side of the PCB…that kind of bugs me, but functionally it’s fine. And the other problem is they are expensive. So I’m not sure what I’m going to do there long-term…I’ll keep looking. In the meantime I had to come up with a technique for bending the contacts back out to make better connection, which requires care because it is easy to dislodge the contacts from the socket header, which is probably not good (it’s a little tricky to get them set back in place when this happens, and I’m thinking it’s probably not a good way for the contacts to be deflected), and at the same time I wanted to avoid bending them too much which would increase the risk of unnecessary wear on the contacts on the modules…so I came up with a way to use a very small flat-blade screwdriver bit to gently pry the contacts out, restoring something closer to the original profile. What I didn’t realize is that sometimes, depending on the angle and how much I deflected the bit, I was pushing an adjacent contact the wrong way. So I can’t tell you how many times I had that module in and out testing the functions and questioning if it was really a contact issue in the edge connectors…ever living mind…losing it I tell you…I finally got wise and realized I could verify it was the edge connector by probing the contact on the module to verify control voltage was present there, and then jumper the contact on the module to the contact in the edge connector to verify proper function. This should have been a step much earlier in my process, but hindsight is 20/20 and I was being an ignoramous. But this is when I finally figured out a way to visualize the position of the contacts in the connector with a flashlight held at the right angle and positioning myself in the right way to actually see the variation between the contacts, and also revise my method of bending the contacts.

To be continued…

 

[sweetbeats]

You might ask why worry about trying to find new connectors if I now have a way to just restore the proper position of the edge connector contacts. Well, the answer to that question is in the answer to the question why did the metal contacts find themselves in the wrong position in the first place? And, of course, the answer to that is they yielded to pressure over time. “Yield” when applied properties of a given type of metal informs us how much pressure a type of metal can take before it yields or bends. And once that occurs it will never be as strong as it was. You can bend it back but then it is much more prone to yielding again, progressively. And this is no surprise. So my concern is a progressive pattern of having to futz with it, and I’d like to get the console to the point I’m not worried if everything is going to work when I power it up. And it’s just generally a little more challenging to get there with a console of this vintage and grade. What I mean by that relates mostly to phenolic resin PCBs (vs glass fiber) and tin plated contacts. The plating doesn’t age nearly as well as something like gold, and the phenolic resin boards aren’t as robust…not as stiff, solder joints can be more prone to failure, and they aren’t as forgiving with application of heat from a soldering iron when executing repairs. It’s more critical to have good equipment and technique when modifying or repairing boards like what’s in the M-__. That doesn’t mean the phenolic boards are “bad” or devices that use them are “cheap”, but glass fiber is better and as a result also more expensive. I think phenolic boards are fine for lots of applications, but motherboards and plug-in boards really benefit from being constructed of glass fiber. It’s all in how the boards are supported too. I’ve mentioned before the M-__’s motherboard and module interface design comes straight from the Teac Model 15/Tascam M-16…the backplane design too. Any of you that have one of those consoles or have worked on one will immediately recognize the design of the M-__. A key difference though is in those production consoles the module PCB with the contacts is coupled to a metal plate that not only supports the PCB, but also has guide fingers that insert into the motherboard chassis to help guide the PCB contacts into the edge connector. All of that is absent from the M-__ design, so care has to be taken when installing or removing the modules. All the stress of those procedures falls entirely on the PCB. So, anyway, these are the things I think about when I’m working on it and making decisions around things to fix or any changes to make, and just general handling.

But…I…digress.

Back to the MONO function in the control room source switchrack for a moment, and the fact it only affects the CONTROL ROOM OUT jacks, not the headphones…since I had the whole Control Module out and all torn apart again (multiple times), I did decide to refresh my memory how that works and where the mono summing amp sits in the signal chain to see if it was possible to easily modify things so the MONO function could be auditioned in the headphones as well as via the CONTROL ROOM OUT jacks. Well, I remember now…the control room source switchrack as well as the studio source switchrack use FET switching. So there’s discrete logic involved in the signal path, and it’s more than I want to get into to figure out what the ramifications might be of trying to get the headphone circuit included in the MONO summing function, and then, once known, sorting that out. So, letting that go completely. And I’m good with it.

So, I finally got all the REMOTE functions firing reliably on all cylinders, but I was also having other issues throughout that whole mind-scramble with things like intermittent signal on one side of the monitor buss as well as the stereo main buss. In one case it turned out to be the same issue with an edge connector contact as the REMOTE signal bussing, and in the other case it was a failed solder joint on a small buss PCB…reflowed all the solder joints there. And THEN I was testing things and one side of one of the PHONO inputs wasn’t passing signal. ARG. Chased that down and that was yet another bad crimp on a mini Molex contact. It felt like all the brewing problems in the Control Module noticed I was pouring some time and attention into it, and collectively decided to capitalize on my attention and fail at the same time. The other thing I recently worked on (again) that I don’t think I mentioned is the test oscillator (the multi-frequency test tone oscillator). For anybody that’s kept up with this thread over the years you’ll recall I have gone 11 rounds with that thing…cleaned the frequency selector switch, replaced all the resistors in the circuit with high-quality metal film parts, and replaced the Mylar film caps with WIMA close-tolerance stacked film parts…basically the entire circuit has ultimately been gutted over several stages when it’s reverted to not working right…the 12K and 6.4K tones weren’t working recently. I had a hunch it was the selector switch again. The contact rails are brass. About 15 years ago when I pulled the switch apart I found the rails to be oxidized, so I polished them up and that helped. There were other problems, but that was one of them. I suspected maybe the oxidation had returned, so I pulled that off the PCB (a total of 36 pins…not all of them terminated, thankfully, but 25 of them are, so not a simple component to remove…), opened it up, and sure enough the rails looked somewhat oxidized toward one end. I realized that end of the rails probably didn’t get as well polished because I wasn’t wiping past the end of the rails with my polishing pad, and that end handles the higher frequencies. I have also learned a wonderful tip in years since that first maintenance effort…Brasso. I have a bunch of ADC TT patch cables with brass plugs. An acquaintance of mine with a career in the broadcast industry shared the tip with me…Brasso quickly removes the dark brown oxide from raw brass surfaces…amazing. So I put some on a cotton swab and within a minute both rails looked new…put it all back together, reinstalled it and the oscillator is working 100% solidly. Here is the frequency selector switch all cleaned up and ready for reassembly:

There was also the master fader problem I detailed in an earlier post…I think there was something else that cropped up too and I’m just not remembering. But, at the end of the day, everything seems to be fully functioning and reliably at that, AND I feel like I’ve gotten very well reacquainted with the Control Module while also learning a lot more about how it works, what is where, how all the signals flow from PCB to PCB, etc., which is comforting because I feel that will make it much easier to troubleshoot when the next thing happens.

Last but not least I am turning my attention back to the I/O modules and fixing a couple isolated issues I’ve noticed over the years. The first was on module #12, the BUSS MASTER level pot has never worked right, barely attenuates the output (which, when installed in slot #12 is the output of the AUX 4 buss). I/O module #12, to refresh those that have been following along, and to inform the rest of you, is clearly the R&D I/O module…like the Control Module, I/O module #12 has controls notated with a marker all over the sub chassis, has a number of jumper wires and cut traces and such across most of the 6 PCBs that make up the module, which are somewhat different than many of the rest of the I/O modules. So #12 is probably the one they mocked up first and used to test and make changes, etc., and then the rest of the modules are version 2.0. Well, it took me about 15 minutes of looking at the buss PCB and following where signal comes from to the level pot and where it goes, testing the level pot to make sure it was doing what it was supposed to do, when it finally hit me…the level pot was simply acting like a potentiometer that wasn’t connected to ground. I had noticed a spot in the ground trace leading up to the pot where the coating was removed and there was just a tinned spot on the trace. I figured at some point a jumper wire was soldered there and moved on, but then it occurred to me to check continuity between the ground pin of the pot and the ground plane…again, probably should have checked that much earlier in the process, but my ignoramousness for the better of me…well, it was very high resistance…should be a dead-short. I looked closer and realized that bared spot on the ground trace was actually a spot where the trace was cut at some point, and then subsequently resoldered. The problem is it’s bad practice to just blob solder on something like that…the solder is prone to cracking over time with heat expansion and contraction. It’s a good idea to take a small piece of wire or a piece of a through-hole component tail trimming and bridge the cut trace. So I did that…level pot is back in business.

I need to do full function testing still on most of the I/O modules…I got through 3 or maybe 4 of them many years ago. Honestly I should probably just start over…but regardless I’m sure I’ll find a couple more things that need straightened out. I’m not sure if I’m going to do that next or create the video series on the history, functions, features, construction, etc…can’t decide.

Anyway, that’s the update. If you’ve read this far, thank you, if not, zero judgment…this is just as much about me leaving bread-crumbs for myself albeit on a public platform.

 

[sweetbeats]

Module #12 without its dress panel:

 

[sweetbeats]

Well, I/O module #1 fully checks out…100%. Module #2 almost…the only thing wrong is the BUSS SOLO function isn’t working right…the SOLO control is working (all audio to the monitors or cans mutes), but the sum of what’s present post BUSS MASTER level control at the BUSS OUT jack is not making it to the SOLO buss. It’s getting everywhere else it’s supposed to go, just not to the SOLO buss. So I’ll need to pull the module apart and see what’s what…shouldn’t be too hard to figure out…

AND…

Need to pull the Control Module apart again…the talkback mic is really muffled and has low output. I have a spare from a parted-out M-500 console. So I need to check the wiring and the mic, likely replace it. It should be a pretty simple fix because the problem exists everywhere the talkback mic propagates so it’s a global problem at or near the source.

 

[sweetbeats]

Okay…Module #2 is now 100%…turned out to be a bad connection from the buss amp PCB to the main module PCB. In the process the wire that carries the left channel of the main buss to the main module PCB from the fader booster PCB broke at the solder joint on the main PCB. Fixed that too. On to Module #3.