MS-16 alingment - 3 weeks later

[pianodano]

Hello,

I am having a heck of a time aligning my Tascam MS-16 recorder. I bought it last year and have since had the heads lapped (they are fine) and have bought a good bit of the test equipment needed and a new MRL tape to align it. I am on my 3rd go around and I am feeling helpless now because I cannot bring the record electronics into spec and I am totally stumped . I started out about 20 odd years ago with a Tascam 38 deck which was very easy to keep in shape but this one is driving me nuts. I am sorry that I have written so much info to read but I know there are lots of folks here that can probably easily point me in the right direction if given enough info and I thank you in advance.

Background

Initally I ran into a tape path / curling problem that took a great deal of time to solve but tape appears to track very nice now. There is very little fluctuation in levels now and I cannot cause much increase in level by either gently pressing the tape against the heads or increasing holdback tension. I am not the sharpest tack in the box but based on what I have outlined so far, the mechanics of the transport seem ok to me.

All of the below listed calibration tests have been done on a track by track basis, plugging and unplugging.

The first thing I did was set the input levels using a 1k 1.23v tone verifying I was getting as close as possible to 1.23v at the outputs. I then set the meters while monitoring input to read 0vu. No problem there.

Tascam then calls for me to put on the alignment tape and reproduce using different reproduce/sync set ups. She comes right into specifications with no problem. Tascam provides a chart of the various frequencies and the acceptable response. All is good there.

Azimuth was tweaked using a dual channel oscilloscope at a coarse and fine setting. Head height seems to be fine. I checked zenith by rubbing a china marker over the heads and watched the wear off. It seemed very even. I did have to adjust the wrap a little to get max output. No problems up to this point.

I then set bias (456) as per Tascams procedure. They call for a 10kz (690mv) tone with bias set at -3 to -4 off peak so I settled on -3.5. I have NOT fiddled with the bias tuning and trap adjustments because Tascam calls for an extender card to be inserted which, I cannot locate anywhere. I can’t infer from the procedure if the required extender card is to be inserted to set the bias trap also, but it appears so.

Record level is set using a 1k tone.

I then started recording tones to verify overall frequency response. And this is where it is totally wacked. It really only has a HF eq which has absolutley NO effect on the response. I understand it (record eq) is optimized for 20 kHz. Whether monitoring INPUT, REPRO or SYNC the response is the same. Again, because the reproduce tones on the alignment tape come into spec, I am pretty confident the heads are ok.

Listed are the tones Tascam calls for to be recorded with my results. Every track reads almost identical.

40 +3.
100 -2
250 -1.5
500 -.5
1kHz 0
4 kHz +2
8 kHz +3.5
16Hz +5
20 kHz off scale.

Overall record frequency response specs call for all tracks to be within 3dB plus or minus when 1kHz is at 0vu.

So, I keep second guessing myself and thinking about what on earth I am doing wrong .

I have come up with one possibility and I hope someone will be able to tell me if it makes any difference. Tascam shows a setup procedure for each test. They call for an oscilloscope to be hooked into the output side along with a meter AND a amp and speaker. But they say nothing about what I should be looking for on the scope. SO - I am thinking now that maybe because I don’t have the outputs loaded that is skewing all of my levels. Normally I would have checked levels with a console hooked up but I just received one of the new TOFT ATB consoles and it has the weirdest metering I have seen. Not really setup for analog tape. It is setup with digital metering levels.

Bottom line is that I really suspect there is something wrong with the bias. Perhaps the trap ( but for all channels to read nearly identical ?) or maybe the overbias. As I said in the beginning, I have been through the entire process 3 times and consistently I have acheived the same results. Again, I apologize for the length of this post but I thought I should provide as much info as possible. Thanks for any help that can be provided.

Danny

[amorris]

are you going -3.5 over peak, not just off peak? sounds like what would happen if you were under biasing by 3.5db. distortion would be shit too.

take 1 k and go to the peak with the bias. then hit it with 10 k and it should be close. over bias to the right.

[pianodano]

Originally Posted by amorris

are you going -3.5 over peak, not just off peak? sounds like what would happen if you were under biasing by 3.5db. distortion would be shit too.

take 1 k and go to the peak with the bias. then hit it with 10 k and it should be close. over bias to the right.

Over. IE all the way down (CCW)- then CW to obtain the highest level then continuing CW until it is reduced by about 3.5 db.

Thanks

Danny

[pianodano]

I spoke with a pretty good friend of mine today that owns a Tascam service center in Florida. He told me that I would go nuts trying to bias the machine using Tascams proceedure and that I needed to forget about the book.

What I have learned on my own is that the bias level, record level and input level are all somehow interdependent or interelated. Tweaking bias upsets record level and both together upsets input level. Evidently the "scientic" method spelled out in the manual doesn't necessaraily work in the field.

I sure can obtain a high bias current by flip flopping back and forth between bias and record level trim but I at that point the input level is totally wacked (1st thing you set) so I cannot see any end to the tweaking using the forget about the book method. Dang I wish I could figure it out.

[stellar]

it sounds to me like you are doing everything right.

I usually play the test tape first. and adjust the output amps to measure 1.73V. This is because my test tape is hot. When I've got 1.73 V I then set the VU meters to read +3. repeat with other frequencies.

Then I remove the test tape, put on a blank tape and do a bias, the way you have described. finally I get a 1khz tone coming out of the computer, which I measure to be 1.23V, and adjust the inputs so that I am reading 0VU. then I hit record, monitor the repro heads, and measure the output. Adjust record amps until output reads 1.23V. At that point I am done.

The eq is never going to be perfect because these are old machines and I recognize that it isn't going to be perfect. one other point is that if your machine isn't clean then it wont function properly. I listened to a tape of a mix I did the other day and couldn't believe the lack of bass. I thought I just did a bad job. I got the mix notes out and rewound and got ready to re-mix the whole thing. I cleaned the heads and the rollers and for whatever reason hit play on the tape from the night before and the bass was back! couldn't believe it was that drastic.

Another point to mention is that each frequency might be reading a different level at the same 'volume'. What you need to be checking is if the level on repro is the same as the level on the input meter. those two readings, at each frequency, should hopefully be within 2 or 3db.

[pianodano]

Thanks man, I appreciate your input. I would just like to get the overall response a little closer. When you say "hot" tape, are you using a 320nw tape for alignment ?

Danny

[stellar]

Yes. Actually the spine says 355 but I think the difference between 320 and 355 is whether you're using 700 hz or 1khz. from what I have been told, that is the MOL of 456 so I actually want to be biased 3db under that. (becuase I am using 456).

so if I am reading your post correctly, if you feed your machine a 4 khz tone and you adjust it to read 0db on input, it reads +2 on repro?

[pianodano]

Originally Posted by stellar

so if I am reading your post correctly, if you feed your machine a 4 khz tone and you adjust it to read 0db on input, it reads +2 on repro?

Yes. +2 would be within the graph Tascam gives. The real problem is @ 8kHz and above being outside of the stated specs.

How about this question ?

I set up my oscillator (HP 204D) to output +4 dBm @ 1.23 volt without being connected to anything. When I plug it into the recorder input circuit the oscillator level is pulled down by about 1dB. (I know this because I have measured the level with my meter). Should I use the before level or should I readjust the oscillator to output EXACTLY +4 AFTER being inserted into the circuit ? I think this is important to know and here's why.

I did quite a few songs on a MS-16 in the 80's in a local studio. I have always been very happy with the sound that of it. It was brand and fresh from the factory. I bought this one to transfer my masters to the daw and I thought I would keep a for a while and then graduate to a JH-24 someday. But what I think I am learning is that one of the cost saving measures Tascam used is in the amplifier circuits or the lack of eq trim adjustments which is causing the extreme difficulties in calibrating this thing.

I know hardly anything (really nothing) about electronics but what I believe is happening is that somehow I am getting the threshold (input) levels wrong which is wacking out all the record calibrations down the line. I say this because as I said earlier, I realized that input levels, record levels and bias are all interelated. It kind of stands to reason to me that the circuitry is capable of working well ONLY when all parameters are set optimally and exactly within the design specs and there is not much of window for error. Maybe I am making it far to complicated and I just need to learn to live with it, but I have owned a lot of Tascam stuff and I have always been happy with it. When someone that works on Tascam gear for a living tells me that I need to forget the manual and use a wizbang method, I don't know, it just makes me wonder.

Danny

[Jim Williams]

Is it a 30 ips version? If so, they don't set up flat. It's a fudge circuit with heads not designed for it. The 15 ips versions set up very flat. Yes, the Tascam method is flawed.

First, bias levels will be different for each track. I use my Audio Precision to set these up as any other way is true guesswork. It may be worth hiring an AP operator to come in and do the line up for you as then you could copy the settings that occur.

First thing is the meters are crap and surely need calibration first. I then set play alighnments with the MRL. Next is bias setting. I use the AP in a THD vs frequency50~10k sweep with a fast rate and re-draw. I then increase the bias until the THD goes minimum, usually around .15 to .55% depending on the tape and operating levels.
Then the record hi EQ is set. With a couple of cap changes GP-9 sets up nicely at +9 db levels. I achieved .15 to .22% THD at 1k this way when the stock machine spes out at .55% THD at +3. The top end was also flat out past 22k hz.

BTW, every track will have a different bias setting as the heads vary a bit.

Jim Williams
Audio upgrades

[pianodano]

Originally Posted by Jim Williams

Is it a 30 ips version? If so, they don't set up flat. It's a fudge circuit with heads not designed for it. The 15 ips versions set up very flat. Yes, the Tascam method is flawed.

First, bias levels will be different for each track. I use my Audio Precision to set these up as any other way is true guesswork. It may be worth hiring an AP operator to come in and do the line up for you as then you could copy the settings that occur.

First thing is the meters are crap and surely need calibration first. I then set play alighnments with the MRL. Next is bias setting. I use the AP in a THD vs frequency50~10k sweep with a fast rate and re-draw. I then increase the bias until the THD goes minimum, usually around .15 to .55% depending on the tape and operating levels.
Then the record hi EQ is set. With a couple of cap changes GP-9 sets up nicely at +9 db levels. I achieved .15 to .22% THD at 1k this way when the stock machine spes out at .55% THD at +3. The top end was also flat out past 22k hz.

BTW, every track will have a different bias setting as the heads vary a bit.

Jim Williams
Audio upgrades

Jim,

Wow. Thanks for the great information. Speed is 15ips. Very interesting the way you set bias. And it is encouraging to know that the top end response should flatten out. Way different from what I am getting. As I said in the beginning, 16K is +5 and 20 is off the meter high. Even 8kHz is out of spec.

Regards,

Danny

[stellar]

that is a good question about whether the tones should be loaded or not. I always did not but I see how that might be messing me up. maybe you should try it on one channel to see if it improves? also, does the bias adjustment go around and around? does it peak several times? you might be on the wrong side of it. my fostex does that.

[pianodano]

Originally Posted by stellar

that is a good question about whether the tones should be loaded or not. I always did not but I see how that might be messing me up. maybe you should try it on one channel to see if it improves? also, does the bias adjustment go around and around? does it peak several times? you might be on the wrong side of it. my fostex does that.

I had heard that Fostex has a contiunous bias pot. The tascam pot has stops fully CCW and fully CW.

The reason I am wondering about the tones being compenstaed for circuit loading is because I keep thinking about a console internal oscillator. Surely that would be compensated to show + 4 when in the circuit - Wouldn't it ?

Danny

[stellar]

Originally Posted by pianodano

I had heard that Fostex has a contiunous bias pot. The tascam pot has stops fully CCW and fully CW.

The reason I am wondering about the tones being compenstaed for circuit loading is because I keep thinking about a console internal oscillator. Surely that would be compensated to show + 4 when in the circuit - Wouldn't it ?

Danny

well within a console it would be loaded wouldn't it?

[pianodano]

Well, it seems so.

[stellar]

let me know if you find out about that.

[pianodano]

Can't say for certain, but I may have figured out the problem. I have suspected for some time that one of those "slam the tape - fat drum sound *** guys" may have got his hands on this thing, screwed it up and ebayed it. When I bought it last year, I did a spreadheet of all of the line ups and it was totally wacked. I have been aware for some time that Tascam installed a +4 and +8 output switch under the amplifier, but I just could not believe someone would be so stupid as to change the output levels to +8. Well, I took out the amplifier tonight and first thing I noticed was nearly all the bottom cover screws were stripped, cross threaded, etc. Finally got the cover off and guess what ? All 16 switches are set to +8. I'll bet that will make a difference. What kind of a person would do something like that and not keep a log ?
dfegadon him. I will let you know the outcome.


Danny

[Jim Williams]

That balancing pcb on the bottom of the rack is a pretty dirty circuit as well. When ever I did an alignment, I took the signals in and out through the RCA plugs, it just lined up better without all that extra crap in the way. The users I did these mods for also used the RCA's for the line outs into the console. The line trims on the console were cleaner than that balancing pcb so it sounded better jacking up the line input trims +12 db to compensate. You can also reset the output trims to get +4 unbalanced off those RCA plugs and the input trims can be lowered to allow +4 levels into the RCA inputs. These machines always sounded better without that balancing pcb used.

Sounds more and more like you need that AP rig in there to line up all the trims, bias, meters, etc. It would save you lots of time and guess work/ re-work.

Jim Williams
Audio Upgrades

[pianodano]

Originally Posted by Jim Williams

That balancing pcb on the bottom of the rack is a pretty dirty circuit as well. When ever I did an alignment, I took the signals in and out through the RCA plugs, it just lined up better without all that extra crap in the way. The users I did these mods for also used the RCA's for the line outs into the console. The line trims on the console were cleaner than that balancing pcb so it sounded better jacking up the line input trims +12 db to compensate. You can also reset the output trims to get +4 unbalanced off those RCA plugs and the input trims can be lowered to allow +4 levels into the RCA inputs. These machines always sounded better without that balancing pcb used.

Sounds more and more like you need that AP rig in there to line up all the trims, bias, meters, etc. It would save you lots of time and guess work/ re-work.

Jim Williams
Audio Upgrades

Jim,

I can't find anyone in my area that has one of the setups like you have. I know that it is very expensive so that's probably why. Unfortunately, I just had all the cables made to get from the Toft ATB 32 to the xlr's on the recorder.

Yes I do see where Tascams graph shows that it will align to within +2 on the RCA's and +3 on the xlr's. I wondered about that.

So what you are saying is I should use the RCA's to calibrate. OK. So I should feed all tones in at -10dBV (0.3V)? My only problem will be is that my meter won't convert to dBV. It's dBm/V.

Also to set bias using the -10's, I wonder what level I would input. I am uncertain on that and here's why. The specs call for 10kHz @ -1 dBM (690mv). But it says "TO THE INPUT CONNECTOR ON THE REAR PANEL. Which connector do you suppose that may be ? They is no way I can infer which from the manual. I have assumed all along they meant the XLR's.

I haven't been able to get on it yet this morning, but I have a feeling she's gonna come around now that I have corrected the XLR outputs.

Thanks

Danny



BTW, I have another project I plan to get on soon. I have given up on the meters and the future meter bridge for the Toft. The reason is that they are calibrated for digital, cannot be changed/calibrated and who knows when they will build the bridge. I have the opportunity to buy enough analog meters for the mixer. Alan Hyatt has said in the past that the meter bridge connector is on the pcb inside the board. I have asked him for the pinouts but he hasn't answered yet.

I am a cabinet maker so I don't forsee it being to big of a problem building and mounting the meters per se, but I will need help to design whatever circuitry that will feed the meters and also allow me to calibrate them to reference 0vu @ +4. I will probably also want to incoporate a PEAK led that will light @ about +15 to line up with the recorder. Does this sound possible and would you be willing to consult on this (for compensation of course)?