MS-16 Puzzle

[bdoughty2005]

Greetings, all.

Hats off to everyone here as I introduce myself. I'm new to the world of semi-pro analog recording and this is my first post here. I have a problem which I've been banging my head on for a few weeks now and decided it was time to share it with others who might enjoy a puzzle.

A few months ago, I purchased a Tascam MS-16 that was reported to be "working" by the seller. He did several recordings on it himself and claims they sounded good.

I have recently had a chance to test it out and found that indeed it seemed to work, at least with a simple record/playback test using a microphone and talking. The speech sounded good on playback, but I found that the input/output levels were way too hot on certain channels.

I don't have a test tape yet, so I thought I would do a poorman's calibration on the hot channels by recording and playing back a test-tone and adjusting the record-playback levels until the input meters matched the input level and the output meters matched the input meters.

I noticed right away when applying a test-tone that there was significant signal degradation on playback: dropouts, and a sort of broad-spectrum noise that sounds like pink noise. This problem occurs on every channel.

Following is a link to a sample, where the first 8 seconds or so are monitoring the input, and the rest is monitoring the repro head. As you can hear, the first 8 seconds sound fine, so there does not seem to be a problem with the input/output amps. The problem is only noticeable when playing back the recorded signal. Replaying the same section of tape yields the exact same signal, so I have concluded the problem is in recording. This was recorded using an Ampex 499 tape the the seller gave me. Again, what you hear on this sample is virtually the same for each channel.

(http) audio.wdoughty.net/499-Tr1-1174hz-8secs_input_remaining_repro.mp3

(sorry the BBS is not letting me create actual links because I haven't made enough posts yet. I'm having to put the parentheses to trick it.)

Next is a link to a sample I did using a brand new RMGI 900 tape. It's about a minute long. You can hear immediately that the problem seems worse. The dropouts are extended and more pronounced and the noise seems louder. If you listen all the way through, you'll hear things get really bad before the end with one dropout that is almost total signal loss and only gradually improves over the course of about 20 seconds or more.

(http) audio.wdoughty.net/RMGI900-Tr1-1Khz-Trimmed.mp3

There are other test samples if you go to audio.wdoughty.net. They include different test-tones and samples from track 4 also. Low frequency test-tones reveal the noise better because the tone is less discernable. Interestingly, the higher the test-tone frequency, the higher the frequency band of the noise.

The test with no test-tone reveals no dropouts as expected, but I find it interesting that it reveals no noise either. I'm wondering if the problem could be with the bias circuitry or the bias oscillator, but those circuits are individual to each channel and it seems unlikely to me that each card would have degenerated in such a consistent fashion. Perhaps not. If I had a known good channel card, it would be easy to test that theory.

I've inspected the heads. Though I'm no expert, I've read the articles on JRF's website and in my opinion, there are no obvious signs of excessive wear or damage on the heads.

I've inspected the tape path also. There seems to be a very slight curl at the bottom of the tape as it passes over the first guide before the record head. When I first noticed this, I thought AHAH! But I've tried pressing the tape to the record head with a queue-tip as I'm testing to ensure good contact and the results are the same. Besides the curl only affects to bottom of the tape which is channel 16, and the problem occurs, of course, on every channel. I don't want to gloss over this prematurely, but my instinct tells me it is not the primary cause of the problem.

I hired a couple of techs to come look at the machine. They weren't able to do much trouble-shooting of the channels, because I don't have an extender card. They didn't have any direct experience with MS-16's either. They thought the problem might be due to the age of certain capacitors. ? They also tested the power supply rails with an oscilloscope. One of the DC voltages showed minor fluctuation. I saw it myself on the scope, but they had to zoom WAY in before it was noticeable. They recommended replacing all of the capacitors in the power-supply but did not have any confidence that doing such would fix the problem. They said it would cost on the order of $1000 to do that. I don't have that kind of money, and even if I did, I'd want to get a second opinion. Does it make any sense that a fluctuation in the DC power would cause a problem only with recording?

I tried adjusting (sweeping) every pot on channel 1's Rec/Playback card. While they all have the effect you would expect, none seem to affect the intensity of the dropouts or the noise.

The machine generally looks to be in good shape. The transport works fine. The cards are generally clean and look to be in good condition (there are no blown-up ICs like Ethan describes in his MS-16 story!).

I don't have any advanced testing gear like an audio-meter or oscilloscope. I wouldn't know what to test if I did but if someone had a suggestion, I would look into acquiring such gear.

The fact that the problem occurs on every channel suggests to me that it might be caused by something common to all channels. The elements that that I understand to be common are the headstack, the transport, the tape itself, and the power-supply. Am I missing anything?

My hope is that one of you veterans with experience might be able to listen to the samples and recognize the problem. Or, perhaps, point me in a good direction for further trouble-shooting.

I'd be grateful for any feedback.

Best regards,
Billy Doughty

 

[sweetbeats]

Bill,

The rabbit hole goes deep, my friend.

The only thing you need an extender card for on the MS-16 is to adjust the bias amp and the bias traps. Don't go there yet, though I believe Tascam still has the extender cards.

I suggest you get a calibration tape before doing anything else.

Suggestion #2: did you get a manual with your MS-16? If it were me, I'd start going through a mechanical adjustment first. If you see tape curling then the tape path is definitely out of alignment. The can cause dropouts. Maybe the bias needs adjusting...its just hard to say (for me, based on the samples you put up) what the cause may be until things are running to spec, and don't just start twisting trimmers...

Another thing that it could be may be the relays on the channel cards that do the switching when you select different input sources (i.e. INPUT, SYNC or REPRO).

Try pulling your channel cards and apply DeoxIT to the multipin connectors and then re-seat them a couple times, and while the cards are out look to see if the relays are covered or not. If not, then its easy to get at them, remove the little plastic covers, spray some DeoxIT on the contacts and gently massage with an old toothbrush and then lay the card on a paper towel or something component-side down to let the relays drain/dry. If they are covered, then same thing but you need to heat up the soldering iron to desolder the tabs on the trace-side of the amp card to pop the metal covers off.

DIDJYA DEMAG THE HEADS REALLY GOOD WITH A REAL DEMAGNETIZER? I'm not yelling...just emphatic.

I've been shocked at what problems can be caused by magnetized heads.

But get the tape path running right, clean-clean-clean those heads and demag and try cleaning the relays. Do you hear nice crisp "clicks" when you switch from SYNC to REPRO?

Does the problem exist whether you are playing back from the SYNC or REPRO heads?

 

[evm1024]

Any chance you cleaned the head to electronics connectors with deox? There are 2 per head. The first is just next to the heads and the other is on the back of the electronics box. Might give them a wiggle to see what happens.

I could send you a known good channel card if you wanted to test it in your deck.

Recapping the power supply is easy. Do it yourself when you are ready. I'll do mine in tandem with you if you like. I would expect 2 volt PS changes to (perhaps) be audable but not mV as you describe.

Ethan

 

[RRuskin]

It sounds to me that you are hearing asperity noise from the tape and that other than cleaning and proper alignment, your machine is in good shape. Check to see if the curling you noticed occurs with both reels you now have. If only one of them curls, you've got an example of "bad slitting," a very common problem with the later Ampex/Quantegy products but it's been known to plague other brands from time to time. Once everything is physically and electronically set up, record some program. If the noise doesn't isn't there, problem(s) solved.

 

[bdoughty2005]

Thank you for your feedback, Sweebeats, Ethan, and RRuskin. I am encouraged.

I called around town to see if I could find a head demagnetizer. No one seems to have one, so it looks like eBay may be the way. Found one from Audio Village for $79.00. Seems reasonable. They also sell 1" MRL tapes for $279. There are two EQ types: IEC or NAB. I don't know which one is appropriate for the MS-16, so I'll look in the manual when I get home.

Sweetbeats, I do hear very pronounced clicks when switching the monitoring source. As I remember, the reproduced signal is the same on playback whether it's from the sync (record) head or the playback, though for most of my tests I was just monitoring the repro head "live" during recording, so I will have to do another test to confirm this.

I did replay the same test multiple times from the repro head and sample the results on my computer. I then compared the waveforms and found they are identical, so the degradation is imprinted on the tape, leading me to believe it's a problem with the recording circuit, not the playback. I will do a similar test comparing playback from the sync head with that of the repro head. If there is a problem with the actual record head, then it might cause even further degradation on playback when used in sync mode. I'll post my results.

Ethan, I will de-ox those head connections as you suggest. Thanks for the tip. And thanks for the offer to send a known-good card. I will try the demagnetization and adjustment of the tape path first and see what results I get. If I continue to have problems, I will take you up on your kind offer. As for re-capping the power-supply, from what you say, it seems that perhaps I should save that as a last resort. I would be happy to do the procedure in tandem should it come to that. If there is any way I can help your project as a fellow MS-16 owner, please let me know. I'd be more than happy to.

I'm going to be out of town for a week, starting tomorrow, so it may be a short while before I get a chance to do more investigation. Thanks again to everyone for your consideration. I'm most grateful and I will post my progress when I get back to it.

Regards,
Bill

 

[sweetbeats]

Bill,

I called around town to see if I could find a head demagnetizer. No one seems to have one, so it looks like eBay may be the way. Found one from Audio Village for $79.00.

Is it the Han-D-Mag? That's the one you want, and that is the going rate for them. You can get them used for maybe $20-30 less...maybe...but it is nice just to get it new and know whatcha got too.

If there is a problem with the actual record head, then it might cause even further degradation on playback when used in sync mode. I'll post my results.

Unless there is a problem with the physical adjustment of the sync or repro heads or an issue with the repro electronics, on a properly running MS-16 the response should be the same from the sync or repro heads. Not sure if you knew that. It is not like some decks where the playback response curve and performance is different between the sync and playback heads...technically you can mixdown from either head on the MS-16...the primary advantage to having the repro head is for calibration activities (being able to monitor recording on-the-fly for record electronics calibration and it makes setting the bias a relative breeze).

They also sell 1" MRL tapes for $279.

Suggestion: find out what the model number is on the $279 tape and compare with the suggested retail on the MRL site, and then compare with the prices at someplace like US Recording Media. You want to make sure that the "good deal" doesn't leave you with a tape that is really short and includes only basic tones unless that's all you can spring for.

There are two EQ types: IEC or NAB

You want the IEC tape.

Hey, put up some pics of your gear when you can, and I didn't say it earlier so I will now, welcome to group here!

 

[bdoughty2005]

Thanks, Sweetbeats. I'm glad to find this forum with such friendly and knowledgeable people. I hope to contribute something of use to someone as I learn the ropes with this MS-16.

As an update, I did several more tests and re-aligned the tape guide before I left for vacation last week.

For one, I verified that the degraded signal is consistent whether I play back from the record of the repro head, so the problem definitely seems to be related to recording and not playback.

I also performed the tape path adjustments according to the manual and eliminated the curling of the tape near the bottom of the center guide. I was disappointed to find that there was no noticeable improvement in the test-tone reproduction.

However, in the course of doing the tape path alignment I removed the headstack. As a result, I was able to examine the heads more thoroughly. I realize now that I did not really know what I was looking for before. After closer inspection and some more research online, I discovered that the heads are quite worn and unevenly at that. There is a flat stripe down the middle of both the record and playback heads where the tape has worn the metal down. Furthermore, the worn region is trapezoidal on both heads, indicating improper zenith alignment of both heads.

After re-installing the heads and aligning the guides, I did some more tests and found that pressing the tape snugly against the record head with a q-tip while recording does eliminate the broadband noise, but not the dropouts. Well, it seems pretty clear that the tape is not making ideal contact with the heads because of improper alignment and wear. I have opted to send the head block to JRF for relapping and alignment. I should get it in the mail tomorrow and hopefully have it back in about a week.

I haven't had a chance to order the demagnetizer yet, but I am going to ask JRF to demag the heads. I'm hoping very much now that the entire problem is with the heads. In any case, I am confident that relapping the heads will at least improve the situation.

I will post updates as I make progress.

Cheers,
Bill

 

[evm1024]

Flatbed scanner

Do you have access to a flatbed scanner? If you do you can put the heads on it and get a nice photo of them.

In any case try to get a good photo of your heads. Perhaps with the flash off and a white card to control the relections (we want to see the cards reflection in the shiny heads).

The goal here is to see if you have worn out heads....

--Ethan
\

 

[sweetbeats]

Right, Ethan...to see if they are gapping yah? No sense in sending to JRF if they're shot...

Bill,

Check out some of the photos on this page to see what I am talking about with "gapping".

If you can, do put up some pics as Ethan suggested. He has relapped heads himself so he knows what he is looking at/talking about.

 

[bdoughty2005]

Thanks, guys. I actually sent the heads off first thing yesterday morning before seeing your posts. I trust JRF will notify me if the heads aren't worth relapping.

I am attaching pictures of the record and repro heads that I took just before packaging the heads. They're not the best pictures, but they are still useful. You can see the trapezoidal wear pattern on the record head. I didn't see any gapping on the heads.

Let me know what you think. I bought an oscilloscope on ebay yesterday from its original owner. Should get here in a week. I'll be able to check out the power-supply myself when it gets here. When the heads get back, I'll re-run my tests and if the record problem gets resolved, I'll look into getting a test tape for calibration.

cheers,
Bill

ps. Sweetbeats, I forgot to tell you that the sync/repro relays on the channel cards all seem to be working nicely - nice crisp audible clicks.

 

[evm1024]

JRF will not waste your time or money. You can trust them.

From here your heads look fine. I can see the head gap and it looks normal - I do a little image processing on those photos to bring it out. A hairs width.

I'm sure that JRF will dress them up for you in no time.

--Ethan

 

[sweetbeats]

Yeah, yeah, Bill...+1 to that. They look totally serviceable and JRF will do a great job and won't take you for anything. As far as I can tell in all my time in this gig they share the top of the heap of vendors with only those you can trust to do reliable work work of excellence and charge a reasonable rates for it.

Keep us posted!

 

[bdoughty2005]

The puzzle continues

Greetings, followers.

I got the headstack back from JRF. But upon replacing it in the machine, the problem is the same as ever.

Here's a reminder of what the problem is:

You may need to turn it up to hear it. If you listen closely, there are mini dropouts or blips occurring at a varying frequency with varying intensity. This occurs on every channel. I have a attached a waveform picture of a single 'blip' during a test using a 2.8 KHz tone. You can see that the attenuation is very brief (one wavelength at 2.8KHz = 0.36ms) , but smooth in character.

Here's what I've discovered. If I short pin 1 of the record head connector to pin 1 of the repro head connector and monitor the repro for channel 1 whilst applying the test tone, I get a pure tone with no signal degradation. This effectively tests the signal path from record through playback without actually going through the heads (to tape and back). This guarantees that the problem is with the heads or the tape, or the tape alignment............. Right?

Again, I just got the the heads back from JRF. They relapped all three heads and did the optical alignment on the whole block. They also said they would demagnetize the heads, so I'm assuming they did. I've given them a thorough visual inspection, and frankly they look great to me.

So I'm left wondering if it is still a tape path problem. I can sometimes see a slight waver in the tape just to the right of the record head during record or playback. Is a very slight tape wavering considered tolerable or is any wavering whatsoever a no-no? I've tried both reels that I have, the old Ampex 499 and the brand new RMGI 900. The problem is just the same with both reels.

I think I'm in over my head here. I've already performed the tape path calibration per the manual. I'm convinced the problem is not with any of the PCBs, amps, power supply or channel circuitry because of the test where I basically short the signal from the record head to the repro head and hear no degradation at all.

Any ideas?

Regards,
Bill

P.S. I've also attached a picture showing four ribbon connectors that aren't attached to anything. I discovered them when I removed the bottom panel to test the power-supply. They were just tucked to the side, not plugged in to anything. I removed the front, back and side panels to try to see where they might go, but didn't see any headers without connectors. Unfortunately, I can't see the other ends of them without doing some major disassembly. It is another mystery to me, given that all of the functions of the machine seem to work (albeit some slightly flawed).

 

[cjacek]

OK, this sounds like either bad tape or the deck needs a bias adjustment, as I'm hearing loads of 'bias rocks', kinda a silly name for the type of sound you'd get of water flowing over rocks. Tape bias seems to affect this type of modulation noise greatly and I'd look into that too.

Try to record a 40hz tone, on one of the channels [for test purposes] and, while monitoring, adjust for the least amount of 'rocks' at 40hz. If this doesn't improve things then I'd look at the tape.

Some of the early RMGI 900 batches, for instance, didn't cure too well and people had problems. RMGI is very good at replacing bad tapes thought.

Also, the Ampex 499, while it doesn't seem to be prone to sticky shed, I'd be suspect of it nevertheless.

Do you have any known 'good tapes' around or is it just your 499 / 900 ?

--

 

[evm1024]

Those connectors are for the meter bridge. IMHO The meters can be installed in number of different places. I've got a set of them too. Lt lease that Is what it appears like to me.

I'll have to listen to the noise to get a feel for it.

Any chance that you have something prerecorded to listen to?

Noise from caps would be hearable regardless of input conditions if it is in the playback path (heads to line out). THis means that you would hear when listening to a tape without regards to what is recorded or not recorded on it.

Noise from caps on the record side would only be hearable for tracks that you laid down on a single or perhaps more than one channel but most likely not all channels. Again this would be without regard to the singal recorded or not recorded.

OK, just listened to the noise. That "clicking" sounds like bad caps or noisy pots. Which is interesting as that it should be different per channel.

Is the DBX in or out?

Oh, On the clip you post I hear it from the start of the clip. Is this clip just the playback? I thought that youhoping that this is said that half of it was the input source. And that the input was clean.

Right at the moment I'm hoping that this clicking is coming from the input signal. Here is my thought:

Cap/pot noise is random and unique per channel - it happens without regard to signal. Thus each channel may have noise but it will not happen at the same time (test - listen to 2 channels stereo and see if the noise unique per channel)

If the noise is being generated in the playback circuit you should hear it in pause mode or with the tape off the heads. (test engage lifters, what happens to the noise)

If the noise is being generated in the record circuit then the pattern of noise should be the same on each playback of the tape and it should be unique per channel. (Test as above in stereo)

We are looking for a common cause here. If the noise is unique then it hes unique sources. Bad caps on each channel for example. If the noise is the same on each channel then we need to find a common cause. The common causes that come to mind are the signal source (the generator, are you using a PC to generate the tone?) and the power supply.

Moving on to dropouts - looks like tape path issues - dirt, tension etc....

Just some thoughts off the top of my head.
Regards, Ethan

 

[evm1024]

OK, this sounds like either bad tape or the deck needs a bias adjustment, as I'm hearing loads of 'bias rocks'.

Try to record a 40hz tone, on one of the channels [for test purposes] and, while monitoring, adjust for the least amount of 'rocks' at 40hz. If this doesn't improve things then I'd look at the tape.

Some of the early RMGI 900 batches, for instance, didn't cure too well and people had problems. RMGI is very good at replacing bad tapes thought.

Also, the Ampex 499, while it doesn't seem to be prone to sticky shed, I'd be suspect of it nevertheless.

Do you have any known 'good tapes' around or is it just your 499 / 900 ?

Yes, that is the thing I forgot to add as well. Without a mod the MS16 cannot bias 499 properly. Just not enough bias.

Being that the test tone is near 1kHz I would not expect to hear rocks like this. Perhaps I'll test it out tonight with 499.

In the MS16 there is a master bias osc that supplies bias amps on each channel. If the master bias osc is faulty then all channels will get that fault (too low output for example)

If the bias freq changed then the bias tuning on the channel cards will all be off.

Try 456 for a test and also try turning the bias level all the way up on one card to see what happens to the sound.

Also as Danial said, adjust the bias level by (I use 12 Hz or so) recording a low freq signal and listening to the playback head signal. Adjust bias for the lowest distortion (rocks).

-Ethan

 

[evm1024]

More thoughts on droupouts.

I'm assuming that you are using a 1 kHz signal. The waveform shows that the dropout lasts about 10 cycles. at 1 kHz and 15 ips we end up with 0.15" per 10 cycle (10*15/1000=0.15"). This is about 1/7th of an inch. pretty big!

If something in the tapepath (dirt, misalignment, riding up on the rollers/guides) should be visible to you.

--Ethan

 

[RRuskin]

I don't think there is anything wrong with the machine. This sounds like the oxide coating isn't totally uniform on the tape. I may have asked this before but have you recorded any program yet? I'm betting you won't hear any "rocks" with anything other than test tones.

 

[trancedental]

Don't try adjusting anything on that or any recorder UNTIL you get a test tape, also you shouldn't trust any tones recorded or level's observed on VU's & / or LED's unless you can confirm the exact levels from other sources! And use new fresh tape on a clean recorder!

I was recording frequency tones a few days back on my Fostex E-16 which was fully serviced 3 years ago but hardly used since, at first I was getting levels on different tone frequencies which were 2-3db's different on certain channels on the Fostex LED meters. Some frequencies were well within spec (less than -+ 1db) on all channels but others were out of spec

However checking my tones coming from a Digi 002R on my Soundtracks Project 8 24 mixer on the VU's & LED's & making sure the gain / levels was exactly 100% right I sent all the tones of various frequencies ( 1khz, 10khz, 15khz, 100hz, 50hz ) to the recorder & played back the recorded tape to the desk tape input channels which were 100% spot on. Then I checked the levels on some of my outboard gear to make sure!

So confirmation of all this testing is that the Fostex E-16 LED levels are not very reliable at all unless you only use 1khz, maybe 10khz ect at a pinch!

I reckon those E-16 LED's would have real problems with a frequency sweep tape.......................................!

 

[bdoughty2005]

OK, this sounds like either bad tape or the deck needs a bias adjustment, as I'm hearing loads of 'bias rocks', kinda a silly name for the type of sound you'd get of water flowing over rocks. Tape bias seems to affect this type of modulation noise greatly and I'd look into that too.

Try to record a 40hz tone, on one of the channels [for test purposes] and, while monitoring, adjust for the least amount of 'rocks' at 40hz. If this doesn't improve things then I'd look at the tape.

Some of the early RMGI 900 batches, for instance, didn't cure too well and people had problems. RMGI is very good at replacing bad tapes thought.

Also, the Ampex 499, while it doesn't seem to be prone to sticky shed, I'd be suspect of it nevertheless.

Do you have any known 'good tapes' around or is it just your 499 / 900 ?

--

Thanks, cjacek. I just have the Ampex 499 and RMGI 900 reels. I tried the 40Hz test you suggested, and while I can definitely make the problem worse, I can't make it any better. It sounds like I should try some 456. If I could find a used cheap reel somewhere, I'd go for it. I don't care if it sheds, I just want to make sure the machine is going to be usable before I spend anymore money on it.

I may try to post a sample of the 40Hz test as it really reveals the ugly nature of this signal degradation.