Tascam MS-16 EQ caps & Noise Level

M

man-bot

New member
Hello All!

I finally took some time to change out the bias caps on my MS-16 cards and change out what I think were the EQ caps on the rec/play cards as I was having some trouble aligning within spec. Can anyone perhaps put in their thoughts about which caps affect the rec and repro eq adjustments on the attached schematic? I changed out C10 and C21.

Part deux - I have noticed the noise of the machine itself seems high (not quantifying evidence... just my ears). If I have the machine fed into my board - not playing tape - in repro mode - and I have all channels muted (off in the cas eof my board) and I flip one on I can hear pretty significant hiss. Now this is related to the repro level as that hiss goes up and down if I adjust the repro level pot.

Ideally I would have as little noise as possible in the amplifiers on the cards... would an opamp or component change help or do you think it inherent in the design of the cards?

Let me know your thoughts!

Mike
 

Attachments

  • tascam_ms-16_manual_rec-play pcb.pdf
    1.5 MB · Views: 13
Why do you suspect the amp card is the culprit for your difficulty aligning the frequency response? And is it both in record and playback you are having trouble getting it into spec?

What shape are your heads in? A head at end-of-life will be difficult to get response in spec.

Is there a difference between the record and sync heads with your frequency response performance?

Regarding the noise are you using the balanced or unbalanced I/O on the Tascam?

What mixer are you using?

What level is the noise? Have you measured?

That amplifier design is not inherantly noisy...it's actually a pretty neat design that minimizes the number of electrolytic caps in the signal path.

So...your turn...you've got a bunch of questions to answer. :)
 
OK -thanks for the reply SB! Here are my answers:

sweetbeats said:
Why do you suspect the amp card is the culprit for your difficulty aligning the frequency response? And is it both in record and playback you are having trouble getting it into spec?

Its definitely in record that I cannot get the 10k signal within spec (signal is -4 to -5 dbu down on the meters)

What shape are your heads in? A head at end-of-life will be difficult to get response in spec.

Heads are in good shape, contour seems ok (wear pattern does not extend to the edge of the head), no keystoning, no open channels. I haven;t had them professionally reiewed but they seem OK

Is there a difference between the record and sync heads with your frequency response performance?

Same performance in both Rec and Sync

Regarding the noise are you using the balanced or unbalanced I/O on the Tascam?

Essentially unbalanced, all connections are XLR and TRS however my patchbay is unbalanced. The recorder is only 5' from the desk.

What mixer are you using?

Soundcraft series II circa 1978

What level is the noise? Have you measured?

I have not measured it - do you have a method you would use? (I;ve always been curious about this). I can hear audible hiss just when I turn the channel on with the recorder on as well (goes away with the MS-16 off)... tape not rolling.

That amplifier design is not inherantly noisy...it's actually a pretty neat design that minimizes the number of electrolytic caps in the signal path.

So...your turn...you've got a bunch of questions to answer. :)
Click to expand...

I'll do a run through again - I've realized part of my issue is my MRL. I have a 1" MRL (new) that I bought for my MCI machine which is NAB, the MS-16 uses IEC which translates to -2.5bd down at 10k.

Let me know your thoughts.

Mike
 
What value caps did you use for the bias caps?

Tell me more about your interfacing between the mixer and tape deck...are you using the XLR I/O on the MS16? And where and how is the connection being unbalanced in the patchbay? It's a TS patchbay and you have TRS plugs going into it?

What tape are you using?
 
I used 220pf caps for the bias cap

It goes like this for the interface:

MS-16 -> XLR to TRS snake -> patchbay (unbalanced) -> TRS snake -> Soundcraft desk (same for ins and outs)

Current tape is Quantegy 407 - it is +3 but I have been pushing it at +6 to try and keep noise down...
 
I'm reasonably certain the cause of your frequency response issue I simply overbias.

In stock form the MS16 can comfortably bias +3 and +6 tapes...the bias amps typically just run out of steam trying to get +9 tape biased. So the stock range is +3 to +6 tape. By doing the bias coupling cap mod you don't extend that bias range, you shift the window of the machine's range. So with the mod you now have a machine that can bias +6 to +9 tape, but now at the low end of the bias adjustment range there is too much bias current and +3 tape is over-biased no matter how far you wind out the bias level. Furthermore, IIRC the ideal value for the mod is 180pF...maybe it's 200pF. But at any rate you used 220pF which will shift your range window even further away from being able to bias +3 tape.

Question: why did you execute the mod if you are using +3 tape?

To test my theory you could put a stock cap back on one of your cards and see if you can bias the 407 with the channel.

Regarding the noise issue:

• did you ever check/adjust your bias traps? I'm not thinking this is a contributor since you notice your perceived elevated noise levels when monitoring the input on the MS16, but I thought I'd ask anyway.

• so is the ring and shield of a TRS plug coupled when plugged into your patchbay? Or is the ring not coupled to anything?

• has the MS16 been recapped at all?

Measuring "noise" is not a simple task if you want to compare apples-to-apples with specs you find in the manuals of your gear...I'm in the middle of learning this with my early 80s Tascam 12x8 proto console. It seems "noisy". Is it just me or is it really noisy? In order to take measurements that you can compare with accepted industry norms, you have to have a very sensitive AC voltmeter, the ability to determine the gain of any preamps in the chain, a means of filtering the output being measured (like a steep 12kHz roll-off low-pass filter), and a means of applying whatever weighting you want to utilize. I want a simple farm-boy git 'er dun method of measuring "noise", but doing anything quick and dirty doesn't really give you results you can compare to anything, which defeats the purpose of doing it in the first place. Now, if you are just taking measurements to compare before and after when implementing modifications or upgrades to a signal path, that's a little different, but again it doesn't give you numbers you can compare to specs of other gear or even the OEM specs for the device being tested...it just gives you relative before and after. I did this with a Soundtracs MX32 32x8 console I used to have. It seems really noisy, so I applied a load to all 32 mic amps (strapped a 150R resistor across pins 2 & 3 of each mic jack), set each mic trim to max, EQ out, channel faders at unity, and each channel assigned to the main L-R buss. With the main faders set to unity I measured the AC voltage present at one of the main out jacks with a true rms voltmeter rated for audio bandwidth (20Hz to 20kHz with reasonably low error specs...I have a Fluke 85 I use for this). Then I did a crap-ton of modifications to the ground scheme, and the regulation and filtering of the audio power rails, and then repeated the test to see if the modifications lowered the noise floor. They did, but my testing methods didn't produce data I could compare to anything else...the data didn't tell me what the noise floor was of the console when compared to accepted industry testing methodologies. But the changes did reduce the noise floor, and it sounded better...usable. So that was good enough for me.
 
Lots of food for thought!

I do appreciate all of the input :) - the issue did exist prior to the bias cap mod though. However it does prove to be concerning going forward...

As far as operating level, I didn't think it had a relation to bias in as far as 407 operates at +3 and 456 operates at +6 but they are both overbiased about 3db. Likely an over simplification on my part though :). Either way the 407 is good - (and I have 5-6 reels of it) but going forward I'd like to get the noise floor down as low as possible without really distorting recordings... thus why I was changing the bias cap value (to go to +9 tape eventually). I obviously didn't fully realize how this was affecting bias range though.

I think I need to take a run at aligning the machine this weekend, taking into account setting proper operating level with the NAB MRL on an IEC machine and see where it all shakes out when I try to get the 10k recording.

The hiss I have could be from trying to set repro/sync levels at 2.5db higher than required at 10k... cranking up the Repro and Sync EQ.

Re - noise questions:

I think the ring is connected at the patchbay - I do have TS snakes I could try to see if the noise level is affected.

As far as measuring noise itself - I don't think I have the wherewithal to do it properly! I do have a good meter (Fluke 123 scope meter), but I'm not sure I have the dedication or understanding to get it all set properly to take measurements to compare to the manual.

I really appreciate the help! I'll attempt to align the machine properly and see what I find and report back.

Mike
 
Did a full alignment this weekend and was pleasantly surprised... the machine set up pretty well. It seems my issues lay mainly on the Repro side:

  • I have 3 channels that just can't get to the reference level @ 1k (about 2 db down)
  • Channel 1 has poor 10k response
  • Channel 3 I can't roll off the EQ enough (i.e. if I was a -3db level @ 10k I'm -2db with it all down)... likely bias

Other than that it was pretty good. Sync response was better than Repro. In record I had no issues aligning.

The bias cap change was a bad move... I barely have any counterclockwise movement... I'm essentially at the peak for most channels... I can still overbias correctly (or so I think).

Noise is fixed - I just use the noise reduction now :)... it is dead quiet with it and with some quick tests I don't find the tone affected in any negative way.

So it looks like I have a bit of digging to do on the Repro side on some channels, but beyond that its pretty good.

Mike
 
how did you do the alignment without setting the bias first?

Yeah if you're not going to be using +9 tape, but mostly +3 and +6 I'd leave the bias coupling cap at stock value.
 
I did set the bias before setting the record levels - it was just a PITA because of the range (as you mentioned). It is a bit wacky that's why I think I did it right... could be off soemwhat though :)

BTW - I found a TapeOp message where I pulled the 220pf value from a Jim Williams post
 
Yeah...Jim Williams...he'd go on the high end so it could bias ATR tape. But that just moves the bar higher and higher to try and reach tape saturation.

When you tweaked the bias you rechecked the level, right? And if you adjusted the EQ then you re-checked the level and the bias again, yah?
 
You must log in or register to reply here.
Back
Top