Tascam 58-OB Story...

[evm1024]

Listening for rocks....

Cory, Really feed a 20 Hz tone into your deck. Listen to the repro head as you are recording the 20 Hz tone. Adjust everything for a confortable listen level.

Do you hear that? Not the 20 Hz tone of course. This is the distortion caused by under or over biasing .

Here is a read: http://home.flash.net/~mrltapes/mcknight_biasing.pdf

Take a look a figure 1. the non-straight line is what you hear when using this method for setting bias.

-Ethan

PS could be 40 Hz, or 15 Hz or ....

 

[sweetbeats]

I'm reading that article on GS, Danny...

If you can't get the meter to go far enough back past the zero mark, then something is not right with the machine's set up.
You should be able to go quite a few dbs PAST the overbias level.

Get out your JH-24 manual do a full calibration starting at the REPRO as a minimum.
If you are brave and smart enough enough to do it correctly you should check the WRAP and AZIMUTH.
In fact, I'd set the TENSION as well.
Tension can greatly effect EQ and can cause many headaches if set wrong.

This is my problem...I'm being redundant here...Not only does the meter not go far enough past the "zero" mark (zero being the peak level when rotating the bias level trimmer CW), it never reaches the zero mark, I hence I agree that something is not right with my 58's setup. I need help. That's why I'm asking if anybody has an idea what kind of response I should be seeing at the above frequency points without bias.

Wrap and azimuth are fine.

You all know I am struggling with tension issues.

You should be able to crank the BIAS pot and get the needle to go WAYYY past -4db.
After a point it starts to saturate the tape and the level quits dropping.
It is increasing actually, but the level falls as the overbias takes effect.

You could check to see that the master bias level is correct.
Setting this requires an extender board, but I've done it without one.
You have to use alligator clips to get to the test points if you don't have one.

Nope...again can't get it to drop at all...and the master bias level is what makes all the difference with just a difference of 0.15V, but I don't want to just assume that I should crank that up and call it good. It is a band-aid fix and I don't want to do band-aids.

I'll also bring up the master bias. We cal'ed the master oscillator because we were having full erase troubles. It turned out to be the culprit. I'm so lucky to have a tech or would be screwed.

I don't have a tech...you guys are it...I'm thankful for you, but sorry for you too.

You know... I have seen a lot of decks screwed up by people who were clueless knob twidlers (I'm not referring to you SoupKing.)

You might try "zeroing out" the pots.
You turn them ALL counter-clockwise twenty turns (to assure they are at the end of their travel) and then turn them back 10 turns.
This is the theoretical starting point for the pot's adjustment.
It doesn't take into consideration the aged components in the machine, but it is a close starting point.
You have to re-align the ENTIRE machine after doing this, but it helps to have a starting point.
I have done this when I encountered a deck where someone had screwed everything up beyond anything that made sense.
It is a "let's just start from scratch" thing.

Hmm....maybe that's what needs to happen...that's what I'm looking for...I feel like I'm chasing my tail...I wonder too if the reason (a few posts up) that I was actually getting a drop in repro level when I would increase the level of the hi-eq trimmer is because the tape was hitting a saturation point...maybe the input hi-eq is too high, the bias is working right but then you add the repro hi-eq, too much of it, and the tape is hitting a saturation point...

I'm using the deck's VU's to set the input level of my tones once I've referenced it and set the meters at 1kHz...I'm doing this because I don't trust my DMM up to 10kHz or especially 20kHz even though it is rated for it. Is that an okay thing to do? The output of my PC-based tone generator and D/A gear drops the level by 1 ~ 2dB when going from 1kHz to 10kHz so I'm boosing that 10kHz tone to get the VU on the deck back up to 0 when monitoring the input before recording...

When setting the bias it is best to drop the Hi Freq tone back from 0vu at the oscillator, otherwise it will probably saturate the tape, thus giving a false reading. If I remember correctly, I used to use somewhere between -4 and -7vu as a starting point, depending on the machine.

You know... I have been a 30 ips analog guy over the years and forgot that you can saturate the tape with 10K @ 0db wehen running at 15 ips on some machines and tape formulations.
The JH-24 manual doesn't tell you to set the 10K tone lower than 0 db for 15 ips.
Still, it isn't a bad idea to lower the 10K signal.

Huh! Does that sound like a good idea with my 58??

When I moved to LA I picked up another trick, roll back the record eq. BEFORE adjusting the bias circuit. Then do level, then eq. This will help with the overall harmonic content.

And another "huh"!

Okay, Danny. That was an awesome thread. thanks for the reference.

 

[pianodano]

Cory, Sorry, you had all kind of replies while I was typing this. Forget this reply.

I need to get the response back to some sort of baseline.

If I pull bias out of the picture, what should I see for input response at:

40Hz
100Hz
1kHz
10kHz
20kHz

What should I expect for repro response at:

40Hz
100Hz
1kHz
10kHz
20kHz

This would help because I'm concerned that in my inexperience I have adjusted eq trimmers based on grossly misadjusted bias leevel trimmers from past ownership...so now the eq gets whacked and now I'm chasing it with bias...one begets the other and I can't find the path anymore.

Cory,

Opps. Seems you'e getting frustrated. Sorry, I never intended that to happen. The proceedure is pretty well laid out in the Tascam manual so you have that going for you. But you only have 2 eq pots to adjust, irrc. I'd almost be willing to bet that you won't ever get the machine closer than 3 dB across the spectrum regardless of what the manual says. You will get 1k, 2k on the money and everything else in the ballpark close. What isn't up to snuff can be fixed with console eq.

Input levels ?? You only set that with a 1K tone at 1.23v. No other tone.

Do you have enough bias current to get to 0db while recording a 10K tone during biasing? If you don't then something is wrong. Turn the bias pot all the way down to start. Then increase to 0db and keep going until it's 3 to 4 db down. That's it.


_____________________________________________________________

What should I expect for repro response at: ???

40Hz + 2 maybe +3
100Hz 0db or maybe +1
1kHz 0dB
10kHz -1 to -2
20kHz -3 to -4

Old machine - relapped heads. My experience only.
Study the MS16 graph here. That's as good as it will get. http://www.endino.com/graphs/





What it's supposed to sound like is CLEAN. Period. You won't get the sound of a JH24 out of Tascam. And that's ok as far as I'm concerned.

 

[evm1024]

This is my problem...I'm being redundant here...Not only does the meter not go far enough past the "zero" mark (zero being the peak level when rotating the bias level trimmer CW), it never reaches the zero mark, I hence I agree that something is not right with my 58's setup. I need help. That's why I'm asking if anybody has an idea what kind of response I should be seeing at the above frequency points without bias.

SNIP!

Corie,

When they are speaking of the zero mark they mean the maximum level returned from tape. (not 0 VU, it is all relative)

Just making sure that is understood.

-E

Duh, of course you do I just reread your post.....

 

[sweetbeats]

Ethan,

When they are speaking of the zero mark they mean the maximum level returned from tape. (not 0 VU, it is all relative)

Just making sure that is understood.

Yeah, yeah...I think we are saying the same thing. I realize that when setting bias the "0" on the VU meter is irrelevant...I am reading and referring to "0" in setting the bias as the repro peak level that is supposed to occur when increasing the bias level...that is the "0", and then as you continue to increase the bias level the repro level is supposed to fall. You stop turning the bias trimmer when the level reaches a predetermined point on the VU meter (i.e. -4 ~ -5dB from the "0" for my 58 according to the manual), or as you, Danny and the good Mr. McKnight are proposing, using yer ears.

I printed the McKnight article you linked by the way, as well as the article on typical response curves and I will read them. Thank you. I think I'm getting it...in...my...head...I used to look at those response curve graphs a couple years ago and didn't understand the significance of them...I do now.

Duh, of course you do I just reread your post.....

And I just typed the above mini dissertation before fully reading your post!



Danny,

Opps. Seems you'e getting frustrated. Sorry, I never intended that to happen.

Oh Danny, no-no-no...I know you didn't, and I do apologize if I wrote words that made you or anybody feel that I did.

I am frustrated but its simply not the fault of any/all of you who are trying to help. I feel the pressure of a pending project, and the lack of continuity and time to work on the 58 and these issues ideally require blocks of time. My life affords small peieces of time here and there which is an inefficient environment for me in which to work. My brain doesn't work well jumping from one thing to the next. I felt I was on the home stretch and then this bias thing...an unexpected adventure.

So that's part of it, and then I assume I'm allowing pride to bite as well...complicated. Anyway, I am just swimming a bit now but I think this is the last hump before this deck is ready to roll, and that is pushing my emotions too...this has been a dream for at least 15 years to have a system like this...to track drums onto a system like this and so I'm impatient now with this last detour, but I need to see it as a necessary educational module rather than a detour.

But you only have 2 eq pots to adjust, irrc. I'd almost be willing to bet that you won't ever get the machine closer than 3 dB across the spectrum regardless of what the manual says. You will get 1k, 2k on the money and everything else in the ballpark close.

Okay...okay. I confused myself...when I talked about "input response" I meant repro response (i.e. when playing back the cal tape)...its supposed to be +/-2dB from 40Hz to 20kHz, and you adjust the hi end with trim pot R202...manual says you should try to hit 0dB on the VU at 18kHz...of course my YTT-1144-2 tape doesn't have 18k...16k and 20k are on there for the high-end so I took a guess. For that matter my cal tape doesn't have 40Hz either...it has 31.5 and 63Hz...Repro response was at -3dB at 31.5Hz, I think +1 at 63 and 100, but flat all the way up to 20kHz. But that seems off from what you put:

40Hz + 2 maybe +3
100Hz 0db or maybe +1
1kHz 0dB
10kHz -1 to -2
20kHz -3 to -4

Maybe That's part of my problem? Compared to what you suggested the low-end is too flat/too low level-wise and the hi-end is too high level-wise...that high-end issue could exacerbate my bias issue could it not?

I need to look at those response curves you and Ethan linked...

Then the record response is supposed to be +/-3dB from 40Hz to 20kHz.

Again, the repro and record response tsts I'm doing are at just select frequencies as specified in the manual, not a sweep...that's probably a whole different story.

Input levels ?? You only set that with a 1K tone at 1.23v. No other tone.

Yeah, input level...right, right, right. I'm running straight again (as explained above).

What it's supposed to sound like is CLEAN. Period.

Right...yeah...I'm not expecting "magic" y'know? But seriously, compared to what I've been dealing with over the years I think it is going to be magical simply by virtue of how the analog path handles the transients...this is all in theory, but my primary instrument is very demanding in that area, and it is where I've felt the most limitation, of course that hasn't been helped by allowing myself to be victimized by the loudness wars...really sad.

Any thoughts on the idea of "zeroing" all the trimmers and starting from a clean slate??? Bad idea?

 

[analogkid]

To be honest here, I am just now getting back to the adjusting Bias settings on this Otari I have, and it has been.... 25 years?!?! since I did these particular adjustments on an ATR. I didn't mean to suggest it was or wasn't a big deal to get it right, just that the proper settings from the manual may not get the job done. I am going to go back and re-read my dusty notes on BIAS setting and theory to make sure my brain is up to speed on this.


AK

 

[analogkid]

OK, going back to my notes and books, I came up with some theory information that might help clear things up a bit - or make them more confusing lol.

I will paraphrase from my old recording manual:

When an audio signal is recorded without bias, the output from the tape will show a large amount of third order harmonic distortion - in other words the output waveform will show a component 3x the frequency of the input signal. As bias is added, the distortion is reduced, as the bias apparently linearizes the the signal transfers characteristic.


Now here is where it gets interesting (at least to me)

The effect of the bias signal is a function not only of its frequency, but also of its amplitude. As the bias amplitude is increased from zero, the transfer characteristic becomes more and more linear, resulting in less distortion. This may lead to the belief that the optimum bias level is that amount which is required to reduce the third harmonic distortion to an absolute minimum.

Unfortunately it may not be that simple. In a typical situation it may be found that at some point before minimum distortion is reached, the high frequency response of the audio signal begins to fall off significantly while low frequency response increases.

Now that is all THEORY - but to me it indicates that there is no 'perfect' setting for bias. Add in older electronic components whose values may have 'drifted' a bit, and getting an older deck back to perfect specs is likely impossible. This does lead me to as another question though - have you or do you have a way to check that the bias frequency is set correctly?

hope this is more helpful than confusing.


AK


Edit: I forgot to say that what that bit of theory means to me is that you must balance the amount of bias against the level of distortion, and that while the manual gives you specific numbers, you may have to live with a bit more distortion to get a decent high end response.... seems like you had said earlier in this thread that you were getting very weak high-end response which MIGHT indicate TOO much bias..... just food for thought, you may curse me at your leisure

 

[sweetbeats]

AK,

just food for thought, you may curse me at your leisure

I am totally on-board with everything you presented both from your mind and your texts.

The article Ethan linked, as well as Danny's link to that Gearsltutz thread have helped to pull it all together, and what you are posting is in no way contradictory.

Now that is all THEORY - but to me it indicates that there is no 'perfect' setting for bias. Add in older electronic components whose values may have 'drifted' a bit, and getting an older deck back to perfect specs is likely impossible. This does lead me to as another question though - have you or do you have a way to check that the bias frequency is set correctly?

Right...it is not a perfection thingy, but adjusting for the best performance possible for each track...I'm on-board with that. I do not have a way to check that the master oscillator is producing the proper frequency...that is one of my problems...I think I need to get something to do this though...Ethan (thank you Ethan BTW...think I neglected to acknowledge earlier...) has kindly offered to loan me his 100mHz analog scope. I'm going to have to get ahold of that or something because it is necessary in order for me to check that the bias oscillator frequency is right, and it is also necessary for setting the trap levels which could be critical to the bias circuitry performance. I'll explain that more in a soon-coming post...I talked to Jim Finch at Tascam this AM...

 

[sweetbeats]

Talked to Jim Finch at Tascam this AM...

Discussed both the tension and bias issues with him. Bias issue first:

• He actually pulled his copy of the 58 manual and read through the bias stuff with me and even referred over to the schematics to confirm what he was theorizing.
• He helped clear up some confusing language issues in the manual.
• He affirmed that I'm going to need something to confirm that the master bias oscillator is producing a 145kHz signal, as well as to confirm the bias trap settings...making sure that bias signal is not leaking excessively into the audio stream...He suggested a "VTVM" (vacuum tube volt meter)...now where do I find one of those??
• He affirmed that the output voltage of the bias tuning modules should be no less than, and close to 0.35V (in other words don't just crank the tuning module output up to 0.5V to make it work...find out what else is going on like checking the bias trap settings. If the traps are set wrong they could be choking off the bias signal causing the need to boost the amplifier setting in the tuning module, and that is not good because an overactive bias trap doesn't just make the bias signal disappear...some of it gets diverted to places you don't want it to go...traps should be set to let enough through, but not too much.)

Tension issue:

• After going through everything I've done and that Jimmy advised, Jim Finch was also stumped...really baffled by my issue with the tension arm bottoming out when shifting the transport from REW SPOOL to STOP or PLAY.
• He basically confirmed I've done all the right things.
• Affirmed that swapping the motors is a good idea...it would offer some form of a conclusive result.
• He suggested that, before swapping motors, to open up the supply reel motor and check the commutator. Carbon buildup in between the sections of the commutator will decrease torque performance. He also suggested that it could be a bad bearing. I was thinking it couldn't be a bad bearing or a dirty commutator because the motor was a NOS replacement...but its not. I now remember that I did order an NOS replacement, but when I received it I wasn't convinced that it was actually new, and it was noisy, so I actually pulled a motor off of my 48 that, with respect to speed and noise, performed like the good motor on the 58...I had to replace a bearing on that motor so it must have some mileage on it. So I'm going to do as Jim suggested, pull the motor, open it up and clear any deposits from in between the commutator sections with a toothpick or some such thing. Then if that doesn't improve things, try swapping the motors and see what happens there.

 

[analogkid]

Sounds like the Tascam guys are helping a ton! I had thought about the Bias Trap issue this afternoon. I now understand what the issue with the .5 volts is, don't know WHERE my head was the last three times I read through it - that obviously can't work. Vacuum Tube VOltmeter.... ermm no clue really, looks like an Ebay hunt to me. I will keep my eyes peeled!



AK

 

[sweetbeats]

Thanks AK...also...

I forgot to respond to this:

you may have to live with a bit more distortion to get a decent high end response.... seems like you had said earlier in this thread that you were getting very weak high-end response which MIGHT indicate TOO much bias

Actually it is the opposite. I think my high-end response is too good. It was at 0VU at 20kHz, and that may be part of my problem. The overall response curve is set to the point where the bias amp can't produce the amount of bias signal needed to deal with the signal strength going to tape...this may be augmented by how I am setting the output level of my tone generator...see my next post.

 

[sweetbeats]

Questions...Answered and Unanswered.

Okay. I've gotten one question resolved that was posted a couple times earlier in the thread and I want to make sure that the info is present in the thread.

I also have several questions from the past few days that have been posted but as of yet have gone unaddressed. I'll number these and hopefully it will help my quest by having them collected in a single post here.

The answered question:

I've been trying to get a handle on what "floating capacitance" is, and whether or not my Fluke 85 DMM meets the requirement in my 58 manual for an "AC level meter with a floating capacitance of 100pF or less" when testing the output of the bias tuning module (the bias amp, L4 on each amp card). Jim Finch stated that basically what the manual writers were getting at is that you want to use a good quality bona-fide true RMS meter, and not a spare VU meter on your workbench with soldered on wire leads, and that this point kinda got lost in the translation. So my Fluke 85 is okay for that part of the bias circuitry testing.

Unresolved questions:

1. When I am setting the input levels to the deck, am I correct that I want to set my tone generator output level with the tone generator connected (loaded) with the deck input(s)?
   
2. My manual says that I should check the bias tuning module (L4) output voltage with the bias level trimmer R211 set to approximately "3 o'clock"...Jim Finch didn't even know the answer to this one...where is 3 o'clock??? It can't be where 3 o'clock would sit with the card as it is when installed in the deck (vertical), because in that case the the trimmer only goes from about 4 o'clock to 2 o'clock. Anybody have any idea if the positional reference is with the card sitting horizontal? That would be 12 o'clock then with the card vertical...
   
3. The Gearslutz thread on bias adjustment that Pianodano linked a few posts ago included a suggestion from one user to "zero out the deck"...turn all the trimmers to their center position and "start from scratch". Any thoughts on this?
   
4. Can anybody recommend a good vacuum tube volt meter like what Jim Finch at Tascam was talking about, or would any of you recommend something else that might work as good or better that might be more readily available (and not cost a fortune)?
   
5. Somewhere these past couple days I read a caution that analog VU meters may not read accurately at high (i.e. 20kHz) frequencies; that they may read even 1 ~ 2dB's low. I have been setting my input levels at all frequencies using the VU's on the 58 (after calibrating the meters with a 1kHz tone at +4dBu), but I notice I always have to boost the tone generator output by 1 ~ 2 dB's when I jump from 10kHz to 20kHz...I've always assumed that it was something about the response of the D/A path from my PC-based tone generator to the deck, but now I'm wondering if I shouldn't trust the meters on the deck for that kind of work, but to use a more accurate level meter to set the output level of the tone generator?
   
6. And lastly is my Fluke 85 good enough for this?? My Fluke 85 is rated up to 20kHz but of course there are accuracy variances shown in a chart based on the band range...I don't quite know how to interpret those variances, and the bottom-line is that I'm trying to determine if my Fluke 85 is a reliable meter for accurately measuring levels at 20kHz. Can anybody help? Here is the accuracy chart:

 

[analogkid]

Well at least that still follows the theory more or less.



AK

 

[sweetbeats]

What about...

this?

 

[cjacek]

Cory, am I to understand that the Tascam 48 and 58 reel motors are one and the same?

Anyway, sorry to not be of help here but recently I turned something of a 'lurker', at least temporarily, until some stuff settles down but I'm still trying to follow this outstanding thread as best as I can.

Keep the thread going. You guys are great!

--

 

[analogkid]

Cory I wouldn't buy that particular O-scope - way to pricey on shipping and too high in general for an unknown condition unit. I believe you can buy a new O-scope - CRT type - starting around $300. I think a VTVM is something different tho......


AK

 

[evm1024]

Vtvm

VTVM in this case justs means "has a high input resistance so as to not load down the circuit". Your fluke if more than up to the job.

They used vacuum tubes in meters before they invented fets (like your fluke uses most likely).

To check the freq of the bias osc you should use a frequency counter that goes that high. My handheld meter goes to 20 kHz so it won't work (unless I put a divide by 10 chip in front ). Typical dedicated counters do 100 MHz.

The scope would give you an approximation of the freq (which should be OK) and a true RMS amplitude ( 0.707 times peak). The key here is that the scope as a freq response way past 145 kHz where as most meters fall far short at that frequency.



-Ethan

 

[evm1024]

Answers

Yes, the voltage measured must be into the specified load for full precision. Measuring then adding a lod pulls the voltage down. Perhaps not much... But on the other hand a 1 or 2% difference is inside your meters accuracy spec. ( meter reads correct voltage + or minus 4% for example)

3 o'clock is relative to 12 o'clock being half way/ Most pots go from about 7 to 5 o'clock. The clock scale is referenced to the pot.

Zeroing out the deck is a drastic move....For when you want to have a fresh start like when replacing the headstack. Your problem is that you do not see a peak then decrease in level as you increase your bias level.

I think that a Fluke 85 would cover your VTVM needs

Trust your Fluke (Luke use the Force)

Measuring a level of a 145 kHz tone wouuld be a problem for most meters....

-Ethan

PS no need to buy a scope. a software scope shold be fine for azimuth and the few times you need a real scope you could borrow one.

 

[sweetbeats]

Thanks again all!

Daniel:

Cory, am I to understand that the Tascam 48 and 58 reel motors are one and the same?

Yep indeed they are! Tascam p/n 5370002701 for both motors in both decks.

Nice to hear from ya, ya lurker!

AK:

Cory I wouldn't buy that particular O-scope - way to pricey on shipping and too high in general for an unknown condition unit.

Perfect...just the clear advice I was hoping for. Thanks!

Ethan:

So if I am reading correctly, your posts provide the following for my 6 questions:

1. Answered! Make sure that them there tone generator is connected to the deck when dialing up the input level as referenced to the AF level meter (Fluke 85)!
2. Answered I think...So the clock face is oriented with the base of the trimmer (i.e. 6 o'clock is closest to the PCB) and therefore when the card is installed in the deck (PCB oriented vertically) 3 o'clock on the trimmer is pointing straight up, right?
3. Answered I believe! Don't get crazy and zero everything out (yet).
4. Answered! Forget the VTVM...use the force...er...the Luke!...............I MEAN THE FLUKE!
5. Still at large...
6. Partially answered...So for my Fluke 85, I can assume that if I am reading a 0.615VAC signal at 20kHz, that the actual voltage may be 0.590 ~ 0.640VAC (0.615 +/-4%) right? That's about a 0.7dB swing...I can swallow that, though 20kHz is at the edge of the rated accuracy specifications...But what is that +4 next to the 4.0% on the spec page (i.e. on the 4.000V range from 5 ~ 20kHz it says there is an accuracy of + or - "4.0% + 4"?

 

[cjacek]

Cory, am I to understand that the Tascam 48 and 58 reel motors are one and the same?



--

Daniel:



Yep indeed they are! Tascam p/n 5370002701 for both motors in both decks.

Cory, that is an eye-opener 'cause, for some reason, I thought the 58 reel motors were different and bigger altogether. I guess my memory fails me..

Also, in the 48 manual, the 38 and 48 share the same part # with no mention of the 58, even though the 58 was introduced a year earlier. Strange but indeed true that the 38, 48 and 58 share the same reel motors. Wow!

I guess it's safe to say then that the 48 is basically a 58 with a bit less robust transport. Cool to know that both machines [which incidentally weigh about the same] are of roughly equal quality.

----