DIY Alignment and Calibration

I'm back....

After waiting quite a while for a RMGI replacement tape due to shedding issues, the tapes were finally shipped and I'm able to resume alignment and calibration. During this down time, I took the head base off, cleaned everything thoroughly and was able to rotate all guides and even the tape lifters to get a nice machine finished side. Setup the tape path and ready to go.

Now back to the electronic adjustments, I went through all the alignment and adjustments again and was discouraged that the same thing was occurring during Repro and Sync EQ adjustment as a couple months ago. That is, the 16kHz tone level would jump all over on tracks 1 and 2 during sync playback, slow increase then decrease, back and forth, unstable levels. Every other procedure checks out except this high frequency calibration.

Previously, when this first occurred, I swapped cards from track 3 to track 1 etc. This time, frustrated, I pulled track 4 card which calibrates nicely on all procedures, pulled track 2 card and swapped them.

Well, its weird, but everything seems to work now. Just swapping track 2 with track 4 cards, tracks 1, 2, and 4 can be adjusted fine now. I went through all the cal steps again just to make sure.

Of course, not being a fan of this kind of fix without an explanation, I will keep testing this over the next week to see how stable this is. If the issue reoccurs, I have a spare 38 I bought for parts but I'd need to know where to start investigating.

I guess the question is, even though I swapped cards a couple months ago, why swapping tracks 2 and 4 cards, fixed tracks 1 and 2 now?
 
Who knows, fstrat...but I still would go after the relays. That just smacks of the exact same issues I was having with my 58 several years ago, though I was able to rectify it by cleaning the relays. evm1024 has that 58 now and I think HE ended up replacing the relays as the the issue resurfaced...which is bound to happen if you attack the problem by cleaning them if the relays are the culprit...its a temporary fix...replacement is the order of the day if the relays are getting sketchy. lo.fi.love had the same issue with his 48 and IIRC replaced the relays on that.
 
OK, well I've gotten this far, I know what a relay is, but hate to sound stupid; where are these located and what do they look like?

Edit: Actually, thinking about it, I bought that other 38 for fixer parts a couple of months ago for a couple of bags of dirt. No guarantee that everything works in that unit, but I could easily pull those cards to try first. I could deal with replacing relays later and get to some real recording usage now.

Much thanks for all the help!
 
Last edited:
You bet...that's a perfectly legitemate course. You may ultimately run up against needing to replace the relays at some point but if you can unplug and plug in and be on your way that sounds real good.

Keep us posted.
 
OK, I'm back! The 38 is running well and I will be attempting to calibrate for higher output tape sometime in the near future after I record some tunes first.

But now, I purchased a real nice Tascam 42 that survived shipping and seems to be in very good mechanical condition. I've already checked torque and adjusted the brake mechanism and at least adjusted input cal for now.

Now on to questions for calibrating this machine:

1) First, this machine requires two alignment tapes, one at 7 1/2 IPS and the other at 15 IPS. Most procedures are done using the 15 IPS tape. The 7 1/2 is used to align the heads. I wonder why I need this extra alignment tape just for this. Can't I just use the 15 IPS to align the heads? It seems kind of crazy to spend big bucks on such a limited use alignment tape.

2) The capstan seems to be always on for this unit and seems a little noisy at high speed. This is OK, but just wondering if its normal. Its just a spinning noise I guess, but I thought direct drive capstan would have been quieter than my belt driven 38. Does this seem correct?

3) When making adjustments to the tension arm positions, I accidentally adjusted the rewind tension pot. I was not intending on calibrating this as the tool needed to check tension is very expensive and frankly, not worth it. But I accidentally changed the pot on this and tried to turn it back to its initial position. Is there another way to test this besides picking up an expensive piece of test gear?

Thanks in advance!
 
Hurray that the 42 arrived well!!! Its one that Vintage Texas refurbed, guys, and it went to a responsible owner from whom fstrat purchased it so its come back full circle to an HR member! :D (hope you don't mind me sharing that, fstrat... :eek:)

  1. AFAIK, if you are always going to use the machine at 15ips then you only need the 15ips tape. You can ALSO get a two-speed tape and they're not much more than the single-speed tape...like around $130 and its got a good selection of frequencies. BUT...still...if you don't plan on using the machine at 7.5ips there's no reason to cal it at 7.5ips and hence no reason to get a 7.5ips tape. Depending on which 15ips tape you get you can do a lot of the calibration using the 15ips tape for 7.5ips operation, you just have to divide the frequency being reproduced by 2 (i.e. the 20k tone on the 15ips tape reproduces a 10k tone at 7.5ips). I'd have to look at the 42 manual but there isn't anything special about 7.5ips for aligning the heads...IOW you should be able to setup the wrap angle and azimuth at 15ips OR 7.5ips so I'm not sure what you are reading...the need for different alignment tapes is to align the electronics at the different speeds since there are different paths the signal goes through when using the different speeds so both sides require separate alignment for many facets of the alignment...but that shouldn't have anything to do with the mechanical alignment...guys? Isn't that right?
  2. I wouldn't judge the motor noise compared to the 38...totally different motors, and a belt drive wouldn't necessarily be noisier or quieter than a direct drive...its all in the individual components...my MM-1000 makes a horrendous racket...belt-drive AC hysteresis capstan motor. My 48 (same capstan motor as your 42) made a nice, smooth and purposeful whine at 15ips (only speed for the 48). My BR-20T is quieter (belt drive BTW) and my 58 was noisier. All different motors and servos. Think about putting up a video or an mp3 or something of the noise...plenty 40-series owners here and I'm sure you'll get some input.
  3. Which pot was mistakenly adjusted (can you give me the R___ number)?
 
1) Azimuth, in the manual it says "Load the reproduce alignment test tape to reproduce at a low speed of 7 1/2 ips; a scope display..........adjust the repro head azimuth screw until the scope shows less than 90 degree out-of-phase at 12.5kHz......

So that says to use the 7 1/2ips tape, but I agree with you, I really don't see why you can't adjust the azimuth using 15ips either. I have no idea? But there are references to using the 7 1/2 tape on head contact tests as well.

2) Yeah, the capstan is not that noisy, the brake bands against the felt were noisier but I adjusted those to spec.

3) I accidentally adjusted R13 under the splicing block, which controls rewind tension. Just wondering if there are any tell tale signs of over adjusting R13 without using an expensive tension meter.


So far, the machine works as described and I was surprised at the weight and build quality of this monster. Some adjustments are required but I guess I expected that since the owner before me had not made any adjustments since acquiring the unit from Vintage Texas 2 years ago.

I think I've learned that it is absolutely worth it to invest the time and expense on alignment and calibration after purchasing a deck. Would I rather be recording? Absolutely. But it's certainly peace of mind that everything is in working condition.

I'd recommend that everyone do these procedures. Why? For example, this 42, while arriving safely, and looking like all the mechanics are working, exhibited several db lower level output than input. That may not mean much, but all of sudden you could be wondering why the output levels of the machine are not producing expected results, in this case, good in, garbage out. Anyway, the point is that it's worth it to eliminate any issues at the tape deck side.
 
1) I'm comfortable stepping out on a limb and suggesting that you can adjust azimuth at either 7.5ips or 15ips, and I believe you are going to have better accuracy at 15ips.

2) Betcher just fine, but do put up a sound sample if you want that extra "yeah you're fine" response.

3) Couple thoughts...watch this segment of a video I made to promote the sale of my 48. This was after setting the tensions using that expensive tool you are avoiding, the Tentelometer. Notice that in REW or FFWD the tension arm on whatever side is the takeup side (depending on whether we are in REW or FFWD) pulls up initially and then as the motors come up to speed it drops slightly...the point at which the arm has dropped is when the tension measurment should be taken. To my eyes the amount of drop and position of the tension arm after it has dropped while in the fast-wind mode is about the same whether it is the left arm in REW or the right arm in FFWD. So I would hypothesize that you could observe the behavior on your 42 and set R13 so that the behavior of the left tension arm in REW pretty closely matches the behavior of the right tension arm in FFWD. Is this the ideal way to do it? Naw...but I think you'll be close enough to not need to worry about anything: https://www.torridheatstudios.com/ftp/share/Movies/Tascam_48_Fast-Wind_Tension_Arm_Behavior.mov
 
I will definitely give that a try on the rewind tension. One thing I did notice was that the left tension roller was not spinning all the time when if FFWD mode. I put a very light touch of lubricant on the spindle and that helped a little but that might be a sign that it is not pulling tight enough in now FFWD mode. I don't see any real major issues with FF or REW, just making sure all is well of course.
 
I think everything is in place and the tension arm seems to spin much better.

I went ahead with rotating the lifters, always a challenge. Checked the azimuth, it was in good shape so I just tweaked it a little. Moved on to the electronics adjustment and there were some things that were really off but I got everything right on.

A couple things different about the 42 over the 38, and I may have to go back and adjust the 38, no big deal, but what does it mean to put a 10k load when you are taking your electronic adjustment readings?
I had the outputs of the 42 plugged into the 10k impedance inputs on my Tascam M35 while taking readings, is this correct? If I pulled the cable from the M35, I noticed a 3mV decrease, not a whole lot anyway.

One annoying thing is that I could not rotate the left and right tape guides. I removed the top screw and top portion but the guides do not turn. I removed the head block and there are no screws or anything underneath holding those guides down. They don't seem bad enough to rotate, but someday I might have to rotate them.

I am almost there. I'm waiting for new tapes to arrive so I can do the final adjustments.
 
I am done with the alignment and calibration of the 42.

What an amazing machine. The transport, electronic access, sturdiness, and most importantly, the thing sounds amazing. Easy to adjust.

I am little confused however, and this happened on my 38 as well, and its probably my lack of electronic understanding. At 1kHz, or 400Hz, the meters read 0VU and output is 0VU and voltage is .316V. Perfect. However, I notice that when higher frequencies (8k, 10k, 16kHz) pass through any system, the meters read 0 VU (-10dbV), that's good, but a voltage level meter reads much lower (closer to -13 to -16dbV). Meters are calibrated to spec.

Why is that?
 
Meters have an accuracy range. Many standard multimeters are only relatively accurate from maybe 400Hz~1kHz...maybe a little higher. They aren't designed to be used with audio equipment and accurately present data across the audio bandwidth. So you need to look at the specs for your meter and see the frequency range over which it is rated to be accurate. My Fluke 85 is rated to be accurate over about a 4% variance from 20Hz to 20kHz.

Also, just to check, you are using a meter that measures AC volts using the RMS scale right (a "true RMS" meter)?

-EDIT-

BTW, nice work. I agree about working with the 40 series...my 48, once I go the hang of some helpful techniques, was really easy to setup compared to my 58, and maintained dead-on tape stability in the tape path under all transport modes, and I really appreciated how the guts were laid out...just a really nice machine to work with.
 
Hmmmm...

RMS "calibrated" does not mean RMS "reading"...and there are no bandwidth specs for the voltage function. My guess is that it is only acurate around 400Hz~1kHz or so and that it is not a true RMS reading meter. What "RMS" means is Root Mean Square...an RMS meter measures AC voltage based on the mean of the energy wave as opposed to the peak voltage...IOW, AC voltage is back and forth + - + - cycles right? And if you measured that on a time scale it looks like a sine wave okay? In fact that's what AC voltage looks like on an oscilloscope...up down up down over time...you can represent a facsimile of this by moving one of your hands from left to right while moving your pointer finger up and down over and over again. Okay...so a standard meter measures the peaks of the waves while a true RMS meter reads a mean of the area UNDER the wave...it is a sum of energy which is more representative of audio energy and therefore that's why its the standard for measurement of audio signals.

Google it.
 
Understood. It now looks like the test tools/meters, some even described here can provide false readings for doing any serious calibration work. I mean, even during my search, apparently there are True RMS meters and then there are really True RMS meters. The True-True RMS meters need to measure and calculate both an AC component and a DC component combined.
 
Since I was using an oscilloscope to do a lot of the cals in our MCI machine and some of the adjustments needed converting to or from RMS voltage, I used these formulas and thought it might help someone here...

RMS voltage = Peak voltage x 0.707

Peak to peak voltage = RMS voltage / 0.707 x 2
 
I still say it is as of equal importance to have a meter that is rated for accuracy across the audio spectrum as well as being a true RMS reading meter. I spent $70 on my Fluke 85. I later had to spend another $20 on an LCD repair kit but it works like a champ, is reasonably tight across the audio spectrum and is a true RMS meter. Yes, more expensive than the cheapy meters but I tried that and it didn't work out for me. I bought one that was $40 and it worked just fine but knowing that I couldn't trust it beyond 2 or maybe 4k at best wasn't good enough for me. I owned both meters for a time and did some comparisons between the two so I knew how far the $40 meter dropped when reading the tones used for calibration, and it was pretty consistent, but I just didn't want to fuss with it...and who knows what the error would be with more complex test material like white or pink noise. I just went the route I went with my degausser and got the right tool (Han-D-Mag)...recently did that with my spring guages too...GCalo sent me a link to a place that has some pretty nice ones for a decent price...got two (covering the range I need) shipped for $40 and they're nice. When you are really really tight on budget you can kind of skimp on these things but honestly in the long run it saves a lot of headaches getting the tools that work for you. My 2p.
 
Tascam 38 Calibration update: Biasing for +9.

I did another alignment and calibration on the 38 because of a noise issue I was having.

Anyway, for the fun of it, I wanted to try biasing for a +9 tape rather than my normal +6 SM911 tapes. I had a Scotch 996 tape, which was my only +9 tape, and tried to bias the 38 with it.

I was able to get enough drop in db for the tape with plenty of room left on the trimmer control, at least on my 38, so it should be possible to bias. The only issue was that the doing it the VU meter way, the meter was at the end of its travel, so it was hard to determine if it was just under, right at, or over the meter's range. It looks like it was OK.

When I use up my SM911's and have time to buy a new +9 tape, I will try the bias again and use the low frequency pulse listen method also. I'll report those results when that happens.

In the meantime, if someone else has successfully biased the 38 for +9, please post your experience with this as I'm am interested in using +9 in the future.
 
Tascam 38 Calibration update: Biasing for +9.

I did another alignment and calibration on the 38 because of a noise issue I was having.

Anyway, for the fun of it, I wanted to try biasing for a +9 tape rather than my normal +6 SM911 tapes. I had a Scotch 996 tape, which was my only +9 tape, and tried to bias the 38 with it.

I was able to get enough drop in db for the tape with plenty of room left on the trimmer control, at least on my 38, so it should be possible to bias. The only issue was that the doing it the VU meter way, the meter was at the end of its travel, so it was hard to determine if it was just under, right at, or over the meter's range. It looks like it was OK.

When I use up my SM911's and have time to buy a new +9 tape, I will try the bias again and use the low frequency pulse listen method also. I'll report those results when that happens.

In the meantime, if someone else has successfully biased the 38 for +9, please post your experience with this as I'm am interested in using +9 in the future.

I've had no problems biasing a TSR-8, 22-2 and 32 for GP9 in the past and I've known people who biased a 38 no problems. Yet I've heard of others not being able to do it with 30-series decks. These were made for so long I'm sure there were some revisions. This may account for differences in success, but not sure. manufacturers do often revise circuitry over time, sometimes for the better and sometimes to cut corners, or if a component becomes unavailable and they have to source another out. It would be interesting to learn which vintage are able to bias up and which aren't. :)
 
Back
Top