Tascam TSR-8 set up (bias...) + Tascam 58 tension arm rollers

[bleachboy]

Hello everybody !

I've had a TSR-8 and 58 for some time waiting to be set up and I'm finally doing it. I'm starting with the TSR-8 which I have used successfully for a couple of years but since I've had all of the heads relapped a while ago I wanted to do a thorough setup so I can finally get recording again.
So far :
- Transport is fine
- I've got the correct MRL 250 nWb/m test tape
- Input levels are all set to 0.316 mV RMS showing 0db on the VU meters
- Repro level is at -3db on all channels (I want to record at +3)
- I've "think" I've managed to set up the erase bias tuning and record head bias tuning
- All of the 0.316mV 1kHz test recordings I've done show at 0db on the machine
- I've recorded 4 stereo stems of a mix I'm doing into the machine, playback levels are pretty identical to the input levels all across the 8 channels

What I'm not 100% sure of is if the frequency response is identical across all channels. It sounded close enough to me on playback, but I'd like to make sure. As far as I know, bias setting on the 58 is much easier than on the TSR-8 because it's a 3 head machine and the TSR-8 is not.

The only thing I can't seem to do is set up the bias trap and record bias correctly, since they are AC and I only have a cheapo FNIRSI 2C23T oscilloscope/signal generator/multimeter. At the required test points the AC multimeter shows nothing, what I'd like to know is can I use the oscilloscope to test the AC voltage between TP6 and TP7 for the bias trap, and TP1 and TP7 for the record bias, and if so: how should I proceed ?

Second question is : on the 58, while transport functions seem to be working fine, I have a little trouble with the two tension arm rollers. According to the attached picture, is this the right way for them to be set up ? The washer on the pic is present in only one of the two tension rollers, I guess that's a problem. I'd like to know the specs of that washer so I can purchase another for the other roller, and also if it's supposed to be placed at the bottom of the assembly or at the top ? That's not shown in the manual. I'm not sure if the two rollers in the middle are supposed to move and roll freely along with the tape movement or if they're juste supposed to be "static" while the tape runs across them. In any case, the quick tests I've done showed that the tape rubs against the rollers but the rollers themselves don't move. Maybe I've screwed up the assembly of those rollers in a way I can't quite understand, but when I screw the top part the pressure of the spring on the bottom makes it impossible for the two middle rollers to move freely when tape is rolling.

Thank you all so much

 

[ecc83]

SANWA N-501 MULTITESTER - ANALOGUE MULTIMETER 500000Ω/V AC/DC V, I , R | eBay

Find many great new & used options and get the best deals for SANWA N-501 MULTITESTER - ANALOGUE MULTIMETER 500000Ω/V AC/DC V, I , R at the best online prices at eBay! Free shipping for many products!

www.ebay.com

If you get yourself something like the above you can make a simple preamp/booster with an op amp or even discrete transistors (FET at the front) That will give you a meter that will read well past 200kHz and you calibrate it (at 400Hz) with a digital meter and tones out of your laptop. Or a good sig genny if you have one of course!

Dave.

 

[sweetbeats]

I'm starting with the TSR-8…

- Repro level is at -3db on all channels (I want to record at +3)

- I've "think" I've managed to set up the erase bias tuning and record head bias tuning…

Just a couple comments…I don’t know what tape you are using, or what your plans are, but if you record hotter like that using a “+6” class tape like SM911 keep in mind you may run into tracking errors using the dbx noise reduction. This is because the closer you run your nominal record signal to the tape’s MOL, the greater the chance of saturation/distortion, and the dbx often can’t accurately track with the distortion. The solution is to just track without the dbx. If you are using hotter tape, like a “+9” class tape, then there is less chance of distortion and you may be fine with the dbx, but sometimes even hotter signal causes tracking issues because the dbx signal path has lower headroom. But often if you are using the higher MOL tape and track hotter, you have greater dynamic range and don’t need the noise reduction. Just food for thought…again, not sure what your plans are or if you’ve already thought through this, but just wanted raise these points just in case.

Also, typically you don’t need to mess with the bias traps and tuning unless you’ve changed components in the bias circuitry. And in fact this should be avoided unless you have the proper spec test equipment to do it. More below.

What I'm not 100% sure of is if the frequency response is identical across all channels. It sounded close enough to me on playback, but I'd like to make sure. As far as I know, bias setting on the 58 is much easier than on the TSR-8 because it's a 3 head machine and the TSR-8 is not.

First of all, you’re talking about two different things in your statement above…while the bias level does impact the record frequency response, adjusting the bias is not how you directly adjust the record frequency response. So is the above supposed to be two separate paragraphs and topics, or is what I’m saying helping to clarify some things?

The electronic alignment/setup procedure always follows the mechanical checks/adjustments, and then the electronic alignment is mapped out in the order it’s supposed to be executed in the manual (I’m referencing the -03 revision manual for the 58):

1. Input level
2. Meter adjust
3. Reproduce level
4. Reproduce frequency response
5. Bias level
6. Record level
7. Record frequency response

Step 3 should be accomplished with your test tape. It should ideally have a multi-frequency tone ladder on it so you can check for proper response from 40Hz to 22kHz (on the 58 in this example). The tape should have at *least* 100Hz, 1kHz and 10kHz tones on it, but that really tells you little about how the machine is performing across the rated response spectrum.

For the record frequency response (which is the “overall” frequency response because it takes into account the record AND reproduce response anomalies), you do steps 5, 6 and 7 and then sometimes have to go back to step 5 and rinse and repeat, because setting the bias does impact the record frequency response and if you have to tweak the bias because the HF response is off that change to the bias level will effect the record level and then the response curve…so you re-tweak the level and the EQ and then re-check the bias and so on until everything is dialed in. But you don’t use the bias level to adjust the response curve per se. It’s interrelated. But setting the bias prioritizes minimization of distortion without over-impacting HF response. That’s the priority reason for setting the bias…that’s what printing bias signal with the recording does. But it monkeys with level and the response curve so that’s why the rinse and repeat until those three things are dialed in.

As for verifying frequency response is the same across all tracks, just follow the procedure in the manual. First, as I indicated above, you play back the tone ladder from your test tape to align the playback frequency response. Then you record the same tone ladder from your signal generator and tweak the EQ trimmers until it’s the same. You can certainly do it by ear if you want just listening to playback of program material, but if you want to be quantitative about it and actually measure it then, after aligning the playback frequency response using your test tape, you record tones to your blank tape and play them back and tweak the appropriate trimmers until all tracks are within spec and close enough to each other. And BTW, to your statement about bias setup being easier on a three-head machine compared to a two-head machine, YES. In fact EVERYTHING about the record-side calibration is WAY easier and faster with a three-head machine. All of it you can inject tone and monitor off the repro head, look at the meters and tweak in real time as you go; no shuttling. If you do use the deck’s meters keep in mind the ballistics on the Teac meters aren’t great at high frequencies. They’re off by 1-2dB at 10kHz+ IIRC. You can check with your audio frequency level meter. But if you take this into account it’s easy to use the deck’s meters to line things up. Also, keep in mind fringing effect…this causes a bass boost and there can also be resonant bleed at HF from track to track and throw off the meters, so I like to calibrate in sets of four tracks (1, 3, 5 & 7, and then 2, 4, 6 & 8). This can hold true with the playback calibration too…check to see if your cal tape is “fringing compensated”…it may say on the documentation with the tape, or if it’s an MRL tape check the website against the catalog number and get the specs. For my 58 I have the original Teac YTT-1144-2 tape, which is fringing compensated. If your tape is not compensated I think there are tables you can use to calculate the compensation.

I pity the tech who has to line up an MSR-24…two-head 24-track. Ugh.

The only thing I can't seem to do is set up the bias trap and record bias correctly, since they are AC and I only have a cheapo FNIRSI 2C23T oscilloscope/signal generator/multimeter. At the required test points the AC multimeter shows nothing, what I'd like to know is can I use the oscilloscope to test the AC voltage between TP6 and TP7 for the bias trap, and TP1 and TP7 for the record bias, and if so: how should I proceed?

Again, why are you tweaking with the bias tuning and traps? Did you change a coil? The heads are the same? Just relapped? I mean, check away if you want, but the manual clearly states you don’t need to do that unless you replace the erase head, sync head, master bias unit or master bias PCB. If you want to do it anyway, again, go for it, but do not touch a tweaker to a trim pot unless you have the right equipment. Yes you can always use a scope to measure AC and DC voltage, but you have to ask yourself if your equipment, even a scope, is accurate/up to the task. I would not use the FNIRSI unit for this. You have to have an audio-grade meter that can accurately measure the bias frequency, which is 150kHz. Most scopes can do this, assuming they are in good repair, but I’m not confident in the FNIRSI unit. I have their component tester (like for transistors, diodes, capacitors, etc.), and I think it’s fine for testing dead vs not dead and doing relative comparisons, but I take test result values with a bit of a grain of salt, like ESR measurements and such.

Second question is : on the 58, while transport functions seem to be working fine, I have a little trouble with the two tension arm rollers. According to the attached picture, is this the right way for them to be set up ? The washer on the pic is present in only one of the two tension rollers, I guess that's a problem. I'd like to know the specs of that washer so I can purchase another for the other roller, and also if it's supposed to be placed at the bottom of the assembly or at the top ? That's not shown in the manual. I'm not sure if the two rollers in the middle are supposed to move and roll freely along with the tape movement or if they're juste supposed to be "static" while the tape runs across them. In any case, the quick tests I've done showed that the tape rubs against the rollers but the rollers themselves don't move. Maybe I've screwed up the assembly of those rollers in a way I can't quite understand, but when I screw the top part the pressure of the spring on the bottom makes it impossible for the two middle rollers to move freely when tape is rolling.

That special washer goes in between the two roller bearings. Yes these are rolling guides…they use standard 626ZZ shielded bearings…two of them stacked to make the 1/2” rolling portion. The special double-shoulder brass or bronze washer, again, goes between the bearings. The top and bottom caps (and BTW your bottom cap on the guide in the foreground is upside-down, which is probably why it’s not working right…) are static. These are height-adjustable guides. All the info you need on how to adjust them is in the manual. See pics below of one of mine…But to adjust you loosen the socket-head cap screw on the top using a 1.5mm hex bit, and then turn the top cap clockwise to lower the guide, counter-clockwise to raise the guide, then lock it in place with the socket-head cap screw.

 

[bleachboy]

Thank you as always !

First of all, yes my MRL calibration tape is fringing compensated, four frequency (1kHz, 10kHz, 16kHz and 50Hz). And the tape I'm planning on using is SM911 RMG tape. I thought about going SM900 at +9 but I thought it could be a little too hot for those heads ? I might be wrong but I thought I read somewhere that the heads on this machine don't handle +9 as well as +6. If not I might very well go down that road later on.

Also, I might add that the rec/rep head azimuth alignment is bang on. One other thing that makes alignment difficult in these machines is the digital VU meter, it's way more difficult to get exact results as opposed to the 58 which has regular analog VU meters (with fresh bulbs I might add !!)

Thanks for the lengthy clarification about bias. I'm gonna avoid messing with it at least until I get the proper tools, that goes without saying ! One thing I have indeed overlooked is the record frequency response, since it comes after bias setting in the manual I didn't want to skip that part and that's why I came to you for help !

The main reason why I was thinking of setting up bias is because I sent both my TSR8 and 58 headstacks to Terry Summers to be relapped at the same time, and asked him to take the least worn heads and to install and align them on the 58 headstack, because that's going to be my main machine and I'm gonna use the TSR8 mainly to dump tracks for downmixes. So at least one of the TSR8 heads was installed on the 58 and vice versa.

As for frequency response adjustment, is it good practice to apply pre-emphasis EQ on a tape machine ? Meaning that in order to potentially reduce tape hiss, the logical thing would be to apply a little more high frequencies on the recording side (let's say +3db at 16kHz for example), and compensating that high freq boost on the repro side by reducing 16kHz by 3db ? Or is the equalization characteristics of the IEC1 taking care of that for me ?

In any case, the mix stems I've recorded on the machine after the initial setup seemed to sound a lot better than the last time I've used the machine a couple of years ago before the head relap. Way more consistent all around, less hiss etc...

EDIT : I've recorded the left side of Supertramp's Breakfast In America on all 8 tracks, and I'm happy to say that on playback it's bang on with the original WAV file, except maybe a little discrepancy in the EQ of track 4 and 5 which sound just a little dull compared to all other tracks.

 

[ecc83]

Regarding "bias", it has been at least a couple of decades since I was setting that (mainly for cassette machines) and had the luxury of a Ferrograph test set to boot!

One critical measurement IIRC, at the root of all others was the distortion at some specified recorded flux level. 3% THD was a value that springs to mind? You don't need an AP systems for such a level of distortion and a perfectly adequate meter can be made employing a tunable parallel T filter and again our friendly op amp.

I recall that most "hi fi" cassette decks came biased for a fairly "soft", modest MOL tape and re-biasing them for TDK AD or even better TDK SA gave substantial improvements in noise performance and HF response, even before Dolby B was engaged.

Dave.

 

[sweetbeats]

First of all, yes my MRL calibration tape is fringing compensated, four frequency (1kHz, 10kHz, 16kHz and 50Hz). And the tape I'm planning on using is SM911 RMG tape. I thought about going SM900 at +9 but I thought it could be a little too hot for those heads ? I might be wrong but I thought I read somewhere that the heads on this machine don't handle +9 as well as +6. If not I might very well go down that road later on.

Ah okay. Cool. That tape is not the basic 100Hz/1kHz/10kHz tape, but also not as comprehensive as the tape with the full tone ladder. But those frequencies are good and will do a reasonably good job of helping you get your response curve setup on the playback side, and of course setting/checking azimuth.

Using higher SOL (standard operating level) tape doesn’t have anything to do with the heads; being too “hot” for the heads. The issue you’re referencing is an issue with the rec/play amp cards in the 58…the bias amp doesn’t produce a high enough signal level to properly bias most +9 class and above tapes. The solution is to replace C42 on each card (a 100pF film cap) with a 180pF cap. This allows a little higher output from the bias circuit in each amp card. So it’s nothing to do with the heads. It’s about the bias amps.

Also, I might add that the rec/rep head azimuth alignment is bang on.

Which one? The TSR-8 or the 58?

Here’s a little tip…on a three head machine, use your cal tape to set the azimuth on the sync head…then record azimuth coarse and fine set tones on a blank tape, and use that recording to set the azimuth on the repro head. It ensures a tighter match in azimuth set between the sync and repro heads.

One other thing that makes alignment difficult in these machines is the digital VU meter, it's way more difficult to get exact results as opposed to the 58 which has regular analog VU meters (with fresh bulbs I might add !!)

Nice on the new lamps!

I’ve never been bothered by the LED meters for calibration. I just consider the LED barely lighting and holding steady as representative of whatever its indicated level is…like, if I’m setting 0VU, I just adjust the trimmer until that 0VU segment just lights and holds steady. If you take your level meter and measure the output and actually look at the difference between just barely lighting and on-steady it’s such a small difference, so I don’t get hung up on the absence of the infinite feedback of the analog meter. That being said, I too much prefer the analog meters. I also like the meters on my Studer console which, rather than 10 segments, have 30 segments over a tighter range…so kind of the best of both worlds.

The main reason why I was thinking of setting up bias is because I sent both my TSR8 and 58 headstacks to Terry Summers to be relapped at the same time, and asked him to take the least worn heads and to install and align them on the 58 headstack, because that's going to be my main machine and I'm gonna use the TSR8 mainly to dump tracks for downmixes. So at least one of the TSR8 heads was installed on the 58 and vice versa.

So…wow…okay…so you had heads re-lapped and you swap-a-ronied heads around. So if you changed the R/P head on the TSR-8, yep…you’ll need to check the bias tuning and trap adjustments. And if that head from the TSR-8 went into the sync position on the 58, then same thing there. Is that the case?

As for frequency response adjustment, is it good practice to apply pre-emphasis EQ on a tape machine ? Meaning that in order to potentially reduce tape hiss, the logical thing would be to apply a little more high frequencies on the recording side (let's say +3db at 16kHz for example), and compensating that high freq boost on the repro side by reducing 16kHz by 3db ? Or is the equalization characteristics of the IEC1 taking care of that for me ?

First of all, understand that tape EQ standards have nothing to do with noise reduction…their intent is to promote flat record/reproduce response. So asking if the IEC standard addresses tape noise is obtuse. I’m not saying that to be judgmental or anything like that, I just want to help you understand your inquiry and its focus on tape EQ standards is disconnected. If you want to try your own pre-emphasis experiment you should try it. I’ve not done this. This is along the lines of the Dolby noise reduction ideology. The problem is tape noise is tape noise. It is semi-broadband across the HF spectrum, and the HF EQ adjustments may or may not (likely not) correspond to tape noise spectral content. Like, the 58 sync/repro HF EQ filters…I have no idea what their specs are…shelving? Peaking? Probably peaking. Q? Center frequency? I have no clue. And does that tightly or only loosely related to innate tape noise? I’m going to wager probably not. So if you add pre-emphasis using the record EQ, and then try to de-emphasize using the sync/repro EQ, it may have limited impact on tape noise. I don’t know. But understand the EQ filters on the, for instance, 58 amp card, are designed to support flat frequency response with broad-spectrum program signal. They are not designed to mitigate tape noise. Try it. Don’t let me stop you. I’ve not tried what you are talking about, but that’s because I’m focused on what the tools are designed for…which could be limiting my view. But that doesn’t change for what they were designed. IMO the best way to mitigate noise is to either properly utilize proper noise reduction pathways, or increase dynamic range (i.e. use higher SOL tape and print hotter).

 

[bleachboy]

1 : Which one? The TSR-8 or the 58?

2 : Here’s a little tip…on a three head machine, use your cal tape to set the azimuth on the sync head…then record azimuth coarse and fine set tones on a blank tape, and use that recording to set the azimuth on the repro head. It ensures a tighter match in azimuth set between the sync and repro heads.

3 : So…wow…okay…so you had heads re-lapped and you swap-a-ronied heads around. So if you changed the R/P head on the TSR-8, yep…you’ll need to check the bias tuning and trap adjustments. And if that head from the TSR-8 went into the sync position on the 58, then same thing there. Is that the case?

4 : First of all, understand that tape EQ standards have nothing to do with noise reduction…

@Dave : thanks so much for all that info ! For now it seems a little over my head with my meagre electrical engineering capabilities !!

@sweetbeats Thanks for clarifying the "too hot for the heads" issue. I knew it wasn't exactly that that was the problem but couldn't think of anything more precise than that assumption.

1 : As for the azimuth, I was talking about the TSR-8 - I haven't done anything besides input and meter calibration on the 58 because of the tape path issues I have regarding the tension rollers in particular, and I don't want to risk ruining my MRL tape on the 58 before those issues are fixed for good. Yep, changing the bulbs was a breeze and did a lot of good ! I had 4 dead bulbs but I've replaced all 8 of them for better looks.

2 : That's very good advice thanks, I'll do that when I'll get around to calibrating the 58. As for coarse and fine tones, what particular frequencies would be best ?

3 : Indeed, only the the Repro head of the 58 has been swapped with the Sync/Repro head of the TSR-8. All other heads on both machines have stayed where they were originally installed.

4 : Again, thank you for that. I'm not gonna mess with it especially with a 2-head machine, too much of a hassle without a simple and direct way to hear the results, if there are any results at all.

As far as the 58 tension roller go, I've installed the left tension roller exactly the way you had it laid out in your picture above and sure enough : the bottom part was upside down on my machine, pretty silly on my part sorry about that...
It's much better now but on the left side the two roller bearings with the washer in between them still don't roll as freely as the right tension roller bearings (which don't have that washer in the middle, either me or the previous owner must have misplaced it). Removing the middle washer on the left side gets it rolling much much smoothly, so I might settle for that, and I can finally get going with calibrating the 58 !

EDIT : I'm very happy to say that after a few additional level and REP EQ tweaks this morning, playback levels and frequency response on the TSR-8 is pretty much as close to perfect as I can make it with my very limited experience (it's the very first machine I've ever calibrated). Comparing the source WAV file to the recorded file on the reel to reel, all three speeds sound great (and very little audible hiss, even on the slowest mode). At normal speed sounds pretty much identical to the source file, on the slowest varispeed setting it is of course a little more hyped on the low end and a little less hyped on the high end, and on the highest varispeed setting the low and is very very close to the original WAV file, and the high end is a little more hyped, as you would expect. I've very happy with the results, I can't wait to put the machine to good use once again. And I think I can now approach the calibration of my 58 with a little more confidence, knowing that it'll most likely be an easier procedure overall.

 

[sweetbeats]

1. Okay that all makes sense.
2. Just follow the procedure in the manual. The 58 manual indicates to use 10kHz. I’m seeing now the manual doesn’t offer coarse and fine set instructions. If your azimuth is way out of whack (which shouldn’t ever be the case with either of your machines because the heads are semi-fixed…maybe that’s why there’s no coarse set procedure now that I’m actually thinking about it…must be thinking of a different machine…), you can initially set it using the 1kHz tone from your test tape, or something like that, then dial it in using the 10kHz tone. But in your case just follow the manual…10kHz.
3. Okay. So you’ll need to, at some point, consider checking/adjusting the bias tuning and traps on the TSR-8. Wouldn’t hurt to check the 58 as well but i would consider it discretionary.
4. Makes sense. My intent wasn’t to squelch your curiosity though…there is *nothing* wrong with experimenting. That’s one of the fun things about using tape and getting to know a particular tape machine; different tape formulations and the intersection of different operating level setups, creative use of bias levels, EQ curves, etc. We tend to setup our tape machines for same-same alignment across all tracks…level, bias, response…but there’s nothing wrong with tweaking those for better performance with particular source material. Like if you typically put the kick drum on track 1, maybe less bias for a little more distortion and bite and and then applying some HF EQ cut to counter the lesser degree of HF attenuation with the lower bias level…this is just one example of countless ways you can use tape and the electronics creatively.

Regarding the tension roller…no, no, no. The washer is not optional. And if it’s not rolling as smoothly with the washer in place then either the washer is misaligned (the shoulder on each side of the washer should nest inside the outer race of each respective bearing, essentially linking the two bearings together) or you might need new bearings. They are 626ZZ bearings. Do not go cheap…get proper Nachi, NSK, SKF bearings or the like. But the reason you can’t exclude the washer is this: the spacing between the upper and lower caps with the bearings and center washer is about 13.1mm. The spacing *without* the washer is about 12.5mm. The NAB standard width of 1/2” tape is 12.7mm. If you exclude the washer the tape will not fit between the upper and lower caps, and will curl at one or both edges. This is not an option.

 

[bleachboy]

Got it !! I'll try to swap around the 4 roller bearings to see if one combination works better than the other. But at the very least I'll need to find a second washer with the correct specs, problem is I can't really measure the washer dimensions accurately (especially the "depth", for lack of a better term).

 

[sweetbeats]

Here is a cross-section i drew up of the washer…it is not to scale. Dimensions are in millimeters:

I believe the material is brass. It is a precision part.

I’m reasonably confident you’re not going to find this off-the-shelf…i suspect this was custom made for or by Teac.

The other alternative is to find some precision shim washers with an ID of 6mm, OD no greater than maybe 12mm, and 0.5mm thick…these are precision parts and can usually be found in stainless steel and brass…and if you can’t find 0.5mm thick maybe you can find something that adds up to 0.5mm, like stacking five 0.1mm washers. These would go in between the bearings. This wouldn’t be as ideal as the factory part because the outer races would not be coupled, but it likely has no impact…regardless, I’d put the modified roller on the takeup side.

 

[bleachboy]

Wow thanks so much, that's gonna help a lot. I'm sure I'm gonna find that missing shim washer by accident in a couple of years but yeah I'll need a replacement in the meantime. I found a couple of places online that sell 06 x 12 x .5 steel washers for nothing, however I'm having a hard time finding brass washers, is that critical ? https://www.bricovis.fr/produit-ron...88-epaisseur-0-5-roncal98805/#RONCAL06/12/0.5

I also might have to go to a local machine shop for another matter (a vintage guitar bridge cover), I could maybe ask them if they would be able to make that particular shim washer from scratch ?

 

[sweetbeats]

The brass is not critical…steel is fine, but I was talking about stainless steel…the one you linked is raw steel if I’m reading it right (or if my browser is translating correctly), so you’ll want to put a light coating of oil on it to prevent oxidation. Otherwise that would work.

Your local machine shop may or pay not be able to make a copy of the factory part, which is not a shim washer but a shoulder washer. It might take specific equipment like a screw machine to make that part, and not every machine shop has one because it’s typically in a shop that specializes in making fasteners that has one of those, but you can always bring the one you have in and ask…I’d bring it in because they’d be able to more accurately measure it, and look at it and right away tell you if they can do it.

 

[bleachboy]

Hello !

I managed to find the missing shoulder washer and installed it, and then I finally went to work on calibrating the 58 ! It's crazy how much faster it went compared to the TSR-8, that was a relief ! Visual check of the tape path was OK, no curling on the edges of the tape. But I guess I'm still gonna get in touch with the machine shop to try and have a batch of these made, just in case.

I eventually chose to start the azimuth alignment with the REPRO head because that's the one that was replaced on the machine, I managed to make it almost perfect at 10k and 16k. Then I checked the SYNC head and it was also almost perfectly in phase without adjusting anything.

1 : Input and meter calibration was a breeze (0.316 Vrms IN/OUT at 0 vu on all channels, no problems there), on with the MRL test tape, I did the same setup as with the TSR-8 which is to get -3db VU at 1kHz on both the SYNC and the REPRO heads, so that I can boost the record level +3db later in the alignment process. Same with the REPRO and SYNC EQ, the manual says to set the EQ to read 0db VU but I set them up to read -3db VU, same reading as the 1kHz tone, I'll compensate later on when it's time to set up the REC EQ.

But the 58 wasn't in great shape at all when I bought it, so it did give me some trouble, namely :

2 : When it came to the bias it was a little harder and weirder than expected : I didn't mess with the bias tuning and bias trap at all, and bias tuning isn't even directly accessible on the channel card anyway. The bias level trim pots all acted strange - in the exact opposite way as what's written in the manual which says "turn the bias adjustment R211 fully CCW, then rotate it CW so that the VU meter rises to a peak reading, then continue the CW rotation so that the VU meter reading drops by 4-5db from the peak..."

What happened to me was that with R211 fully CCW the needle on the VU showed maximum reading, way in the red over +3 at pretty much the extreme end of the scale. Turning R211 CW made the needle drop back to minimum reading on the VU. So now I'm not quite sure what to do exactly to get a proper bias level ? That sort of weird inverted pot behaviour happened to me on the TSR-8 also, where the record level trim pot acted opposite to what it was supposed to do, on all 8 channels.

3 : I made a test recording after adjusting the REC level and REQ EQ so that both read 0VU. Those 20 second recordings can be found here on this link : TASCAM 58 TEST.
Levels and EQ on some channels still need to be fine tuned.
I understand that since the bias setting wasn't conclusive at all, that test recording might be futile. Most channels played back okay, except channel 8 (only on REPRO, SYNC playback was fine all around). As you can hear on the recording Channel 8 on REPRO is way quieter than all others and sounded kinda muffled, it had a lot of noise and sounded to me like there was a bad connection somewhere. I switched channel cards 8 and 7 so see if the issue followed the card (which it did) or if it was an issue with the RCA output 8 (but since I had no problem playing back the tape on SYNC output, it couldn't have been a faulty or dirty RCA connection). So my guess is that it's either a bad setup on channel card 8, which I hope it is and seems more likely, or perhaps an issue with the REPRO head and I really hope it's not that. I could also be the connection between the REPRO head and the head PCB.

EDIT : as I was writing the above I decided to check the REPRO head connection to the HEAD PCB and sure enough, it was half-disconnected out of its socket at the right end of it which corresponds to the 8th channel. I'll double check later with another test recordings to make sure that's where the issue was.

4 : Transport seemed fine when using the MRL tape, but it's only 4 minutes long so pretty light weight. When I put a full 10 inch roll of SM911 to make test recordings, I noticed an intermittent issue with the rewind speed which was abysmal most of the time (at least when most of the tape is on the supply reel). Distance between the two E points of the tension arms in EDIT mode with the same quantity of tape on both reels is supposed to be approx. 112mm according to the manual, and it was approx. 115-116mm on my machine which seems close enough.

 

[sweetbeats]

So, if your cal tape is 250nWb/m, but you’re setting up your record side to print hotter relative to 0VU (by reducing the reproduce level by some amount, -3dB in your case), don’t you have to drop your input level by that same amount when setting the bias? Like take whatever the manual says you’re supposed to do and reduce it by the same amount as your repro offset? The bias setup, when you’re using the over-bias method, is relative…it’s not about what the meter actually reads, it’s about whatever the peak is and then continuing turning up the bias level until the meter reads -4 to -5dB below whatever that peak was. It might vary track to track and that’s fine. But the peak is going to be significantly over 0VU on a normal setup. So I think what you’re experiencing is, because you’re pumping signal +3dB higher than a normal setup, the peak level is sitting above the meter range, as well as the level when the bias level is backed all the way off. See what happens if your 10kHz tone you’re inputting is at -3VU rather than 0VU. And am I correct in understanding you checked your reproduce frequency response and got that all within spec, but at -3dB across the response range? So -1VU to -5VU from 40Hz to 22kHz? Probably best to use your level meter to actually measure this because the beach meter ballistics are skewed at high frequencies. I wouldn’t use the meters for measuring signal at 10kHz+. Also, if you’re trying to do all tracks at once, like you’re injecting 10kHz tone to all 8 tracks at once and monitoring and tweaking, I guarantee hitting the electronics with 10kHz tone at +3dB over 0.316V is going to make the meters be weird because of resonance between adjacent channels. I think I mentioned to staggger it (do 4 tracks at a time, the odds and then the evens). So take these things into account and post back when you know more.

Regarding the rewind speed, what happens, when it’s sluggish, if you manually move the tension arms closer together?

 

[bleachboy]

Great advice as always, thanks a lot. Quick thing : I understand that bias levels on the VU will be out of range because of the +3 db record level (I’ll try what you’ve advised) but my main concern was that the bias level potentiometer was acting opposite to what was described in the manual, and I didn’t know what to make of it.

I've spent another day on it, starting the calibration procedure from scratch once again, and I'm very happy to say that this time it worked great all across the 8 channels, IN/SYNC/REP levels are spot on, EQ as well."Wow" is almost non-existent, which was great because on the other hand Wow was a bit more noticeable on the TSR-8 for some reason.

About that intermittent 8th track, it occured to me that the issue was more likely to come from the SYNC/REC head, since I had no problems playing the MRL test tape in REPRO mode. I've cleaned and re-seated the connections from the heads to the head PCB and the problem was gone, what a relief !

Regarding the rewind speed, it still acts a little weird but I've adjusted R1 through R4 according to the manual and it did improve things a bit. Sometimes there's no issue at all, but I seem to recall that the issue seems to occur when the machine has been on for some time. I reinstalled all the front panels and controls, as well as testing the transport control with my RC 51 remote, and all is well, time to get recording ! I will let you know how this goes during the next few weeks, but in the meantime I want to thank you SO MUCH for your invaluable help guiding me though the procedure. This will come it handy later on I am sure ! It was the first time I'd ever calibrated a machine (let alone two...)

Cheers !

Louis

 

[sweetbeats]

I understand that bias levels on the VU will be out of range because of the +3 db record level (I’ll try what you’ve advised) but my main concern was that the bias level potentiometer was acting opposite to what was described in the manual, and I didn’t know what to make of it.

What I’m trying to say, but maybe I’m not making myself clear or I’m misunderstanding what you’re saying, is because of your modified setup, the sweep range during the bias setup is outside the VU meter range. I’m suggesting the behavior of your bias level trimmers and what you are seeing on the meters is normal. With a normal setup, bias trimmers fully CCW, the meter might show - something VU, and you rotate the trimmer CW and the level rises, and then eventually drops, and you keep turning until the level on the meter shows -4 to -5VU below whatever that peak was for that track. But I think maybe between your non-standard setup, and HF meter ballistics issues compounded during frequency response setup, your starting point with the bias trimmer at full CCW is above the meter limit. So all you see is, eventually as you rotate the trimmer CW, the drop in level that happens after the peak. So it looks like when you are rotating the trimmer CW the level is going down instead of up. It is. It’s just you never see the rise because it’s all happening above the VU meter limit. Another reason I think this is the case is because you are having the same issue with both decks, and you are applying the same non-standard setup to both decks (right?). You probably should be using your AF level meter connected to the output jack rather than referencing the VU meter. And, as I said earlier, make sure you are reducing the level of your bias set tone to account for your non-standard setup…and before all that, until you get used to the flawed HF meter ballistics (and I need to refresh my memory on which direction the meters are flawed…whether they read 1-2dB higher than actual at 10kHz and up, or 1-2dB lower…), I’d be using the AF level meter instead of the VU meters throughout the setup. Frequency response diminishes with increased signal level. That’s one reason that, often times, frequency response checks are done using signal at maybe -10dB from the nominal 0dB (0VU). And I don’t know if you’re doing this or not, but you should be “loading the outputs” during setup and calibration. I don’t know about the TSR-8 manual, but the 58 manual directs you to do this…all this means is your tape machine outputs should be connected to a load; connected to the inputs of another device with appropriate impedance level inputs (like 10kohm or more). So you have to connect a ‘Y’ adapter at the tape machine output and then connect the cable to the load to one side of the ‘Y’ and you connect the other side to your AF level meter. It has impact, albeit minor, to the output level. Even though it is minor, that minor amount becomes less minor with the compounding of the multiple stages of setup and calibration procedures. And if all else fails another suggestion is to setup one track (not an edge track) using the standard setup to see if the process follows what you expect. You’ll at least know there’s not something weird with your machines and that it’s anomalies of the non-standard setup to iron out using my suggestions and any other helps you acquire. I know this is a lot of information, but I feel these are all details that matter, and when they go unaccounted for can add up in a way that can cause one to chase their tail.

About that intermittent 8th track, it occured to me that the issue was more likely to come from the SYNC/REC head, since I had no problems playing the MRL test tape in REPRO mode. I've cleaned and re-seated the connections from the heads to the head PCB and the problem was gone, what a relief !

Great. We like the easy solutions!

Regarding the rewind speed, it still acts a little weird but I've adjusted R1 through R4 according to the manual and it did improve things a bit. Sometimes there's no issue at all, but I seem to recall that the issue seems to occur when the machine has been on for some time.

I’d keep an eye on that. If performance suffers with power-up time, you may have an issue with bad solder joints in the motor power path. It’s normal for the power transistors to get hot…these are up on the heat sink or sinks above the motors…I’d have to refresh my memory on the rest of the design, like joint PCBs and what’s flowing through there…but heat causes expansion causes higher resistance causes more heat causes loss of continuity which sometimes causes simple failures and other times zorching. I’d test your theory. Fast-wind a tape when the machine cold, time it, wait an hour, do it again and compare.

I reinstalled all the front panels and controls, as well as testing the transport control with my RC 51 remote, and all is well, time to get recording !

I didn’t realize you had the RC-51 full-function remote…thought it was the basic RC-50 transport remote…cool!

Glad things are coming together.

 

[ecc83]

I have as I said, done a great many line ups and bias settings, mostly on cassette (where a 3 head machine was a HUGE luxury!) but a few OR machines. The various makes all had one thing in common in their service manuals...the internal meters* were NEVER referred to except to set them to the correct level. All other measurements were made with a wide band external AC mV meter.
Usually connected to a line out as SB says but sometimes to a test point on the PCB.

*Not to be graced with the term "VU" meters since in terms of size, ballistics and scale details they never were!

Dave.

 

[sweetbeats]

Really good point, Dave. I agree. “Back in the day”, and still presently I’m sure, techs at professional studios, typically with professional-grade machines in operation, would use the meters when lining up their machines. BUT, these were/are machines in use daily, often undergoing *daily* calibration (SOP, power up the signal chain, wait awhile until the tape machines are up to operating temp, and then line them up before the first session), and the tech staff knew/know the machines very well. Meter ballistics characteristics vary by meter/manufacturer/machine, and some are definitely better than others, but most have unique characteristics to account for, and the tech staff would know these idiosyncrasies objectively, and could therefore account for them when using the meters for calibration procedures, as well as any particular tweaks employed by the studio, like to the response curve…all part of knowing and operating the machines in that professional environment. Anybody can do the same thing, but the important thing is it starts with following the procedure in the respective manual using the prescribed test equipment, and learning the machine first.

 

[ecc83]

Thank you Sweetbeats, if I understand your last post correctly you mean that 'tape jockeys' who are on intimate, day to day terms with their machines CAN line up with the machines* meters?

I am quite sure that is the case now I think about it. My experience of course was almost all "one offs" a Sony here, a Yamaha there. Once in a bluey a Revox! I was trained to take service manuals as gospel and indeed, 'back in that day' we were knee deep in truly excellent manuals sent, unbidden and free from the companies...OH! HOW I wish I had hung onto a few kgs of those!

*I remember reading reviews of last generation multitrack recorders where each track module had a switch to put the meter into a range suitable for line up. My son, years ago was experimenting driving the **** onto our A3440 and I feared for the life of the VU meter as it was being pinned hard right so I reduced its sensitivity by 10dB on track one.

Dave.

 

[sweetbeats]

Yes, correct. But the scenario I describe is one where it’s the same tech and the same machine day in and day out…your circumstance is one where you are servicing different machines on an ongoing basis. In your case I consider it more important to follow the instruction in the manual. In the scenario I describe I think one can easily get away with using the meters *after* learning the machine and acclimating to the meter ballistics using an acceptable AF level meter.