Calling Reel-2-Reel Tape experts - TEAC A3440 15ips speed problems

[ToTo Man]

I've just bought a set of 10.5" Ampex reels to load new tape stock onto. The reels already have tape on them and according to the seller they are ex-studio, but he was unable to test them. Most of the reel boxes are unlabelled but one has a Red Bus Recording Studios label on it with Chris Hunter written as the artist. I'm curious to have a listen before I trash them, but I don't want to risk mucking up the heads with sticky she'd syndrome, so I intend to bake them in the oven at 50'C to be on the safe side. I've never baked a tape before. Many of the reels only have around 10 mins of tape on them. Will this affect the impact of the heat on the reels and how long I should bake them for? (I presume 50'C isn't hot enough to warp the metal flanges?!)

 

[ToTo Man]

A couple more observations about the A3440:

- When I rewind a tape back onto the supply reel, the wind is loose and sits forward against the front metal flange.

- When I tighten and undo the NAB adaptors, the surface of the metal flange gets scuffed by the adaptor.

Does this happen with all R2R decks?

 

[Tony L]

Is the deck level? Gravity certainly plays a part here. No way an undamaged (i.e. flat) reel should scuff against anything. Bent reels are not that uncommon.

Most studio decks are angled backwards or even flat, though I suspect more due to ergonomics than anything to do with spooling, e.g. way easier to do splicing in that orientation. Probably helps with spooling though.

 

[ToTo Man]

Is the deck level? Gravity certainly plays a part here. No way an undamaged (i.e. flat) reel should scuff against anything. Bent reels are not that uncommon.

Most studio decks are angled backwards or even flat, though I suspect more due to ergonomics than anything to do with spooling, e.g. way easier to do splicing in that orientation. Probably helps with spooling though.

The reels don't scuff when they turn, the scuffing I'm referring to is the action of locking and unlocking the NAB adaptors when I put a reel on and take it off, which leaves faint scuff marks around the inner circle of the flange. Judging by the cosmetic condition of some of the used 10.5" reels for sale on eBay it seems to be a pretty common issue, but I'm wondering if there's a way to avoid it?

As for the deck being level, the table it sits on is level so I assume the deck is also level. The tape winds onto the supply reel perfectly central when playing, it's just when FF or REW that the alignment of the wind moves towards the front flange.

I might try raising the front of the deck up a little and see if that makes a difference.

I've also noticed something else about this deck ever since I've had it. The suspended wooden floors in our house are particularly springy, and whenever you walk across the floor when the deck is on and the capstan shaft is spinning, you can hear a gentle 'knocking' sound coming from inside the deck. I suspect it's the wheel moving forwards and backwards on the bearing (see 18min 06sec to 18min 30sec of the video below for the part I think is the culprit). Again I'm not sure if this is normal behaviour when the deck is in an upright position. @337alant did you ever notice this 'knocking' sound with your A3440?

PS - Given the movement in the capstan wheel shown in the video clip, wouldn't operating the deck when laying flat cause the capstan wheel to drop down to the lowest part of the bearing thus causing increased friction and wear?

 

[Barrymagrec]

Out of interest, what would happen if you played back a 2-track stereo tape recorded on a 2-track stereo deck on a "4 tracks all in the same direction" 4-channel deck, and vice versa?

Playing a 2-track Stereo tape on a 4 track machine you would get left channel from tracks 1 and 2 and right channel on tracks 3 and 4. Depending on if your 2 Stereo tape was true Stereo (0.1 inch track width) or twin track (0.08 inch track width) you might get a lower signal level on tracks 2 and 3 due to track positioning.

Playing a 4 track tape on a 2 track machine would be the same in reverse, depending on what had been recorded you would be listening to tracks 1 and 2 combined and tracks 3 and 4 similarly combined.

 

[ToTo Man]

Playing a 2-track Stereo tape on a 4 track machine you would get left channel from tracks 1 and 2 and right channel on tracks 3 and 4. Depending on if your 2 Stereo tape was true Stereo (0.1 inch track width) or twin track (0.08 inch track width) you might get a lower signal level on tracks 2 and 3 due to track positioning.

Playing a 4 track tape on a 2 track machine would be the same in reverse, depending on what had been recorded you would be listening to tracks 1 and 2 combined and tracks 3 and 4 similarly combined.

So a 2-track deck distributes audio frequencies evenly across each track? Meaning that, assuming the tracks all line up correctly, a 4-track deck would still be able to play a 2-track tape even if only tracks 1 and 3 were being monitored?

Again, assuming the tracks all line up correctly, would there be any difference in the quality of the sound output by replaying a 2-track tape through only channels 1 and 3 of a 4-track machine versus replaying it with all 4 channels monitored?

 

[Barrymagrec]

There should be no difference in frequency response but the noise figure will be slightly worse as you are not using all of the original track width. If you mix the relevant tracks together you should improve the noise figure slightly, probably not enough to worry about.

 

[ToTo Man]

I've been admiring the smooth tape spooling and perfectly centred packing on my Revox B77 since Alan worked his magic on it. I'd like to improve the packing alignment on my A3440 if possible, as the tape winds a little too close to the front flange of the metal reel for comfort. The tape isn't being physically deformed by the flange, but you can hear it rubbing against it. Is there a simple way to 'bodge' a fix for this without having to loosen off the allen screws on the internal reel hubs behind the tables and pulling them forward slightly? I thought about shimming between the reel table and nab adapter but doing this would presumably make it more difficult to lock/unlock the centre spindle?

 

[337alant]

Richard
If you remove the wooden side panels, 4 screws in each
There is a window where you can see the spooling motor shafts and the back of the reel table there are 2 grub screws in each.
If you slacken those grub screws the reel table can be moved in and out to set the table at the desired height.
So do it with tape on so you can get the tape to run in the middle of the spool
Alan

 

[ToTo Man]

Richard
If you remove the wooden side panels, 4 screws in each
There is a window where you can see the spooling motor shafts and the back of the reel table there are 2 grub screws in each.
If you slacken those grub screws the reel table can be moved in and out to set the table at the desired height.
So do it with tape on so you can get the tape to run in the middle of the spool
Alan

Thanks Alan, that sounds reasonably simple, I'll get my dad to give it a go.
On closer inspection I can see that the tape winds nicely onto the centre of the take-up reel when in play mode but pulls to the front edge when FF'ing. Does this suggest that it might only be the supply reel that needs adjusting?

PS - Does the reel table height affect recording and playback alignment/azimuth at all, I'm assuming not?

 

[337alant]

Richard,
During play, the tape guides either side of the heads keep the tape accurately in the correct position, so the reel table height does not effect the tape path over the heads.

There are 2 tape tensioners on the 3440 and its these that guide the tape onto the spool and should line up with the tape head guides, these tensioners can get badly worn or bent so it would be worth checking these first

The Play rew and FF should spool the tape in the same position, if its in the correct position in play mode then its correct dont adjust the take up table find out why the tape is drift out of position on FF
The take up tensioner had a micro switch and very little in the way of damping it is just spring loaded but it should not have any play or be bent and its arc should keep the tape in the correct position.

The supply tensioner is damped and and provides tape tension again it should not have any play and should not be bent.
Alan

 

[ToTo Man]

Richard,
During play, the tape guides either side of the heads keep the tape accurately in the correct position, so the reel table height does not effect the tape path over the heads.

There are 2 tape tensioners on the 3440 and its these that guide the tape onto the spool and should line up with the tape head guides, these tensioners can get badly worn or bent so it would be worth checking these first

The Play rew and FF should spool the tape in the same position, if its in the correct position in play mode then its correct dont adjust the take up table find out why the tape is drift out of position on FF
The take up tensioner had a micro switch and very little in the way of damping it is just spring loaded but it should not have any play or be bent and its arc should keep the tape in the correct position.

The supply tensioner is damped and and provides tape tension again it should not have any play and should not be bent.
Alan

Whilst investigating the above I discovered another issue with the deck...

I mentioned in an earlier post that the deck has a noticeable hiss in playback mode. I recorded this out to my ADC to check the level and, when no tape is playing, the noise floor of the playback circuitry is about -58dBFS when my ADC is set to capture 0dBVU peaks from the deck at -6dBFS. The hiss is audible but not intrusive, I can live with it.

I have however found a more concerning issue with the deck's input circuitry. When my ADC is connected to the A3440's outputs, I can hear RF noise (Radio Clyde 2 to be exact, mixed in with other high frequency shit) from Ch1 of the deck, but only when the deck is in Source Monitor mode. Oddly, the level of the RF noise is highest when the input knob is set to Min, it gradually reduces as I turn the input knob all the way to Max (I'd have expected the opposite??!!). Also, the RF noise is quietest on the Line Input at around -40dBFS, it's about 10dB louder if I switch to Mic input.

If I disconnect my ADC from the A3440's outputs the RF noise of Ch1 goes away, but there is still a lot of background noise on this channel (it sounds like white noise). Disconnecting the inputs actually makes the background noise louder, especially if I switch to the MicAtt input. Ch2, Ch3 and Ch4 are essentially silent by comparison.

Does what I describe sound like a faulty cap, resistor or transistor on the Ch1 input board? AIUI, the A3440's input boards are accessed and removed by unscrewing the bottom panel and simply pulling them out. Would it be worth removing the board to inspect it for signs of damage?

I could of course forget about Ch1 and just use the deck as a 3-track, but I'd like to get it fixed if possible.

 

[Arkless Electronics]

It would certainly seem that there is a fault yes. Or it could even be down to the grounding used for channel 1. The radio station could be coming from anything before the 3440 as well of course!
Do you have a scope?

 

[ToTo Man]

It would certainly seem that there is a fault yes. Or it could even be down to the grounding used for channel 1. The radio station could be coming from anything before the 3440 as well of course!
Do you have a scope?

I'm pretty sure it's coming from the ADC as it's still present even after disconnecting the source input. Sadly I don't have a scope.

 

[Arkless Electronics]

There's so many if's, but's and possibilities as to make this a potential nightmare and not having a scope triples it!
It could be as simple as a screw that holds a pcb down and grounds it to chassis is loose, it could be a myriad of other things and the noise on that channel may or may not be related to the radio station breakthrough..

 

[ToTo Man]

Revisiting the problem the next day it had jumped from Ch1 input circuit to Ch3 playback circuit, which is bizarre. The radio station seems to have disappeared, it's now just a combination of hiss and high frequency noise.

After trying lots of input/output permutations, I lost track of how each one affected the noise floor and RF interference, however my latest permutation is definitely the quietest:

- I have all 4 inputs on the deck connected (L output from amp split to Ch1 and Ch2 inputs, R output from amp split to Ch3 and Ch4 inputs).
- Ch2 and Ch4 outputs are connected to ADC.

This produces a noise floor of -70dBFS when Source Monitor is selected and -54dBFS when Tape Monitor is selected, MUCH better than before.

If I output Ch1 and Ch3 to the ADC instead of Ch2 and Ch4, the Source Monitor noise floor increases to -63dBFS and the Tape Monitor to -46dBFS, which isn't as good but still a significant improvement on what I reported in post #94.

I think I'm going to leave it be for the time being as I suspect troubleshooting further will just open a can of worms!

 

[ToTo Man]

I recorded some test tones and part of a song at 15 ips using a variety of tapes as a distinctly unscientific approach to see if I can determine which tape the A3440 is most accurately biassed for so that I can avoid calibrating it (I'm sure @Werner will be shaking his head at this! ).

If you fancy partaking in a blind test I've included a link to the recordings at the bottom of this post (the download is about 540MB in size and the link expires in 7 days).

The test track consists of sine tones recorded at -10dBVU (30Hz, 40Hz, 50Hz, 60Hz, 80Hz, 100Hz, 200Hz, 400Hz, 1kHz, 2kHz, 4kHz, 8kHz, 10kHz, 12kHz, 15kHz, 16kHz and 20kHz), followed by Periodic Pink Noise recorded at 0dBVU, followed by part of a song recorded at a level where the VU meters briefly touch but never exceed 0dBVU.

Each sine tone lasts for 5 secs and is bookended by 2 secs of silence. The Periodic Pink Noise lasts for 10 seconds.

Note that on the "Tape D" file the 12kHz and 16kHz tones are missing and some of the other tones have been shortened to 2 seconds, this is because this particular tape was short so I had to reduce the running time of the playlist to fit.

The runners and riders are:
- Maxell UD 35-90 (bought sealed, I opened it 10 years ago but never recorded on it, so I’m classing it as virgin tape stock),
- Maxell XL1 35-90B (bought sealed, I opened it 10 years ago and only recorded on part of it, I used the remaining unrecorded section for this test and am again classing it as virgin tape stock),
- Ampex 456 (bought used as part of a job lot, I baked it in my oven last week so I could listen to was on it before trashing the tape and re-using the metal spool, but it had already been erased so I recorded on it for this test out of curiosity),
- RTM LPR-90 (bought new a few weeks ago, virgin tape stock).

I have also included two 'Control' files, the actual digital master file of the test track, and the test track after being passed through my recording chain (DAC>Amp>A3440>ADC) so that you can judge the transparency and channel matching of the A3440's input and output circuitry without any tape stock involved.

I set my ADC’s input level at 12 o’clock, this records 0dBFS peaks from my DAC at approx -5.8dBFS. I have therefore reduced the output level of the digital master file included in the download bundle by about -5.8dBFS to level-match it to the ADC captures as best I can.

The digital master file is 24-bit/44.1kHz. The ADC captures are 24-bit/48kHz. (48kHz is the lowest sampling rate my ADC captures at).

The same record and playback levels were used for each tape, however some tapes have higher MOL than others. I have not corrected for this.

At the start of each of the four tape files you will hear approx 5 secs of Input noise, followed by 5 secs of virgin tape noise, followed by 5 secs of recorded tape noise. At the end of each file you'll hear the same but in reverse order.

Maybe my ears are growing old but I think all four tapes sound remarkably similar, I was expecting to hear larger tonal variations between them. I'm interested in what you think?

https://we.tl/t-BMx9EodoBz

WARNING: I SUGGEST PLAYING THE TRACKS IN SOFTWARE THAT DISPLAYS THE WAVEFORMS SO THAT YOU HAVE ADVANCE WARNING OF THE 0dBVU PERIODIC PINK NOISE AS IT'S BLOODY LOUD!

Have fun!

 

[ToTo Man]

I also did a comparison of 15 i.p.s. vs 7.5 i.p.s.

This download contains two files and is 205MB in size:
https://we.tl/t-pR6B5aQ8ND

I didn't know if there was any point on doing this test on all 4 tapes so I only did it using "Tape B" which was the last tape I had on the machine.

The 15 i.p.s. recording has slightly superior high frequency extension to my ears but the subjective noise floor of the recorded tape is actually lower on the 7.5 i.p.s. recording because there is more LF noise and less HF noise. Is this what you would expect?

Left spectrogram = 7.5 i.p.s. / Right spectrogram = 15 i.p.s.

 

[Werner]

which tape the A3440 is most accurately biassed for so that I can avoid calibrating it

One would expect on such a deck to set bias for optimal distortion and headroom, and then correct the frequency response with adapting the record equaliser.

15 ips: All four tapes lose some treble compared to the original. The best of the lot is tape B. There are some broadband channel imbalances, but more so in the treble. The left channel needs some attention there. But overall not too bad. With a bit of care this deck should be capable of fine results. Assuming, of course, that distortion is low and even in both channels.

Tape D looks worn.

I did not investigate the noise patterns. Maybe later.

however some tapes have higher MOL than others

You mean higher sensitivity. MOL is the tape output level where a specific distortion is reached.

 

[ToTo Man]

One would expect on such a deck to set bias for optimal distortion and headroom, and then correct the frequency response with adapting the record equaliser.

15 ips: All four tapes lose some treble compared to the original. The best of the lot is tape B. There are some broadband channel imbalances, but more so in the treble. The left channel needs some attention there. But overall not too bad. With a bit of care this deck should be capable of fine results. Assuming, of course, that distortion is low and even in both channels.

Tape D looks worn.

I did not investigate the noise patterns. Maybe later.



You mean higher sensitivity. MOL is the tape output level where a specific distortion is reached.

The recordings I uploaded were Ch2 and Ch4, but I also recorded Ch1 and Ch3. I can upload Ch1 and Ch3 from Tape B to see if there is any difference in performance?

The tapes are as follows:
Tape A = RTM LPR90,
Tape B = Maxell UD 35-90 (my A3440 instruction manual says the deck is factory-adjusted for Maxell UD, so you were correct in identifying the best one ),
Tape C = Maxell XL1 35-90B,
Tape D = Baked Ampex 456 (this tape performed excellently the first recording pass I made, but when I recorded over it again I noticed drop-outs, so I guess the advice that baking a tape is only good for playback and not for recording is true!)

PS - Yes I meant sensitivity, not MOL!

Would we have heard greater differences between the tapes if I had set the recording levels higher? In hindsight I think I might have set them too low as according to this thread the 3440's VU meters need to be slammed into the red to maximise S/N ratio?

I'm going to record the same test tracks on my B77. It'll be interesting to compare to a fully restored deck.