old, crappy tape deck advice...

[rdenney]

Have you used this software wow and flutter meter?

No, but I've heard of it, and I'm familiar with A.N.T. I need to purchase one of his tapes for setting the speed on my Nakamichi cassette deck, but just haven't done it yet. But the test tape one needs for accurate wow and flutter testing probably requires more precision than the tape I'm using for frequency and level setting. I could probably record a 3-KHz signal, but the same wow and flutter that would be present during recording would come back during playback, with unpredictable results.

So, I've done more testing with the pinch roller reversed where it works better. With a 10 KHz signal at a nominal -8 dBm (0VU), and with my ADC set to level that at -6 dB digital level, and when recording the Tape Monitor output (i.e., what's actually coming off the playback head) to Audacity, the loudest noise I'm seeing is a power hum at 60 Hz of -47dB, with its harmonics below that. The second harmonic of the signal is at -72 dB. There are lots of noise products in the -60's and lower. Specified performance is 45 dB S/N, but with the exception of power hum it's better than that. I could notch 60 Hz on playback and remove that one altogether. Hiss is definitely audible, of course, and those noise products do add up to hiss.

When recording the Source Monitor to Audacity (which is just what is coming in from the signal generator through the inputs to the deck, but includes a partial trip through the electronics to get to the outputs of the deck), all the power hum and its harmonics are there, but a few dB lower. But the hiss, of course, is not.

With 1 KHz at 0 VU on the Teac and the Source Monitor recorded at 0 dBFS, power hum is the loudest noise at -52 dB. The power hum's third harmonic is nearly as loud as that. The Tape Monitor output enlarges the power hum to -46 dB, with the signal still at 0 dBFS. Harmonic distortion of the input signal is at -47 dB (second harmonic) with others much lower. Spec'd S/N is -48 dB, which is is pretty close to. The specifications for this deck claim 1.5% THD when recorded at specified input level, which this is. With the Maxell tape, this deck beats that substantially. There is some close-in spectrum right around the signal frequency which is probably product of wow and flutter. Those seem to be no higher than the -30's, but they are so close in that I doubt they are audible. My signal generator is only spec'd to 1% at full voltage output (which this is not by any means). I don't have a distortion analyzer beyond what I read from these spectrograms.

Recorded frequency response is 30 Hz to 17 KHz, +/- 1 dB, recorded at 0 VU (nominally -8 dBm at the inputs). Recorded at -8 VU (-18 dBm at the inputs), the frequency response is 27 Hz to 24 KHz, -1 to +2 dB. This is on Maxell UD tape at 7-1/2 ips in high bias.

I'm calling this one ready for my project, and I'm going to put the covers back on it. But I will still try to find a (really, this time) new Teac roller or will have one rebuilt when Terry Witt is back in service.

Rick "old but no longer as crappy" Denney

 

[MakeMineVinyl]

I've found REW very good at analyzing FR, distortion etc. I actually use it a lot in developing electronic filters.

 

[antennaguru]

The problem with "new" pinch rollers in TEAC packaging is that they are by necessity "NOS". Over many years the rubber can develop a bump from an internal bubble, almost like a pimple, that you may or may not be able to see/feel. I had this happen with a Sony ES Cassette Deck, when I was replacing all of the rubber and the "new" pinch roller was causing a problem that wasn't there with the old one. When I examined the small pinch roller more closely, under magnification, it actually looked like a tiny little pimple. I binned that one and tried another one and the problem was immediately resolved.

There is another brand of replacement polyurethane (yellowy clearish material) pinch rollers that are actually brand new recent manufacture, ATHAN, but they are a bit expensive and they need to be cleaned with a special mild cleaner. I found those ATHAN pinch rollers to be pretty reliable and have them on two decks. I don't know if they will last as long as regular rubber pinch rollers though. They have an extra set of ball bearings and are always a tight fit on the deck's shaft.

 

[rdenney]

The problem with "new" pinch rollers in TEAC packaging is that they are by necessity "NOS". Over many years the rubber can develop a bump from an internal bubble, almost like a pimple, that you may or may not be able to see/feel. I had this happen with a Sony ES Cassette Deck, when I was replacing all of the rubber and the "new" pinch roller was causing a problem that wasn't there with the old one. When I examined the small pinch roller more closely, under magnification, it actually looked like a tiny little pimple. I binned that one and tried another one and the problem was immediately resolved.

There is another brand of replacement polyurethane (yellowy clearish material) pinch rollers that are actually brand new recent manufacture, ATHAN, but they are a bit expensive and they need to be cleaned with a special mild cleaner. I found those ATHAN pinch rollers to be pretty reliable and have them on two decks. I don't know if they will last as long as regular rubber pinch rollers though. They have an extra set of ball bearings and are always a tight fit on the deck's shaft.

I don't think that was the case here. The rubber on this roller was fresh and soft, compared to the concrete of the old pinch roller.

Teac, by the way, still makes these pinch rollers. The same roller as used on this deck was used on most of their later decks as well, including all of the X series. But buying directly from Teac is like buying directly from Ford--if you don't have a connection, you are forced to go through a dealer or an auto parts store. And stuff made by third parties is sometimes portrayed as being factory parts when they aren't. This is almost assuredly an ersatz part misrepresented as a genuine factory part.

I have not heard good things about those clear polyurethane rollers, but some love 'em. The pinch roller is easy to remove, clean, and relubricate on a routine basis, so the bronze bushing on which they are mounted seems like a good design to me. Just from an engineering perspective, ball bearings impose bearing precession that results in a rumble (tapered roller bearings can avoid this). Reducing friction (the point of ball bearings) isn't that important here, but the bushing is already pretty good at that. When I spin my roller, it spins for a while with no noise that I can detect. Roller bearings are never that silent.

I've resolved the issue with my deck for now (though I have not measured wow and flutter), and will have one of these rebuilt by Terry Witt when he's healthy again. In the meantime, I'd love to find a connection into Teac parts, and probably what I need to do is call them on the phone. I'll try that today.

Rick "fun with rollers" Denney

 

[rdenney]

I've found REW very good at analyzing FR, distortion etc. I actually use it a lot in developing electronic filters.

Well, duh. I'll have to mess with that tonight.

Rick "too obvious" Denney

 

[rdenney]

Took a few minutes this afternoon to run some REW measurements. Based on that, my frequency response isn't quite as good as what I measured using a microvoltmeter. But it's still not too bad and it's well within spec.

The output is the measurement sweep from REW, through a Presonus Studio 24C USB audio interface. Data rate was 24/96. The main outs from that device were plugged into the Line In jacks on the Teac. I swept from 20 Hz to 24 KHz, but even though there was output above 22 KHz, I was hearing difference tones in the audible range that suggested the signal might be more distortion and noise than signal.

The Line Out jacks, being fed by the Tape Monitor (meaning: off the playback heads, which should surely have confused REW's timing), were wired to the Presonus line inputs, and used for the measurement input. The test sweep was nominally -12 dB FS, which I take to be -12 dB with respect to digital clipping. The level-setting signal in REW is pink noise, which sounded identical to me input and output. I set the measurement level to be as close as I could get it to the test level, but clearly I didn't get it exact.

I think I can report that the frequency response is 25 Hz to about 22 KHz, -1/+3 dB. There is a channel imbalance of half a dB or so, but that's likely just me not matching levels properly. It would be -1/+2 over the range of my hearing.

The roller coaster in the bottom four octaves is interesting, and I suspect that's just the effects of record and playback EQ that is made up of resistors and capacitors.

In terms of distortion, I tested each channel separately. Here's the left channel:

Here's the right channel:

Specified THD is 1.5%, and this is well below that. Note that the noise floor is pretty high at a little better than -40 dB in the right channel. That's not as good as it should be, but it's within spec, pretty much.

The Maxell tape records a bit hot, to be sure. But it's predictable and a bit of EQ would fix that. But it's certainly not suffering from limited response.

I'm thinking this was about as good as consumers could get in the late 70's, and it would take a very expensive cassette deck to approach this performance. I suspect I could make it better by replacing all the caps, etc.

Subjectively, recorded music sounds great in headphones, but I want to test that again using a better headphone amp than is provided on the deck. The headphone output of the Presonus is designed for listening to tracks while recording other tracks, so it's not useful as a general-purpose head amp. I do have a JDS Atom amp that is new to me, however, so I'm going to do that subjective comparison again. The headphone amp on the Teac is designed for 8-ohm headphones.

It's time to clean up the mess on the floor of my listening room. My wife will be ecstatic.

Rick "fun project" Denney

 

[MakeMineVinyl]

The lifted high end suggests that you need to overbias a bit until that flattens out (regardless of what the manual says). There is no guarantee that Teac wasn't using a lower biasing tape when they wrote the manual and did not take into consideration newer tape formulations. In general, older tape machine manuals, even for professional units, need to be taken with a huge grain of salt in light of how tape formulations and other considerations changed after the machines came out.

At any rate, around 0.6% THD is typical at operating level at 1kHz. The variations in response in the low end are likely head bumps.

As for tape, I don't know if you're familiar with Capture Tape, but it seems like a good prosumer offering.

This is the REW measurement file for my Ampex 354. Its at 15ips and CCIR EQ using SM-468 tape. 7.5 ips should be similar but roll off a bit sooner.

 

[rdenney]

The lifted high end suggests that you need to overbias a bit until that flattens out (regardless of what the manual says). There is no guarantee that Teac wasn't using a lower biasing tape when they wrote the manual and did not take into consideration newer tape formulations. In general, older tape machine manuals, even for professional units, need to be taken with a huge grain of salt in light of how tape formulations and other considerations changed after the machines came out.

At any rate, around 0.6% THD is typical at operating level at 1kHz. The variations in response in the low end are likely head bumps.

As for tape, I don't know if you're familiar with Capture Tape, but it seems like a good prosumer offering.

I have a couple of reels of Capture 914, a standard-bias formulation. But it seems to have a high noise, and adjusting bias for it proved quite challenging. If I do any recording, it will be on high-bias tape. I can get old Maxell UD reels for considerably less than the Capture tape. They have a high-bias version, of course, but it's more expensive.

And there are a couple of other manufacturers as well, at a higher price.

Maxell has never been associated with sticky shed syndrome or other tape aging faults, and it seems to work very well for me.

But yes to the notion of biasing it a bit further. That's an easy last step. Plus, I find I'm having fun with this project and I'm not quite ready to clean up the mess yet The original 4300 manual suggested using 400 Hz and overbiasing by half a dB, even in the high-bias setting. The 4300SX manual suggests biasing at 7 KHz and overbiasing it 2.5 dB. I think I'll bias it at 10 KHz and overbias it until it reads the same as 1 KHz, just to see what happens.

Rick "no, honey, the project isn't quite done yet..." Denney

 

[MakeMineVinyl]

I think I'll bias it at 10 KHz and overbias it until it reads the same as 1 KHz, just to see what happens.
Denney

Using the REW spectrum analyzer (RTA) while recording pink noise (monitoring the tape playback) is the quickest way to adjust either EQ or bias; the results can be seen in real time. Then do regular sweeps to confirm. Done.

 

[rdenney]

Note to file: Teac can no longer supply pinch rollers for reel-to-reel tape decks of any vintage. Teac was owned by Onkyo, which last year was sold to Klipsch, which created a new distributing company (11 Trading Company), which referred me to a shop in California, which told me nothing was available and just search the internet. Sigh.

So, I'll have to wait for Terry Witt to get better.

The reports that I have seen on those polyurethane rollers is that they turn to goo after some relatively short period of time less than ten years, and often much less. They become sticky with apparent suddenness and leave residue on tapes. But I may get one just to use until I can have these others rebuilt.

But what I now understand is this: There is no factory source for pinch rollers for open-reel tape decks, and any claims of such are ersatz. Thus, the one I bought, which looks and feels good but is clearly not.

Rick "the mighty have fallen" Denney

 

[antennaguru]

Yes, that is what I suggested.

I have six reel to reel decks, with three of them quarter track 3.75/7.5ips TEAC decks, so similar to yours mechanically though none of mine are auto-reverse. One has a NOS TEAC black rubber pinch roller. One has a pinch roller that I rebuilt myself by installing a new black rubber tire that came with its new capstan belt and counter belt set. One has had an ATHAN polyurethane pinch roller for over three years. The one with the ATHAN roller is the only one used for making recordings, and all three are used for playback. I am using the ATHAN pinch roller on the most critical of those three decks. All three decks perform as well as or better than the quarter track tape format allows, and they all play vintage pre-recorded tapes nicely. I have a 7.5ips NAB EQ MRL reference tape and have aligned them all with this. The collection of tapes I have recorded myself from high resolution sources are on ATR MDS36 and Maxell UD 35-90 tapes, and are all recorded with only one pair of stereo tracks which eliminates the cross-talk limitation of the format's standard of two pairs of interlaced stereo tracks.

I also have three 2 track 7.5/15ips decks, two TASCAMs and one Crown. One TASCAM has a NOS TASCAM black rubber pinch roller. The Crown has a professionally rebuilt black rubber pinch roller. The other TASCAM has had an ATHAN polyurethane pinch roller for over two years.

I use the ATHAN pinch roller mild cleaner for cleaning all of the pinch rollers, along with non-lint Kimtech Kim-Wipes.

The only cases I have ever heard of where people ruined any pinch rollers, regardless of type, were from using harsh chemicals to clean them. Some of those went gooey I believe in reaction to use of the harsh chemicals, but possibly it was environmental or age.

I really like my two ATHAN polyurethane pinch rollers, and think the rumors you heard about them are either propaganda or from people that used harsh cleaning chemicals. The two of my decks that the ATHAN polyurethane pinch rollers are on both perform extremely well.

I see no issues with the polyurethane material and often make my own round profile tape/turntable belts from round polyurethane belt stock.

 

[MakeMineVinyl]

The reports that I have seen on those polyurethane rollers is that they turn to goo after some relatively short period of time less than ten years, and often much less. They become sticky with apparent suddenness and leave residue on tapes. But I may get one just to use until I can have these others rebuilt.

From what I have heard, that problem was solved. That could still be wrong however.

I have a roller which was done by Terry, and under strong light it looks a bit urethane-ish, and possibly is urethane with black coloring. It certainly doesn't have the solid, deep black color of real rubber. Doesn't mean its urethane, but who knows. The one I have has worked for about 8 years now with no problems though.

He refuses to tell what the exact composition of his rubber is FWIW.

 

[rdenney]

The reports were a couple of guys on Tapeheads, one of whom was using Athan’s own cleaning fluid.

I’ve identified a source in Italy for rebuilt rollers, and I've ordered one. I’ll report on it when it arrives. His “Super Roller” is on a ball bearing but he uses rubber for the tire, and he was recommended on Tapeheads as a backup to Terry Witt.

I got my reference tape from a guy in Germany who goes by Ural91 on eBay. He’s highly recommended, and the tape is on BASF stock and is an exact duplicate of the Teac YTT-1003 test tape, including the 185 nWb/m flux and the NAB EQ.

Rick “who’ll tackle the 4010 when done with the 4300” Denney

 

[antennaguru]

I had read that tale on Tapeheads since I'm a member there, and that's part of the reason I came to the very conclusion I stated. I currently have two ATHAN pinch rollers in service for several years on two of my own decks. I would not hesitate to put another one on one of my decks.

Meanwhile, I have round profile polyurethane belts in service on equipment for well over a decade, with no stretch and no degradation. Occasionally, I make otherwise unobtainable belts for rare equipment, and have never had one fail in any manner. I frankly think it's a superior material to rubber - unless you are going to subject it to harsh chemicals and then it probably breaks down faster than rubber, which also breaks down or we wouldn't be having this discussion.

Note that rubber fuel lines made before they started adding alcohol to fuel are ticking time bombs and need to be replaced with alcohol resistant rubber fuel lines. This is important on vintage cars and boats.

You're braver than me to trust an ebayer for a reference tape. I bought my MRL reference tape from ATR in Pennsylvania, who I regularly do business with as they make and sell the best tape currently available.

 

[rdenney]

The problem with MRL is that they don't record to the same flux that Teac used in their calibration tapes, which means a lot of math or calibration to a different standard. But the guy I bought from has a good reputation with the long-standing techs on Tapeheads.

In any case...here's the final outcome.

I decided to go back through the whole adjustment again, this time using what I'd learned up until now to really get it right. That has paid off. I used the signal generator and the microvoltmeter to adjust the bias by over-biasing it to bring 10 KHz down to the same level as 1 KHz. This was easy and quick--just clicking the range buttons on the generator. I tried using the pink noise into REW, but I could not find where to make the RTA display show equal octaves rather than equal decades, and therefore the pink noise had a slope which precluded me really stretching out to the scale to read a range of just a couple of dB. I used the white noise generator (which is flat on a log frequency scale) to provide a reference for adjustments, but still use frequency sweeps to really check it. I needed just two or three additional nudges to center the high-frequency roller coaster on the same horizontal trend as the lower frequencies. That really paid off. Here's the frequency response for the two channels:

The bump at the low end is caused by a weird calibration on the soundcard, I think, but I'm still trying to puzzle through that one. In any case, the frequency response of both channels as recorded fits in a +/- 1 dB window from 40 Hz to over 20 KHz, which is MUCH better than specified. It did require slightly more overbiasing than the 4300SX manual suggested. The difference in channels is probably not quite adjusting the input levels on the sound card precisely. The outputs of the deck are calibrated to a somewhat smaller fraction of a dB.

Getting all that tweaked properly meant that the channel distortion went down. Here is the left channel response and distortion:

Distortion peaks at 75 Hz and again at 6-8k at just below 1%. At 1 KHz, it's 0.2%, far less than the 1.5% specification.

Here's the right channel, which performed similarly:

Something I don't understand is what's happening with the spectrogram. I'm assuming that what appears to be ringing in the higher frequencies is really hiss and noise on the tape, as shown below. Any ideas how to interpret this?

Note the power hum, which I can interpret.

Wow and flutter is probably wrong--I recorded the tape and then played it into the WFGUI software many times to make sure any recorded wow wasn't somehow aligned or aliased against playback wow. I'm seeing wow and flutter in the range of better than 0.04%, which is about a fifth of the specified performance. I'm reasonably sure that wow and flutter isn't a problem with this deck.

Finally, I recorded and listened to music--excerpts from Liszt's tone poems for two pianos, Rick Wakeman's Red Planet, Wynton Marsalis's Think of One, Mike Oldfield's Tubular Bells, and the St. Louis Symphony playing Bruckner's 7th. I piped the outputs into my new JDS Labs Atom headphone amp and listened on my AKG K371 headphones, and I could not easily distinguish between source and recording. Getting that high treble toned down to where it should be really subdued the tape hiss. Of course, I could hear that it was tape between the "tracks", but the hiss was effectively masked by the music, even with headphones. Making one more pass through the deck made the difference between pretty good and excellent, and once again I'm reminded that even a signal/noise ratio of not quite 50 dB can still play music really well.

Now, it really is time to put the covers back on this one.

Rick "time for bed" Denney

 

[MakeMineVinyl]

Sounds good! One thing I would mention though is that the fluxivity of the test tape that TEAC supplies is just a number they pulled out the air (I was going to say 'their ass') and it became 'their standard' - there's nothing magic about it. The 'real' and very old original 'standard operating level' was that level which produces 1% harmonic distortion. That's obsolete now obviously.

Different studios, manufacturers and such use levels which can depart from 'the standard'. These levels are almost always based on such and such db above the original Ampex standard of 185nWb/m. Sometimes 200 is used. Personally, I'd recommend you set your operating point as +3db above the standard level; this is a very common setting (when I was a recording engineer at Sound City they used +2 db for some strange reason). Maxell tape can surely work well at +3 as its a 'modern' formulation (vs Scotch 111 which is an old formulation). To set this level, set the VU meters so that the test tape standard level reproduces as -3db and turn up the record level so that the playback is '0' on the VU meters (then set the record level calibrate).

 

[rdenney]

Sounds good! One thing I would mention though is that the fluxivity of the test tape that TEAC supplies is just a number they pulled out the air (I was going to say 'their ass') and it became 'their standard' - there's nothing magic about it. The 'real' and very old original 'standard operating level' was that level which produces 1% harmonic distortion. That's obsolete now obviously.

Different studios, manufacturers and such use levels which can depart from 'the standard'. These levels are almost always based on such and such db above the original Ampex standard of 185nWb/m. Sometimes 200 is used. Personally, I'd recommend you set your operating point as +3db above the standard level; this is a very common setting (when I was a recording engineer at Sound City they used +2 db for some strange reason). Maxell tape can surely work well at +3 as its a 'modern' formulation (vs Scotch 111 which is an old formulation). To set this level, set the VU meters so that the test tape standard level reproduces as -3db and turn up the record level so that the playback is '0' on the VU meters (then set the record level calibrate).

Yes, I've read recommendations of some folks to use calibration tapes normally to a flux of 250. But I would rather the deck be in factory-spec condition, but with the bias tweaked for my preferred tape (Maxell UD). If I ultimately decide to sell it after my project is done, that's what I'd like to claim. I've learned when restoring all manner of old things to start with the factory settings and depart from those only with due care.

This particular deck provides two meter ranges, the normal range that goes from -20 to +3, and a high range that goes from -17 to +6. Rather than set the meters to read higher, I simply record such that peaks bounce the needles above +3 only rarely, using the high range position, and I'm fine if sustained fortissimos sit on +3. My Nakamichi BX-300 cassette deck is calibrated similarly from the factory--and it's customary to record 3 dB higher on those decks, as I understand it.

Rick "note to self: do not break the rules before learning them" Denney

 

[rdenney]

Final home:

Now I just need to figure out how to wire it into a system with two tape loops, one of which is for the Nakamichi cassette deck and the other for the ADC and DAC. I think I’ll Y out of the Tape Out bus to the input to this deck and the Nakamichi, and then wire the output of this deck to the the last unused audio input.

Rick “too heavy and big to go anywhere but it’s own table” Denney

 

[antennaguru]

There are switch-boxes for this, and naturally some will be vintage. The Sony SB-300 "TapeCorder" switch-box is a nice one and will take three tape decks.

 

[rdenney]

There are switch-boxes for this, and naturally some will be vintage. The Sony SB-300 "TapeCorder" switch-box is a nice one and will take three tape decks.

Well, that's handy-dandy. Just bought one. I figured something like that had to exist, and it wouldn't be that hard to make one. But this looks well enough finished and clean for hanging tape decks off of one tape loop while keeping the ADC/DAC out of the other tape loop and away from all this switching. Thanks for the pointer.

Rick "who remembers these now, but had forgotten they existed" Denney