Tandberg 10XD measurements (reel to reel tape recorder)

[peniku8]

This is my 'review' of the Tandberg 10XD. It's my dad's machine, who purchased it in the 70's when it came out. It's a Stereo half track machine, which means you can't turn the tape around to record/play different audio material. You'll have to rewind it instead. You can get a used one for 500-1000 USD it seems.

The first time I've interacted with this machine was when I was 6. I liked winding and rewinding the tape. Long story short, all the tape was off the reels after some time and ended up in the trash later and we had no tape left. When I got into recording some 15 years ago, I used very old tape, which disintegrated after a while, so I bought new tape.
I also recorded music onto it directly from the PC, which I always found to sound just like the original recording to me.

I did not bother to clean the heads, since it hasn't seen much use in the recent years. I tested it with the RMG LPR 35 tape I bought 10 years ago. The machine is 45 years old and has not been serviced (except some cleaning), so take these measurements with a grain of salt.

Testing procedure: in contrast to Amir's tape deck review, I did not test the machine with a pre-recorded test tape. Instead, I fed it from the PC, recorded that onto the tape and monitored the tape while recording, which I fed back to the PC.
Since I wasn't sure at which levels to test (it doesn't go to 2V), I decided to test its performance through direct monitoring first. This means the PC sends the signal through the input and output stage, without of the tape in between.

Direct passthrough (no tape), 66,7dB THD+N:

Performance drops to ~65dB with 1V out (maximum output levels are 1,2V according to the manual).

I still tested 2V in 2V out, which it didn't seem to mind too much:

For the tape tests I went with 1V unity gain. The tape itself had the levels set to be at -6dB on the VU meters during these tests (produced the best figures). All tests were run at 15ips and I recorded silence over the tape once before the tests, to get rid of influences from prior recordings.

1V unity gain, tape -6dB, 46,4dB THD+N:

46dB THD+N sure is disappointing from today's point of view, but so are the limitations of magnetic tape. Distortion is dominated by the 2nd harmonic, followed by the 3rd. Higher orders are 20db down from that, which is good. The higher the order of the harmonic distortion, the less we associate it with the actual signal and start percieving it as a problem/defect, so it's good to see mostly 2nd here. Noise is high, but we have Dolby noise reduction for that, so let's have a look at that next.


1V unity gain, Dolby Normal, 46,1dB THD+N:

The numbers didn't improve, but we see an almost 10dB lower noise floor in the top two octaves. This reflects what is written in the manual.

Here is the noise floor itself on a regular dB scale with 1V@0dB, which gives us a dynamic range of 67,7dB:

We have 3 different Dolby implementations at hand, which I'll get to now.

Frequency Response:

It's a bit fuzzy (smoothing was 1/48th) and hard to read with the overlays, but there are no obvious defects. Here is the response of only the Dolby Norm setting:

If you disregard the fuzziness, it's pretty flat. -1dB window is ~30-15k. -3dB is ~25-20k.

THD vs Frequency 1V:

THD vs Frequency Dolby Norm 1V:

THD vs Frequency Dolby Filter 1V:

THD vs Frequency Dolby FM 1V:

So in conclusion, Dolby NR increases distortion but also increases SNR, which makes sense when you realize that it's basically just an EQ boosting high frequencies before recording. That also makes sense, since music is not white noise (same levels at every frequency), it's more like pink noise (lower intensity the higher the frequency goes), so we're wasting headroom in the high frequencies in theory, if we don't use something like the Dolby implementation here. A lot of thought has gone into those devices in the past, about how to overcome its technical limitations and improve audio quality, I like it.

Well, time to enter chaos town, with the THD vs level graphs. The no-Dolby one looks fine:


THD vs level (unity gain):

The harder you push it, the more 3rd harmonic you'll get in relation to the 2nd harmonic. That should give drums some bite. 2V is 0dB on the VU meter in this case.


THD vs level, Dolby Norm (unity gain):

THD vs level, Dolby Filter (unity gain):

THD vs level, Dolby FM (unity gain):

The random noise spikes are maybe caused by the old tape, you can probably dismiss them as measurement artefacts.


Conclusions: Noise levels seem to be dominated by mains hum, which I didn't chase after to get rid of. If it wasn't for this, we might've seen 50dB THD+N. That's still not very impressive, considering that top of the line devices nowadays approach or even surpass the 120dB figure (keep in mind this is a recording+playback solution, so if you want to compare it to devices today, you'll have to compare to a loopback from your DAC back into an ADC and can't directly compare to pure DAC figures on an audio analyzer).
The inputs and outputs of the machine are unbalanced RCA, which means I already hate it (jk). Mic inputs are balanced but only accept a 20mV input or something. I tested it, but wasn't really worth the effort. There is also a "radio jack" input, which you might know as MIDI connector. This was being used to transmit bi-directional unbalanced stereo audio a long time ago. I have a Yamaha CR-600 (Stereo reciever/amp) from the same era, which also has these connectors (I will also test it and post the results here).
How does it sound? Well I haven't really tested it thoroughly lately, but it sounds good. An unexperienced listener might not be able to pass a DBT, depending on the type of content. Orchestral performances with super high dynamic range will have a constantly audible noise floor, but your average pop song will do just fine. I got a mastering request once, where the band insisted on having their album put through a tape machine, so it's convenient to have it.
It's fun to use and makes very satisfying mechanical sounds. And looks cool in my studio, which is its main purpose at the moment
Oh and one last thing: I love VU meters!

 

[MakeMineVinyl]

This machine appears to be not working very well. The distortion spectrum should be almost entirely odd order with ideally no even order. High amounts of even order distortion indicates either magnetized heads, a leaky coupling capacitor going to the record head or other causes. The distortion should look like this, which is a vacuum tube tape machine from the early 60s:

You don't mention what speed you are using, but generally frequency response tests at 7 1/2 ips are taken at -10db on the VU meter. At 15 ips you can record at the full 0db on the meter without running into saturation. Your machine should have relatively flat response at 15 ips from around 40Hz to at least 15kHz at the least. 7 1/2 ips response shouldn't be much worse.

In any event, switching from source input to tape playback in real time shouldn't change the signal all that much, especially at 15 ips. Your machine being a half track, I wouldn't bother with Dolby NR since the tape noise shouldn't be all that bad if the machine is working correctly, and Dolby introduces its own problems.

Or you can just leave it as-is if you're not going to be using it for any critical recordings. I know Tandberg tape machines were excellent, so what you are getting is suspicious.

 

[peniku8]

This machine appears to be not working very well. The distortion spectrum should be almost entirely odd order with ideally no even order. High amounts of even order distortion indicates either magnetized heads, a leaky coupling capacitor going to the record head or other causes. The distortion should look like this, which is a vacuum tube tape machine from the early 60s:

You don't mention what speed you are using, but generally frequency response tests at 7 1/2 ips are taken at -10db on the VU meter. At 15 ips you can record at the full 0db on the meter without running into saturation. Your machine should have relatively flat response at 15 ips from around 40Hz to at least 15kHz at the least. 7 1/2 ips response shouldn't be much worse.

Or you can just leave it as-is if you're not going to be using it for any critical recordings. I know Tandberg tape machines were excellent, so what you are getting is suspicious.

My distortion spectrum looks similar to Amir's results on his machine. Judging by the tape bypass measurements, most of the THD seems to come from the electronics, while the high noise floor is a result of the medium (the tape). The machine is 45 years old and has never been serviced (except me cleaning the heads with alcohol and cleaning the faders which had contact issues), so a leaky cap sounds plausible.
I know my way around electronics and have serviced a few devices in the past, do you think I could indentify a bad cap visually? In that case I could open it up and see if anything needs to be replaced.

Good point about the speed, it was running at 15ips at all times and I forgot to mention that in the post. That's fixed now! As you can see in my THD vs level sweeps, the machine performs best at around -7dB on the VU meters and starts saturating from there. I confirmed this by increasing the input gain and decreasing the output gain (and playing around with them in general to find a sweet spot). 0dB on the meters always showed worse results, no matter the input and output voltages.

The measurement you show is way worse than the one I got, so I'm not sure exactly what to make of that. Is it just to illustrate the relation of 2nd and 3rd harmonic?
2nd is approaching -60dB in your pic, while it's at -66 in my measurement. 3rd harmonic is 30dB lower in relative level on the Tandberg, maybe you're confused by the scale of my graph, which isn't ideal indeed (I noticed the volt axis too late, I usually export to 0dB SPL reference).
For better comparison: THD in your pic is ~0.7%, while it is 0.05% on the Tandberg.

Frequency response graphs are included in the post, which confirm (and exceed) your guess

 

[watchnerd]

Nice!

Is it NAB or CCIR EQ?

Nice ReVox reels, too...they'd match my ReVox decks.

 

[peniku8]

Nice!

Is it NAB or CCIR EQ?

Nice ReVox reels, too...they'd match my ReVox decks.

I read through the manual a few days ago and can't remember reading anything about EQ at all, they just explained how the Dolby system records higher frequency at hotter levels to get better SNR, but not what type of EQ is used. I could measure it if you're interested.
I have some aluminium reels somewhere (I think revox too), but don't use them, since they're slightly bent and scratch the machine

 

[watchnerd]

I read through the manual a few days ago and can't remember reading anything about EQ at all, they just explained how the Dolby system records higher frequency at hotter levels to get better SNR, but not what type of EQ is used. I could measure it if you're interested.
I have some aluminium reels somewhere (I think revox too), but don't use them, since they're slightly bent and scratch the machine

Looks like it's NAB:

Tandberg 10XD Stereo Tape Deck Manual | HiFi Engine

www.hifiengine.com

 

[MakeMineVinyl]

My distortion spectrum looks similar to Amir's results on his machine. Judging by the tape bypass measurements, most of the THD seems to come from the electronics, while the high noise floor is a result of the medium (the tape). The machine is 45 years old and has never been serviced (except me cleaning the heads with alcohol and cleaning the faders which had contact issues), so a leaky cap sounds plausible.
I know my way around electronics and have serviced a few devices in the past, do you think I could indentify a bad cap visually? In that case I could open it up and see if anything needs to be replaced.

Good point about the speed, it was running at 15ips at all times and I forgot to mention that in the post. That's fixed now! As you can see in my THD vs level sweeps, the machine performs best at around -7dB on the VU meters and starts saturating from there. I confirmed this by increasing the input gain and decreasing the output gain (and playing around with them in general to find a sweet spot). 0dB on the meters always showed worse results, no matter the input and output voltages.

The measurement you show is way worse than the one I got, so I'm not sure exactly what to make of that. Is it just to illustrate the relation of 2nd and 3rd harmonic?
2nd is approaching -60dB in your pic, while it's at -66 in my measurement. 3rd harmonic is 30dB lower in relative level on the Tandberg, maybe you're confused by the scale of my graph, which isn't ideal indeed (I noticed the volt axis too late, I usually export to 0dB SPL reference).
For better comparison: THD in your pic is ~0.7%, while it is 0.05% on the Tandberg.

Frequency response graphs are included in the post, which confirm (and exceed) your guess

Amir's Otari machine wasn't performing very well either, and there were a good number of posts as to the possible culprits.

The electronics-only distortion plot shows the 2nd harmonic around -70db which is too low in level to influence the off-the-tape distortion; the tape distortion will swamp the electronics distortion and the much stronger odd order harmonics from the recording process should predominate. You will note some 2nd harmonic distortion in my plot (Ampex 354) which cannot be trimmed out by balancing the bias oscillator. Still, this plot should be something to compare against. The overall THD should read about what I show, or <1% at 0VU at 15 ips. Mine is .6905% which is typical for a +3 tape on a vacuum tube machine.

You should probably buy a reel of good, current production tape. Some people here use vintage Maxell which is good. The graph I show is with Recording The Masters SM-468. I've got a couple reels of Capture recording tape which is a good consumer grade tape. All tape is pretty expensive now but about the same price as in the old days considering inflation.

As far as checking capacitors, you would need a good meter which is capable of measuring capacitance to weed out bad ones, and even then, you'd need to take most of them out of circuit to test. That might not be worth it to you. I've been working with tape machines as a specialty since the late 60s so I don't mind gutting orphan machines, but I understand that most people are not as insane as me.

Tape machines can be a monumental time-suck, but if you have any questions, feel free to ask.

 

[peniku8]

Amir's Otari machine wasn't performing very well either, and there were a good number of posts as to the possible culprits.

The electronics-only distortion plot shows the 2nd harmonic around -70db which is too low in level to influence the off-the-tape distortion; the tape distortion will swamp the electronics distortion and the much stronger odd order harmonics from the recording process should predominate. You will note some 2nd harmonic distortion in my plot (Ampex 354) which cannot be trimmed out by balancing the bias oscillator. Still, this plot should be something to compare against. The overall THD should read about what I show, or <1% at 0VU at 15 ips. Mine is .6905% which is typical for a +3 tape on a vacuum tube machine.

You should probably buy a reel of good, current production tape. Some people here use vintage Maxell which is good. The graph I show is with Recording The Masters SM-468. I've got a couple reels of Capture recording tape which is a good consumer grade tape. All tape is pretty expensive now but about the same price as in the old days considering inflation.

As far as checking capacitors, you would need a good meter which is capable of measuring capacitance to weed out bad ones, and even then, you'd need to take most of them out of circuit to test. That might not be worth it to you. I've been working with tape machines as a specialty since the late 60s so I don't mind gutting orphan machines, but I understand that most people are not as insane as me.

Tape machines can be a monumental time-suck, but if you have any questions, feel free to ask.

Very interesting, thank you for the insights!
I'm still a bit confused by your distortions. You say mine should look close to yours, at 0.7%, but mine look way better to me, so I don't understand the concerns. Let me Summarize below:

Tandberg Off-tape (only electronics):
THD 0,03%
2nd -73,4dB
3rd -74,1dB
(Everything is about 10dB worse in the 2V measurement)

Tandberg On tape 15ips, no Dolby:
THD 0,05%
2nd -65,0dB
3rd -76dB
(THD was about 0,15% at 0dB on the VU meter)

Your measurement:
THD 0,69%
2nd -57dB
3rd -43dB

You say distortion from the tape should swamp electronics distortion, but 3rd harmonic is actually lower in level than off-tape (note 3rd harmonic at -76dB on tape), on my machine, what ever wierd reason that might have.
Also, 3rd harmonic is 20dB lower in level on my machine compared to yours, both running about the same settings.
I'm really confused by the fact that you say my machine might not be working properly, but then compare it to a machine that measures way worse.

I just checked and the tape I have is the RMG LPR 35. According to the internet it's the same as the RTM LPR 35. Bought it in 2010. It was 30€ for the 1100m reel (plastic). Edit: I added this info to the original post as well

 

[MakeMineVinyl]

Very interesting, thank you for the insights!
I'm still a bit confused by your distortions. You say mine should look close to yours, at 0.7%, but mine look way better to me, so I don't understand the concerns. Let me Summarize below:

Tandberg Off-tape (only electronics):
THD 0,03%
2nd -73,4dB
3rd -74,1dB
(Everything is about 10dB worse in the 2V measurement)

Tandberg On tape 15ips, no Dolby:
THD 0,05%
2nd -65,0dB
3rd -76dB
(THD was about 0,15% at 0dB on the VU meter)

Your measurement:
THD 0,69%
2nd -57dB
3rd -43dB

You say distortion from the tape should swamp electronics distortion, but 3rd harmonic is actually lower in level than off-tape (note 3rd harmonic at -76dB on tape), on my machine, what ever wierd reason that might have.
Also, 3rd harmonic is 20dB lower in level on my machine compared to yours, both running about the same settings.
I'm really confused by the fact that you say my machine might not be working properly, but then compare it to a machine that measures way worse.

I just checked and the tape I have is the RMG LPR 35. According to the internet it's the same as the RTM LPR 35. Bought it in 2010. It was 30€ for the 1100m reel (plastic). Edit: I added this info to the original post as well

Your reference levels from which the distortion levels are shown don't appear to be referencing an actual level of fluxivity, i.e. 0VU equals 185nWb/m or something like that which is typical. That's why a playback calibration tape is useful to establish a known playback level, then all the other measurements in record are referenced to that. If you were to calibrate the playback level, and record to that level, I think you will find that the THD (not the harmonics) will be somewhat close to what I show. This is tape dependent mostly. Of course its entirely possible that your machine records with a bit less distortion than the one I show, but I'd be suspicious of 0.15%. An Ampex ATR-100 would be lucky to get that level of distortion, and that's probably the best recorder ever produced.

Again, this much trouble might not be worth it for you, and calibration tapes are about $120, so there's that.

Mostly however I included the picture to show that the spectrum of distortion should be almost exclusively odd order harmonic distortion, and after the 5th harmonic or so, the harmonics should drop off to essentially zero. Mine has a bit of the 7th harmonic and a stronger 2nd, but basically nothing else - the 2nd harmonic does not dominate. That you are getting strong 2nd harmonic is a red flag that either the heads are magnetized or that a bit of DC is getting to the record head because of a leaky coupling capacitor from the constant current record amp. There is also the possibility that the bias oscillator is not balanced, and if this is adjustable, it should be looked at. You basically adjust the balance pot for lowest 2nd harmonic distortion and noise.

I make a point of heads being potentially magnetized because magnetized heads will erase the high frequencies from you tapes as you play them if they are magnetized enough. This is not ideal!

A basic adjustment which should be made before doing any measurements is to set the bias. This is easy. Place the machine in record with the VU meters a bit below 0VU with a 1kHz signal. Monitor the tape playback on the meters (or an external meter if you wish) and adjust the bias level pot for maximum playback level. After this is done, you should do a record level calibration so that the machine records at the 0VU point established from your calibration tape. The playback high frequency EQ should have been set with the calibration tape. After the record level calibration is set, then adjust the record high frequency EQ for flattest high frequency response. All this should be done at 15ips.