Magnetic tape head voltage output

I am currently working on a cassette tape deck project trying to build a unit entirely from scratch. I have the microprocessor controlled motor and control systems completed and have partially completed tone control and balanced output stages done and am now starting on the tape head amplifier and frequency compensation stage.

Unfortunately I am as of yet unable to get any accurate measurements from the tape heads I have been using (not sure if they are still in working condition) and am still trying to hunt down some good replacements.
In the meantime I am keen to keep working on the project so was wondering if someone could give me a rough mV (I am presuming that is the range I should be looking at) values for what I should be expecting from the tape heads so I can start the circuit design?

Many thanks in advance for any help, also any general advice on tape head equalisation would be appreciated
Tom
 
Millivolts sounds right. I'm pretty sure there were plans for a tape preamp in Wireless World back in the '70s, possibly by Linsley-Hood. A university library should have bound back-issues, or access to microfilm. I'm pretty sure National Semi. included a tape playback circuit with the app note for the LM381 or the LM387. IIRC, the normal curve is NAB, but I think that's only for regular ferric (low noise), not metal and chrome.

Making a cassette player seems a little quixotic, though, when there's stacks of decent cassette decks in thrift stores for 1/10 to 1/100 of their original selling price. $15 got me a Harman-Kardon that was so recent, h-k had a pdf of the manual on their site.
 
Thanks again for the response, I checked out the LM381/387 and as you said they have some great typical application examples. The only problem I have now is that as it is an old IC I cant get any, also the searches I have done to find an equivilent all bring up obsolete IC too.

Would the LM833 be a suitable replacement for it? Or maybe the LM4562?
I can’t work out if the LM381 can be substituted with a standard op-amp or if it has totally different characteristics
 
Hello again all,
I am nearing completion of my project and am basing the analogue tape compensation section on a circuit i have found (hopefully attached) from the NE5532 application notes. I was initially planning on using the design from the LM381 but it is a very hard to find IC and has some extra internal capacitors which I believe mean it cannot be directly replaced with a standard op amp.

My first question is : on the circuit attached there is a input cap labelled as being 0.22 which I presume is farads which seems a freakishly large capacitance to use for an input cap, could anyone recommend a value that could be used instead?

Secondly there is a input resistor RSL, does anyone know its function? I am trying to work out a suitable value (or if indeed it is necessary?)

Lastly the feedback capacitor is stated to be 0.003 which is 3mF which is again a very large capacitor, is it possible there might be a u or p missing from the end? I only ask as I have now ripped open a fair few tape decks and have never seen a capacitor this large outside the power supply, perhaps they are just using a better compensation circuit.

Thanks in advance for any advice
Tom
 

Attachments

  • screenshot002.gif
    screenshot002.gif
    15.7 KB · Views: 850
I was planning on using OPA2134s as I have a few spare from a previous project, do you think the NE5532A would be more suitable?

As for the transport I’m using a old 90s Sony deck with modified spools (not driving the capstan but the spool instead), I won’t go into the drive details as I’m sure my approach will upset some analogue purists, safe to say it’s not a normal implementation.
Thanks again for the help it really has been very useful
 
Administrator
Joined 2004
Paid Member
Steerpike said:
Well those 2 ICs are different; each has its own plus points as far as figures go. I guess it all becomes a matter of personal preference.
The NE is quieter, but has poorer slew rate.

Are you saying you aren't using a capstan+pinch roller *at all*?

Presumably speed accuracy and stability during the course of playback is not an important consideration? Tape speed with spool drives is a function of the diameter of the tape pack on the take up reel, and unless you have some way to measure the velocity across the heads I am not sure how you intend to maintain a steady speed which is required for compatibility with existing standards, and just as importantly a tape head is velocity sensitive, change the velocity of the tape across the head and you significantly change its frequency response. I suspect no matter what you do controlling wow and flutter with a spool drive will be near impossible even if you can overcome the varying velocity profile as the tape plays.

It's not even really a question of us analog(ue) purists being outraged... :whazzat: :D

(As an aside I have designed and built playback only cassette players in the past, and have restored more than a few ReVox G36 tape decks.)
 
Having built many tape decks in the past with both playback and recording with Dynamic Noise Reduction etc., may I add my two cents.

Philips used to make an an IC, think it was TDA1522, specifically meant for tape amplification. I've used it and it sounded almost very good.

I have also used the JLH circuit which uses an FET in the input and an opamp for the NAB correction part. This also sounds good.

However, I found that some of the discrete circuits from Sony sounded the best. Many have a switch to toggle between 120uS and 70uS; so you can play both normal and metal/chrome tapes.

I have also seen some Pioneer circuits with dual JFETs in the input stage but have never tried them due to component unavailability.

All the best for your project.

By the way, if at all you are using the circuit you posted with the NE5532s, you wouldn't need the RIAA correction part of the circuit in the NFB part, but only the 1.1M, 16K resistors and the 3nF cap. You could also use another RC circuit for 70uS time constant that you can switch to for metal/chrome tapes.
 
switching out tape heads

Are 8 track tape heads and cassette tape heads the same voltage, could I interchangeable switch them out. I also plan on just spooling regular cassette tape into a 8track cartridge?


Why? Cause I bought a relic from the 70's. Its a guyatone echo tape, 8 track tapes arent made any more, heads neither. It works well actually but introduces just a little more noise than I'm comfortable with. So I'm thinking just remove the head assembly three of them actually, and shotgun new cassette heads in there.
 

PRR

Member
Joined 2003
Paid Member
You can still buy 8-track tapes, used, possibly with Tom or Frank singing but you will erase that. Even blank sealed carts 3/$15.

It needs to be 1/4" and it needs special lubrication. Cassette tape is not even close.

8-track (also 4-track and broadcast cart) machines can be found, "for parts", or even "works great!"
 

Attachments

  • 8tracks-on-eBay-42.gif
    8tracks-on-eBay-42.gif
    172.2 KB · Views: 118
I can only speak from my own experience regarding the voltage output of 8 track cartridge heads. With my scope set to the 2 mV/div scale, I get barely any noticeable movement of the wave on typical program material. Thus, I'll say the output of the head itself is in the high microvolts, or at least far less than 1 mV according to the Tektronix 2465 CT scope.
 
  • Like
Reactions: 1 user
As someone playing arround with RTRs and cassette decks I feel obligated to give a final word.

The voltage on your heads during playback is frequency dependant. The function is 20dB/dec or 6db/oct
Reason being its an inductor.

For standardisation of your measurements you should measure head output levels at 315Hz for cassette decks (unaffected frequency by both HF rolloff and LF shelf) and for reel to reels at 1kHz. (For speeds at or above 19cm/s)

If you were tomeasure 1mV at 1kHz on a reel to reel head, youd find that at 100Hz youd be measuring 100uV.

The head impedance thus is very important, and you can relyably measure it only with a calibration tape with known record levels.

For what its worth, cassettes usually are recorded 250nWb/m 0dB or 320nWb/m 0dB
This flux will then produce a certain volte in the head and so you can measure by a refference, how is your head doing.

Couple of noteworthy things:
DO NOT MEASURE WITH MULTIMETER OR OSCILLOSCOPE.
These devices have some amounth of leakage current flowing back and mere nano amps are enough to magnetise your heads permanently. The only way to measure the head output level would be with a tube AC voltmeter or a specialised low high input impedance test meter.

Just plug any cassette deck or reel to reel head preamplifier into LTSpice. Youll find that 315Hz most cassette head preamplifier run roughly 60dB gain, thats a lot. And pretty much checks out with the expected voltages on the head. Remember, 20dB / dec voltage gain. The lower the frequency the more gain you need. My preamps for my reel to reel go 85dB gain at 10Hz.

With this knowledge youll find out measuring voltage across your heads is pointless, in fact it does harm because you magnetise your poor heads.

Youre better off building a head preamplifier with known parameters and a Jfet opamp that has input current specs in the picoamp regions. Or CMOS, whichever is at hand. And you need lots and lots of gain to get your head voltage to line levels.

Please for the love of god do not use the NE5532. Its input bias and offset current are so high they will magnetise your heads, and that is no good. Use a TL072 or something with JFETs. If you want something hella good look for OPA1656 opamps. Replacing the old TBA opamps in a revox A700 has proven to be the best decision ever. No noise without tape running across the heads compared the old stuff that was barely below the noisefloor of the tape.
 
  • Like
Reactions: 1 user