Tube Tapedeck
Page 1
I have been working on tube electronics for an old Tandberg tapedeck for some time. The original boards were all very dead - probably because of a lightning strike. At least all the transistors were dead...

But instead of repairing the original electronics I thought it would be a lot more interesting to design tube electronics for it. So far the playback amp design is completed. The playback amp is a three-stage triode amp. I will describe the design spread accross a couple of pages. The circuit description is below. The part list and construction notes are on page 2. I will add measurements later.

To the right the schematic of the first stage can be seen. The first stage uses the 5842 triode (also known as WE417A). The 5842 has a very high transconductance (around 25mA/V), and is very low in noise. The stage has a flat frequency response and a gain around 35 times.

The four 150 Ohm resistors are grid stoppers that are there to prevent oscillations. The 5842 has four grid connections, and it's necessary to use a grid stopper on each pin.

As can be seen from the voltage figures, the current is quite high at about 22mA. That increases the transconductance of the tube and lowers the noise. But the power dissipated by the tube is only around half the rated power, so tube life should be fine.
Input stage
Second stage
And here is the second stage. As the first stage this stage is also "just" an amp stage with a flat frequency response, but this time the gain is lower at around 25 times.

The second stage uses half of the 6N1P double triode. The 6N1P is a medium mu triode with higher transconductance and mu than the "standard" tubes like ECC82 and 6SN7.

The 6N1P also offers low distortion (and a great price), which makes it a good choice IMHO.
 
Third stage
The third stage is where most of the "interesting" stuff happens. All the filters are placed in this section.

The amp in this section is the other half of the 6N1P. The stage is similar to the second stage. One difference is that the cathode resistor R2 is not decoupled by a big electrolytic cap. Instead there is a series resonant filter consisting of L3, C7 and C8 that resonates at 21kHz. This creates a small peak in the frequency response around 21kHz, that is there to compensate for losses in the playback head. The resistor R2 decreases the Q of the filter making the peak less sharp.

The second filter is the bias filter. This filter is placed between the second and third stage and consists of the parts L1, L2, C5 and C11. It creates are lowpass filter that removes any residual bias signal from the tape, to avoid problems like high-frequency instability in the following amp. The lowpass frequency can be adjusted by the trimmer C11. The -3dB frequency can be adjusted between 35 and 40kHz.

The third filter creates the NAB eq (de-emphasis). It consists of the parts R13, C6, C12 and either R8+R9 or R10+R11. This filter creates a low pass that becomes active above 50Hz and "stops working" around 4kHz. The high frequency part can be adjusted by the trimmers R9 and R10. The reason there's two of them is to provide settings for two different speeds. In my case I will need 3.75 and 7.5ips. The frequency response probably won't be perfect at 3.75ips, but since I only have old tapes recorded at that speed, I don't care much about that. By making the 21kHz peak filter switchable the response could easily be made perfect though. 15ips uses the same eq as 7.5ips, so 15ips should work just fine - I don't have any decks to test that with though.

The output is taken from the plate (anode) of stage three. That makes the output impedance quite high, so the input impedance of the following preamp or poweramp should be 100kOhm or higher. Is that is not possible (or if long cable runs are needed) a buffer stage like a cathode follower should be added. I will probably add one later.

Here is the complete schematic. I have made both a low resolution file (72dpi) for viewing, and a higher resolution file (200dpi) for printing.

On to page 2 >


Home

Comments? Questions? Just send me an e-mail!

MailMail: mcs@post5.tele.dk