Monday, 26 September 2016

Quad 44 pre-amp. Minor restoration, and some sensible modifications

Along with the FM tuner I got last week, I also obtained a matching 44 pre-amp and a 405-2 power amplifier. Very nice kit!

The 44 is a very interesting pre-amp. It's got a lovely design of tone control. Quad went their own way with adding tone controls, and made theirs with 3 tone filters and an adjustable slope control.

The look of the thing is even unique.

The pre-amp has a flexible card system. Each input is selected by a CD4066 switch, which connects the selected card. This unit has the standard line-up of "Radio", "Disc" for a Moving magnet cartridge (an MC card is/was available), a CD/AUX card and two tape loops. Excellent.

Not only that, but the Disc card allows differing gains to be selected to accommodate for most carts, and switchable capacitance load too. The tape loops are similarly configurable for level and impedance. Nice....

This later "MKII" model has a re-designed main board and ALPS pots....


CD4066? Known for introducing distortion into the signal path. Thankfully Maxim produce the MAX4066. A much better specified, and pin for pin compatible device....

So, before I start, let's make a baseline measurement of THD (Total harmonic distortion) at 1KHz....  0.059% on the CD/AUX channel.

Not too shabby....

Out with the old, and in with the new....

All 4 ICs are changed.

Whilst I'm in there I evict the Red electrolytics, as they're know to fail, although these all read OK.

Let's have a look at the disc card...

It consists of TL071 op-amps.... which are mediocre. I'll swap them out for some better NE5534, which should lower noise a bit, as well as distortion. These are the better NE5534's with 18V/uS slew rate.

... and here is the main issue for me....

The CD/Aux card....

300mV? ... 300mV!!!!

Since when?

Most CD players put out at least a volt...

All this means practically, is we'll need to turn our volume control down everytime we switch to CD. An inconvenience. We *may* overload the input.

I've seen some CD players output as much as 2V, so 300mV is not going to cut it....

It's also originally fitted with a TL072, which I'll change for an NE5532..

I do some googling, and there's various recommendations about changing the resistors in the feedback loop of the card to reduce the gain to unity... but none of the modifications seem to  relate to the input card I have fitted, an M12815:1. I don't have a schematic, so I sketch one out....

... well, it's already a unity-gain amplifier!

So, what to do? I could trace the signal through and change the gain on the main board, but that will leave me with the same discrepancy when switching between CD and disc or radio.... Nothing for it but to add a bit of attenuation in.

Simply adding a 180 ohm resistor across the existing 820 ohm resistor will divide the voltage down to a more respectable level.

... and it's easy to implement...

Don't forget to do both channels.

So after all out mods... how does it measure?

... The THD has dropped to 0.014%....

How does it sound? Noticeably quieter (noise-wise!), but to my cloth ears, I can't really tell much difference in the distortion.

Right, now to the 405-2 Power amp!

Watch this space....

Thursday, 22 September 2016

Quad FM4 drifting and amnesia.

So here it is, the wonderful Quad FM4 tuner. A thing of beauty.

This one is an early example and dates from around 1982-83.

It's suffering from amnesia (it won't store stations after a power off) and even in manual tuning the frequency falls after a few minutes use....

Let's see what's wrong...

It's a simple layout. First off the memory battery is shot, and leaking. Thankfully it hasn't damaged the board, and a replacement (Varta Mempac 4.8V, 150mAH) restores it's memory... as for the frequency drift...

The tuner operates by a voltage controlled tuner, there's no synthesiser in here! The memories work by digitising the tuning voltage and storing it, recreating it with a DAC when the preset is selected. The fault is the tuning voltage is falling off. Turns out it's our best friend (and foe!) the elctrolytic capacitor. The guily parties are all 100uF 6.3V types. I changed them all. I dislike 6.3V electrolytics, they seem to give more problems than higher voltage types, even though they were all working well within voltage limits. Perhaps it's their physical size that makes manufacture difficult? ....  (perhaps they just don't like me?!)

Zoom in on the picture and you can see where they are physically leaking.

This cured the drift, but one of the 220uF 16V caps that smoothes the regulated 12V supply looked like it had seen better days, so I changed both of those too... even though they measured OK. One is adjacent to the 12V regulator transistor, which runs a little hot. Never an ideal location for an electrolytic!

Sound quality is great, and will give many years of service now!

Friday, 16 September 2016

Nakamichi RX-202E

This arrived in the workshop.

It's a Nakamichi RX-202E Unidirectional Auto-reverse cassette recorder. A thing of rare beauty and quality. I must say the quality of reproduction on this unit, knocks the Pioneer CT-F1250 we looked at earlier in the year into a cocked hat, but it is substantially younger!

It unidirectional, because the tape only ever travels in one direction, so no need for two capstans, two pinch rollers, and a complicated and compromised head which will play both sides of the cassette or a complicated mechanism to rotate the head.

Oh no, this uses a complicated mechanism to remove the cassette from the transport and turn it over!

A video is worth a thousand words....

All this machine wanted was a service, and a repair to the position sensor pot.

Monday, 12 September 2016

Arduino Astronomical Clock (and pond pump controller) - now with master clock update!

Well, the DS1307 clock module has a tendency to drift about a bit on my Pond pump controller. It can lose a few minutes each month. Irritating at best.

Now the GPS master clock as been created, it should be fairly easy to synchronise the DS1307 with the 433MHz data from the master clock.

The receiver hardware is connected to 5 volts, GND and it's data line is connected to pin 9 of the Arduino. The antenna is simply 17cm of stiff wire.

Once again, the wonderful Virtualwire library is used, and we simply call the receiver routine (remoteClockSet) in the main loop to see if any data has been received. If it has, update the DS1307, and recalculate the sun rise and set times.

There's a few more variables to declare to deal with the incoming time, and, to be honest the whole thing was written a while ago, and is now horribly un-weildy, and could do with a damn good tidy up and re-write, but it works.

No more difting clock!

Here's the code.