Fitting 'Aux-in' connector to a Stereo Radio

[Old Sparky]

As far as I know, this is straightforward.  I've just made up an adaptor to allow an I-Pod or other device to be used with a Radiomobile 108SR car stereo  8 Track player.  Although this unit is stereo, it includes an AM radio, which is obviously mono, coupled to the o/p stages via series resistors.  I know that, when fitting such an adaptor to a mono radio, series resistors of between 1 & 10k are needed in each channel, but haven't fitted them in this case.  Preliminary tests with an AF Generator indicate that this will be OK so I'll probably leave it that way, unless anyone thinks that resistors should be added.

[Terry]

Your I-Pod will be stereo, so what sort of connection are you making here?

If it is stereo, then the left and right outputs will go to their respective inputs, which is ok.

However, I'm concerned from your mention of mono that you are making a mono connection to a stereo amplifier by connecting both left and right inputs together, thus both outputs from your I-Pod will be bonded together.

So, what happens when attempt to play something in stereo? When the left and right outputs are in antiphase, as they frequently will be when playing a stereo source, you will effectively be putting a dead short across both output stages - which won't do your I-Pod any good at all!

Also, as you presumably don't have a dedicated input position on the amplifier, both inputs will still be bonded together when you play stereo tapes. This won't have such disastrous consequences for the stereo preamp but why condemn a perfectly good stereo set up to play in mono  for evermore?

Make sure your new input is stereo and incorporates suitable switching to isolate the original signal sources when you plug the I-Pod in.

If the I-Pod, set to a normal listening level, is much louder that the tape or radio, then you can fit series resistors to equalise the levels afterwards.  

[radioman]

If your source is stereo and the destination is mono then you definitely need to use series resistors - otherwise you will be joining two low impedance outputs together -  but I think you already know that.
However,  you say the unit has a stereo amplifier so I don't think the fact there's a AM radio in it is relevant.
I would simply keep both channels separate and connect the inputs to the tops of the volume controls via suitable switching.

Andy

[Mark Hennessy]

It might help us to help you if you could post a copy of the schematic.

If nothing else, I'd be interested to see it

[Old Sparky]

The attached files show (1) My sketch of how I wire in a 'radio/aux' switch.  I use a 4-pole toggle switch so that unwanted inputs are grounded, preventing radio breakthrough on the 'Aux' i/p, which may be due to between a mere ans 1.5M of screened cable acting as an aerial
(2) is a copy of the 108SR N/P circuit diagram, though it may need enlarging to read clearly

[Mark Hennessy]

Wow - that's hard to read! Do you have a higher resolution copy?

It looks like they aren't actively switching the audio signals between the radio and the cassette pre-amp - rather, they are joined together, and power to either the head preamp or the radio receiver is switched according to mode. Hard to be sure from the schematic though.

From there, the signals travel via a transistor pre-amp before being applied to the volume control and hence the power amps.

Now, it would be nice to know what the signal levels are. It might be that the iPod doesn't require the extra gain afforded by the preamp, in which case you'd inject the signal at the top of the volume control. However, the headphone output of a portable device is often a fairly low voltage, so perhaps the preamp is useful? Impossible to say without the thing in front of me. If the schematic was clearer it might be possible to determine the "gain structure", but as I can't read any of the component values...

Having determined where to put it, I can't imagine why the 4-pole switch would be required. Obviously I'm aware of this type of configuration, but it's only used when separation of 80dB or more is required. Just think about the amount of signal switching in a sound desk with 48 channels and 8 groups, 2 mains, 6 auxes, 50+ insert points, etc, etc. Total isolation between channels is required, and that is usually the case. With conventional single-pole switching. Read some of Douglas Self's writings about this, where he deals with the subject comprehensively.

So, given that we don't need anything like 80dB of isolation in a car, I suspect that a previous case of breakthrough was caused by some other undiagnosed factor (leaky switch, perhaps?), but since a 4-pole switch cured it, it's become a routine.

Personally, I would rely on the double-throw contacts within a 3.5mm socket to do the switching. You can get sockets that have a pair of contacts that are separate from the input connections, which give you the flexibility to switch signals at any level, or to switch rails to disable the radio and the tape preamp. Lots of options, and no separate switch needed.

In answer to the original question, no, you don't need resistors for the usual reason (to stop the output stage of the portable device "fighting"). If the signal levels seem about right to you, then you don't need resistors to attenuate the signal. I assume you are going in at the same place where the signals merge?

However, I would recommend the following:

As the signal comes in, I would put resistors to ground. This is to ensure that any output capacitor the device has is able to charge to the expected DC voltage. Some use bootstrapping that requires a DC path to ground. Something like 1k would be OK.

Then I'd put a series resistance from there. This is to protect the input of amplifier, should a DC transient appear (like when the device is initially plugged in). Something between 1k and 10k is a good starting point. If you go too high, you'll start to lose level (which you might want anyway) - only you can determine this.

After this series resistance, I'd add a cap to ground. This is to keep out RFI. It would be a few hundred pF, perhaps more depending on the resistance "seen" by it. Choose it to have a turnover frequency of 15 to 20 kHz. If you omit this, your set might suddenly become much more prone to cell-phone interference and similar problems. To stop re-radiation within the set, keep the connections between the input and this cap as short as possible - otherwise a cap in parallel with the input socket contacts might be needed.

Next, you'll need a series capacitor to couple the audio in to the base of the audio preamp, just like the radio and tape signals do. Probably sensible to make this the same as those already used by Radiomobile.

Next comes the switch. I presume you'll be putting this in series with the base of the preamp, after the junction of the two electrolytic capacitors?

If you're going in at the volume control (rather than into the base of the preceding transistor), then you can omit the series coupling cap, but I'd still include the other components.

Does this make sense?

[Joe]

I haven't got an i-Pod but the output of my generic MP3 player is 0.22V p-p.

— Joe

[Old Sparky]

Sorry about the poor resolution, Mark. Having to limit the file size to post on the forum doesn't help. In the meantime, thanks for all the advice. Over the years I've made up quite a few 'Radio/Aux adaptors for other people, and fitted them to several different makes & models of radio. AFAIK, there have been no problems, but I will consider the mods. you recommend. One thing puzzles me with the circuit as it stands, which is, having, as I usually do, disconnected the radio/tape o/p from the top of the volume control, and inserted the c/o switch at this point, I expected to read a DC resistance between the 'aux-in' point and ground, since, with the switch in the 'aux' position, this is connected directly to the top of the Volume control, which, in the case of the 108SR(an 8-track player incidentally)is 25k Log, with a 1k fixed r. to ground from the loudness tap (At least that's what's shown in the diagram!). With a DMM I don't read a constant DC resistance, but rather a rising resistance going up to infinity as if a capacitor was charging, Why would this be? If I inject, say, a 1kHz sinewave at around 200mV, the volume of this can still be controlled in the usual way, so no real problem, just a puzzle for me, and I've been servicing audio gear for almost 45 years, and used to repair 108SRs 'back in the day' (early to mid 1970s)!! Incidentally, the way I connect these 'aux-in' devices -i.e. via a switch to the top of the volume control is essentially identical to the method used by manufacturers back in the 70s to allow a car tape player or recorder to be used via the o/p stage of the radio. Apart from the source medium, the only difference is that the tape players had an o/p at nominal 'line' level & impedance, whereas of course most, if not all portable CD< MP3 players, I-pods, etc., only, AFAIK, have a headphone level o/p.

[Mark Hennessy]

No worries with file sizes on this forum - they are pretty generous

Not sure I can immediately explain the strange resistance measurement. Certainly, with the volume at minimum, and the previous stage disconnected by the switch, there are no large-ish capacitors that can influence the meter. The unit was definitely powered down when you did this test?

I did notice the resistor on the volume tap. They're using it to "bend" the law of the volume control, though not entirely sure why as you say it's already log.

Audio levels are a difficult topic. Even the definitions of "line level" and "headphone level" are incredibly wide. Ultimately, it's all about the volts

[Refugee]

I converted a headphone output to 200mv by simply connecting a 33 ohm resistor across the terminals in the plug on the end of the connecting cable.
The input of the amplifier was 1 meg and there was enough gain to get the output up to full power.