Philips N2401 - help needed

[Radio_Dave]

I've just finished removing and checking every component on that PCB. All resistors are now within 5% tolerance and the transistors seem to test OK The electrolytic capacitors C303 and C304 are a little high in value, around 120uf and C301 is around 12uf, but I've left them in place as their ESR is very good. Unfortunately the problem is still there and messing around with Sk312 does nothing.

I've noticed that the voltage at the base of Ts302 is around 200mV, at idle and not the 50mV as shown on the circuit. This voltage then varies up and down, around 100mV to 300mv when playing the start of the tape. When playing towards the end of the tape it's around 100mV to 350mv but every few seconds it shoots up to 1.6V, triggering the eject.

David

[JohnBHanson]

When the switch closes the capacitor C301 charges and feeds a current into the base of Ts301. This makes it conduct and short out cap C302. When the switch opens the Ts301 turns off and the capacitor discharges through R302 and R303. At this point the base of Ts301 will be negative, hopefully not beyond its breakdown voltage. This discharge time could be speeded up by placing a diode in parallel with R303 (Anode towards the base).

Thus Sk312 opening and closing (ie the take up spool turning) prevents C302 from charging.

When the take up spool stops C302 will start to charge up. This will cause TS302 to turn on which will cause Ts303 to start to turn on.. The action of Ts303 turning on will turn on the relay/solonoid Re301 to turn on. As Ts303 turns on current will flow through C303 and R308 into the base of Ts302. This makes the turn on faster and also behaves like a monostable eventually reducing the drive on Ts302 and Ts303 (and hence reducing the current through Re301 - although I am not sure if this will go below the holding current.

The low value of C301 could well be limiting the base charge into Ts301 and preventing it fully discharging C302 - this would account for the higher voltage on C302. Replace C301 (or connect another cap in parallel with it and see what happens). ESR is not everything. (Assuming C301 should be 68uF and not 6.8uF).

[Radio_Dave]

C301 should be 6.8uf. I've tried 4.7, 10 and 22uf in its position and there was no improvement.

I've been randomly swapping components, with ones of different values and noting any improvements. So far I've found that changing R305 from 1M to 1.2M very nearly cures the problem. Changing C304 from 150pf to 470pf makes a small improvement as does changing C304 100uf to 220uf. Hopefully there's some logic to these changes?

Tomorrow I'll try all the "improvements" together and see if it works.

Thanks
David

[Nickthedentist]

Is the switch actually making and breaking properly?

I spent ages chasing my tail with a similar arrangement on a Pioneer CTF-2121, only to find that the rotation sensor itself wasn't closing properly. I replaced it with a reed switch salvaged from an old reed relay, and normal operation was instantly resumed. More modern decks used a Hall effect senor, of course.

Nick.

[ricard]

Thanks John for the excellent analysis. It makes sense now.

However, one thing though: as noted earlier in the thread C302 is not 150µF but 150nF. However, the majority of capacitance in this part of the circuit would actually be C303 with it's low series resistor R308. Since the relay is non-energized, the right hand end of C303 would actually be grounded until the relay is energized by Ts 303.

Looking at the circuit though, this means that C303 will normally be reverse biased, albeit by a small amount. I wonder if the nominal lack of polarizing voltage would cause eventual failure of the component, even though it apparently checks out normally. I've had problems in the past with electrolytics being subjected to very low voltages (in the half volt range), the capacitor basically needs its polarizing voltage to maintain its function. In this circuit of course as soon as Ts 303 energizes the relay the C 303 will get forward biased, so one would expect it to reform fairly often ... still... worth trying to replace this particular one?

Since increasing R305 helps, it would indicate that something is causing the time constant of the R305-C302-C303 network to be too small.

Of course, the simplest cause would be Nick's suggestion that the Sk 312 is not operating properly.

[Radio_Dave]

I really sorry but I've made a some mistakes in my earlier post. It was C303 (not C304) that I changed from 100uf to 220uf and it is of course C302 that I changed from 150pf to 470pf.

Yes, the switch has been cleaned and is working fine.

Thanks
David

[ricard]

More guesswork: you mentioned that the collector of Ts 302 is at 200mV rather than 50mV. That might indicate that the saturation voltage of Ts 302 is outside spec (I looked in a datasheet for BC547 which specifies it as 250mV max but that is at a collector current of 10mA, and it's much lower here so I'd expect the saturation voltage to be lower too), so perhaps try to replace this part?

BTW, not that I think it matters, but was C302 150nF or 150pF ? Initially in this thread it was noted that the 150µF on the schematics was a typo, but that it was in fact 150nF, but it has since been referred to as 150pF.

[Radio_Dave]

Originally Posted by ricard

More guesswork: you mentioned that the collector of Ts 302 is at 200mV rather than 50mV. That might indicate that the saturation voltage of Ts 302 is outside spec (I looked in a datasheet for BC547 which specifies it as 250mV max but that is at a collector current of 10mA, and it's much lower here so I'd expect the saturation voltage to be lower too), so perhaps try to replace this part?

That voltage was actually measured at the base of Ts302. I have now managed to get that voltage to fluctuate around 35 to 60mV, with the machine at idle, by adjusting the gap between the switch Sk312. I don't know if the fluctuating voltage is significant? TS302 has previously been changed for another BC147A and a BC547, without improvement.

Originally Posted by ricard

BTW, not that I think it matters, but was C302 150nF or 150pF ? Initially in this thread it was noted that the 150µF on the schematics was a typo, but that it was in fact 150nF, but it has since been referred to as 150pF.

Yes sorry it's a typo, C302 should be 150nf. I changed it to 220nf, resulting in a slight improvement.

Thanks
David

[Radio_Dave]

It's early days but I'm pleased to say I think I've fixed it.

I've changed R305 from 1M to 1.2M and C303 from 100uf to 220uf. Now when playing at the start of a C90 tape the voltage on the base of Ts302 varies between 90 and 190mA and at the end of the tape it's varying between 100 and 380mA. There's no sign of the 1.6V randomly showing up and starting the stop/eject (well not until the tape stops).

Thanks
David

[Nickthedentist]

Well done, David, we know you would!

I wonder why you had to alter values though.

Nick.

[Radio_Dave]

Thanks Nick, I've no idea why I had to change the value of those parts, or what they are doing to make it work? I just did loads of random changing and making notes.

The circuit diagram does show a BC147 for Ts302 and mine had a BC147a. I don't know if there's enough of a difference there to cause it the problems it had, or maybe it never did work on C90 and longer tapes

Thanks
David

[Sideband]

Should work on C90 but never ever try a C120 in these.....You'll end up with a mess to clear up!

[Luiz Claudio Vi]

Hi. Are you still in the need for understanding the auto-change circuit? If so, I can go ahead here or by e-mailing you straight. Let me know what you want the most. Cheers. Luiz (Brazil) (after editing) Oops! I did not read the last posts, I thought there was still some clearance needed. Cheers. Luiz

[Radio_Dave]

Hi Luiz,

Thanks for the offer but that N2401 is still working fine and I've also picked up a full English service manual along the way.

However by some strange coincidence I picked up a second N2401 just this Monday. I'd totally forgotten all about this thread and the problems I had. It will be interesting to see if this one has the same problems with stopping too early.

Thanks

David

[Radio_Dave]

Well how strange! My second N2401 has exactly the same fault as my first one. It stops and ejects the cassette 5 to 10 minutes before the end.

Like the first one I've swapped C303 100uf for 220uf and it's working fine. I wonder if all these N2401's have the same problem?

David

[Luiz Claudio Vi]

Hi! And congratulations in successfully taking this second N2401 into working. Would that be a good idea to post some considerations of mine regarding the auto-changing circuit operation? If so, I believe I can shed some light
Cheers.

[Radio_Dave]

Hi, Yes please tell me everything you know.

David

[Luiz Claudio Vi]

Hi.
I'm not home right now, only after this Friday, and all my notes, pictures and related data are stored in my local, homebuilt NAS (FreeNAS).
Anyway, what I remember by now about C303 (100uF) is that it acts like a 'saturation booster' (kinda positive feedback). If you follow the diagram in n2401b.jpg, as long as the circuit is steady Ts301 is saturated which provides TS302 being kept in cut state, hence Ts303 is in cut state, too (not conducting, Vbase = Vemitter), and solenoid coil Re301 is not energized. When Ts301 is cut (Sk 312 is either open or close for more than 2 seconds), Ts302 saturates, hence Ts303 conducts, too.
The problem is the transient current after Ts303 energizes Re 301, which may reduce overall Vcc voltage in the circuit (the +29V is obtained from power supply with a voltage duplicator) after the current peak through R310 and cut Ts 303 before Re 301 is properly fed. So, when voltage raises over Re 301, C303 acts as a current booster for Ts 302 and forces it to keep conducting because it is actually being charged by the voltage applied to Re301 trough R 308. This arrangement makes sure the solenoid will be fed with enough current long enough to make sure the changing gear stars properly.

I found one mechanical problem related to part 160 (see N2401Deck.pdf) that also caused this very behaviour of premature changing, but I need to find my notes prior to write about it.

Hope this helps for now. Please, if anyone finds any flaws in my reasoning, I'd gladly accept corrections that may lead to better explanations in the future.
Cheers.
Luiz

[Luiz Claudio Vi]

Just to add some extra info (I lost my 30 minutes editing limit).
I guess a 'Schmitt Trigger' might also apply here as functional related circuit to the C 303 arrangement. I also missed mentioning I followed John B Hanson's explanation which is more complete than mine, and I thought I could add some extra tips.
If you want to take a look at some of my Philips recorders, please feel free to come by @ http://s1212.photobucket.com/user/el...?sort=6&page=1.
Cheers.
Luiz

[Luiz Claudio Vi]

Hi. Please, consider that I'll refer to parts present at N2401Deck.pdf and surround all their numbers with brackets, like [1], OK?

So, engine [208] is energized and [165] (main pulley?) starts rotating at the moment the recorder is turned ON. The right-hand turntable [160] also starts rotating because pulley [184] is mechanically coupled to [165] and is kept pressed to it by spring [159]. Pulley [184] is a metal pulley with a rubber ring, and we are getting back to it right ahead. By carefully inspecting [160], we see that it is actually a complete assembly built with two 'half', upper-and-lower turntables that are mechanically coupled to eachother by a clutch assembly (it has an internal spring and its internal shafts are easily broken if trying to disassemble it... Please, hear the voice of an experienced wiseass...). The rotating speed of the lower half of [160] is steadily kept by means of [184] and [159], unless the 'PAUSE' key is pressed while the system is in PLAY/REC position. This same lower half also has an eccentric cam underneath of it (the eccentric cam is part of it, actually) which constantly and steadily pushes [227] and makes sure [232] is moved towards [235] (SK312) and closes it in constant frequency. [237] should pull [232] back to its original position, hence opening SK312, right after the eccentric cam turns back to the opposite, initial position, BUT [232] will only do this if the upper half of [160] is also rotating and pushes [230] counter-clockwise. You can observe [230] moving 'upwards' and 'downwards' when the deck is turned ON and the mechanism is in 'resting' position, i.e., after ejecting a cassette and not arming itself back to play another. Making sure there is no cassette in the deck, just look through the right-hand small hole in the cassette deck, providing enough light, and you'll see [230] 'dancing up and down'. There is a small 'lead' cap over it that does not show up in the mechanical drawing, and I cannot tell you for sure why it is there (I can only guess it is for stability, but this is just a wide guess). In normal operation, the signal generated by SK312 resembles a PWM-coded, constant-frequency signal, with a closed-to-100% duty cycle. As the system arms itself and starts playing a cassette, the frequency of the signal does not change, but percent rate varies because the upper half of [160] slows down while the lower half of it keeps its rotating speed. So, SK312 is kept ON a little longer then it is OFF, and the ON/OFF ratio increases as the upper half of [160] slows down even more as the right-hand cassette pulley is filled with tape. At the end of the tape, the upper half of [160] stops, [230] does not move counter-clockwise and [232] keeps SK312 closed so the autochanging circuit triggers.

Now for the mechanical issue itself . In two of the seven N2401 units I have (yep, seven units, two originally assembled, one being heavily modified and four uncompleted others) I faced problems too close to the ones you describe. In both of them I found two dried, glassy-surfaced rubbers in both [184] pulleys. Because of this, [184] was constantly sliding over the lower half of [160], which rotated as a whole piece instead of two parts, both upper and lower halves together with no sliding over eachother. So, instead of keeping in constant frequency with a varying duty cycle, SK312 slowed down its frequency too much and the autochanging circuit started triggering prematurely before any C90 cassette reached their end (sometimes it happened even with C60 cassettes). I found no replacement for the [184] rubber rings so I used a thin sandpaper and removed the slightest surface coat out of it I could. All I can tell you is that the signal from SK312 got back to the expected behaviour in both units, and I did not need to change any of the components in the circuit.

Please, forgive me for writing too much, and I'd also like to apologize for taking too long to write back. I did not find my 'ancient' notes (they are here, somewhere) so I had to write it all by heart. In this case, please allow me to know if there is any flaw in my description. Also, as I am not a native English speaker and I am writing in a UK-based forum, I'd gladly thank you guys for helping me to write better by pointing any typos, miss writings and any kind of aberrations you find in my text (I'd appreciate corrections, too ).

Hope this is of any help.

Cheers.