[Ron Ramirez]

A few weeks ago I picked up this Fisher 4400 receiver:

[Image: http://www.philcoradio.com/images/phorum...r_4400.jpg]

It is a very unusual model - there is basically no info on it available online, not even at AudioKarma.

The right channel was dead, but the left worked fine.

After replacing some transistors in the right channel (many thanks, w4rtc ), I now have sound in the right channel but it is very low - if I crank the volume to max, the right channel is about at normal volume.

Here is a partial schematic showing the 4400's audio circuitry.

[Image: http://www.philcoradio.com/images/phorum...io_Amp.jpg]

I bought a transistor tester, and it confirmed that the original driver transistor was bad. According to the tester, the replacement GE-76 output transistors are good, as is the replacement GE-3 driver. So now I am looking at something else that is pulling down the 24 volts (shown as 22.5 volts on the partial schematic above). I arrived at the 24 volt figure by taking several voltage measurements in both channels and comparing the results of the working left channel to the non-working right - the left channel has 24 volts where 22.5 is shown at the outputs; the right channel has 19 volts. I have 45 volts where the schematic shows 46/46.2 volts.

I am suspecting capacitor C20 on the schematic (1500 uF, 50V) between the right channel output transistors and the right speaker output as leakage in this capacitor would pull the +24 volts down.

Any thoughts, comments, second guesses, etc.?

[BrendaAnnD]

That is a possibility, you can verify by checking to see if there is DC across your speaker. Or see if you have DC on the speaker terminals with no speaker connected. Also please check both of the diodes and associated resistors in the bias network, these are famous for shorting/going out of tolerance and messing up the bias voltages... not good for the outputs.

[morzh]

Ron.

1. Do the test without load. This will tell you clearly where the current goes when your voltage shifts.
2. if it is OK, just swap the caps. Or better yet, change the cap - you should have something like this in your stash somewhere.
3. Test the output transistors. Q8.

Last. You do not really need a transistor tester. I never ever needed it. Analog ohmmeter is all you ever need. THhis makes it quicker.
test transistors only outside the circuit, having unsoldered the wires if it is a power one.
Keepeing it on x1, Test BE forward-reverse, should be NO LEAKAGE (the arrow should not move at all when reversed), do the same with BC, and then EC both ways (no visible movement should be present). This pretty much means the transistor is good.
That goes for any transistor, power or small signal.

[Ron Ramirez]

Brenda - thanks, I'll try that this evening.

Morzh -
1. Thanks, I will do that this evening also.
2. Actually, no, I don't - I have plenty of capacitors for the old Philcos, but not for solid-state equipment as I have not worked on solid-state electronics for many years. I am going to order some new electrolytics for a couple other receivers soon, and will also order some new electrolytics for the 4400 as well.
3. Tested, replaced, retested. The driver Q6 was definitely bad and was replaced. I replaced the outputs, Q8 and Q10, as well based on my testing with the multimeter. But now that I have the transistor tester, I find that the original outputs test OK. (I retested the original driver with the transistor tester and it was bad without a doubt, so it went into the trash can.) I still have the replacement outputs in the 4400.

I tried using my multimeter to test the transistors. Worked OK for silicon but I was getting inconclusive results for the germaniums. This is why I made the investment in a Heathkit transistor tester. It was cheap, really - $19.99 with free shipping, and it only needed a new battery and the switches needed to be cleaned with DeoxIT. It works fine.

[morzh]

Ron

Also, those 330 Ohm resistors, the R36/R38.....those gotta be large - they constantly dissipate at ;east 1.5W so they are probably high wattage, like 4-5W types. See if one of them has drifted. There is no DC feedback in this amp (not that I see) and there is nothing else to ensure the 45V voltage divides equally (BTW not sure where they came up with those 46V and 46.2V numbers if the power is 45V with no active conversion) into 22.5V except those resistors (there is the inter-stage transformer concept for you) so you dissipate some extra.

So, it is either the resistors or the cap leaking. The leaking cap is readily eliminated by unplugging the load (BTW make sure your load is 0 Ohm when there is no speakers - like your phone jack is shorted or has something in it...conducting....a dead roach....). And it has to be some leakage too to overpower 75mA of current.

If it is still the case, measure the resistors then.


PS> just an idea - for seeing if a regular good cap makes a difference you could use one of them tube radio caps - just use 20 to 100 uF cap. You will also get sound with the base attenuated.

[Ron Ramirez]

Okay, I do have a few 100 uF, 100 WVDC electrolytics so I can try one of those after I make the other tests mentioned above.

[Eliot Ness]

As Brenda mentioned I'd check those diodes, the Fischer GET50825-1 translates to a 1N2326, and those old germanium diodes didn't always age very well. Plus you know every electrolytic left in there is suspect by this point in time along with the resistors, especially any that are passing very much current.

The advantage of working on a stereo with one good channel is the ability to compare measurements between the two.

I know you don't want to hear this, but keep the thought in the back of your mind that the output xsitors you subbed in might have different characteristics than the originals, and matching on these is often fairly critical.

I've checked a ton xsitors with a meter over the years, but don't recall testing any germaniums, except for the 1N34 diodes in my little rocket crystal sets. Anyway, the testers are quick, cheap, and an easier way to match up gain. Good luck, that's a nice looking receiver.

[morzh]

To be honest I do not remember any difference in testing with a meter. Ge has a lower forward drop but that did not affect this type of testing.

I had an old analog meter with a transistor tester that could test static/differential Beta which was of some interest but not much....it was all listed in datasheets.


The diode, if one of them is open, could affect the centerpoint by about 1V. If shorted - by about 0.4V.

[Mondial]

The bias point on these old OTL transistor amps is not well controlled, as there is no DC feedback from the output as Morzh previously mentioned. Centering the output stage voltage really depends on the match between the top and bottom sections of the circuit. Sometimes I have seen a variable resistor in place of one of the 330 ohm resistors so the bias point can be accurately set to half the supply.

If the series output 1500 uF cap is dried out, this could also cause low output in the bad channel, although the signal would most likely be distorted when you tried to turn the volume all the way up.

In any case, the low audio problem may not be associated with the output stage voltage. Have you checked that the signal level is identical into each channel's power amp input? The low audio problem may be located further upstream. You can try switching the wires from the volume control slider into each input and see if the low audio is then on the same or opposite channel.

[morzh]

That is easily found out by simply disconnecting the "bad channel" wire and then shorting the inputs together with "good one" and seeing if the volume is the same.

But...let's start with imbalance - this may or may not indicate something.

[Ron Ramirez]

Please refer to the revised image below.

[Image: https://philcoradio.com/images/phorum/Fi...ltages.jpg]

The above revised partial schematic contains actual voltage readings I took tonight in red.

I did not measure any appreciable voltage across the right speaker terminals, with speaker connected or disconnected.

I tried removing C20 from the circuit and using a 100 uF cap in its place. Volume was as low as before.

I carefully unsoldered two joints so that I could properly measure resistors R36 and R38 - they are nearly perfect. I also unsoldered one leg of diodes CR6 and CR8; they test good.

So I am now puzzled as to why the voltages, which are 25.1V at the base of Q7 and 25.3V at the emitter of Q7 and the collector of Q9 are 19.3V at the base of Q8, 19.6V at the emitter of Q8, and 19.7V at the collector of Q10.

Oh, and I also removed the right channel wire which leads to the volume control and bridged it to the left channel - the volume is still much lower in the right channel.

I think this one has me stumped for now.

[Mondial]

Ron, I don't think the differences in voltages are significant with regard to the low audio issue. One channel is a few volts high, the other a few low, but not really out of the ordinary. This balance issue really does not affect the AC gain of the amp. Otherwise the voltages look reasonable with regard to biasing of the individual transistors.

On thing that will affect the AC gain is the driver feedback bypass cap C26. This cap reduces the AC feedback from the driver stage collector, preventing loss of AC gain while maintaining the DC feedback. If this cap is open, It will reduce the AC gain, which may cause the low volume. You can try bypassing it with the 100 uF cap you have and see if it makes a difference.

Electrolytics are always suspect, so this is where I would look first.

[462ron]

Ron this a shot in the dark but maybe the reason for the voltage differences between the good channel and the weak one might be attributed to the differences between the original transistors and the universal replacement ones as far as their operating characteristics are concerned. Out of curiosity I would check caps C14 and C26. If they were open I would think that would reduce the signal across the primary of T2 thus reducing the drive to the output transistors creating less output to the speaker.

Ron

[Ron Ramirez]

Thanks, all.

I should mention that this schematic was taken from a Fisher TX-100 amplifier and revised by me to match the audio output circuitry of the 4400, since I do not have a 4400 schematic nor do I know where I might find one. (Credit goes to Larry Derouin of AudioKarma for pointing out the similarity between the TX-100 audio output section and the 4400.) So, the printed voltages are for the TX-100 and I do not know what the factory voltages were for the 4400.

I'll check C14 and C26 tomorrow, or sometime this weekend.

My plan is to replace all of the electrolytics in this set, but I just wanted to hear the right channel match the left first, before I give it a complete rebuild.

Did I mention that it has been more than a decade, maybe two, since I did any work on solid-state equipment?

[morzh]

Ron

If everything tests OK I think it might be the impalance of rthe Vf between the diodes that are there (I think) to remove the crossover distortion (to open the transistors slightly). Because transistors might conduct differently, the voltage might distribute unevenly.

I do not think this is a problem.

I vote for C14 and C26 as they control AC gain of the driver.