ARCHIVED FORUM -- March 2012 to February 2022READ ONLY FORUM
This is the second Archived Forum which was active between 1st March 2012 and 23rd February 2022
Hi All,
I decided to finally fix a strongly humming Beomaster 3000-2, which I was given about a year ago. It looks pretty good on the outside and even has the little presets plastic cover. So far so good. I opened it up and replaced the three big capacitors for output stages and power supply (they were leaking and probably causing the humming) and then started it up. It came on, and spontaneously gave me some decent FM reception. All presets function and the tuning meter works, and its readings nicely coincide with the perceived station strength etc...Maybe the output volume is a bit weak (but this may have to do with the 6 Ohm speakers I have on my work bench...)
However, the "S lights" (arrow lights) do not work, and neither does the 'stereo' indicator. The lamps are o.k. I cannot figure out what is wrong, but then this is the first tuner I am working on...so far all my B&O experiences were rather focused on the power supplies and amplifiers (seems the tuners work most of the time), i.e. my conclusions may be of poor judgement...;-).
Anyway, here is what I checked so far around the S-lamps:
There is zero voltage at the collector of TR9 at all times independent on the field strength reading on the meter, which explains why the lamps are off. The base voltage of the TR9 is just below the turn on point (emitter is at 15.2V, and the base is at 14.7V, i.e. close to the voltage differential one would expect). If I understand the service manual correctly, the transistor is turned on when there is enough field strength, which pulls the base of TR9 down a bit via the RF signal coming out of the filter(?) (at measurement point D) fed in between the two Ge diodes D1/D2, which act as a rectifier/voltage doubler generating a negative voltage:
I hooked up my oscilloscope to D, and all I get there is a constant DC bias (independent of the field strength on the meter), i.e. the capacitor has nothing to transmit into the diodes.
After this I did a little experiment: I put a 1k resistor across the 4.7k base resistor of TR9 to pull the voltage a bit towards the 14V on the other side of the resistor. This produced a collector voltage of ~10V and one of the S-lams came on. I guess this suggests that TR9 and the rest of the S-lamp circuit may work.
Since the tuner is basically working the absence of a signal at D really baffled me...I played a bit more, and tried to follow the 19 and 38kHz signals in the decoder section. The amazing thing: I only saw the circuit diagram-prescribed DC biases on the scope, but nothing of the pilot signals. While this would explain why there is no stereo indicator, but how can the tuner work in absence of these signals?
What am I missing here? Am I making these measurements the wrong way (I connected the oscilloscope between GND and the measurement points)? Any input would be appreciated. I guess I need to brush up a bit on my RF knowledge...;-). Thanks much in advance!
http://beolover.blogspot.com
http://beolover.com
A too weak IF level to the discriminator and MPX circuits is usually caused by one of the transistors in the tuner.Most often the culprit is TR8 (BF194).Diagnose by measuring its B-E voltage.I think, you will find it higher than the usual 0.7V. Maybe as high as 2V (or more).
Martin
Hi Martin,
Thanks much for the suggestion. I will check tonight! If TR8 needs replacement, do you have a suggestion for an equivalent modern replacement? It seems that the BF194 is not available anymore...Thanks!!
Rudy
I measured the bias voltages in the tuner section. The only value that I could find to significantly deviate from the specced values in the diagram is Vcc (pin 10) of the CA3012 wideband amp. There the sheet specifies 9.2V but I measured 7.8V. This also dragged the output bias of the amp down to 5.9 instead of 6.7V. I guess this means the CA3012 draws to much current...what do you think? Thanks!
A dead CA3012 wouldn't be a first but it's far more often to see a dead TR8.What is the voltage on the collector of TR8 ?
BF194 can be substituted by BF240 or BF 254. It's not a very critical position, pure amplification.
Misaligned RF/IF stages could also cause wrong DC readings.
And have you checked the rectifier in the corner ?They can sometimes break down (even physically!) causing ripple on the DC supply.
Great advice! I guess I forgot that one...;-). The collector voltage on TR8 is indeed too low: 6.3V vs. 7.6V. The rectifier seems to work. I get a very stable 15V on the oscilloscope. Could I possibly use a 2N2222 (http://www.fairchildsemi.com/ds/PN/PN2222A.pdf) as replacement? It seems fast enough (If=300MHz). Thanks!!
With TR8 apparently driven permanently on, I tend to agree that things point towards the IC in this case.
Perhaps a stupid question: How could the potentially defective CA3012 drive the base of TR8 too high? There is a decoupling cap between them (#90).
Is there a commonly available equivalent type for the CA3012? I tried to find the CA3012 as well as the NTE726, which appears to be identical, but the best I seem to be able to do is ebay from Hongkong for the CA3012...
An interesting case at any rate! I am learning a lot! Thanks much for your advice!
No,TR8 is not permanently on.A transistor doesn't know if it's driven by 6 or 7 or 8 volts.It happily functions down to below a few volts.Yes,there are changes in noise,overload margin,gain and so on,but it has no problem amplifying at say 6 volts.
I would try replacing the ceramic filter(s) first.These can sometimes fail and cause weak reception.
Sometimes they can be tested by temporarily bridging a simple ordinary 1-10n cap from input to output.
You are right about the series cap but have you looked at the signal at the base of TR8 ?
The collector voltage at TR8 is a bit on the high side, pointing towards a transistor that is a bit more open than supposed(or a bad transistor).
It's a coupling cap, anyway. :-)
I measured the IF signal at the TR8 collector (finally I get maybe a bit of a grasp how a FM tuner is working...;-). I get a ~5.8V 10.8MHz signal when tuned into a local station (signal strength: ~4 on the meter). Then after the filter (#101) at point D this signal vanishes...all that is remaining is a very weak 30mV 10.8MHz signal. This explains why there is nothing going on at the base of TR9. Does this mean the filter has a problem? Thanks!
Either that or something is loading down the voltage at the right side of the filter.
A filter of this type normally has a loss of around -6dB(half voltage)
I would try a new filter.But just for testing almost any old 10.7 Mhz CF will do.Even a cap can work as a temporary solution.
I think the circuit is o.k. around the filter, i.e. I would like to try replacing it. Since I do not have a filter laying around, the capacitor sounds appealing...;-). Are you suggesting to put a cap between the filter in/outputs and GND to form a 10.7MHz RC lowpass with R97 and R98? C=1/(2piRf)=20pF? I know, I probably just know enough about RF to be 'dangerous'...;-). Thanks much in advance for advice!
PS.: With regard to replacement filter choice, I found this datasheet for Murata 10.7MHz IF filters: http://www.jameco.com/Jameco/Products/ProdDS/1914365.pdf
Are these adequate? What bandwidth and "center frequency/tolerance" should I choose?
One more: The service manual also makes this cryptic statement:
Is this related to 1972 component tolerances, or to be taken seriously? Thanks!!
Excellent advice!! I put a 5 pF cap across the suspected filter, and things got much better! The S-lights now work, and also the volume is much stronger now. The stereo light works, too. The question remains, which is the correct replacement filter, and should one exchange all three? Thanks for guiding me through this!
HarryPierce:The question remains, which is the correct replacement filter, and should one exchange all three?
If you don't have the equipment & skills for a complete RF/IF alignment, replace as little as possible. I would expect modern components to have close enough tolerances so that you don't necessarily need to touch anything else.
--mika
Yes...excellent advice! But I definitely would like to replace the filter. It seems there is not much to align in the IF section (they are actually proud of this engineering feat in the manual...;-)...The signal reduction across my emergency capacitor is definitely not like the Martin prescribed -6dB more like -15dB. Unfortunately the manual does not give any spec on the 3dB Bandwidth of the filter and the center frequency/tolerance parameters given for 'modern' IF filters (http://www.jameco.com/Jameco/Products/ProdDS/1914365.pdf). Any advice with regard to selection, or whatever part has worked as replacement in the past would be very welcome! Thanks much in advance!
I presume the filters are Murata.What exactly does it say on the filter and what colour is the dot(if any)?With these data it should be possible to find out both bandwidth and center frequency.
In those days it was not possible to make CFs so that all resonated exactly on 10,7Mhz.That's why they are divided into groups with the same center frequency and a corresponding colour dot.Red dot is spot on 10,7 if I remember correctly.
I have catalogs on Murata filters.
And yes,all three must come from the same group.
So-find an exact replacement for the non-working(NOS or new),or replace all three.
They have red dots. The info I found on Murata filters groups them by capital letters, i.e. the 'dot color code' is not transparent to me right now. So, any advice from your catalogs would be much appreciated!!
Can one say that the necessary minimum bandwidth of the filters depends on the frequency precision/stability of the TR3 oscillator that is used to generate the IF carrier f(RF)-F(TR3)? If TR3 is off by some margin then the IF frequency could be filtered out if the bandwidth is too aggressive?
Thanks much for your input!
Hi
If it helps I have a red set of filters x 3 ( I think red ) that work the amp had a issue if you want them they are from a later model but look the same
PM if of any help Regards
Again-what does it say on the filters?
Sorry...I do have a day job...;-). It says SFC 10.7MA on one side and MURATA X (or I) on the other. Below are a couple pics of the filter. I downloaded the current Murata filter catalog (http://www.murata.com/products/catalog/pdf/p50e.pdf#SFELF). If the current naming convention is the same as in '71, then this may mean the filter has a 50kHz band width and the center frequency is exactly 10.7MHz (see page 2 "CERAFIL for FM"). Thanks!
For some reason I couldn't open your link.
Unfortunately my catalogs start with the year 1975,and by then the SFC types were gone and the new smaller SFD types had come.
But I found this interesting pdf:
http://html.alldatasheet.com/html-pdf/193810/ABRACON/ASFC10.7MA/297/1/ASFC10.7MA.html
Could this be applied to old Murata filters?
If so,your filters are 280kHz which seems OK since there are three- the real bandwidth becomes narrower when cascading filters.
According to my catalogs you could also use the later SFE10.7MA15-Z,or SFE10.7MA5.Both are 280kHz.
The color codes:
I think you are right: A 280kHz band width filter is what fits. I verified this by extracting one of the two working filters and measuring its frequency characteristics. This is the circuit I used. Both my waveform generator and spectrum analyzer have 50 Ohm out/inputs, i.e. I used 270 Ohm resistors for R1/2 to match the 330Ohm impedance of the filter (it apparently has 330 Ohm - see radiomuseum.org: http://www.radiomuseum.org/tubes/tube_sfc-107ma.html):
In the flesh:
The measured spectrum (x-axis goes from 10.2 to 11.2 MHz, i.e. it is 0.1 MHz/decade; y-axis from 0 to -50 dBm):
We can see here that the filter is actually not exactly centered at 10.7 MHz but rather at about 10.77 MHz. The -3dB width of the spectrum ranges from about 10.65 to 10.92 MHz, i.e. the width is about 270kHz. This suggests that the filter is of the 280kHz variety.
I found in the Murata filter sheet their CERAFIL "Standard Lead Type SFELF Series". They come in 180, 230, and 280 kHz widths:
And their frequency response:
It appears the FA type has a pretty similar response like the SFC 10.7MA. I will obtain a few and report back after I put them in!
I got the new filters (SFELF10M7FA00 from Digikey). They have the 280kHz band pass. The first thing I did was to measure their frequency response and compare them with the old filter. This is what I got:
The old filter is shown in black, and the new one is in red. It is obvious that they do not have the same center frequency. The new one is much better centered at 10.7 MHz. I was worried that using these new filters would result in an increased loss in the IF stage. To test whether the shifted pass window would affect the gain in the IF stage, I first replaced the first two filters between the initial two amplification stages (TR5 and TR6). Before the exchange I measured the 10.7 MHz amplitude at the collector of TR5 (point B) and after passing through the filters at the base of TR6. When tuned into a strong station (>4 on the meter) then the TR5 collector signal was about 250 mV and after the filters I measured about 80 mV.
After exchanging the two filters I measured pretty much the same values. This gave me confidence that replacing the third (broken) filter should work. I did, and the radio plays now loud and clear. Also the S-lamps are now much more responsive...no surprise considering that they get their signal after this filter. The amplitude at point D after the filter measures now about 2.5 V compared to only about 50-70 mV before, i.e. a significant improvement in signal strength resulted from exchanging the filter.
However, a new phenomenon occurred: Now the stereo light is always on, even between stations. I measured the 38kHz subcarrier signal at the collector of TR21:
This photo shows the signal when a strong station is tuned in (it is a bit stronger than what the manual suggests - 12.4 vs. 10V peak-to-peak voltage):
This is what I measured in-between stations (only noise in the speakers, meter at ~2)
The signal fluctuates, but s still pretty strong. TR22, which seems to control the stereo light, is always on no matter if a station is tuned or not.
When I pull the antenna, then the stereo light does go off (strong stations are still audible at a very low level, but the meter only shows a level of approx.1).
I wonder: Are today's stations just too close/strong? I looked at the relevant adjustment steps in the "adjustment plan" of the manual, but it seems the decoder is in pretty good alignment...the subcarrier is too strong if anything.