A question of natural bass? - UK Vintage Radio Repair and Restoration Discussion Forum

YC-156 · 28th Feb 2006, 9:42 pm

Fellow valve heads,

The question for this evening: What, in general, is the bass reproduction actually like in better, late fifties VHF capable sets?

As some of you will recall, then I restored a Grundig 3033/W3D last year. However I also complained about the audio, excessive bass in particular. It seems I'm not the only one with that problem.

I don't know is Stewart found a problem in his radiogram, but today I decided to do a bit of detective work, trying to be as objective as possible in the process.

I started out by simply sweeping the frequency response of the two stage audio amp (EABC80 + EL84) at low volume, feeding a function generator into the gram socket and hooking the scope across the connector for the external speaker (which is always in parallel with the four speakers inside the set).

What I found surprised me a bit. With feedback enabled and the bass and treble adjustments set to 'neutral', the amp was amazingly even in gain up to 12-14KHz or so. The gain was perhaps dropping off slightly above 10KHz. I cannot hear much above that range, so that seemed OK.

Making a few, deliberate shorts in the feedback path showed the expected response of a feedback-less amp: Gain continuously dropping down from a maximum of a few hundred hertz.

However there was one anomaly: At low bass, starting around 200Hz going down, the response quickly rises to a peak around 80Hz before dropping down again to near zero around 25Hz. The peak was about a factor of 4 in voltage, or 12dB.

Hmm...?

Wearing my thinking cap I sat down to delve in dusty tomes from the ancient ones (Radiotron Designer's Handbook, RCA, 1954), and revisited the sections about loudness. What it said, audiophools be damned, was that for the audio impression of reproduced music to be identical to a live performance, the frequency response of an audio amp must have a certain amount of volume dependent bass boost, maxing out at around 12 dB centered at 75Hz. The reason being that the sensitivity of human hearing drops dramatically below 1KHz, especially at relatively low playback volumes.

Since most live performances are way louder than most radio playback, the radio must compensate accordingly by providing the missing bass, if the perceived sound balance is to match a live performance.

To put it differently: The frequency response of human hearing depends on the volumen. It is never completely even and only nearly so at extremely high volumes. Compensating for this effect is the purpose of the loudness control, which is part of the volume control on the 3033.

Hmmm...!

Next step was to grab the best set of headphones the house had to offer, my Sennheiser HD600s, and make a bit of wire to feed the audio signal from the CD player into the gram socket of the 3033.

At this point it has to be mentioned that these headphones are a very, Exceptionally, EXTREMELY good set of cans IMNSHO. I can hear things when using these that I cannot find on my CDs any other way. Particularly the bass, treble and midrange is worth listening to.

After an hours worth or so of back and forth between the Grundig and the Sennheiser cans, listening to the most 'extreme' music in my collection (don't ask, there is probably a law against playing it through vintage valve amps...), I am forced to conclude that the RCA and Grundig engineers knew what they were doing.

To within the limits of a 50 year old set, my ears and my living room, then the audio balance impression of the 3033 is quite close to that of the Hi-Fi headphones at 'reasonable' listening volumes for music, ratteling furniture or not. It appears to be a trick of the mind when I feel there is too much bass from the radio. Well, in a way there is, since the wooden chassis and the room surrounding the radio can only barely cope with the amount present.

This appears to be a simple problem of scale. My objective mind wants to claim that if I had the sound pressure impressions of the headphones filling the room, then the windows would be at the point of shaking out of their frames. It just so happens that todays manufacturers of consumer electronics apparently have decided consumers don't like their bass en naturel, and thus most modern Hi-Fi equipment sounds tame compared to a simple set of properly balanced headphones.

I would be very interested in hearing the experiences in this regard from my fellow valve set enthusiasts. I cannot recall hearing about this issue, or the uneven audio passband response, before, save for the report from Stewart.

Yours for powerful and natural bass response.

Frank N.

paulsherwin · 28th Feb 2006, 10:46 pm

There's a large amount of fashion involved in what constitutes good sound reproduction, and fashions change. The whole area of loudness contouring has been much discussed over the last 50 years, and the consensus since the late 70s has been that it is a Bad Thing. High fidelity systems should reproduce, as closely as possible, the sound heard by the recording engineer in the studio. This means high levels of accuracy and a flat frequency response.

People didn't always think this way, and in the 50s and 60s they liked a *lot* of warm, woolly, muddy bass. This possibly dates back all the way to the replacement of horn speakers with paper cones. It was quite normal for people to listen to radios with the tone control turned fully in the 'bass' direction. My Ecko A277, a reasonable FM only set from about 1956, only produces what I think of as acceptable levels of treble with the tone control almost completely at the 'treble' end.

Ordinary people who don't know anything about hifi still seem to like a lot of bass boost, but that's because they're ignorant

Best regards, Paul

YC-156 · 1st Mar 2006, 8:52 am

Thanks Paul, that makes perfect sense, though I'm not sure if this explains everything.

Judging from a quick browsing of some of the circuits for some 'high end' Grundig sets from the late fifties, then a loudness control was a very common feature. That makes me wonder why I only heard about the 'problem' now, after several years worth of valve set restorer's apprenticeship.

Do people actually either like, or do they believe the strong bass is natural and as it should be?
Few sets of this complexity is actually being used?
Emperor's new clothes? ("This is supposed to be a good set, so by definition I like it!")
Why is this not mentioned in any of the FAQs that I have read?

Even more interesting, then I discovered by accident how to disable the loudness, as I had to repair the loudness tap on the volume potentiometer of the 3033. When the tap was disconnected, then I found the radio to be very dull and boring, bordering on 'tinny'.

One of my early suspicions was that the bass boost was there to offset a low efficiency in the reproduction of low frequencies in the internal speaker units. After having seen the audio balance of the audio amp across the spectrum, then I suspect this may be the reason behind the low treble reproduction (which I agree is a bit of a problem). The amp itself does provide almost equal amounts of treble power to well over 10KHz.

But all this just meakes me wonder how the 'flat as a pancake' amps and their makers deal with the loudness phenomenon of human hearing, which is quite real. Everyone, who have heard them, seems to love the HD600s, and noone complains about excessive bass. So the explanation must have something to do also with habits and psychology, and what we expect to hear in different situations.

For me the solution will probably be that I back off the bass boost a bit by fiddling with the size of the loudness tap capacitor. 'Perfectly flat' will probably not sound that great in this set IMHO, even if that statement marks me ignorant.

Best regards

Frank N.

paulsherwin · 1st Mar 2006, 12:45 pm

Your point about comparing headphone listening and speaker listening is interesting, but the sound contour of headphones does vary a lot, even with expensive high end phones. My old Sony MDR-V7s also have quite a bassy sound while many Sennheisers I've tried sound a bit bass light (I don't know yours though).

The attitude of the hifi community towards loudness contouring is that a recording of an orchestra played quietly should sound like an orchestra heard from a distance, so loudness compensation is not required. There is also the issue that loudness contouring (and any frequency dependent gain modification) muddies the sound and makes it more difficult to achieve accurate reproduction. The Marantz PM-66SE which I use as my main amplifier has no tone controls, and this is common with serious hifi amplifiers even at the bottom end of the market.

I think your experience with valve radios has been biased by concentrating on expensive high end sets, and assuming that these sets have been deliberately designed to deliver high fidelity sound with a loudness contour. The fact is that most large sets from the 40s and 50s have a similar sound, it's just that your Grundigs are high quality examples. Manufacturers made them sound like that because it sold radios

The general public had no real awareness of the concept of high fidelity sound until the 70s. They would describe a radio or record player as having a 'nice tone' or 'good tone', which most people find dull and over-bassy today. When middle aged people bought their first stereo systems in the 70s it was quite common for them to operate them with the bass control at maximum and treble at minimum!

Best regards, Paul

Steve_P · 1st Mar 2006, 5:12 pm

Not always the case. In some cases this is a fault. I was repairing a Murphy Radiogram (well, helping Mike Barker repair one!) in 2003 and he actually had this problem with Excessive Bass.

Check the power line - not a quick check, but really check it.

Put in a Brand New rectifier. BRAND NEW. See if this reduces the bass. Seriously. Also, check the O/P Transformer for a shorted turn or two.

A few volts down on the HT can actually cause this.

Regards,

Steve P.

Stewart · 1st Mar 2006, 8:14 pm

Hmmm, this is all very interesting....

I've yet to return to my radiogram that helped raise this topic.

Even though I'm confident that back when it was new a warm rich bass was pretty much what makers aimed for and buyers expected, I'm also confident that there is a problem on my set! As I suggested in my original thread I believe this to be down to an error by me when fitting new caps. I've got some diagrams now to help me so when I have the time to remove the chassis again (and its not done lightly!) I expect to find I'm a factor of ten out on one of my replacements.

Even then I imagine the set will still give more than enough bass, but we will see!

SteveP comments about the power supply are interesting. The HT rail on my set is down from 272V on the diagram to about 258V...I would never have expected that to make much difference - and confess I can't quite see how it does - but who am I to question the combined efforts of Steve P & Mike B! So...something else for me to try if I can't find any cap problems. Thoughts of a failing output transformer frighten me - I've re-wound a few but this one has more taps than the Plumb Centre!

With regard to the measurements made on the output of the Grundig table set that started this thread, I wonder if measurements with a sound level meter would have been instructive? Measuring the output from the amplifier is surely only part of the story? The colouration added by the speakers - and the cabinet - surely need to be considered too? With valve equipment cooling vents would probably have had priority over acoustic properties?

When and if I get my set sorted I may have a play along such lines. Sadly although I still have the sound level meter I no longer have access to an anachoic chamber...but results in a typical room would be more useful anyway.

Stewart

YC-156 · 1st Mar 2006, 10:00 pm

Paul,

The HD600s have a very even Sound Pressure Level spectrum, there is a curve on Sennheiser's homepage.

Your comment about listening to an orchestra from a distance was to me a novel way of looking at the problem. I have a hard time coming up with a counter argument to that.

Nor is it any deep secret that I might have a few additional sets by Grundig, and the 3033 just might be the least complicated of the lot. However only the 3033 is in fully working order, and I don't have any useful basis for comparison. The other sets I have restored have all been low end AM only configurations.

That said, then I'm not sure I agree with your suggestion about my expectations regarding the performance of my sets. The core issue here I believe is that my personality is that of being an optimizer. I cannot help trying to make my sets perform as well as technically possible. All these exercises have the single goal of trying to determine if and when I'm done restoring a set. If the bass response of the 3033 is factory original, then nothing needs fixing. If it is faulty however...

There is still the small issue of the IF response and the proper adjustment of the ratio detector, but I'm working on those as well.

Steve,

The rectifier in the 3033 is new, 4x 1N4007 plus 2x 600/5W ohm dropper resistors. All paper and electrolytic caps are new as well. I will have a closer look at the HT rail in the near future.

The set is on a Variac, so it would be easy for me to increase the mains voltage a bit and see what happens even though I, like Stewart, cannot see how this could make a difference.

How do you check an O/P transformer for a single shorted turn or two? I'm not being sarcastic, I genuinely would like to know.

Stewart,

A swept sound level measurement would definitely have been in order, but I don't have one of those in my junkbox.

I also worried about the non-constant impedance of the speaker voice coils and the resonances of the chassis. But the raw voltage measurement was the best I could do at the moment.

I would definitely like to hear about your results if you get the opportunity to do a SPL sweep of you radiogram. Remember that a wide open field away from fixed structures is a passable substitute for an anachoic chamber, assuming you can find a mains outlet nearby to power the set.

Best regards Frank N.

paulsherwin · 1st Mar 2006, 10:13 pm

Quote:

Originally Posted by YC-156

That said, then I'm not sure I agree with your suggestion about my expectations regarding the performance of my sets. The core issue here I believe is that my personality is that of being an optimizer. I cannot help trying to make my sets perform as well as technically possible. All these exercises have the single goal of trying to determine if and when I'm done restoring a set. If the bass response of the 3033 is factory original, then nothing needs fixing. If it is faulty however...

I am sure the rising bass response is original. Most 1950s single ended valve radios have some sort of negative feedback mechanism that causes an elevated bass response at low volume levels. I suspect this has little to do with loudness contouring and more to do with maximizing output volume at high volume levels, where an elevated bass response would cause clipping distortion (and make the cabinet rattle

)

Paul

YC-156 · 5th Mar 2006, 11:02 am

Just a quit note on the progress I have had with dampening the height of the 'loudness hump', perhaps someone else can make use of my approach, if not the solution.

The good news is that I believe I have found a completely satisfactory, though perhaps non-intuitive, solution. The bad news is that it will probably not apply to most other sets, as it involves tweaking the feedback network of the audio amp.

Realizing that it would take a man better than me to do a proper analytical analysis of the audio amp, I resorted to a combination of pen, paper, pocketcalculator and SPICE simlulations on the computer.

What I learned was that the loudness hump comes about as a seemingly conscious combination of two distinct frequency response curves: That of the core amplifier, and the one resulting from the negative feedback/loudness network.

The core amplifier contains no less than three high pass elements, ensuring a rather steep, low frequency cut-off somewhere below 100Hz. The coupling cap between the EABC80 triode and the EL84 grid is actually only just barely large enough for the job. When it is combined with the anode resistor of the triode, it forms a high-pass filter with Fo = 36Hz.

The EL84 cathode bypass cap is only 50uF, sitting across a 170 ohm resistor. At low frequencies the capacitor will be ineffective in decoupling the cathode for AF, and the gain of the output amp will suffer as a result of the increased, local negative feedback provided by the cathode resistor. The -3dB point here is 18Hz, not too bad.

Finally it appears obvious that the output transformer seriously begins to run out of inductance at low audio frequencies.

Next step was wrapping the feedback and loudness network around a pair of Op-Amps in SPICE. What I got from that was a flat passband, except a steep increase in gain/reduction in feedback as the frequency went below a few hundred hertz.

Knowing all of the above, the solution seemed to be to alter the feedback response to increase the feedback at low frequencies, thus lowering the gain of the whole circuit in that region. Surprisingly in this case the best way I found for accomplishing this, was to increase the size of the loudness coupling cap. The simulation showed the shape and placement of the loudness hump would remain unchanged by doing so.

A quick test was performed on the radio, and increasing the cap from 22nF to 47nF made everything much more agreeable. It is now trivially possible to regulate the bass to a proper level, or even below 'neutral'.

Simply removing the loudness coupling cap, as I had previously suggested, is not a good idea. This will certainly increase the volume of the mid range in comparison to the bass hump, but at the expense of ruining the overall linearity of the frequency response. The treble suffers in particular.

So there you have it. Not a generic solution, but alas the best I can do for now.

Best regards

Frank N.

paulsherwin · 5th Mar 2006, 1:25 pm

Glad to hear you've found a solution you're happy with. Many radios have some frequency selective negative feedback between the output valve and the bottom of the volume control, and adjusting component values here will change the amount of bass boost at low volume levels.

Some people object to this because it reduces the radio's originality, but most people would agree that it's more important to have a radio that sounds good to a modern listener. Paul Stenning went as far as adding this arrangement to a radio that didn't originally have it: http://www.vintage-radio.com/recent-...ps-b2g81u.html

Best regards, Paul

Tim · 5th Mar 2006, 9:05 pm

Hi all.
I am currently working on a Grundig 3028, and was worried there was insufficient Bass! (Electronics all done, including ALL low value capacitors. HT electrolytics are original and perform and test fine.
HT OK. BRAND new EL84 and newly rewound O/P transformer.
Fourway tone controls seem to work, and sound is OK, I would just be happier with more Bass.
Any one else have one of these sets?
Cheers
Tim

Gordon · 5th Mar 2006, 9:24 pm

Quote:

Originally Posted by YC-156

However there was one anomaly: At low bass, starting around 200Hz going down, the response quickly rises to a peak around 80Hz before dropping down again to near zero around 25Hz. The peak was about a factor of 4 in voltage, or 12dB.

Could this be the natural (self) resonance of the loudspeaker? Do you get the same result with a resistive load?
Gordon.

YC-156 · 5th Mar 2006, 10:05 pm

Quote:

Originally Posted by Gordon

Could this be the natural (self) resonance of the loudspeaker? Do you get the same result with a resistive load?
Gordon.

I didn't test with a resistive load, would have required a bit of disassembly. Yet if you read my previous follow-up, I suspect the frequency response of the feedback network can explain the phenomenon. Also I can change the amplitude by fooling with the component values. This should not have had any influence, if the speaker unit was to blame.

Best regards

Frank N.

YC-156 · 6th Mar 2006, 10:00 pm

Grundig engineers: 1 -- Frank: 0

Sometimes you cannot see the forest for all the trees.

Today I hooked the radio up to the cable network, and even though the VHF tuner only officially tracks to 100MHz, it just barely manages to tune Denmark's Radio Classic Channel (in English) at 100,8MHz. As I type this, the station co-transmits a Mozart concert from London, courtesy of the BBC.

Never mind my talk about excessive bass and hissing sibilances. What I failed to realize was that I have been completely spoiled by low grade sand state FM receivers and local, ad driven headbanger radio stations, where the compressor has one setting: Overload.

Once I tuned in a proper Hi-Fi station with a sound engineer at the helm, and with the music playing, all my nitpickings went away. Yes, the bass is rumbling when the male commentators speak, the s'es still his. But it just doesn't matter when the music is playing. If I turn down the bass, some of the instruments in the orchestra turn up missing. When the women sing, the sound is Just Right(TM).

I no longer have any reservations regarding the set's ability to reproduce a properly balanced, high quality FM transmission. Maybe I should have. Maybe I'm half deaf. Maybe the audiophools will want my head on a platter. Maybe I have no idea what I'm talking about.

But right here and now I cannot think of a single thing I want improved, no matter what era of receiver I could choose from.

End of rant.

Frank N.

Nickthedentist · 6th Mar 2006, 10:10 pm

Quote:

Originally Posted by YC-156

But right here and now I cannot think of a single thing I want improved, no matter what era of receiver I could choose from.

So, a happy ending, then!

Quote:

Originally Posted by YC-156

...it just barely manages to tune Denmark's Radio Classic Channel (in English) at 100,8MHz. As I type this, the station co-transmits a Mozart concert from London, courtesy of the BBC.

I'm listening to it on-line now, well worth knowing about. Coincidentally, BBC Radio 3 is also playing Mozart, a piano concerto I think, in the very same key!
Nick.

paulsherwin · 6th Mar 2006, 11:02 pm

Quote:

Originally Posted by Nickthedentist

I'm listening to it on-line now, well worth knowing about. Coincidentally, BBC Radio 3 is also playing Mozart, a piano concerto I think, in the very same key!
Nick.

I think the same concert was broadcast live on R3 and Denmark Radio.

Paul

YC-156 · 6th Mar 2006, 11:39 pm

Quote:

Originally Posted by paulsherwin

I think the same concert was broadcast live on R3 and Denmark Radio.

You are probably right about that. After the concert the commentator here listed the countries, where the transmission was heard live. Sounded pretty much like all of Europe plus Canada.

Frank.

Off topic fun fact: I have no stations on cable worth listening to on any frequency higher than DR Klassisk at 100,8MHz. Meaning that I don't in the slightest mind vintage sets with a 100MHz frequency limit.

The other end of the scale has BBC World Service at 87,7MHz, which the radio also tunes without problems.

28th Feb 2006, 9:42 pm	#1
YC-156 Retired Dormant Member Join Date: May 2005 Location: Aarhus, Denmark Posts: 281	A question of natural bass? Fellow valve heads, The question for this evening: What, in general, is the bass reproduction actually like in better, late fifties VHF capable sets? As some of you will recall, then I restored a Grundig 3033/W3D last year. However I also complained about the audio, excessive bass in particular. It seems I'm not the only one with that problem. I don't know is Stewart found a problem in his radiogram, but today I decided to do a bit of detective work, trying to be as objective as possible in the process. I started out by simply sweeping the frequency response of the two stage audio amp (EABC80 + EL84) at low volume, feeding a function generator into the gram socket and hooking the scope across the connector for the external speaker (which is always in parallel with the four speakers inside the set). What I found surprised me a bit. With feedback enabled and the bass and treble adjustments set to 'neutral', the amp was amazingly even in gain up to 12-14KHz or so. The gain was perhaps dropping off slightly above 10KHz. I cannot hear much above that range, so that seemed OK. Making a few, deliberate shorts in the feedback path showed the expected response of a feedback-less amp: Gain continuously dropping down from a maximum of a few hundred hertz. However there was one anomaly: At low bass, starting around 200Hz going down, the response quickly rises to a peak around 80Hz before dropping down again to near zero around 25Hz. The peak was about a factor of 4 in voltage, or 12dB. Hmm...? Wearing my thinking cap I sat down to delve in dusty tomes from the ancient ones (Radiotron Designer's Handbook, RCA, 1954), and revisited the sections about loudness. What it said, audiophools be damned, was that for the audio impression of reproduced music to be identical to a live performance, the frequency response of an audio amp must have a certain amount of volume dependent bass boost, maxing out at around 12 dB centered at 75Hz. The reason being that the sensitivity of human hearing drops dramatically below 1KHz, especially at relatively low playback volumes. Since most live performances are way louder than most radio playback, the radio must compensate accordingly by providing the missing bass, if the perceived sound balance is to match a live performance. To put it differently: The frequency response of human hearing depends on the volumen. It is never completely even and only nearly so at extremely high volumes. Compensating for this effect is the purpose of the loudness control, which is part of the volume control on the 3033. Hmmm...! Next step was to grab the best set of headphones the house had to offer, my Sennheiser HD600s, and make a bit of wire to feed the audio signal from the CD player into the gram socket of the 3033. At this point it has to be mentioned that these headphones are a very, Exceptionally, EXTREMELY good set of cans IMNSHO. I can hear things when using these that I cannot find on my CDs any other way. Particularly the bass, treble and midrange is worth listening to. After an hours worth or so of back and forth between the Grundig and the Sennheiser cans, listening to the most 'extreme' music in my collection (don't ask, there is probably a law against playing it through vintage valve amps...), I am forced to conclude that the RCA and Grundig engineers knew what they were doing. To within the limits of a 50 year old set, my ears and my living room, then the audio balance impression of the 3033 is quite close to that of the Hi-Fi headphones at 'reasonable' listening volumes for music, ratteling furniture or not. It appears to be a trick of the mind when I feel there is too much bass from the radio. Well, in a way there is, since the wooden chassis and the room surrounding the radio can only barely cope with the amount present. This appears to be a simple problem of scale. My objective mind wants to claim that if I had the sound pressure impressions of the headphones filling the room, then the windows would be at the point of shaking out of their frames. It just so happens that todays manufacturers of consumer electronics apparently have decided consumers don't like their bass en naturel, and thus most modern Hi-Fi equipment sounds tame compared to a simple set of properly balanced headphones. I would be very interested in hearing the experiences in this regard from my fellow valve set enthusiasts. I cannot recall hearing about this issue, or the uneven audio passband response, before, save for the report from Stewart. Yours for powerful and natural bass response. Frank N.

28th Feb 2006, 10:46 pm	#2
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,957	Re: A question of natural bass? There's a large amount of fashion involved in what constitutes good sound reproduction, and fashions change. The whole area of loudness contouring has been much discussed over the last 50 years, and the consensus since the late 70s has been that it is a Bad Thing. High fidelity systems should reproduce, as closely as possible, the sound heard by the recording engineer in the studio. This means high levels of accuracy and a flat frequency response. People didn't always think this way, and in the 50s and 60s they liked a lot of warm, woolly, muddy bass. This possibly dates back all the way to the replacement of horn speakers with paper cones. It was quite normal for people to listen to radios with the tone control turned fully in the 'bass' direction. My Ecko A277, a reasonable FM only set from about 1956, only produces what I think of as acceptable levels of treble with the tone control almost completely at the 'treble' end. Ordinary people who don't know anything about hifi still seem to like a lot of bass boost, but that's because they're ignorant Best regards, Paul

1st Mar 2006, 8:52 am	#3
YC-156 Retired Dormant Member Join Date: May 2005 Location: Aarhus, Denmark Posts: 281	Re: A question of natural bass? Thanks Paul, that makes perfect sense, though I'm not sure if this explains everything. Judging from a quick browsing of some of the circuits for some 'high end' Grundig sets from the late fifties, then a loudness control was a very common feature. That makes me wonder why I only heard about the 'problem' now, after several years worth of valve set restorer's apprenticeship. Do people actually either like, or do they believe the strong bass is natural and as it should be? Few sets of this complexity is actually being used? Emperor's new clothes? ("This is supposed to be a good set, so by definition I like it!") Why is this not mentioned in any of the FAQs that I have read? Even more interesting, then I discovered by accident how to disable the loudness, as I had to repair the loudness tap on the volume potentiometer of the 3033. When the tap was disconnected, then I found the radio to be very dull and boring, bordering on 'tinny'. One of my early suspicions was that the bass boost was there to offset a low efficiency in the reproduction of low frequencies in the internal speaker units. After having seen the audio balance of the audio amp across the spectrum, then I suspect this may be the reason behind the low treble reproduction (which I agree is a bit of a problem). The amp itself does provide almost equal amounts of treble power to well over 10KHz. But all this just meakes me wonder how the 'flat as a pancake' amps and their makers deal with the loudness phenomenon of human hearing, which is quite real. Everyone, who have heard them, seems to love the HD600s, and noone complains about excessive bass. So the explanation must have something to do also with habits and psychology, and what we expect to hear in different situations. For me the solution will probably be that I back off the bass boost a bit by fiddling with the size of the loudness tap capacitor. 'Perfectly flat' will probably not sound that great in this set IMHO, even if that statement marks me ignorant. Best regards Frank N.

1st Mar 2006, 12:45 pm	#4
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,957	Re: A question of natural bass? Your point about comparing headphone listening and speaker listening is interesting, but the sound contour of headphones does vary a lot, even with expensive high end phones. My old Sony MDR-V7s also have quite a bassy sound while many Sennheisers I've tried sound a bit bass light (I don't know yours though). The attitude of the hifi community towards loudness contouring is that a recording of an orchestra played quietly should sound like an orchestra heard from a distance, so loudness compensation is not required. There is also the issue that loudness contouring (and any frequency dependent gain modification) muddies the sound and makes it more difficult to achieve accurate reproduction. The Marantz PM-66SE which I use as my main amplifier has no tone controls, and this is common with serious hifi amplifiers even at the bottom end of the market. I think your experience with valve radios has been biased by concentrating on expensive high end sets, and assuming that these sets have been deliberately designed to deliver high fidelity sound with a loudness contour. The fact is that most large sets from the 40s and 50s have a similar sound, it's just that your Grundigs are high quality examples. Manufacturers made them sound like that because it sold radios The general public had no real awareness of the concept of high fidelity sound until the 70s. They would describe a radio or record player as having a 'nice tone' or 'good tone', which most people find dull and over-bassy today. When middle aged people bought their first stereo systems in the 70s it was quite common for them to operate them with the bass control at maximum and treble at minimum! Best regards, Paul

1st Mar 2006, 5:12 pm	#5
Steve_P Dekatron Join Date: Sep 2003 Location: Bolton, Greater Manchester, UK. Posts: 6,644	Re: A question of natural bass? Not always the case. In some cases this is a fault. I was repairing a Murphy Radiogram (well, helping Mike Barker repair one!) in 2003 and he actually had this problem with Excessive Bass. Check the power line - not a quick check, but really check it. Put in a Brand New rectifier. BRAND NEW. See if this reduces the bass. Seriously. Also, check the O/P Transformer for a shorted turn or two. A few volts down on the HT can actually cause this. Regards, Steve P. __________________ If we've always had it, why is the Car Boot open? You're not sneaking another Old TV in are you...?

1st Mar 2006, 8:14 pm	#6
Stewart Retired Dormant Member Join Date: Jul 2003 Location: N.E. Surrey, UK. Posts: 361	Re: A question of natural bass? Hmmm, this is all very interesting.... I've yet to return to my radiogram that helped raise this topic. Even though I'm confident that back when it was new a warm rich bass was pretty much what makers aimed for and buyers expected, I'm also confident that there is a problem on my set! As I suggested in my original thread I believe this to be down to an error by me when fitting new caps. I've got some diagrams now to help me so when I have the time to remove the chassis again (and its not done lightly!) I expect to find I'm a factor of ten out on one of my replacements. Even then I imagine the set will still give more than enough bass, but we will see! SteveP comments about the power supply are interesting. The HT rail on my set is down from 272V on the diagram to about 258V...I would never have expected that to make much difference - and confess I can't quite see how it does - but who am I to question the combined efforts of Steve P & Mike B! So...something else for me to try if I can't find any cap problems. Thoughts of a failing output transformer frighten me - I've re-wound a few but this one has more taps than the Plumb Centre! With regard to the measurements made on the output of the Grundig table set that started this thread, I wonder if measurements with a sound level meter would have been instructive? Measuring the output from the amplifier is surely only part of the story? The colouration added by the speakers - and the cabinet - surely need to be considered too? With valve equipment cooling vents would probably have had priority over acoustic properties? When and if I get my set sorted I may have a play along such lines. Sadly although I still have the sound level meter I no longer have access to an anachoic chamber...but results in a typical room would be more useful anyway. Stewart

1st Mar 2006, 10:00 pm	#7
YC-156 Retired Dormant Member Join Date: May 2005 Location: Aarhus, Denmark Posts: 281	Re: A question of natural bass? Paul, The HD600s have a very even Sound Pressure Level spectrum, there is a curve on Sennheiser's homepage. Your comment about listening to an orchestra from a distance was to me a novel way of looking at the problem. I have a hard time coming up with a counter argument to that. Nor is it any deep secret that I might have a few additional sets by Grundig, and the 3033 just might be the least complicated of the lot. However only the 3033 is in fully working order, and I don't have any useful basis for comparison. The other sets I have restored have all been low end AM only configurations. That said, then I'm not sure I agree with your suggestion about my expectations regarding the performance of my sets. The core issue here I believe is that my personality is that of being an optimizer. I cannot help trying to make my sets perform as well as technically possible. All these exercises have the single goal of trying to determine if and when I'm done restoring a set. If the bass response of the 3033 is factory original, then nothing needs fixing. If it is faulty however... There is still the small issue of the IF response and the proper adjustment of the ratio detector, but I'm working on those as well. Steve, The rectifier in the 3033 is new, 4x 1N4007 plus 2x 600/5W ohm dropper resistors. All paper and electrolytic caps are new as well. I will have a closer look at the HT rail in the near future. The set is on a Variac, so it would be easy for me to increase the mains voltage a bit and see what happens even though I, like Stewart, cannot see how this could make a difference. How do you check an O/P transformer for a single shorted turn or two? I'm not being sarcastic, I genuinely would like to know. Stewart, A swept sound level measurement would definitely have been in order, but I don't have one of those in my junkbox. I also worried about the non-constant impedance of the speaker voice coils and the resonances of the chassis. But the raw voltage measurement was the best I could do at the moment. I would definitely like to hear about your results if you get the opportunity to do a SPL sweep of you radiogram. Remember that a wide open field away from fixed structures is a passable substitute for an anachoic chamber, assuming you can find a mains outlet nearby to power the set. Best regards Frank N.

5th Mar 2006, 11:02 am	#9
YC-156 Retired Dormant Member Join Date: May 2005 Location: Aarhus, Denmark Posts: 281	Re: A question of natural bass? Just a quit note on the progress I have had with dampening the height of the 'loudness hump', perhaps someone else can make use of my approach, if not the solution. The good news is that I believe I have found a completely satisfactory, though perhaps non-intuitive, solution. The bad news is that it will probably not apply to most other sets, as it involves tweaking the feedback network of the audio amp. Realizing that it would take a man better than me to do a proper analytical analysis of the audio amp, I resorted to a combination of pen, paper, pocketcalculator and SPICE simlulations on the computer. What I learned was that the loudness hump comes about as a seemingly conscious combination of two distinct frequency response curves: That of the core amplifier, and the one resulting from the negative feedback/loudness network. The core amplifier contains no less than three high pass elements, ensuring a rather steep, low frequency cut-off somewhere below 100Hz. The coupling cap between the EABC80 triode and the EL84 grid is actually only just barely large enough for the job. When it is combined with the anode resistor of the triode, it forms a high-pass filter with Fo = 36Hz. The EL84 cathode bypass cap is only 50uF, sitting across a 170 ohm resistor. At low frequencies the capacitor will be ineffective in decoupling the cathode for AF, and the gain of the output amp will suffer as a result of the increased, local negative feedback provided by the cathode resistor. The -3dB point here is 18Hz, not too bad. Finally it appears obvious that the output transformer seriously begins to run out of inductance at low audio frequencies. Next step was wrapping the feedback and loudness network around a pair of Op-Amps in SPICE. What I got from that was a flat passband, except a steep increase in gain/reduction in feedback as the frequency went below a few hundred hertz. Knowing all of the above, the solution seemed to be to alter the feedback response to increase the feedback at low frequencies, thus lowering the gain of the whole circuit in that region. Surprisingly in this case the best way I found for accomplishing this, was to increase the size of the loudness coupling cap. The simulation showed the shape and placement of the loudness hump would remain unchanged by doing so. A quick test was performed on the radio, and increasing the cap from 22nF to 47nF made everything much more agreeable. It is now trivially possible to regulate the bass to a proper level, or even below 'neutral'. Simply removing the loudness coupling cap, as I had previously suggested, is not a good idea. This will certainly increase the volume of the mid range in comparison to the bass hump, but at the expense of ruining the overall linearity of the frequency response. The treble suffers in particular. So there you have it. Not a generic solution, but alas the best I can do for now. Best regards Frank N.

5th Mar 2006, 1:25 pm	#10
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,957	Re: A question of natural bass? Glad to hear you've found a solution you're happy with. Many radios have some frequency selective negative feedback between the output valve and the bottom of the volume control, and adjusting component values here will change the amount of bass boost at low volume levels. Some people object to this because it reduces the radio's originality, but most people would agree that it's more important to have a radio that sounds good to a modern listener. Paul Stenning went as far as adding this arrangement to a radio that didn't originally have it: http://www.vintage-radio.com/recent-...ps-b2g81u.html Best regards, Paul

5th Mar 2006, 9:05 pm	#11
Tim Dekatron Join Date: Jan 2003 Location: Bradford on Avon, Wiltshire, UK. Posts: 3,310	Re: A question of natural bass? Hi all. I am currently working on a Grundig 3028, and was worried there was insufficient Bass! (Electronics all done, including ALL low value capacitors. HT electrolytics are original and perform and test fine. HT OK. BRAND new EL84 and newly rewound O/P transformer. Fourway tone controls seem to work, and sound is OK, I would just be happier with more Bass. Any one else have one of these sets? Cheers Tim __________________ "Nothing is as dangerous as being too modern;one is apt to grow old fashioned quite suddenly."

6th Mar 2006, 10:00 pm	#14
YC-156 Retired Dormant Member Join Date: May 2005 Location: Aarhus, Denmark Posts: 281	Re: A question of natural bass? Grundig engineers: 1 -- Frank: 0 Sometimes you cannot see the forest for all the trees. Today I hooked the radio up to the cable network, and even though the VHF tuner only officially tracks to 100MHz, it just barely manages to tune Denmark's Radio Classic Channel (in English) at 100,8MHz. As I type this, the station co-transmits a Mozart concert from London, courtesy of the BBC. Never mind my talk about excessive bass and hissing sibilances. What I failed to realize was that I have been completely spoiled by low grade sand state FM receivers and local, ad driven headbanger radio stations, where the compressor has one setting: Overload. Once I tuned in a proper Hi-Fi station with a sound engineer at the helm, and with the music playing, all my nitpickings went away. Yes, the bass is rumbling when the male commentators speak, the s'es still his. But it just doesn't matter when the music is playing. If I turn down the bass, some of the instruments in the orchestra turn up missing. When the women sing, the sound is Just Right(TM). I no longer have any reservations regarding the set's ability to reproduce a properly balanced, high quality FM transmission. Maybe I should have. Maybe I'm half deaf. Maybe the audiophools will want my head on a platter. Maybe I have no idea what I'm talking about. But right here and now I cannot think of a single thing I want improved, no matter what era of receiver I could choose from. End of rant. Frank N.