Improving the frequency response on a Roberts RM33

[stevehertz]

Having got my little Roberts RM33 working properly now, I'd like to be able to tweak the circuitry to provide a tad more bass and treble. There are no tone controls on the set. Yes, I realise that the small, elliptical speaker has its limitations, but nonetheless I'm sure the present 'plummy', yet not too nasty tone can be improved even if its at the cost of less output by way of reducing the middle part of the frequency response. Unfortunately I don't have the design capability to mess around with component values myself, but I'm sure lots of others do. Can anyone with a knowledge of this set make any suggestions please? - specific components and values named please, I can't work to generalisations . I have the schematic that is available here: http://www.service-data.com/section.php/7352/1/rm33

[G6Tanuki]

What are you using it to listen to?

Medium/Long-wave broadcasts are by nature subject to merciless HF-cut at source (to restrict the bandwidth in order to comply with 9KHz broadcast channel-spacing) and you're on to a loser if you're expecting to hear anything much above 5KHz.

On FM - there are possibilities to tweak the "De-emphasis" components after the discriminator [which impose a roll-off at high frequencies] in order to get better pass-through of treble.

Must admit, my daily-listener Roberts R606 spends its life with the bass control turned to 'min' and the treble turned to 'max'. I like my treble!

[Sidenote: I have historic explosion-induced hearing damage meaning anything below about 200Hz is always sounding horribly distorted, so for me it makes sense to minimise all audio-content below that point. Your ears may differ.]

[Synchrodyne]

Firstly, I’ll request immunity from prosecution as I am clearly not respecting the OP’s perfectly reasonable request for a quantitative solution. But then we don’t know how much bass lift and treble lift the OP wants to hear, nor how much the speaker would handle. But it seems to me that a Baxandall network set for mild bass lift and moderate treble lift might do the job, perhaps more so than say a middle tone control set for moderate middle reduction.

The RM33 schematic I am looking at has, in simple terms, the AF output from a TDA1071 IC (an elusive beast) feeding a TDA2611A audio output IC via a 100k volume control, with a switched de-emphasis capacitor for FM just after the TDA1071. I’d guess that the TDA1071 delivers somewhere around 200 to 300 mV of audio, not too low for additional processing. So one way of achieving the goal might be to interpose a unity gain Baxandall network with presettable treble and bass controls between the volume control and the TDA2611A. It would need to have a buffered input, as I don’t think one would want to have a 100k volume control looking into a low-impedance Baxandall network. Actually, the nominally 45k input impedance of the TDA2611A would probably modify the volume control curve, so probably the buffer should have the same input impedance to achieve the same effect. I’d guess that there’d be room in the box for a small board containing a Baxandall circuit with preset controls and buffer built around a suitable dual op-amp (TL072?) and configured for single-ended supply from the RM33 power supply. I am sure that others here could provide quantitative data for such a unit. It’s more elaborate, but I think more elegant than possible passive solutions and also readily adjustable via the presets, desirable I think because there will likely be a trade-off between what is wanted and what the speaker can handle.

My guess is that the TDA1071 delivers quite respectable quality audio not only on FM but also, within its bandwidth limits, on AM, likely not having that element of background distortion that is found quite often on transistor equipment, and which probably comes from the final IF stage and/or diode demodulator. (Even the Beolit 707, so good on FM, suffers a little from this problem on AM.) So maybe it’s worth installing a decent equalization circuit in the RM33.


Cheers,

[stevehertz]

Thanks so far guys. I listen to FM mainly, and some AM. In both cases the aforementioned bass and treble lift is required/desired. Look, I'm not after getting hifi out of this thing (the limitations of the speaker and everything else are clear), just a 'noticeable' improvement in the plummy frequency response by way of swapping or adding a few components internally. I don't listen loud, so that will obviously help when boosting the bass. Thanks again.

[Edward Huggins]

Perhaps the most non-interventionist solution might be to swap the speaker out for one with a longer throw cone. Is there room for for a 6" x 4" unit?

[mhennessy]

Originally Posted by Synchrodyne

The RM33 schematic I am looking at has, in simple terms, the AF output from a TDA1071 IC (an elusive beast) feeding a TDA2611A audio output IC via a 100k volume control, with a switched de-emphasis capacitor for FM just after the TDA1071.

That's different to the schematic (on Paul's DVD). That one uses the TDA5700, and has a transistor pre-amp prior to the volume control. Same output IC.

Therefore, before proceeding any further, I need to know which version Steve has, and then - crucially - I'd need to see which version my one is. Hopefully we have the same one.

[stevehertz]

Just checked, the chip in my RM33 is a TDA1071.

[Herald1360]

You could try a network like this (see pic) between original connection to top of volume pot and the pot itself. It should give about 4dB midrange suckout.

[mhennessy]

Frustratingly, I can't locate my RM33 (nor my RM30 - they're in the same box). I'll have another look tomorrow...

[stevehertz]

Fantastic, thanks Mark, I'll try that out in due course.

[mhennessy]

Credit to Chris, but I'd certainly expect to come up with something very similar

Once I've found mine and confirmed it's the same version, I'll measure it and see what can be sensibly be achieved. A passive loss network is the obvious option in this case - the version with the transistor pre would offer more flexibility, but we'll see what can be done.

Chris, there's a few passive between the IC and volume pot already (de-emphasis), so it would be fun to include those in the simulation..

[Herald1360]

Hi Mark,


I don't have access to a TDA1071 version circuit- Pauls CD only has the TDA5700 version. If you've got the other details I'll happily play with them unless you have a simulator to play with yourself.

[mhennessy]

Hi Chris,

The circuit of the (presumably later?) model can be found here: http://www.valve-radio.co.uk/literat...ervice-manual/

Feel free to play around - I won't give it any more thought until I've located mine and measured it - I want to see what the speaker is like

Steve, if you want to play, I'd recommend getting a breadboard, and 3 lengths of wire that will reach inside to wherever you put your breadboard. You need to intercept the top of the volume control, and also bring the ground connection out. Start with what Chris suggested, but expect to experiment. For example, if the LS has a spike in the 4-6kHz region (very common), then you'll want to reduce the value of C2 so that it starts to lift at a higher frequency. Once you get going, it's good fun!

Mark

[stevehertz]

Thanks Mark, understood. As I'm always busy with one thing and another (not VW) I may wait a while and see if you come across your RM33 and perform a frequency sweep to see what the status quo is. I'm in no hurry to get this done but it will be fun to see how/if it can be improved a tad. Cheers.

[Herald1360]

Hmmmm, I'll have a play. Looks like the whole setup is probably designed to produce a noise that someone thought was "right". The load on the volume control is probably rather less than 45k (the input impedance to the 2611A with pin 9 floating) so the setting of the volume control will vary the loading on my suggested network and change its effect depending on volume.


I think Steve may find a few of the cheap cap and resistor substitution gizmos worth playing with to see what sounds right to his ears!

Edit: Did anyone else think Volks Wagen rather than Vintage Wireless whilst reading post #14?

[stevehertz]

Originally Posted by Herald1360

Hmmmm, I'll have a play. Looks like the whole setup is probably designed to produce a noise that someone thought was "right".

Well yes, the plummy tone that it has is not particularly objectionable, but it does lack any form of bass, and the treble too could do with a lift. All within reason and the scope of the small elliptical speaker fitted. Like I say, not looking for a massive improvement, just a modicum of such to make it more enjoyable to listen to. And I don't want to get into 'big mods', changing the speaker, that kind of thing. Just some selective filtering will be nice.

[Herald1360]

OK, I've had another play.

See pic attached. Original circuit and modified circuits shown with response for both at 100% volume and 25% volume pot settings.

You can zoom in viewing the pdf to actually read the text!

[stevehertz]

Brilliant! I will give it a go, thanks!

[stevehertz]

Originally Posted by Herald1360

OK, I've had another play.

See pic attached. Original circuit and modified circuits shown with response for both at 100% volume and 25% volume pot settings.

You can zoom in viewing the pdf to actually read the text!

Chris, I've swapped the components as per your suggestion. The treble response is definitely 'up', but frankly I am not hearing much if any improvement in the bass area. In truth, the sound is now bordering on the proverbial tinny transistor set - kinda 2Kish, harsh. I don't mean to be rude saying that! but I'm sure you prefer to have accurate feedback for it to be meaningful. I'm thinking that if the bass could be brought up, the overall balance would be better.

Given that the theoretical curves seemed to have 'cracked it', I'm thinking that the problem is exactly that; those curves are theoretical and based on the electrical response ie taking no account of the loudspeaker? Maybe if Mark can take a look at what the speaker in the cabinet is doing, it may add another dimension to add to the number crunching and yield a better result.

[GeoffK]

The speaker coupling capacitor should be increased in capacity, 220uF or more will increase the lower frequency to the speaker. The input coupling 100nF on each side of the volume control could be increased to 150nF. And before anyone asks, no I don't know how many dB difference it will make, no doubt it could be worked out.