Wiring cartridge directly to preamp

[Nickthedentist]

Thanks, Synchrodyne, that's great!

[sp10mk11]

This is the circuit I built

Gary


[URL="http://users.belgacom.net/hamradio/schemas/Bass%20EQ%20pre-amp%20correction%20for%20high%20impedance%20cerami c%20crystal%20PU.gif"]

[Radio Wrangler]

Great!

Well worth preserving.

The Bailey-Burrows preamp was one I thought rather good back in the day. It was almost a shame that it arrived at around the time magnetic cartridges had started to become affordable, so many people were changing and the Deram and 9TAHC were no longer kings of the castle. They still worked as well as they ever did, but the new Shures on the scene promised better.

I rate this the best ceramic era preamp in WW, then Doug Self did the job for magnetics several years later.

Hugh Walker did a decent magnetic RIAA stage with three transistors, then Peter Baxendall did a better fit to the RIAA curve with an NE5534 opamp with an added R-C pole after the stage. Doug Self later came along with his 'DIY opamp' sort of preamp.

Wireless World used to be very fertile ground for interesting circuits. (Remember the ring-of-two reference, anyone?)

David

[G8HQP Dave]

Originally Posted by Synchrodyne

Anyway, given the difficulties and that source has been acknowledged, it would seem to be not unreasonable to attach the article itself (size-reduced), which I had downloaded earlier.

Thanks.

It is amusing to see that the author continues the long WW tradition of saying "everyone else has got it wrong!". Despite what he says, I still believe that high impedance loading is correct for a ceramic cartridge. Low impedance loading requires frequency correction which needs to be different for each cartridge. However, he usefully reminds us that flat frequency response and good efficiency can be mutually exclusive in a transducer, as the latter requires tight coupling between electrical and mechanical domains while the former often requires loose coupling.

[Synchrodyne]

Originally Posted by G8HQP Dave

It is amusing to see that the author continues the long WW tradition of saying "everyone else has got it wrong!"

Yes, there was an element of that. I wonder if it was part of the WW style guide of the era.

Clearly, Burrows preferred the low impedance-with-preamplifier-as-integrator option, and if nothing else his article presented a detailed treatment of this less-travelled road. Anyway, Linsley Hood had pretty much staked out the high-impedance territory in 1969, and Burrows did acknowledge his work in a positive way.

Originally Posted by G8HQP Dave

Low impedance loading requires frequency correction which needs to be different for each cartridge.

True, although a given single frequency correction might serve reasonably well for a range of cartridges. Quad chose that pathway for its Quad 33 control unit. The C1 position of the disc input board presented 100 k impedance to the cartridge, and the frequency correction was said to be suitable for cartridges with internal capacitance in the range 450 to 900 pF, and with an output range (at 1 kHz and 1 mV/cm/s) of 25 to 80 mV. At a guess I’d say that the frequency correction was exactly right at around 640 pF, approximately the geometric mean of the range. Thus there would have been some (probably moderate) errors at the extremes. I imagine that Quad would have supplied upon request circuit details (for wiring into the S position of the board) for cartridges outside of this capacitance range, and for those like the Connoisseur SCU1 that did not include inbuilt mechanical equalization. This approach was actually a step change for Quad, in that for the preceding QCII and 22 control units, the crystal and ceramic cartridge input adaptors were basically intended for cartridges that were not self-equalized, and so step-type equalization was provided, namely the magnetic curves (of which several were provided) rotated π/2. Probably when the QCII was released, in 1953, most crystal cartridges, at least those with hi-fi pretensions, were not self-equalized, and ceramics were just emerging. E.g. see the attached EMI advertisement from Wireless World 1954 April. By the time the Quad 33 was designed, self-equalization for ceramics was the norm, albeit not universal.

Both Burrows and Linsley Hood observed that ceramic cartridge inbuilt equalization was not always fully effective, and made provisions for variable step equalization. Once one got to that level of complexity, then the need for variable correction with the low-impedance approach might have been more tolerable. Interestingly, whereas Linsley Hood used a jfet buffer for his simple high-impedance circuit, for the variable step equalization case, he reverted to his shunt feedback RIAA stage, but with 4M4 on the input leg, which I wonder about.

Provision for ceramic cartridge matching and equalization must have been something of a conundrum for the amplifier makers back in the late 1960s and early 1970s. For higher quality units with which magnetic cartridges were more likely to be used, there was probably a reluctance to divert too much resource to the ceramic option, which was provided more as a courtesy. In that sense, the high-impedance approach, which typically required a dedicated additional active device, would have been less attractive than a low-impedance solution built around the RIAA stage.

Thus quite a few makers, I think including Leak and Rogers, adopted the “bung it into the RIAA input – maybe with reduced gain – and then look the other way” approach. Then it was up to the more fastidious users of ceramic cartridges to install cartridge-specific decompensation networks, if in fact they knew that such were desirable (although they may have been mentioned in the cartridge operating manuals). I wonder how many dealers were able and inclined to help customers in this way, and whether or not ready-made plug-in adaptors were available at all. The Quad approach, low-impedance with correction according to cartridge capacitance was as far as I know not common. Apparently plug-in adaptors were available for the Disc 2 input of the Sugden C51 control unit, the range of these being inclusive of those for ceramic cartridges, but the details are unknown. Most likely they would have been low-impedance solutions, though.

One maker who did take the high-impedance route was Tripletone. Its amplifiers were aimed at that level of the market where ceramic cartridges such as the Decca Deram were most likely to be used, so evidently some care was taken. Initially it used a bootstrapped emitter follower to obtain the necessary high impedance, changing later to the Mullard TAA320 IC, which had a mosfet input stage.

Ferrograph also used the high impedance approach in its F307 amplifier, which was more upmarket than the Tripletone models. It did this by including a jfet source follower as the input buffer for all of the high-level/line inputs. (A separate RIAA stage was used for magnetic cartridges). In fact this approach appears to have been copied over from the Series 7 tape recorder, which had a jfet buffer for its line input. Although this allowed direct connection of ceramic cartridges, I doubt that this was its primary raison d’être. More likely Ferrograph saw that its first generation solid state tape recorders would sometimes be used in systems that included valve amplifiers and control units, and that the tape recording outputs of these were almost always unbuffered, and needed to see load impedances of at least 500k, sometimes 1M minimum being specified. Hence the 2M2 line input. (In the same vein, with its A77 tape recorder, Revox used bootstrapped bipolar circuitry to achieve 1M for the auxiliary input.)

Outboard high-impedance pre-amplifiers evidently were available, e.g. see the attached Wireless World advertisement from 1969 June, but I suspect neither well-known nor widely distributed.

Originally Posted by Radio Wrangler

Hugh Walker did a decent magnetic RIAA stage with three transistors, then Peter Baxendall did a better fit to the RIAA curve with an NE5534 opamp with an added R-C pole after the stage. Doug Self later came along with his 'DIY opamp' sort of preamp.

In hindsight, it is perhaps surprising that High Walker – and for that neither matter Bailey and Dinsdale before him - did not include the added passive R-C pole on order to correct the HF error in the RIAA curve that was characteristic of series feedback circuits. Baxandall I think had prior association – perhaps “ownership” of – this technique from his connection with the Quad 33 design, as its RIAA stage did include a passive R-C pole. It was a two-transistor not a three-transistor circuit though, having been designed when active device costs were such that the total count had to be watched carefully. Effectively the emitter follower was still there, but outside of the feedback loop and serving not only as the output load for the RIAA stage, but also as the input buffer for the Radio inputs and output buffer for the tape recorder outputs. Had the 33 been designed a couple of years later, one imagines that a much relaxed device count obtaining from lower device costs would have allowed for a three-transistor RIAA stage, and a sidechain tape output buffer.

Originally Posted by Radio Wrangler

Wireless World used to be very fertile ground for interesting circuits.

Indeed, and for some interesting debates, particularly where amplifiers were concerned. Not always were the proceedings gentlemanly, either. I seem to recall that there was a time in the 1970s when Stan Curtis got into a stoush with Reg Williamson and Peter Baxandall, and he ended up referring to them both as “old men of the industry”. Baxandall was a model of clarity when it came to technical writing, and usually quite calm. But I had the impression that he was a bit miffed back in 1969 when Shaw beat him to publication at least when it came to the use of a diode to iron out quasi-complementary output stage asymmetry. Still, that circuit technique is usually referred to as the “Baxandall diode”, or sometimes as the “Baxandall/Shaw diode”, but I have never seen it as the “Shaw diode”. He also wrote about the Quad 303 output triples in more detail than did Quad itself, but I suspect that in this case the idea was not his, but came from within Quad (maybe from Collinson).

Now for some perspective though on just how far it is necessary to go when it comes to matching cartridges to amplifier inputs. At one of the late 1970s London hi-fi shows, Quad featured a new disc input board for its 33 in which the M1 position was specifically configured to best match the Shure V15 III. Thus it was 3.8 mV, 47 k with a bunch of parallel capacitance (I forget just how much) in place of the standard 2 mV, 68 k, virtually pure resistance (within ±5 degrees I think Quad used to advertise). Over on the Shure stand, the V15 III was being demonstrated via a Quad 33/303. So naturally enough, I asked the question about the 33 special input board. The answer, presented laconically and with a Gallic shrug, was that although they knew about it, they were not using it. Left unsaid was that they thought it didn’t matter that much.

Cheers,

[Radio Wrangler]

And there is one of the biggest questions of our time. How small does an error in frequency response or any other characteristic need to be before it doesn't matter?

Currently hifi aficionados have taken the approach that no error is inaudible and that all differences produce life-changing differences in sonic performance. Looking from outside, I see the size of the hifi market falling steadily, and the prices going even higher. I think the whole thing is mirroring gravitational collapse. Those inside seem not to be able to notice things outside their event horizon, though, which seems the reverse of a true gravitational singularity.

Meanwhile, the escape for those who just want to get on with enjoying some good music is to find some nice, classic hifi gear from a golden age when things had improved, but before things went silly.

Equally meanwhile, the majority of people listen to music on car radios, iPods and TV sets. The days of a significant proportion of homes having a hifi system have gone. But that era has left us a lot of nice gear to play with.

Hugh knew of the discrepancy between exact RIAA and the usual two-capacitor network, so he probably thought it not worth pursuing. I'll try to remember to ask him the next time we bump into each other.

Back in the day, I suppose the ceramic cartridges would be worth doing a decent job of correcting frequency response, but did the crystal ones change characteristic with humidity?

David

[Synchrodyne]

Originally Posted by Radio Wrangler

Hugh knew of the discrepancy between exact RIAA and the usual two-capacitor network, so he probably thought it not worth pursuing. I'll try to remember to ask him the next time we bump into each other.

I’d guess that to be the case. I took another look at Walker’s Wireless World 1971 May article. The RIAA curve certainly shows the upper turnover characteristic of series feedback circuits. But the error at around 15 kHz looks to have been negligible, and it was probably still minor at 20 kHz.

The gain of the RIAA stage I think would be a factor here, in that the lower the gain, the greater would be the error. The Walker circuit had a gain of around 42 dB (240 mV out for 2 mV in), so was quite high gain. The Quad 33 RIAA stage had a gain of 34 dB (M1 position, 100 mV out for 2 mV in) or 25 dB (M2 position, 100 mV out for 5.6 mV in). So at least in the M2 position, the HF equalization error was likely large enough to warrant the use of passive R-C correction. The same R-C circuit was used for both sensitivities, which was theoretically inexact, but evidently the residual error either way was small enough not to worry about. I suppose one advantage of the passive correction is that by keeping to the 75 µs curve, it provides a low-pass filter to dispose of any supersonic detritus that may be around, and being part of the standard curve, it is legitimate and so presumably acceptable to those who otherwise see high pass filters, etc., as the work of the devil.

Walker’s approach to ceramic cartridge equalization and matching is pertinent to this thread, and he had a foot in each camp.

For lower output ceramic cartridges, he used the RIAA stage, with 200k input impedance and with correction for about 600 pF self-capacitance, which was stated to include the majority of the better cartridges. This I think tends to validate the choice made earlier for the Quad 33. It seems that the 200k input impedance was chosen to give a turnover frequency (1.5 kHz) with 600 pF cartridges that aligned with the treble tone control turnover frequency, thus allowing the latter to be used to correct for different cartridge self-capacitances and imperfections in the inbuilt mechanical equalizing.

For higher output ceramic and for crystal cartridges Walker used a “flat” 2-transistor amplifier (that looked a bit like a conjugate pair) with a 2M input impedance.

Given its relevance, I have attached a copy of the WW article.

Originally Posted by Radio Wrangler

Back in the day, I suppose the ceramic cartridges would be worth doing a decent job of correcting frequency response, but did the crystal ones change characteristic with humidity?

With both humidity and temperature I expect. But when they were used for hi-fi applications, some amplifier makers at least seem to have taken the approach of correcting against their nominal characteristics, so that humidity and temperature errors were then attributable to the cartridge, not the amplifier.

Some of the crystal cartridge makers seem to have argued otherwise, as the attached Acos advertisement from 1949 shows. Although this might have been more in the nature of saying something like: “They’re actually not as bad as they are.”

Cheers,

[robbo18]

Hi,
I stumbled across this forum looking for something else.
It may be a bit late to add anything of interest but nevertheless...
I noticed some mention of the 9TAHC cartridge back in post 23. I assume that this was the 9TAFHC manufactured by Sonotone.

The following link is to a page on the Sonotone cartridges:
http://www.roger-russell.com/sonopg/sonopc.htm

There is a small diagram of the preferred method of terminating the Sonotone cartridges overlaid on the frequency response of the cartridge.

As a matter of interest I owned (and still do own) a couple of Micro-Acoustics 'ceramic' cartridges, specifically the 630 and 830 cartridges from the early 1980s. These cartridges were excellent performers and outperformed almost every other cartridge I ever owned including the strain gauges and moving coils. One of their great advantages was that unlike all the moving coil cartridges I owned there were immune to temperature and humidity issues - important if you live in a sub-tropical climate. There were also the best tracking of all the cartridges I ever owned (once compensation for their lack of mass was applied).

The Micro Acoustics cartridges has compensation circuits integrating into cartridge to convert the output signal back to a velocity characteristic. I did write to Micro Acoustics urging them to put out a special series without the compensation network so that a much simpler preamp could be utilized but they were not interested. Web page on Micro Acoustics:
http://www.roger-russell.com/sonopg/micropg.htm

Regards,
Bob Nicholson

[Synchrodyne]

Thanks very much for that posting and the links – most interesting reading!

Re the Micro-Acoustics electret cartridges, it is possible that their natural output levels would have been too low to allow their use with typical unequalized amplifier ceramic cartridge inputs (or line/high-level inputs for that matter), which had sensitivities of say 50 mV and upwards. The inbuilt compensation networks probably did attenuate to some extent, but whether without them the natural output was then high enough for such inputs is unknown.

Also, assuming that the natural output of the electrets was an unmodified amplitude response, then step equalization – with a 12.5 dB step - would still have been required to achieve an overall flat response. Amplifiers equipped to do this – outside of the home construction field – would have been quite rare. Whilst step equalization circuitry might have been included in the cartridge, perhaps it was thought easier to convert the amplitude response to a velocity response (by simple differentiation) and so make it suitable for the ubiquitous RIAA-equalized magnetic cartridge amplifier inputs. Effectively the standard RIAA velocity curve then “undid” the differentiation and provided the step.

That Sonotone offered plug-in “velocity equalizers” as early as 1960 is indicative. Apart from a growing unavailability of dedicated ceramic cartridge inputs, it is possible that even when they were available, they were seen as likely to have been less well engineered than magnetic cartridge inputs. (The comments in the Burrows article attached upthread seem to support this notion.)

The Sonotone article does though include the almost customary conflation of amplitude response and lack of need for equalization – about which Burrows was unhappy - thus: “Another unique feature of the ceramic cartridge is that it's response is amplitude based instead of velocity based like magnetic cartridges. This means that phono equalization is not required.” The first sentence is correct, but it does not provide the basis for the second. Rather, the equalization was not required because (in this case) it was done mechanically within the cartridge. Were that not so, then for a pure amplitude response, external equalization would be required. As Linsley Hood noted, and in which Burrows concurred, inbuilt mechanical equalization was not always fully effective, such that some external correction was sometimes desirable.

The plug-in velocity equalizers effectively undid the built-in mechanical equalization, and the lowish impedance presented by the network in conjunction with the RIAA stage input did the differentiation, thus generating a velocity-type signal. But it was surely the “long way around”, as the signal was equalized, de-equalized and then re-equalized.

Cheers,

[robbo18]

Hi Synchrodyne,

Unfortunately, I never really found out how Micro Acoustics (MA) equalized the cartridge. I always assumed it was partly mechanical as this occurs to some extent in most (all) cartridges. Short of destroying the cartridges I have no way of determining the nature of the electrical compensation unit, or indeed the raw output or impedance of the transducers. What is interesting is the claim of a purely resistive output impedance while still managing to create the shelf in the middle of the RIAA curve - normally this would require a pole zero pair which would have consequences on the output impedance of a passive network.

I have only glanced through the article you reference but the above claim of purely resistive output impedance has raised some interesting questions. The author makes some statements that from my experience do not apply to these much later cartridges. Also, on first glance there are some statements that I don't agree with. Further,I am not sure that I agree with your last paragraph as there were two basic equalizations rather than one being applied, defeated and then reapplied. When I get some time I will have a read and a long think or what passes for it these days.

Each MA 630 and 830 cartridge came with an individual frequency response curve which from memory was flat from below 20Hz to well past 20kHz within .2 db (perhaps less). Obviously these were created using a mechanical test rig rather than a record but they do show how well integrated the entire mechanical and electrical system was. I would dig up the curves but we are currently traveling Australia in a motorhome - not the best place for a turntable even if I were still using such technology. I have not been successful in finding any independent testing of the cartridges, although I do remember that I read it before purchasing them.

I think the subsequent performance of the cartridges was a tribute to what must have been gifted mechanical and electrical engineers. Sadly, I cannot say the same about many of the moving coil and other cartridges I have owned.

Using the cartridge without the equalizer would have required a purpose built preamplifier but this wasn't that unusual in the mad seventies. I build quite a few of them, several for strain gauge cartridges which are also displacement devices. The strain gauges not only required custom equalization (given the variability between examples - individual equalization!) but also a constant current source to derive a signal. I believe at least one company is still playing with strain gauges albeit at obscene prices.

Back to MA. I suppose it was naive of me to think that a relatively large company such as MA would have been interested in such a small niche market. Simplification of the preamp was actually one of the lesser reasons I would have liked the compensation network removed from the cartridge. The main ones were to further reduce the size and mass of the cartridge and perhaps to improve the cartridge body resonance. At the time I was playing with low mass tonearms and had started to get very interested in the complex mechanical resonances of the tone-arm/cartridge combination.

Cartridges were my Achilles heal, I could accept every part of my system except cartridges until the Micro Acoustics and strangely enough the B&O MCC20CL which was a very good cartridge although perhaps not quite as 'dynamic' as the Micro Acoustics. The B&O also came with an individual frequency response plot which was almost ruler straight. The B&O was rather different in that it used a moving micro-cross system that resulted in a very compact (though heavy) cartridge body. The B&O was another cartridge that was immune to weather conditions. I had loaned the MCC20Cl installed into a Beogram 4002 to a friend with a massive record collection, he used it for all his LP listening for at least 12 years until the turntable stopped. I know what is wrong with it and one of these days I will fix it (simple enough fix) although the cartridge is probably in need of a new stylus.

Enough rambling for the moment. When I get a chance I will read the referenced article. I can't say that I will have a great deal to add, my engineering career (Telecommunications) came to an end 13 years ago after a CVA (Cranial Vascular Accident), subsequently abstract mathematics and quite a lot of 'engineering style' thought are foreign concepts to me these days.

Regards,
Bob

[robbo18]

Hi Synchrodyne,

I have had a chance to read the Burrows article but I did not find much was of interest to me. Most of it was a 'debunking' of claims made in the 50s which did not really relate to later periods, the 2nd article was even less of interest to me.

I have located some information of the Micro Acoustics System II cartridges which include the 630 and 830 (SQ) series. There is a lot of 70/80s hype from the marketing dept in the documentation but there is some good information and an excellent diagram of 630 series cartridge.

http://www.vinylengine.com/library/m...ystem-ii.shtml

You will have to register to download the information. Unfortunately one of the articles is missing which is most unfortunate as it was on cartridge/arm interactions.

There is also a catalogue of the earlier cartridge series with some information on the production process at:

http://www.vinylengine.com/library/m...rtridges.shtml

Also, one of the patents:
http://www.google.com.au/patents/US3952171

The diagrams in the Series II documents show the complex internal structure of the cartridge especially in respect to the damping system. Whether the cartridge is mechanically equalized and if so by which damper(s) is not stated. I have read postings from MA employees that state that the complexity and thus expense of the cartridges was probably one of the reasons for the company's demise although I am not sure whether these people were in a position to know that information.

As you can see from the catalogues, the MA cartridges were not cheap but in my opinion their performance was outstanding especially when it came to tracking - totally the opposite to that spawn of the devil - the Decca London cartridge - with which one of my friends persisted to tinker.

The Panasonic strain gauge cartridges were also relatively complex (although nowhere near as much so as the MA) compared to MC and MM cartridges. They also were relatively poorly made (but cheap). One peculiarity was that the semiconductor strain gauges were partly exposed and like most semiconductors photo-sensitive so strange hum problems were often tracked back to AC turntable lights. Sadly, the strain gauges were so variable that getting them to perform requires a test record and considerable work to get them equalized, channel imbalance was also quite common and could be considerable (>3db).

Patent for the Panasonic strain gauge is pretty much as they were put into production:

http://www.freepatentsonline.com/3975025.pdf

One day, just as an experiment, I might dig out all the cartridges I still have that are superficially okay, mount them up and see how they have fare. The MAs are about 35 years old so I don't hold much hope for the condition of all the dampers although the suspension (which is separate from the dampers) may still be okay. The MCs are all toast, last time I looked the stylus of one of them was at about 3 or so degrees off centre - another suspension failure. First I will have to service my Pioneer PL L1000A turntables (a matched pair), cross my fingers and pray that they don't go up in smoke when I turn the power on.

Regards,
Bob